Team:Shanghai SFLS SPBS/Poster

Poster: Shanghai_SFLS_SPBS

Not hair die, hair DYE: Biosynthesis of natural hair dyes with engineered bacteria

Presented by Shanghai_SFLS_SPBS

Crystal Liu1, Olivia Feng2, Monique Wang2, Kevin Bao3, Chuwen Cheng1, Ivy Ding2, Chenyu Gu1, Oscar Hang4, Olive Li5, John Ling6, Jeff Shan7, Minhao Qin8, Chang Yuan2, Mingjiu Zhao1, Ken Zhu9, Jiaheng Li§, Jianzhao Yang10, Cris Ding11, Boqin Yang12, Xinyue Yu§, Shiyuan Li§

1Shanghai Foreign Language School affiliated to SISU, 2Shanghai Pinghe Bilingual School, 3Qibao Dwight High School, 4the Hun School of Princeton, 5Guanghua Cambridge International School, 6Strathallan School from Scotland, 7Stevenson School, 8the High School Affiliated to Shanghai Jiaotong University, 9Cranbrook Kingswood School in Michigan, 10Institute of Plant Physiology & Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 11Imperial College London, 12Ross School, §Bluepha Co., Ltd.

Hair-dyeing is becoming increasingly popular, and the demand for harmless dyes is constantly rising. However, more common synthetic dyes may damage the hair cortex and cause allergies. Despite increasing attention to natural dyes, its production is limited and products are expensive. We propose using engineered bacteria to mass-produce natural, harmless hair dyes. We successfully synthesized melanin, indigo, dopaxanthin, and indoline-betacyanin and dyed hair into black, blue, and red. We used Vibrio natriegens to increase the rate of production. V. natriegens could produce melanin faster than E. coli. Furthermore, considering that synthetic dyes are composed of oxidants and pigment precursors, we envisage combining oxidases and pigment precursors to dye hair. We have expressed bacterial laccase and tested its activity. Next, we will try to optimize dyeing protocols and discuss the safety aspects of potential products. If successful, our products could bring dramatic changes to the market and introduce substantial social benefits.


Since the last century, hair dyeing has become a part of our daily life. However, numerous problems still exist. Having found that the color range is still limited, we targeted dyes in blue, yellow, red, and black to mimic a color printer. We shall be able to acquire almost any color by mixing the pigments.

We then proposed using Vibrio natriegens, a faster-growing chassis organism, to increase the rate of production. Synthetic hair dyes in the market utilize the combination of the oxidant, hydrogen peroxide, and pigment precursors. We propose using laccase, an oxidase, and natural pigment precursors to replace this method.

Figure 1. Our targeted pigments, melanin, indigo, dopaxanthin, and indoline-betacyanin. We targeted dyes in blue, yellow, red, and black to mimic a color printer.


To synthesize our targeted pigments, we found these pigments' metabolic pathways. Melanin is synthesized from L-Tyrosine after numerous steps of oxidation catalyzed by the enzyme tyrosinase. Indigo can be produced from L-Tryptophan when enzymes tryptophanase and FMO are present. For dopaxanthin and indoline-betacyanin, the enzyme 4,5-DODA is used to catalyze the transformation from L-DOPA to 4,5-seco-Dopa. It then spontaneously forms different betalains when different precursors are present.

Figure 2. Metabolic pathways of (A) melanin, (B) indigo, and (C) betalains (dopaxanthin and indoline-betacyanin) production.

We found the genes coding for these enzymes from various species to construct respective gene circuits. We would first use E. coli as the chassis organism to produce these pigments. Additionally, we read about Vibrio natriegens, the fastest-growing non-pathogenic bacterium known to date and an alternative chassis organism to E. coli. We envisage using V. natriegens to increase the rate of production.

Figure 3. Genetic circuit for (A) melanin, (B) indigo, and (C) betalains (dopaxanthin and indoline-betacyanin) production.


To help quantify the production of synthesized pigments, we measured the standard curves of melanin and indigo. We first measured the absorbance spectrum of these pigments and identified the peak absorbance. Then, we prepared solutions of different concentrations and measured their absorbance at the peak absorbance wavelength. Finally, we drew the standard curve. We were unable to acquire standard dopaxanthin or indoline-betacyanin, so the standard curve could not be constructed.

Figure 4. Standard curves of (A) melanin and (B) indigo.

Shake Flask Experiments

After several trials, we successfully produced melanin, indigo, dopaxanthin, and indoline-betacyanin with E. coli in shake flasks. The graphs on the first-row show pigment production against time. Melanin and indigo concentrations were calculated with the standard curve. Melanin production reached 11.9 g/L, and indigo production at 72 h was 37.95 mg/L. For the betalains, absorbance is graphed on the y-axis. The second-row photos show more intuitively the production of these pigments with the bacterial solution moved from the flasks and stored every 12 hours.

