To produce and verify the functionality of our protein, we first built up a gene construct and then conducted SDS-PAGE to confirm that our protein was correctly produced. To prove that our protein crosslinks with dityrosine bonds. We confirmed the formation of dityrosine bonds from FT/IR and CD measurements of our sample. Later, we explored the application of our Ultrabithorax (Ubx). Viscosity test results showed that our sample is a potential bioink. We later 3D printed our sample and acquired the ideal outcome.

Mechanism Introduction

Our sequence consists of Y167 Y240 region and mRFP. mRFP was aimed to prove protein keeps its functionality even after crosslinking.

Figure 1.Tyrosine locations in composite part K4377009 amino acid sequence. (Sequences marked in blue are Y167 Y240 region.)

Crosslinking can be induced by hydroxyl radicals, forming dityrosine covalent bonds with two tyrosines (Figure 2 ).

Figure 2.Dityrosine forming process
Figure 3.Artistic representation of cross-linking

Gene Construct

Cloning Result

We successfully amplified Ubx plasmid in E. coli DH5α and transformed into E. coli BL21 for expression.

Figure 4.The digest check result of Ubx plasmid extracted from E. coli BL21. M - Marker. 1 - Ubx plasmid extracted from E. coli BL21 digest check result, backbone (pSB1A3, 2114bp) / insert (composite part K4377006, 1475bp)

With the same process of building composite part K4377006, we successfully built composite part K4377009.

Figure 5.Y167 mRFP Y240 plasmid extracted from E. coli BL21 digest check result. M - Marker. 1 - Y167 mRFP Y240 plasmid extracted from E. coli BL21 digest check result, backbone (pSB1A3, 2114bp) / insert(composite part K4377003, 1066 bp)

Protein Expression


After we expressed Ubx plasmid, we conducted SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) to verify whether the protein was produced (result in Figure 6).

Figure 6.Composite part K4377006 SDS-PAGE result. M - Marker. 1 - Control, pSB1A3 pure plasmid. 2 - Ubx plasmid, target band Ubx 42.56 kDa

After the expression of composite part K4377009, we conducted SDS-PAGE to verify the protein was produced (result in Figure 7).

Figure 7.Composite part K4377009 SDS-PAGE result. M - Marker. 1 - Control, pSB1A3 plasmid. 2 - mRFP control, E1010 on pSB3K3, target band mRFP 25.4 kDa. 3 - Y167 mRFP Y240 plasmid, target band Y167 mRFP Y240 30.89 kDa

Functional Test

In order to prove that our peptide monomers form feasible cross linkages, we used FT/IR-4700, J-1700 Circular Dichroism Spectrophotometer to examine whether our material forms dityrosine bonds. Finally, to demonstrate our product as a potential 3D printing material, we measured viscosity of our product with Viscometer and further confirmed shear thinning behavior of our material.

Crosslinking Reaction

According to the paper, Fenton reaction increases the dityrosine bonds in the hydrogels but has little effect on the rheological or mechanical properties. Therefore, we used H2O2 and EDTA-Fe to prepare our materials through oxidation of tyrosine residues in our protein-based material, leading to dityrosine crosslinking.

Figure 8.Fenton reaction for dityrosine formation.

We mix H2O2 and EDTA-Fe in eppendorfs of each concentration for them to react before we add 25 μL of our protein sample. H2O2 were prepared in the following concentrations of 0%, 0.1%, 0.15%, 0.2%. While EDTA-Fe were added by 0μg/μL, 0.5μg/μL, 1μg/μL.

Figure 9.Macroscopic testing with H2O2 and EDTA-Fe. mRFP on the left and Y167 mRFP Y240 on the right.


To further confirm the existence of dityrosine bonds in our product, we used FT/IR-4700 to analyze additional and reduced bondings before and after crosslinking. As shown in the reaction, C-H bond is reduced and C=C bonds could be observed when dityrosine bonds form.

Figure 10.Dityrosine forming process
Table 1.Frequency range of C-H and C=C in aromatic rings
Bond Type of Compound Frequency Range, cm-1 Intensity
C-H Aromatic rings 690-900 Strong
C=C Aromatic rings 1500-1600 Variable
Figure 11.(a) Transmittance at 875 cm-1 of sample processed with 10μg/μL EDTA-Fe and different concentrations of H2O2 . “Before'' represents samples without crosslinking. (b) Transmittance of sample processed with 20μg/μL EDTA-Fe and different concentrations of H2O2 . “Before'' represents samples without crosslinking.

