The engineering and design cycle was of instrumental importance to the successful creation of the D-tector and the composite part BBa_K3957000.

Our target for the project is to create a cheaper and less machinery intensive vitamin D sensor that could accurately measure the concentration of vitamin D in a blood serum sample such that the deficiency can be classified.

Preliminary research was done into the current methodologies to tackle vitamin D detection of measurement. It was determined that, especially for northern British Columbia, the major limiting factor was an unavailability of scientific equipment to perform such tests; therefore our system must make use of minimal and accessible scientific equipment. After some research, two possible ideas to address the problem were formulated: an expression system that takes advantage of natural metabolic pathways, and a chimeric GPCR.

The Engineering Cycle for GPCR

GPCR Cycle

Research
Imagine
Design

The Human Metabolic pathway

Cycle 1

Preface
Research
Imagine
Design
Improve

Megalin, Cubilin and the purify step

Cycle 2

Preface
Research
Imagine
Design
Improve

25-hydroxyvitamin D-1 alpha hydroxylase and the Activate step

Cycle 3

Preface
Research
Imagine
Design
Improve
Learn
Testing

VDR-FRET and the Detect step

Cycle 4

Preface
Research
Imagine
Design
Improve
Learn
Testing

Plasmid design

Cycle 5

Preface
Research
Imagine
Design
Improve

Protein production management

Cycle 6

Preface
Research
Imagine
Design
Improve
Testing

At home vs benchtop

Cycle 7

Preface
Research
Imagine
Design

1. Wacker, M; Holick, M. F. Vitamin D-Effects on Skeletal and Extraskeletal Health and the Need for Supplementation. Nutr. 2013, 14(5), 111-148. 10.3390/nu5010111

2. Cheskis, B.; Freedman, L. P. Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol. Cell. Biol.. 1994, 14(5), 3329-3338. https://doi.org/10.1128/MCB.14.5.3329

3. Protein. https://www.ncbi.nlm.nih.gov/protein(Accessed 27-09-2021).

4. Vemulapati, S.; Rey, E.; O’Dell, D.; Mehta, S.; Erickson, D. A quantitative point-of-need assay for the assessment of vitamin D3 deficiency. Nat. 2017, 7, 14142-14150. https://doi.org/10.1038/s41598-017-13044-5

5. Sawada, N.; Sakaki, T.; Kitanaka, S.; Takeyama, K.; Kato, S.; Inouye, K. Enzymatic properties of human 25-hydroxyvitamin D3 1alpha-hydroxylase coexpression with adrenodoxin and NADPH-adrenodoxin reductase in escherichia coli. Eur. J. Biochem. 1999, 265(3), 950. doi: 10.1046/j.1432-1327.1999.00794.x

6. Imperial 2007. T7 promoter. http://parts.igem.org/wiki/index.php?title=Part:BBa_I719005 . (Accessed 27-09-2021).

7. Anderson, J. Ribosome Binding Site Family Member. Part:BBa J61100 - parts.igem.org. (Accessed 27-09-2021).

8. Shetty, R. Double terminator. http://parts.igem.org/Part:BBa_B0015. (Accessed 27-09-2021).

9. Klupt, K. Vitamin D receptor binding protein with cysteine modification(s) to bind to a biosensor and FRET to. http://parts.igem.org/Part:BBa_K3515011. (Accessed 27-09-2021).

We looked into various GPCRs in yeast and found that the yeast pheromone receptor STE2 was most suitable, as it had previously been targeted for directed evolution in the literature.

We need to develop a mutant with preferential binding to DBP.

- The natural ligand of STE2 is alpha-factor (a small peptide); as such, it is more feasible to target DBP (also a peptide) than vitamin D directly.

1) A Monte-Carlo system would be developed to optimize the binding affinity of DBP to the yeast pheromone receptor.

2) This system would include a loop of the following activities:
-A program to randomly mutate amino acid regions of the STE2 GPCR that are key to binding activity
-A program/webservice to model the mutated STE2 GPCR receptor region, designating three dimensional atomic coordinates
-A program/webservice to check the binding affinity of DBP in direct comparison to the binding affinity of alpha-factorA program to receive scoring values for vitamin D / alpha-factor binding and determine which version of the GPCR will move forward to the next iteration.

