Team:Victoria Wellington/Results
Team
Victoria
Results
Results Sections
Aim
As we wanted to build an enzymatic fuel cell, a key component of our project is to characterise those enzymes - of particular interest were their kinetic parameters and pH optima.
The enzymes we tested were c. boidinii formate dehydrogenase and m. thermophila formate dehydrogenase. The main experiments conducted were as follows:
- Michaelis-menten kinetics of each enzyme with formate
- Optimum pH assay with each enzyme
- Crude kinetics of each enzyme with mesoxalate
Methods
Since these enzymes use NAD+ as a cofactor, we were able to follow catalysis by examining increases in absorbance at 340nm with a spectrophotometer as NAD+ was reduced to NADH.
NADH was quantified in formate assays using a standard curve. This contained formate, phosphate buffer and a range of NADH concentrations. Using the resulting regression equation, NADH concentration (and quantity) could be calculated. Due to the reaction stoichiometry, we could infer that for each mole of NADH produced, 1 mole of formate had also reacted.
Due to the double-substrate nature of these enzymes' kinetics, there were additional considerations for the assay design. When measuring kinetics parameters for the substrate of interest (i.e. formate), its concentration is varied, but [NAD+] must be held constant.
Reaction conditions for each test were as follows:
- Formate kinetics - pH 7.5, 30℃, [NAD+] = 10mM, in 100mM phosphate buffer. Formate varied from 1-216mM
- pH assays - 30℃, [NAD+] = 10mM, [formate] = 200mM in 100mM phosphate buffer(s)
- Mesoxalate kinetics - pH 7.5, 20℃, [NAD+] = 5mM in 100mM phosphate buffer
Reactions were done in triplicate at 200𝜇l, warmed to temperature before adding enzyme and incubating in a thermocycler. Absorbance was read in a cuvette or 96-well plate.
Control reactions were included throughout; one without enzyme, one without NAD+ and one without formate. No enzyme controls were selected as blanks, and their absorbance subtracted from the other data.
For pH assays, as series of 100mM phosphate buffers at pH 6.5-8.5 were made up.
Results
cbFDH kinetics
A Michaelis-menten model was fitted in R, and kinetic parameters were calculated for cbFDH with formate. It’s residual standard error was 0.222
substrate
Vmax (μmol/min/mg)
Km (mM)
Kcat (min-1)
Specificity constant (M-1s-1)
Conditions
Formate
4.19
28.4
169.3
99.4
pH = 7.5, T = 30℃, 10mM NAD
Results were similar to those obtained by Guo, Qi et al. [1]
pH assays:
Each FDH enzyme was tested at pH 6.5-8.5, and resulting data was plotted in R, along with a polynomial model fitted to each dataset:
mtFDH:
Equation: Activity = -1.06(pH)2 +17.5(pH) -65.6 (R-square = 0.94 - strong correlation)
Derivative: da/dpH = -2.12(pH) +17.5
Optimum pH = 8.26
cbFDH:
Equation: Activity = -0.63(pH)2 +10.16(pH) -38.62 (R-square=0.96 - strong correlation)
Derivative = -1.26(pH) +10.16
Optimum pH = 8.06
The main finding here is that mtFDH has the best activity at a higher pH than cbFDH (even if only slightly).
It is important we acknowledge the shortfalls of this particular experiment. In each case, no substantial decrease in activity was observed, so a polynomial equation is not necessarily reflective of what happens outside the pH range we tested.
Of additional concern is that the cbFDH had been frozen longer than mtFDH before the assays were done. That mtFDH has a generally high activity here, is therefore not necessarily reflective of the enzyme’s properties - cbFDH is likely to have degraded.
Later kinetic assays also demonstrated mtFDH activity at pH 7.5 to be considerably lower - there appears to have been a systematic error in this experiment. We have not found an explanation to this, and as such the pH pattern, but not activity values, are the key outcome from these assays.
mtFDH kinetics
An initial assay was done, and a Michaelis-Menten model fitted in R:
Residual standard error was 0.273. Of note was that the activity shown here did not align at all with what was observed in the pH assay (for pH 7.5). As such, a repeat experiment was done:
Again we found activity to be lower than in our pH assay. Due to the repetition from these kinetics assays, we decided these must be accurate.
The second model fitted had a residual standard error of 0.172 (< model 1) so this was selected for calculation of kinetic parameters:
substrate
Vmax (μmol/min/mg)
Km (mM)
Kcat (min-1)
Specificity constant (M-1s-1)
Conditions
Formate
4.53
32
197.2
102.8
pH = 7.5, T = 30℃, 10mM NAD
Overall, under the same reaction conditions mtFDH performed better than the cbFDH. It also seems to function better at high pH than cbFDH. Both these characteristics are desirable for our enzymatic fuel cell, and so if we were to select one enzyme it would be mtFDH.
One caveat with this conclusion is that cbFDH assays were done long after expression, so it may have degraded (which was not the case for mtFDH). Repeat experiments after defined amounts of time would be recommended.
Mesoxalate kinetics:
Although values are very small, we can still make qualitative conclusions from these data. Both formate dehydrogenases have a detectable reaction with mesoxalate - which is not their natural substrate. The mtFDH reacts more quickly than cbFDH, but as cbFDH had been idle in the freezer much longer, it is possible this is due to degradation.
References
[1] Guo, Qi et al. “Structural and Kinetic Studies of Formate Dehydrogenase from Candida boidinii.” Biochemistry vol. 55,19 (2016): 2760-71. doi:10.1021/acs.biochem.6b00181