Team:Stanford-Brown/SB16 BioMembrane UV

Stanford-Brown 2016

Float team member Taylor introduces the radiation resistance subproject

The Problem

The high energy content of ionizing radiation in space poses a severe threat to life, technology, and synthetic materials alike in an extraterrestrial context. Energized subatomic particles generated from nuclear events carry over large distances in space, and are capable of tearing through organic materials and altering molecular bonds. On Earth, the vast majority of space radiation is deflected by the magnetosphere. Much of the radiation that passes through Earth’s magnetic field is then filtered out by the atmosphere, allowing only UV, visible light, and some infrared wavelengths to reach the surface. Extended UV exposure can disrupt the molecular interactions within thin membranes, creating micropores that decrease the material’s structural integrity and impermeability.
Figure 1: Radiation Damage Degrades DNA.
Designed by team member Taylor Sihavong in Adobe Illustrator
Balloons flown in Earth’s stratosphere are offered limited UV protection due to thinning of the atmosphere with increasing altitude. Chief Scientist of World View Enterprises Alan Stern stated in a meeting with our team that the materials used in his near-space exploration vessels cannot be reused due to the radiation damage they sustain during flight. The Stanford Space Initiative’s Balloon team expressed to members of our team that radiation exposure was the most significant factor in limiting the lifespans of their balloons. The unmet need for reusability of balloon materials flown in high altitude conditions generates a lack of sustainability that our team aimed to tackle through UV resistance.
On Earth, the main source of concern for radiation damage is ultraviolet radiation (UVR). These wavelengths are between 290 and 400 nm when they hit the Earth’s surface, and do more damage the shorter they get. UV wavelengths between 320 and 400 nm are called UV-A, and are the culprits behind sunburns and cataract formation. UV-A radiation creates indirect DNA damage through the generation of reactive oxygen species (ROS), forming strand breaks in DNA and DNA-protein crosslinks. [1] Shorter wavelengths from 290 to 320 nm are called UV-B wavelengths. UV-B is absorbed directly into DNA and causes damage to DNA at the molecular level. [2] Excessive UV-B exposure can induce significant DNA mutation and cell death, and our bioballoon will be exposed to dangerous amounts of radiation due to its high elevation during flight. Ideally, our balloon coating would have included an interweaving of zinc oxide, a white pigment that excels at blocking and reflecting UV-A and UV-B rays. [3] Zinc oxide is used in many commercial sunscreens, as a solution with as low as 25% zinc oxide content results in protection of at least SPF 20. [4] However, we wanted to produce our UV protector entirely biologically rather than attempting to incorporate an inorganic compound into our membrane. As such, we decided to focus on creating a method of UV protection for our balloon that utilizes a melanin incorperated via a binding agent directly into our membranes.

Our Solution

Figure 2: Chemical schematic of melanin.
Designed by team member Taylor Sihavong in Adobe Illustrator
To install radiation resistance in our biomembranes, the team investigated the absorptive properties of melanin and designed a novel binding mechanism to incorporate the pigment directly into materials. Melanin is the primary source of naturally-occurring UV protection in human skin. In response to human skin damage through UV exposure, keratinocytes release cytokines to stimulate melanocytes’ production of melanin, which functions as a UV absorbent and generates the phenotypic response of darker skin. [5] Melanin when used as a sunscreen has been shown to absorb between 50-75% of UVR. [6] Two main types of melanin exist: eumelanin and pheomelanin, with eumelanin being darker and browner in color and more effective at photoprotection. After significant UV-A absorbance, melanin tends to produce ROS and create single-strand DNA breaks, with pheomelanin being significantly more susceptible to photodegradation than eumelanin. Thus, brown eumelanin was chosen as a good candidate for our balloon’s UV protection.

Melanin Production


Melanin is synthesized by many biological organisms from the aromatic amino acid L-tyrosine through the activity of tyrosinases. These enzymes catalyze the hydroxylation of L-tyrosine to L-DOPA, as well as a further oxidation to produce the cyclical dopachrome. Eumelanin is formed in vivo by the non-enzymatic oxidation and polymerization of dopachrome. [7] The MutmelA gene endogenous to Rhizobium etli encodes a feedback resistant tyrosinase that has been shown to be effective when transfected into E. coli. [8] Based on the work of MI Chávez-Béjar, our team sought to increase the production of L-tyrosine in E. coli through metabolic engineering to shuttle the flow of carbon from the cells’ central metabolism into the shikimate pathway.
Figure 3: Simplified Schematic of our approach to producing melanin.
Designed by team member Taylor Sihavong in Adobe Illustrator

Increasing L-tyrosine Levels

Figure 4: The complete Shikimate pathway.
Designed by team member Taylor Sihavong in Adobe Illustrator
L-tyrosine, along with many other common aromatic compounds, is produced endogenously in E. coli through the shikimate pathway. In this pathways initial reaction, DAHP synthase catalyzes the conversion of Phosphoenolpyruvic acid (PEP) and Erythrose 4-phosphate (E4P) to DAHP. PEP and E4P are naturally derived from glucose, and are present at high concentrations in the cytoplasm of E. coli grown in glucose containing medium. E. coli naturally contain three isoenzymes of DAHP synthase, each of which are inhibited by the increasing concentrations of specific amino acids produced via their activity. This negative feedback limits the production capacity of L-tyrosine in wild type E. coli. To upregulate L-tyrosine production, we investigated and biobricked multiple feedback resistant variants of the DAHP synthase isoenzymes AroG (BBa_K2027013, BBa_K2027014) and AroF (BBa_K2027008).
Genomic AroG and AroF are inhibited in vivo by the presence of phenylalanine and tyrosine respectively, both of which are synthesized through the shikimate pathway. [9] In AroG, two single codon point mutations (Pro150Leu and Leu175Asp) can be altered to generate feedback resistant forms of this isozyme. KU Leuven’s 2013 iGEM team biobricked the natural AroG enzyme and identified the same two mutation sites (found here BBa_K1060000). Accordingly, our team ordered the part from iGEM headquarters, and designed primers to perform site directed mutagenesis at these two locations. Before biobricking these parts, we inserted our two AroGfbr mutants into a pSB1C3 backbone derived from our composite rAIP plasmid ( BBa_K2027000), which included a constitutive promoter, strong RBS, and double terminator that flanked our genes. A similar point mutation to wild-type AroF (Asn8Lys) can instill tyrosine feedback resistance in the enzyme. [10] Having never previously been biobricked, we ordered our AroFfbr gene as two gBlocks from IDT and assembled the construct using Gibson Assembly Protocol into the same composite backbone as our AroGfbr genes.

To further stimulate the shuttling of carbon into the shikimate pathway, we aimed to decrease the metabolic consumption of PEP and E4P, the substrates of DAHP synthase. PEP and E4P are consumed by the glucose phosphotransferase transport system (PTS). Inactivation of this system has been shown to increase the yield of these compounds in E. coli grown in a glucose medium. [11] Instead of spending hours in the lab performing genomic deletions in our host cells, we contacted Dr. Guillermo Gosset of La Universidad Nacional Autónoma de México in request of a live sample of his PTS- tyrR- glucose+ strain VH33tyrR. [10] This strain, derived from W3110 E. coli contained genomic knockouts of PTS and the tyrR negative transcriptional regulatory protein, decreasing the cellular consumption of PEP and E4P and increasing the biosynthesis of and transport of tyrosine. Dr. Gosset readily agreed, sending us samples of his VH33tyrR strain lacking an AroGfbr containing plasmid pRW300, which he could not distribute due to patent regulations. He sent us a secondary sample of W3110 E. coli containing his pMmelAtyrCpheACM plasmid, which included genes for the tyrC protein and mutMelA tyrosinase necessary for melanin production. We extracted the plasmid from the secondary cell sample with a Miniprep kit, but ran into vast difficulties transfecting his VH33tyrR cells with his plasmid and our engineered AroGfbr and AroFfbr plasmids. After numerous attempts at chemical transformations and electroporation (see our lab notebook), we settled on transfecting T7 Express E. coli from New England Bioworks with our AroGfbr and AroFfbr plasmids, along with our plasmid containing MutMelA (BBa_K2027012, see below).

Binding Agent


Coating a balloon with a protective agent requires strict homogeneity to be useful. Kirill Safin, the team lead of Stanford Space Initiative’s Ballooning project, provided insight into the adverse effects of uneven coatings on balloons. Irregular coatings can shorten balloon lifespans compared to uncoated balloons by creating inconsistencies in a membranes tensile properties and allowing the material to stretch more in one location than another. In balloons with flexible membranes, changes in ambient pressure and temperature lead to fluctuations in volume that stretch the balloons’ membranes. This can cause an inelastic coating to become uneven, exposing weak spots susceptible to tearing. Instead of designing a melanin coating for our biomembranes, we looked into alternate routes to incorporate melanin directly into our materials.

Melanin Binding Peptides

Melanoma cells share the characteristic of increased melanin production. Accordingly, many treatments have highlighted melanin as an ideal target molecule for radionuclide therapy. By utilizing melanin binding peptides linked to radioactive subunits, melanoma cells can be localized and delivered doses of radiation to eradicate harmful malignant cells. Researchers pursuing this approach to melanoma treatment have made use of phage display libraries to identify multiple small peptides capable of binding melanin. We sought to create a binding domain that would retain large quantities of melanin by engineering a repetitive gene construct containing many copies of the most successful small melanin binding peptides from various studies. Our team decided to pursue the decapeptide known as 4B4 and a hepaptide called 4D, along with irrelevant control peptides PA1 and P601G for reference. [12] [13]

Repetitive Nucleotide Sequence Assembly

Figure 5: Electrophoresis Gel of the Melanin Binding Proteins.
Assembling large chains of repetitive DNA sequences is a difficult process due to the high probability of improper annealing during construction. In 2015, our advisor Kosuke Fujishima published a novel method for repetitive DNA strand assembly using small dsDNA blocks with overlap-based redundancies. [14] We designing small chunks of DNA, each coding for a single deca or heptapeptide, with identical 3' and 5' single stranded six nucleotide overlaps. By phosphorylating the DNA blocks' 5' ends with T4 Polynucleotide Kinase (NEB), and allowing the strands to ligate overnight with T4 DNA Ligase (NEB), we were able to generate up to 30 repeated segments per strand (shown in gel image, see protocol). Elongation is halted when a building block has a deletion, insertion, or substitution on its 5' overhang, which prevents further DNA blocks from annealing to its sticky end. These errors led to much higher concentrations of strands with fewer repeats than our target length, which we defined as around ten repeated binding domains. Although we were able to form repetitive strands that could be visualized in a gel with all four of our peptides, along with a randomized hybrid of 4B4/4D, we ran into difficulties in gel extracting all but 4D and P601G at a high enough concentration and purity for insertion into plasmid backbones via Gibson Assembly Protocol.
Figure 6: Repetitive DNA block assembly using blunt ended linker as initiation site.
Designed by team member Elias Robinson in Adobe Illustrator
Repetitive 4D and P601G strands were assembled into pSB1C3 backbones with an upstream flanking constitutive promoter and strong RBS. Downstream of the genes, a cellulose binding domain (CBD) extracted from the 2016 distribution kit part BBa_K1321357 via PCR was included as a proof of concept membrane binding domain, separated from the melanin binding module by a GS linker to preserve each domain's function. Following the CBD, we added a FLAG tag, Lumio tag, and His tag for protein presence verification, extraction, and purification. Finally, a double terminator was included before the biobrick suffix to insure termination of transcription.
Figure 7: BBa_K2027011, Melanin-Binding Heptapeptide (4D) with CBD and Flag-Lumio-HisTag
Figure 8: BBa_K2027009, Control Sequence Peptide (P601) for Melanin Binding (Nonpolar-Protic)

Results/Future Directions

Although the 4D Melanin Binding Peptide was the only construct we were able to purify with a Cellulose Binding Domain and a Flag-Lumio-His-Tag, we were determined to increase the radiation resistant properties of our materials. Since cellulose has been a common proof of concept materials to plug into with our chromoproteins and FQ sensor, we decided to build off the work of previous iGEM groups by further validating the cellulose binding domain in a new application. The Cellulose-Melanin 4D linker was incubated with melanin in ethanol since melanin is a relatively insoluble compound. The mixture was then coated on cellulose sheets (and glass filter paper for a control). This linker protein could help serve as a coating for other materials like elastin and latex down the road. We acknowledge the tradeoff in mass necessary to provide enough UV protection; however, dealing with the radiation is an unmet need in both ballooning and space applications, so any steps in the right direction are promising.
Figure 7: Melanin attached to Cellulose Paper.
Figure 8: Melanin coated filter paper left under UV light for a week to test integrity of Melanin under prolonged radiation exposure.

In order to simulated our UV protected material under the harsh conditions of space, we used the NASA Ames Solar Simulator which uses up to 1300W of power to provide radiation comparable to UV bursts during solar flares. Likewise, we wanted to interrogate the long-term UV exposure effects on materials, so we place our melanin bound cellulose filter paper under our lab's hood's UV lamp for one week.
To test the efficacy of our melanin linker, we washed each sheet with water three days after the initial attachment to the cellulose sheet and collected the runoff. In theory, since we added the same initial amount of melanin to each sheet, the concentration of melanin in the runoff should inversely indicate the linker's binding activity. We quickly analyzed these results with a UV-Vis Nanospectophotometer. Interestingly, the sheets without any linker had the highest absorption readings on a range of 200-300nm, while the runoff from the sheet with the linker had the lowest. This confirmed our expectations that the linker was latching on to the cellulose sheet and retaining the melanin.
Figure 9: UV-Vis spectra of runoff generated by washing melanin coated membrane. Higher UV absorbances corresponds to lower binding efficiency of construct used to retain melanin.
  1. Brenner, M. & Hearing, V. (2007). The Protective Role of Melanin Against UV Damage in Human Skin†. Photochemistry And Photobiology, 84(3), 539-549.
  2. >
  3. Ultraviolet Radiation: How It Affects Life on Earth : Feature Articles. (2016). Retrieved 9 July 2016, from
  4. Soni, R. & Toetia, M. (2016). Impact of Zinc Oxide on the UV Absorbance and Mechanical Properties of UV cured films. Presentation, Department of Chemistry, C.C.S. University of Meerut.
  5. ZINC OXIDE | ZnO - PubChem. (2016). Retrieved 9 July 2016, from
  6. Miller, A.J & Tsao, H, 2009, ‘New insights into pigmentary pathways and skin cancer’, British Journal of Dermatology, 162(1):22-28.
  7. Brenner, M. & Hearing, V. (2007). The Protective Role of Melanin Against UV Damage in Human Skin†. Photochemistry And Photobiology, 84(3), 539-549.
  8. García-Borrón, J. C. and Solano, F. (2002), Molecular Anatomy of Tyrosinase and its Related Proteins: Beyond the Histidine-Bound Metal Catalytic Center. Pigment Cell Research, 15: 162–173. doi:10.1034/j.1600-0749.2002.02012.x
  9. María I Chávez-Béjar, Metabolic engineering of Escherichia coli to optimize melanin synthesis from glucose. Microbial Cell Factories, 2013 Nov 13;12:108. doi: 10.1186/1475-2859-12-108.
  10. Hu, C., Jiang, P., Xu, J., Wu, Y., & Huang, W. (2003). Mutation analysis of the feedback inhibition site of phenylalanine-sensitive 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase ofEscherichia coli. Journal Of Basic Microbiology, 43(5), 399-406.
    10. Jossek, R. (2001). Characterization of a new feedback-resistant 3-deoxy-?-arabino-heptulosonate 7-phosphate synthase AroF of Escherichia coli. FEMS Microbiology Letters, 202(1), 145-148.
  11. Báez, J., Bolívar, F., & Gosset, G. (2001). Determination of 3-deoxy-D-arabino-heptulosonate 7-phosphate productivity and yield from glucose inEscherichia colidevoid of the glucose phosphotransferase transport system. Biotechnology And Bioengineering, 73(6), 530-535.
  12. Ballard, B., Jiang, Z., Soll, C., Revskaya, E., Cutler, C., Dadachova, E., & Francesconi, L. (2011). In Vitro and In Vivo Evaluation of Melanin-Binding Decapeptide 4B4 Radiolabeled with 177 Lu, 166 Ho, and 153 Sm Radiolanthanides for the Purpose of Targeted Radionuclide Therapy of Melanoma. Cancer Biotherapy & Radiopharmaceuticals, 26(5), 547-556.
  13. Howell, R., Revskaya, E., Pazo, V., Nosanchuk, J., Casadevall, A., & Dadachova, E. (2007). Phage Display Library Derived Peptides that Bind to Human Tumor Melanin as Potential Vehicles for Targeted Radionuclide Therapy of Metastatic Melanoma. Bioconjugate Chem., 18(6), 1739-1748.
  14. Fujishima, K., Venter, C., Wang, K., Ferreira, R., & Rothschild, L. (2015). An overhang-based DNA block shuffling method for creating a customized random library. Sci. Rep., 5, 9740.