3) We tried using computer-aided directed evolution of enzymes (CADEE). We attempted to use the CADEE program developed by the Kamerlin lab from Uppsala University; unfortunately, we were unable to get the program to function.

4) At this point, we decided to shift focus to our other idea: a biosensor that mimicked vitamin D metabolism.

Biomimicry was a natural place to start our research, and thus we began by studying the metabolic processes involving vitamin D.

We started by researching and understanding the metabolic pathway vitamin D follows in the human body. There are into four major steps^{1, 2}:

- 25-hydroxyvitamin D (25(OH)D) is bound to a storage protein in the bloodstream
- 25(OH)D enters a kidney cell through large multidomain proteins called megalin and cubilin, which separates the vitamin from its storage protein
- 25-hydroxyvitamin D-1 alpha hydroxylase
(CYP27B1) converts 25(OH)D to 1,25-dihydroxyvitamin D (1,25(OH)₂D)
- 1,25(OH)₂D binds to vitamin D receptor’s (VDR) ligand binding domain, which then induces downstream effects by dimerizing with retinoid X receptor (RXR) and then binding to DNA enhancer sequences.

1) We could express all the proteins necessary for vitamin D gene regulation in the human body in E. coli with the endpoint enhancer sequences promoting a colorimetric or fluorescent protein. Change in colour or fluorescence could be the determinator of vitamin D concentration.
2) We initially wanted it to be an at-home test, and so we considered a colour change or agglutination. Fluorescence requires lab equipment (which precipitated our shift to a benchtop system)

1) The following proteins (and associated DNA base pair length) would have to be expressed in 1-2 plasmids due to total length³
- Megalin (13 965 bp)
- Cubilin (10 869 bp)
- 25-hydroxyvitamin D-1 alpha hydroxylase (1 524 bp)
- VDR (1 281 bp)
- RXR (1 386 bp)
- Fluorescent protein (mCherry: 708 bp)
- Each protein would individually need to be tested for functionality
- BioBrick, Golden Gate, and/or Gibson assembly could be used.

To follow this pathway we would need to express, ligate and transform a multitude of proteins; this system was inelegant and unfeasible, especially given our resources and timeline. Therefore, a better system was needed to be developed to simplify the signal transduction process.

1) In the preceding plan, megalin and cubilin are the two largest proteins. The combination of these two would use more than the 20 000 free nucleotides IDT gives each team. Additionally, it was unknown how these large cell membrane proteins would interact within the gram negative bacterial outer/inner membrane system.
2) A new way to separate vitamin D from its storage protein was needed.

1) We were unable to find alternative protein complexes that could separate the vitamin D.
2) Research into the current vitamin D testing methods revealed that the separation of vitamin D from blood serum was not the cost and equipment limiting step.
Organic solvents, such as acetonitrile, are currently used to separate vitamin D from its binding protein.

Vitamin D could be purified from a blood serum sample and applied to our cells, which produce a signal.

Using organic solvents, we no longer have to express megalin and cubilin and the only change to the project design is a new purify step in the protocol.

No further improvements have been made and this step is now deemed the “purify” step in vitamin D detection.

In the human body VDR interacts with 1,25(OH)₂D; therefore, for our project to sense the storage form of vitamin D (25(OH)D), it must be hydroxylated.

Sawada, N et al. (5) showed that 25-hydroxyvitamin D-1 alpha hydroxylase could be expressed in E. coli without additive proteins or molecules.

We could express 25-hydroxyvitamin D-1 alpha hydroxylase in E. coli so that the storage form of vitamin D can be converted such that it can interact with VDR.

1) The genetic code of 25-hydroxyvitamin D-1 alpha hydroxylase was taken from NCBI by converting the protein amino acid sequence to nucleotides; this way, introns would not need to be removed⁴.
&nbsp- The genetic code was codon optimized to 14% in SnapGene based on E. coli codon tables.

2) A His-tag was added for protein purification.
3) Based on available restriction enzymes, BioBrick was chosen as a gene insertion method
- Multiple restriction enzyme sites had to be removed so that this gene would be BioBrick compatible. This was done by manually changing the nucleotides while maintaining the codons adjacent to the restriction enzyme site.