Figure 11 shows our result with FT/IR, at 875 cm-1,where the frequency range of the C-H bond, the transmittance changed significantly after crosslinking. C-H bond is reduced when dityrosine bonds form. Thus, higher transmittance represents a greater amount of dityrosine bonds.

Based on results from Figure 9 and Figure 11, we took each sample that contains the highest amount of dityrosine bonds from both precipitation and homogeneous groups for future tests. Sample 1 (adding 10μg/μL EDTA-Fe, 0.2% H2O2) from precipitation group and Sample 2 (adding 10μg/μL EDTA-Fe, 0.1% H2O2) from homogeneous group.

Figure 12.(a) FT/IR graph of sample 1 before and after processing with 10μg/μL EDTA-Fe and 0.2% H2O2 (b) FT/IR graph of sample 2 before and after processing with 10μg/μL EDTA-Fe and 0.1% H2O2

Figure 12 shows the significant influence of EDTA-Fe and H2O2 on 875 nm(C-H bond) transmittance. (a) Sample transmittance increased 12.22% after adding reagent. (b) Sample transmittance increased 6% after adding reagent. Data above proves crosslinking reaction came from dityrosine bond formation with the obvious decrease in C-H bonds.

It's observed from Figure13, 14 that C-H bonds were reduced at 690-900 nm, and that C=C bonds were produced at 1500-1600 nm. Based on data acquired from FT/TR measurements, we have successfully proved the crosslinking reaction came from dityrosine bond formation.

Figure 13.FT/IR graph of mRFP before and after processing with (a) 10μg/μL EDTA-Fe and 0.2% H2O2 (b) 10μg/μL EDTA-Fe and 0.1% H2O2

To confirm Y167 mRFP Y240 crosslinking surely occurs from dityrosine bonds in Y167Y240 region and not mRFP protein, we did FT/IR measurements of sole mRFP. Observed from acquired data, results barely changed after adding different concentrations of EDTA-Fe and H2O2 , showing the change of C-H bonds in our protein sample did come from tyrosines in protein residues Y167Y240 region and not mRFP.

Circular Dichroism

In addition to observing changes in bonding, we also measure absorbance of Sample 1 with J-1700 Circular Dichroism before and after crosslinking. After crosslinking, redshift occurs in the band of absorption values. Showing that two benzene ring bonds together after dityrosine bond formation. Electron delocalization of benzene rings leads to the change of wave pattern in graphs.

Figure 14.Absorbance graph of Y167 mRFP Y240 before and after processing with (a) 10μg/μL EDTA-Fe and 0.2% H2O2 (b) 10μg/μL EDTA-Fe and 0.1% H2O2

Observed from Figure 14, redshift occurs after reagents (10μg/μL EDTA-Fe and 0.2% H2O2) were added to Sample 1 and absorbance was higher than unprocessed samples. This could be the consequence of scattering caused by precipitation. Sample 2 (adding 10μg/μL EDTA-Fe and 0.1% H2O2) decreased in 275 nm absorbance and increased in 310 nm absorbance. Showing that electrons shifting from dityrosine formation changed the wave pattern.

Based on our Circular Dichroism result, we chose Sample 2 (which shows better results from absorption values) to excite our protein sample with 275 nm UV light, observing our sample giving out red and blue fluorescence at the same time. Red fluorescence came from mRFP and blue fluorescence came from dityrosine bond formation.

Figure 15.Fluorescence of our sample (a) before and (b) after processing with EDTA-Fe and H2O2 at 275 nm excitement.


We envision our product to be a biomaterial of great use. In professor Ming-Chia, Li's lab, we did measurement of viscosity for our protein sample and confirmed its shear thinning behavior. This proves that our Y167 mRFP Y240 biomaterial has great potential as 3D printing materials.

Figure 16.Viscosity vs. Shear rate of Y167 mRFP Y240

The result of viscosity vs. shear rate shows that the viscosity decreases when the shear rate increases, illustrating the properties of shear thinning, which shows high potential for the future development of bioinks for 3D printing.

Operation of 3D printer

To prove that our Ubx biomaterial can serve as a bioink for 3D printing, we learned to operate 3D printers from professor Ming-Chia, Li and printed a graph as we expected. View more details in Proof of Concept. Result shows that the Y167 Y240 region separated from Ubx is a successful bioink material. Aside from cross linking monomers, Ubx biomaterial also ensures functional proteins maintain their functionality after crosslinking. Providing a brand new and advanced biomaterial with great potential in medical, industrial, and agricultural fields.


  1. Choi, J., McGill, M., Raia, N. R., Hasturk, O., & Kaplan, D. L. (2019). Silk hydrogels crosslinked by the fenton reaction. Advanced healthcare materials, 8(17), 1900644.