Supplemental Information – Data limitations
The resolution of our analyses is limited by the diversity of assay methodologies used to generate appKm and appVmax data and by uneven coverage across enzyme classes and ecosystem types. Measurements were conducted over wide ranges of temperature and pH. In some studies, assay conditions were selected to mimic those in situ; in other studies, conditions were selected to maximize assay sensitivity. These choices have differential effects on appVmax and appKm. For example, Stone et al. (2012) measured five hydrolase activities (α-glucosidase,cellobiohydrolase,β-xylosidase,β-glucosidase,β-N-acetylglucosaminidase) in two unamended and two N-amended temperate forest soils at seven temperatures (4-40 C), showing that appVmax was more responsive than appKm with Q10 values of 1.64-2.27 and 1.04-1.93, respectively. A parallel study by German et al. (2012) found similar results for five soils along a longitudinal gradient from Alaska to Costa Rica, with Q10 values of 1.53-2.27 and 0.90–1.57 for appVmax and appKm, respectively. Other studies have found no consistent relationship between appKm and temperature (Cartes et al. 2009, Christian and Karl 1998). In either case, the general effect of above ambient assay temperatures is to increase apparent turnover rates by increasing appVmax relative to appKm.
Assay pH also introduces variation in appVmax values. Most studies that measure the kinetics of multiple enzymes use a common pH, often pH 5 for soils and pH 8 for surface waters. Studies that focus on a single enzyme often measure activities at an optimal pH, which may be displaced from the ambient pH of the samples. Because the pH optima of the enzymes included in our data set range 3-9,appVmax may vary by 1-2 orders of magnitude relative to appKm.
In contrast to temperature and pH, substrate choices have greater effects on appKm than appVmax. Three classes of substrates are commonly used to assay hydrolase activities: p-nitrophenyl,β-naphthyl and 4-methylumbelliferyl. For soil hydrolases, the mean appKm value for assays conducted with 4-methylumbelliferyl substrates, which we refer to as the high affinity series, is 41 μM while the mean for p-nitrophenyl substrates, which we refer to as the low affinity series,is 4900 μM. Marx et al. (2001), for example, measured phosphatase and β-glucosidase activities in the same soils using 4-methylumbelliferyl and p-nitrophenyl substrates. The appKm values were 138X and 14X greater, respectively, for p-nitrophenyl substrates while appVmax values differed by 1.02X and 1.62X, respectively (Table 1).
Sample preparation also introduces large effects on appKm. When samples, particularly soil samples, are disaggregated, homogenized or dispersed the appKm values decrease because competing substrates are diluted and confounding diffusion and sorption processes that impede enzyme-substrate access are reduced. Independent of sample handling, non-competitive inhibition and sorption of enzymes contributes to the sublinear scaling of appVmax and appKm. These non-specific interactions between enzymes and other reactive molecules, including humic substances,exopolysaccharides and organomineral surfaces restrict access to active sites and increase the energy needed to effect the conformational changes required for catalysis (Quiquampoix et al. 2003, Wetzel 1991). The net effect is to reduce appVmax and increase appKm.
This continuum of interactions that progressively compromise the activity of extracellular enzymes, termed a kinetic cascade (Sinsabaugh and Follstad Shah 2012), combined with the diversity of enzyme sources within a microbial community, produce the multiphasic kinetics typical of ecological systems (Williams 1973, McLaren 1978). For example, Marx et al. (2005) estimated high and low affinity appKm for four size fractions of soil using 4-methylumbelliferyl substrates. For cellobiohydrolase,β-glucosidase,β-xylosidase, phosphatase and leucine aminopeptidase the low:high ratios were 3000, 360, 200, 155 and 70, respectively.
The stoichiometry of microbial growth also contributes to variance in the kinetic regression models. Positive or negative displacements of appVmax relative to appKm may reflect up or down regulation of expression in response to microbial nutrient requirements rather than the dynamics of environmental substrate availability. Differential expression of extracellular enzymes in relation to biomass stoichiometry optimizes microbial growth (Westerhoff et al. 1983,Sinsabaugh and Follstad Shah 2012). These effects are likely a major source of variance in appVmax and appKm relationships for local or short term data and may account for the low correlation between appVmax and appKm in some studies.