Figure 5. Production of melanin (A-B), indigo (C-D), dopaxanthin (E-F), and indoline-betacyanin (G-H) in E. coli.

We were honored to have the opportunity to use an HPLC-MS machine from Shimadzu Enterprise Management (China) Co., Ltd. With LCMS-9030 quadrupole time-of-flight (Q-TOF) mass spectrometer from Shimadzu, we found the substances of correct molecular weights and determined that we had synthesized the targeted products. We were not able to quantify the melanin bioproduction because melanin is a polymer.

Figure 6. HPLC-MS results of indigo, dopaxanthin, and indoline-betacyanin.

Hair Dye

Having synthesized these pigments, we tried to dye hair with them. The coloration effects with synthesized melanin and indoline-betacyanin were very satisfying. The effects of indigo and dopaxanthin were not as significant. We hypothesize that indigo concentration was too low, and the color of dopaxanthin was too close to the original, bleached hair. We believe that by concentrating the solutions and changing the conditions, we will optimize the dyeing effects.

Figure 7. (A) Hair dye results with synthesized pigments, with positive and blank controls. (B) Our dyeing protocol.

Production with Vibrio natriegens

To lower the production cost, we proposed replacing E. coli with Vibrio natriegens, the fastest-growing non-pathogenic bacterium known to date, to increase the production rate. We found that V. natriegens produced faster than E. coli at the same temperature during the first 36 hours, but after 72 h, the production in E. coli was higher. We hypothesized that this was due to a lack of nutrients or substrates. We believe that by optimizing the growth conditions, we will be able to produce more melanin in V. natriegens than in E. coli.

Figure 8. (A) Production of melanin in E. coli BL21(DE3) and Vibrio natriegens at 25℃ and 37℃ in 72 h. (B) Production of melanin from 0-72 h. Top left: E. coli BL21(DE3), 25℃. Top right: E. coli BL21(DE3), 37℃. Bottom left: Vibrio natriegens, 25℃. Bottom right: Vibrio natriegens, 37℃.

Laccase & Pigment Precursor Dopamine

Inspired by synthetic hair dyes, which are composed of oxidants and pigment precursors, we proposed combining bacterial laccase with natural pigment precursors to dye hair, hoping that it can last longer in the hair.

We constructed T7-laccase using the gene coding for laccase in Bacillus sp. SDS-PAGE protein electrophoresis was performed and found an overexpressing protein at around 65 kDa. We also carried out an enzyme activity assay with a commercially-available kit and acquired positive results. We also tried to synthesize a common pigment precursor, dopamine. Since dopamine forms brown polymers when oxidized, we hypothesized that it might be combined with laccase to dye hair. When tested with HPLC-MS, we identified dopamine particles with the correct molecular weight.

Figure 9. (A) Verification of laccase using SDS-PAGE protein electrophoresis. (B) Laccase activity assay results.

Figure 10. HPLC-MS results of (A) standard dopamine and (B) biosynthesized dopamine.


We have successfully synthesized the pigments of interest and dyed hair into black, blue, and red. This could lay a foundation for a future mass-production method, possibly in bacterial fermenters.

In the future, we would like to optimize our dyeing methods, such as concentrating the pigment solution before dyeing. We will try washing the dyed hair to see how long the color would stay in the hair. We will try mixing the pigments to see if we can reach a more comprehensive color range, and we will do more experiments to make sure that these potential products are safe to humans and harmless to the environment. We also plan to try combining laccase with both synthesized and standard pigment precursors and possibly provide a more permanent alternative to directly dyeing with pigment precursors.


To optimize our bacteria's growth conditions, we built a model with an Artificial Neural Network to predict bacterial growth curves given a set of environmental parameters. With this model, we can optimize any bacteria species' growth conditions after collecting some experimental data.

We measured the growth of E. coli and V. natriegens at different temperatures, oxygen levels, growth media, and ion and glucose concentrations. We measured the OD600 of each sample every hour. We fitted the data to a logistic growth model.

Figure 11. Examples of our experimental data and the best fit growth curve.

We used the Artificial Neural Network (ANN) for machine learning. The environmental factors were the inputs, and the parameters on the growth curve were outputs. In terms of inputs, we defined the inputs as decimal values for ion or glucose concentrations and binary values for growth media or shaking. We divided our data into 80% training and 20% testing.

We used the Pearson correlation to evaluate our model. The environmental capacity and growth rate parameters achieved a correlation of 0.722 and 0.702, respectively.

Figure 12. Model evaluation with Pearson correlation.


1. Choi, H. et al. A novel flavin-containing monooxygenase from Methylophaga sp. strain SK1 and its indigo synthesis in Escherichia coli. Biochemical and Biophysical Research Communications 306, 930-936 (2003).