4) Since we no longer needed the IDT free nucleotides for megalin and cubilin, we were able to make “functional” units of 25-hydroxyvitamin D-1 alpha hydroxylase. This means that in one BioBrick part we had:
- BioBrick prefixes and suffixes
- A T7 promoter site (BBa_I719005)⁶. The promoter was chosen based on the engineering cycle for protein production management.
- A ribosome binding site (BBa_J61100)⁷
. The RBS was chosen based on maximal efficiency.
- The gene
- A His-tag
- A terminator (BBa_B0015)⁸
. A double terminator was chosen it was well characterized on the iGEM parts page

1) Once the BioBrick was completed, the part was sent to IDT and it was determined the complexity score was too high. To create our project, we split the part down the middle and ordered two separate parts with lower complexity scores individually.
- We then had two separate parts that needed to be combined; we determined that PCR SOEing would be an effective technique to accomplish this.

1) After testing, an effective PCR SOEing protocol was determined. Unfortunately, the product was unable to be amplified through regular PCR.
- After careful consideration of the goals of the project and a workable timeline it was determined that there was no need for the creation of a successful PCR SOEing product amplification protocol, and that the original PCR SOEing product would be enough for future BioBrick ligations.

1) To confirm the ligation and transformation of the 25-hydroxyvitamin D-1 alpha hydroxylase BioBrick functional unit, a colony PCR was performed
- Colony PCR suggested correct ligation of 25-hydroxyvitamin D-1 alpha hydroxylase BioBrick functional unit in pSB1C3

2) Sequencing was done to confirm DNA sequence
- There was a deletion/frameshift upstream from the His-tag invalidating the part
- The 25-hydroxyvitamin D-1 alpha hydroxylase BioBrick functional unit was set aside so that we could focus on testing and characterizing BBa_K3957000.

To detect the concentration of vitamin D in a sample following the natural mechanism, three proteins would need to be expressed and tested: VDR, RXR, and an expressible fluorescent protein. To transform, express and test all these proteins would be an unreasonable undertaking for our team so a better solution would need to be found.

1) After we searched the iGEM Parts Registry, we found a chimeric protein consisting of the ligand binding domain from VDR, mCherry, and mNeon green.
- This chimeric protein worked under the foundations of Förster resonance energy transfer or FRET for short.
- The chimeric protein was made in the program Chimera and had not been expressed before, but after careful consideration we decided it was the best move forward for the project.
- Multiple communications were made with the creators of the part from the 2020 Queen's iGEM team to get a deeper understanding of its inception and intended use.

2) Inspection of the measurement abilities of FRET systems was evaluated, and it was determined that the FRET protocol could be able to measure the levels of vitamin D in a sample.

After the 25(OH)D is converted to 1,25(OH)₂D in the activate step it can immediately interact with the VDR-FRET to produce a signal thus creating the “detect” step in our project.

1) The genetic code of BBa_K3515011 was copied from its iGEM Part page.⁹
- The genetic code was codon optimized to 14% based on E. coli codon tables.
- On the parts sequence there were three stop codons in the reading frame that were removed. These stop codons were after each domain used to create the chimeric protein: mNeon green, the ligand binding domain of VDR, and mCherry.

2) A His-tag was added for protein purification.

3) Based on available restriction enzymes, BioBrick was chosen as a gene insertion method
- Multiple restriction enzyme sites had to be removed so that this gene would be BioBrick compatible. This was done by hand changing the nucleotides while maintaining the codons adjacent to the restriction enzyme site.

4) Since we no longer needed the IDT free nucleotides for megalin and cubilin, we were able to make “functional” units of our chimeric part. This means that in one BioBrick part we had:
- BioBrick prefixes and suffixes
- A T7 promoter site (BBa_I719005)⁶. The promoter was chosen based on the engineering cycle for protein production management.
- A ribosome binding site (BBa_J61100)⁷
. The RBS was chosen based on maximal efficiency.
- The gene
- A His-tag
- A terminator (BBa_B0015)⁸
. A double terminator was chosen it was well characterized on the iGEM parts page

1) Once the modifications to the sequence were completed the part was sent to IDT, and it was determined the complexity score was too high. So we split the part down the middle and ordered two separate parts with lower complexity scores individually.
- Now that we had 2 separate parts that needed to be combined, we determined that PCR SOEing would be an effective technique to use.

2) After testing an effective PCR SOEing protocol was determined. Unfortunately, the PCR SOEing product was unable to be amplified through regular PCR.
- After careful consideration of the project goals and a workable timeline, it was determined that there was no need for the creation of a successful PCR SOEing product amplification protocol, and that the original PCR SOEing product would be enough for future biobrick ligations.

1) After some testing it was revealed that the fluorescence reader we were using, the Synergy^TM 2 Multi-Mode Microplate Reader , was unable to test the fluorescence of mNeonGreen with the filters available to us. It can detect the fluorescence of mCherry, but it was unable to measure at the maximal wavelength. As for the FRET system, the plate reader can accurately excite mNeonGreen and measure the emission of mCherry.
- This means we could only use the data from FRET to describe our protein rather than each individual fluorophore.

2) It was revealed that VDR-FRET was able to provide a detectable background fluorescence; however, the FRET signal did not change with increasing and decreasing 1,25-dihydroxyvitamin D concentrations, suggesting that the VDR-LBD is not able to interact with the 1,25-dihydroxyvitamin D to cause conformational change.

1) To confirm the ligation and transformation of the VDR-FRET biobrick functional unit, colony PCR was performed.
- Colony PCR suggested correct ligation of the 25-hydroxyvitamin D-1 alpha hydroxylase BioBrick functional unit

2) Sequencing was done to confirm the DNA sequence.
- There was a codon change in mNeonGreen at the 72nd codon: His → Asn

With limited resources available, the decision on which plasmid to use was based on what was available to our lab.

1) Looking at our inventory, the 2070 bp plasmid pSB1C3 was available and included a BioBrick prefix/suffix zone
- We had a pSB1C3 stock with a red fluorescent protein (RFP) from a previous year's distribution kit.
-- The RFP was inserted in the BioBrick site, so we planned to use restriction enzymes and gel purification a linearized pSB1C3 with the correct cut BioBrick sites for the addition of our BioBrick functional.

2) We also had a stock of chloramphenicol from previous iGEM years which is the same antibiotic resistance given by pSB1C3.

We could create two plasmid stocks, one with the VDR-FRET BioBrick inserted into pSB1C3, and another with both VDR-FRET and the 25-hydroxyvitamin D-1 alpha hydroxylase. This allowed us to test the effects of 25(OH)D and 1,25(OH)₂D on the whole system and the individual functionality of VDR-FRET.

1) We removed the RFP from pSB1C3 by digesting the plasmid with SpeI and EcoRI.
2) VDR-FRET was prepared for ligation into the plasmid by digesting with EcoRI and SpeI
3) 25-hydroxyvitamin D-1 alpha hydroxylase was prepared for ligation into the plasmid by digesting with EcoRI and SpeI.

We tried a 10 minute ligation combining pSB1C3 and VDR-FRET, but it was unsuccessful.
- It was determined that since pSB1C3 is only 2070 bp and VDR-FRET 2426 bp, the ligation will be less efficient than expected, so we switched to an overnight ligation. The overnight ligation provided satisfactory results that allowed us to progress with the project.

This project revolves around the production of large quantities of proteins from which we can purify and use to test samples. A strong promoter made sense to use in our cell cultures.

1) Research for this engineering goal came in two parts: what could be done, and what can be done.
2) Previous iGEM teams had a supply of competent DH5α cells
3) Through generous donation by the Murray and Rader laboratories at UNBC, the iGEM team was able to obtain both DH5α and BL21(DE3) non-competent E. coli cells.
4) Based on this, the viral T7 promoter was chosen to enhance the production of our protein. This is due to the fact DH5α cells lack the ability to produce T7 polymerase; whereas, BL21(DE3) cells have the T7 polymerase under inducible control.

With two cell lines and differential abilities to express the proteins, the DH5α cells could be used to create a storage cell line and the BL21(DE3) cells could act as "factories" to produce our proteins.

1) It was determined that the cloning and transforming effort would be done in the DH5α cells since they cannot produce the desired proteins.
2) The induction of protein production would be done in the BL21(DE3) cell line.

