Cross-laboratory comparison of fluorimetric microplate and colorimetric bench-scale soil enzyme assays

There has been growing interest in fluorescence-based microplate methods to measure enzyme activities due to
the sensitivity of fluorimetric detection and the potential for simultaneous and rapid assaying of multiple enzyme
activities in the same soil suspension. However, micro-scale methods could introduce considerable operator error
such as: 1) the requirement to put soil samples into a suspension; 2) the very small amounts of soil placed in each
microplate well; 3) pipetting error because μL volumes are required; and 4) the need for standard curve calibration
with every sample to account for quenching. For valid data comparison and interpretation, there is
clearly a need to have a strict and agreed-upon enzyme assay protocol to standardize the microplate-based
method. Therefore, the objectives were to: 1) determine the reproducibility and comparability of the standard pnitrophenol
bench-scale and 4-methylumbelliferone microplate enzyme assays measured by five laboratories for
β-glucosidase (EC 3.2.1.21) and acid phosphomonoesterase (EC 3.1.3.2) on the same soil samples; and 2) determine
the degree and the sources of variability associated with the assays within and among the laboratories.
The results showed that overall variability was highest for replication on the microplate (n=4), whereas suspension
replication had low CVs. This suggests an important source of variation is from pipetting not variability
from soil suspensions. A major effort was made to control for methodological differences by using air-dried soils
(therefore more stable over varying storage periods) and operator consistency for each task across the labs (e.g.
preheated reagents, microplate reader sensitivity set to the highest standard, readings taken within an hour of
reaction termination, and controls for substrate autohydrolysis). As a result, the differences among labs were
much smaller than differences due to soil type for the microplate method, indicating operator error can be
minimized by following the same strict protocol. At the molar level, enzyme activity rates measured across the
five labs were not the same between bench and MUF microplate methods (although they were within an order of
magnitude), but were quite similar in terms of ranking of soil management treatments and soil types (Table 2).
Correlations between bench and microplate assays were strong for both enzymes, although slightly stronger for
acid phosphomonoesterase (r=0.93) than β-glucosidase (r=0.81). Additionally, for both acid phosphomonoesterase
and β-glucosidase, correlation r values were mostly similar for MUF microplate and PNP bench
method correlation with EL-FAME biomarkers, suggesting both methods were measuring activity originating
from the same microbial groups. We conclude that different labs using the same MUF microplate protocol tested,
gives reasonably similar absolute activity values, variability, and ranking of treatments (highest to lowest). We
propose that the MUF microplate method described in this study be considered as a standard protocol for assaying
soil enzyme activities, providing that the buffer pH for the incubation be adjusted to the optimal pH
according to the enzyme of interest.