2. Das, A., Verma, A. & Mukherjee, K. Synthesis of dopamine in E. coli using plasmid-based expression system and its marked effect on host growth profiles. Preparative Biochemistry & Biotechnology 47, 754-760 (2017).

3. Eagon, R. Pseudomonas natriegens, a marine bacterium with a generation time of less than 10 minutes. Journal of Bacteriology 83, 736-737 (1962).

4. Guerrero‐Rubio, M., López‐Llorca, R., Henarejos‐Escudero, P., García‐Carmona, F. & Gandía‐Herrero, F. Scaled‐up biotechnological production of individual betalains in a microbial system. Microbial Biotechnology 12, 993-1002 (2019).

5. Jeon, J. et al. Laccase-catalysed polymeric dye synthesis from plant-derived phenols for potential application in hair dyeing: Enzymatic colourations driven by homo- or hetero-polymer synthesis. Microbial Biotechnology 3, 324-335 (2009).

6. Lewis, D., Mama, J. & Hawkes, J. A Review of Aspects of Oxidative Hair Dye Chemistry with Special Reference to N-Nitrosamine Formation. Materials 6, 517-534 (2013).

7. Mohammadian, M., Fathi-Roudsari, M., Mollania, N., Badoei-Dalfard, A. & Khajeh, K. Enhanced expression of a recombinant bacterial laccase at low temperature and microaerobic conditions: purification and biochemical characterization. Journal of Industrial Microbiology & Biotechnology 37, 863-869 (2010).

8. Murdock, D., Ensley, B., Serdar, C. & Thalen, M. Construction of Metabolic Operons Catalyzing the De Novo Biosynthesis of Indigo in Escherichia coli. Bio/Technology 11, 381-386 (1993).

9. Retrieved July 2020, from

10. Retrieved July 2020, from

11. Retrieved June 2020, from

12. Retrieved August 2020, from

13. Roessler, A., Crettenand, D., Dossenbach, O., Marte, W. & Rys, P. Direct electrochemical reduction of indigo. Electrochimica Acta 47, 1989-1995 (2002).

14. Tschirhart, T. et al. Synthetic Biology Tools for the Fast-Growing Marine Bacterium Vibrio natriegens. ACS Synthetic Biology 8, 2069-2079 (2019).

15. Wang, Z. et al. Melanin Produced by the Fast-Growing Marine Bacterium Vibrio natriegens through Heterologous Biosynthesis: Characterization and Application. Applied and Environmental Microbiology 86, (2019).


We worked closely with our partner, SZU-China, throughout the project. We met SZU-China at CCiC and learned that they had hit a bottleneck in the experiment of producing bio-indigo to dye jeans. We held numerous discussions and shared some lab data to troubleshoot the bioproduction of indigo. We borrowed their gardenia blue to dye our hair and lent them bacteria with the TnaA-FMO plasmid to produce indigo. We established a partnership: iDYE and co-organized the iDYE workshop to popularize our projects, scientific experiments, and synthetic biology. This event drew the interest of many students at Shenzhen University. Participants of the workshop used food-grade Gardenia blue to dye cloth, and various dyes from our project were put on display.

Figure 13. Partnership between SZU-China and Shanghai_SFLS_SPBS. Left: Hair dye results with SZU-China's Gardenia blue. Right: co-organized iDYE Workshop.

Human Practices

To identify current problems, we interviewed a hairdresser, two businesses that extract pigments from henna, and distributed a questionnaire to the mass public. The hairdresser introduced that although synthetic dyes may harm the hair, they are cheaper and more popular. He also finds it difficult to acquire precisely the desired color. For natural hair dye companies, their major challenges include mass-producing the dyes and the limit in the color range of their products. Respondents to our questionnaire also expressed their hope to have a harmless, non-allergenic, and even natural hair dye product. These experiences enabled us to better target our project.

Figure 14. Major demands of hair dye products, data from our questionnaire.

Hoping to make sure that we got our targeted pigments, we contacted Shimadzu Enterprise Management (China). We visited them in late August and were allowed to use their HPLC-MS equipment in September.

We used social media and a TEDx event to popularize our project. We have written 12 passages counting over 20,000 words in total on our WeChat public account. These passages include an introduction of our team, more detailed explanations of our project and an introduction to synthetic biology and microbiology. Furthermore, we had the precious opportunity to be invited to TEDx as one of the speakers. We introduced synthetic biology, our project, and the iGEM competition in that presentation. To further promote our project and popularize synthetic biology, we also created a bilibili account. We made a video about Vibrio natriegens, our chassis organism.


Team Members

Our PIs and instructor helped us troubleshoot and answered our questions about experiments, HP, public engagement, etc. Cris Ding provided us with numerous opportunities and ideas in Human Practices and public engagement. We would also like to thank Bingjing Cheng for guidance during experiments, Weixu Wang for his instruction on our model, and Jiaoyun Ma and Man Liu for their help with coding our wiki.

We thank our sponsors: