Part:BBa_K2682000

CBD Nanobody Anchor Binder

This Biobrick codes for a nanobody that specifically binds to cannabidiol (CBD). The part is intended to be used with a dimer binding nanobody, together creating a chemically induced dimer (CID) system. This CID system has applications in a range of projects, as CID can be used for creating biosensors, translational control, and a number of other processes.

Team Washington iGEM 2019: Characterization of protein structure and kinetics

Note: The characterization of this biobrick is heavily tied to BBa_K3123002, which is biobrick based on this part, but with a mutation at protein position 32. As a result, the background information, methods, and experiments are identical.

This characterization was performed to better understand the protein structure and binding affinity of this CBD nanobody binder. Our simulations team created Rosetta models and found a possible folding structure, along with a possible location and structure of the binding pocket for CBD. We sought to verify this model by performing site-directed mutagenesis to alter tyrosine-32, which was identified as being a key residue in the binding activity of the binder within the binding pocket, to alanine. This purpose of this mutation was to alter the binding pocket enough that it would affect the binding affinity (K_D) of the binder to CBD without causing any major structural changes. We hypothesized that if our protein model was indeed correct, then this mutation would affect the binding affinity of the nanobody binder significantly compared to an non-mutated version of the binder. We expected to see a decrease in binding affinity for the mutant binder based on our hypothesis.

Fig. 1: Anchor Binder Model View 1 (CBD in red)

Fig. 2: Anchor Binder Model View 2 (CBD in red)

Methods and Experiments

Our experimental design can be found on our wiki in the Project Design tab. To summarize, we performed site-directed mutagenesis on this CBD anchor binder to change tyrosine-32 into alanine. We verified this mutation through Sanger Sequencing, and transformed the mutated plasmid into E. coli strain WK6. We transformed this mutated plasmid in parallel with the non-mutated plasmid, yielding a mutant anchor binder and "wild-type" anchor binder.

The protein was expressed in WK6 overnight and released from the periplasm of the bacteria using a TES buffer solution. The periplasm extract was then run through a nickel his-trap column for protein purification. After protein purification, the protein was biotinylated and the protein concentration was quantified using a spectrophotometer.

The binding affinity of the mutant and wild-type binder was then quantified using biolayer interferometry (BLI). The specific machine we used was the OCTET Red96 plate reader, using Super Streptavidin Sensors. This procedure, outlined in our experiments page, measures the association and dissociation rates for the binder to CBD at different CBD concentrations. Using these two values, the equilibrium dissociation constant (K_D) can be calculated. This K_D is a inversely proportional to binding affinity, thus a high K_D would indicate a low binding affinity, and vice versa.

Results

The K_D for the anchor binder was calculated by taking the ratio of the K_off (1/s) value to the K_on value (1/(M*s)). The K_on value for the anchor binder was found to be 2.76E03/(M*s) and the K_off value was found to be 8.71E-02/s. From these values, the K_D value was calculated to be 31.6uM.

Fig. 3: Biolayer Interferometry raw data sensorgram of the anchor binder with CBD. As the sensor was exposed to CBD, the wavelength measurement increases as the bound nanobody associates with the CBD. The wavelength measurement then decreases as the nanobody dissociates from the CBD; from these changes we can measure the K_on and K_off.

In figure 3, the "Nanobody Bound (CBD)" data line indicates the data from the nanobody-bound sensor that is exposed to CBD. This is the raw data that measures the K_on and K_off. The "No Nanobody Bound (CBD)" data line indicates the data taken from a sensor that is not bound to any nanobodies, but still exposed to CBD. This data is used as a reference to the "Nanobody Bound (CBD)" data to remove background activity. The "Nanobody Bound (No CBD)" data line serves as a control, as this sensor runs the same assay as the "Nanobody Bound" sensor. The sensor is coated with nanobodies but is not exposed to CBD. The "No Nanobody Bound (No CBD)" is analogous to “No Nanobody Bound (CBD)” and serves the same function for the "Nanobody Bound (No CBD)" data line. In making final measurements, all three of these last data lines serve as a control for different conditions and is subtracted from the "Nanobody Bound (CBD)" data.

Fig. 4: This BLI sensorgram of the data from Fig. 3 (after subtractions), aligned on the x-axis. The trend line, in red, is calculated using all four concentration measurements and final K_on and K_off is calculated from this trend line. The R² value for this global fit is 0.9517, indicating a strong fit between the trend line and the data. The x² value is 0.0422, indicating a low error in the fitting.

Multiple BLI assays were run with this nanobody, but most of our results came back inconclusive due to user error. The most likely reason is that the test plate was contaminated during set up or during transport to the machine, as it was in another building apart from our lab. An example of these failed assays can be viewed on our results page. For more results on protein structure, please see our 2019 biobrick [BBa_K3123002for the mutant anchor binder and its characterization.

Discussion

Overall characterization of the binding affinity of this nanobody binder reveals that it does not have a relatively high binding affinity, compared to some other binders like antibodies¹. Further work should be done to improve the binding affinity of this binder; this could come in the form of informed mutations from models or from directed evolution.

Based on the results kinetics results for this anchor binder and the mutated anchor binder, we suspect that the protein model this mutation was based off of is incorrect. Given that the K_D remained relatively the same despite the mutation (K_D of mutant = 35.6uM) suggests that the residue at that position was not important to binding, and thus the entire model may be wrong. It is also possible that the results found here are flawed, as the concentration of the wild type binder tested were much higher than the concentration of the mutant binder; this was not a conscious choice, but rather a result of low overall protein yield for the mutant.

References

1. Stubenrauch K, Wessels U, Essig U, Kowalewsky F, Vogel R, Heinrich J. Characterization of murine anti-human fab antibodies for use in an immunoassay for generic quantification of human fab fragments in non-human serum samples including cynomolgus monkey samples. Journal of Pharmaceutical and Biomedical Analysis. 2013;72:208–215. doi: 10.1016/j.jpba.2012.08.023.

Sequence and Features

Assembly Compatibility:

• 10
  COMPATIBLE WITH RFC[10]
• 12
  COMPATIBLE WITH RFC[12]
• 21
  COMPATIBLE WITH RFC[21]
• 23
  COMPATIBLE WITH RFC[23]
• 25
  COMPATIBLE WITH RFC[25]
• 1000
  COMPATIBLE WITH RFC[1000]