Transformation of the DH5α cells with the previous iGEM team's competent DH5α cells did not produce results.
- We tested the antibiotic resistance on the plates, the plasmid itself, and the protocol used, and it was determined that on the cell's creation their competency was too low to transform with reasonable amounts of plasmid. To push through this, the donated DH5α and BL21(DE3) non-competent cells were used to create fresh competent cells which had a high enough competency to transform with our plasmid.

During protein production a multitude of cell lines were grown up in an LB broth: an IPTG induced transformed BL21(DE3) cell line, an uninduced transformed BL21(DE3) cell line, and an uninduced untransformed BL21(DE3) cell line. After growing up the cell lines, both the uninduced and induced transformed BL21(DE3) cells had protein production. Further research suggested that the LB broth we used for the culture growth medium contained lactose, which was enough to induce the cultures.

Originally, an at-home test would be the gold standard for project design; However, one of the main reasons behind the cost of vitamin D testing in northern British Columbia is the complications associated with getting one. An at-home test could side step these; however, the concept of a benchtop kit for laboratory use was not completely taken off the table.

Originally, an at-home test kit was able to fit into our project when megalin, cubilin, RXR and VDR were still on the table; unfortunately, after the decision to move to VDR-FRET and the necessity for a fluorescence reader, the idea of an at- home kit was no longer feasible.

The project could be a premade kit sold to a local health facility. Purified protein samples could be stored and frozen in a well plate, such that a set volume of blood serum sample would be added to a well and the fluorescence could be measured.

The project never made it to a proof of concept as VDR-FRET proved to be fallible.

Team:Northern BC/Engineering

The engineering and design cycle was of instrumental importance to the successful creation of the D-tector and the composite part BBa_K3957000.

Our target for the project is to create a cheaper and less machinery intensive vitamin D sensor that could accurately measure the concentration of vitamin D in a blood serum sample such that the deficiency can be classified.

The Engineering Cycle for GPCR

GPCR Cycle

Research

Imagine

Design

The Human Metabolic pathway

Cycle 1

Preface

Research

Imagine

Design

Improve

Megalin, Cubilin and the purify step

Cycle 2

Preface

Research

Imagine

Design

Improve

25-hydroxyvitamin D-1 alpha hydroxylase and the Activate step

Cycle 3

Preface

Research

Imagine

Design

Improve

Learn

Testing

VDR-FRET and the Detect step

Cycle 4

Preface

Research

Imagine

Design

Improve

Learn

Testing

Plasmid design

Cycle 5

Preface

Research

Imagine

Design

Improve

Protein production management

Cycle 6

Preface

Research

Imagine

Design

Improve

Testing

At home vs benchtop

Cycle 7

Preface

Research

Imagine

Design

1. Wacker, M; Holick, M. F. Vitamin D-Effects on Skeletal and Extraskeletal Health and the Need for Supplementation. Nutr. 2013, 14(5), 111-148. 10.3390/nu5010111

2. Cheskis, B.; Freedman, L. P. Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol. Cell. Biol.. 1994, 14(5), 3329-3338. https://doi.org/10.1128/MCB.14.5.3329

3. Protein. https://www.ncbi.nlm.nih.gov/protein(Accessed 27-09-2021).

4. Vemulapati, S.; Rey, E.; O’Dell, D.; Mehta, S.; Erickson, D. A quantitative point-of-need assay for the assessment of vitamin D3 deficiency. Nat. 2017, 7, 14142-14150. https://doi.org/10.1038/s41598-017-13044-5

6. Imperial 2007. T7 promoter. http://parts.igem.org/wiki/index.php?title=Part:BBa_I719005 . (Accessed 27-09-2021).

7. Anderson, J. Ribosome Binding Site Family Member. Part:BBa J61100 - parts.igem.org. (Accessed 27-09-2021).

8. Shetty, R. Double terminator. http://parts.igem.org/Part:BBa_B0015. (Accessed 27-09-2021).

9. Klupt, K. Vitamin D receptor binding protein with cysteine modification(s) to bind to a biosensor and FRET to. http://parts.igem.org/Part:BBa_K3515011. (Accessed 27-09-2021).