Genetic Perturbation Alters Functional Substates in Alkaline Phosphatase

Abstract

Enzymes inherently exhibit molecule-to-molecule heterogeneity in their conformational and functional states, which is considered to be a key to the evolution of new functions. Single-molecule enzyme assays enable us to directly observe such multiple functional states or functional substates. Here, we quantitatively analyzed functional substates in the wild-type and 69 single-point mutants of Escherichia coli alkaline phosphatase by employing a high-throughput single-molecule assay with a femtoliter reactor array device. Interestingly, many mutant enzymes exhibited significantly heterogeneous functional substates with various types, while the wild-type enzyme showed a highly homogeneous substate. We identified a correlation between the degree of functional substates and the level of improvement in promiscuous activities. Our work provides much comprehensive evidence that the functional substates can be easily altered by mutations, and the evolution toward a new catalytic activity may involve the modulation of the functional substates.

Figure 1

FRAD for measuring single enzyme activity. (A) Schematic image of the single-molecule assay in the FRAD. In each well or “reactor” patterned on amorphous perfluoro polymer (CYTOP), a single AP molecule hydrolyzes non-fluorescent FDP substrate molecules to release fluorescent product molecules (fluorescein), which accumulate in the reactor. Red and magenta spheres in the crystal structure of dimeric AP-wt (PDB ID: 1ED8) represent Zn²⁺ and Mg²⁺ ions in the active site. (B) Typical time-course measurement of fluorescence intensity in reactors. Time-course images were obtained at 20 min intervals for 240 min. The histogram shows the distribution of fluorescence intensity at 240 min. The histogram typically presents three peaks: peak 0 (enzyme-free reactors), peak 1 (single enzyme activity), and peak 2 (containing two or more enzyme molecules in a single reactor). The dotted red and black lines indicate fitting with Gaussian functions and threshold to exclude peak 2 from the analysis. The top panels exhibit snapshots of time-course measurements of catalytic activities observed by microscopy. Reactors containing AP-wt molecules exhibit an increase in fluorescence intensity. (C) Illustration of the aims of this study. The first aim is to investigate how readily single-point mutations modulate functional substates (left). The second aim is to evaluate possible correlation between variance of the native activity and promiscuous activity (right).

Figure 2

Alteration in catalytic activity and functional substates by single-point mutations. (A) Catalytic activity and functional substates (CV_n) (%) of AP-wt and mutant APs were plotted (means ± SD). The turnover rate of the mutants was normalized to that of the AP-wt, and the x-axis was converted to a logarithmic scale. The data points are colored according to the distance (Å) between a given mutation and the ligand (phosphate) present in the active site to the crystal structure of AP. The red-filled circle represents the AP-wt. (B) Dependence of activity on distance from the ligand to mutation sites. The turnover rate of the mutants was normalized to that of the AP-wt. The dotted line shows a linear fitting of the plot (slope = −0.13, R² = 0.24). (C) Dependence of CV_n on the distance from ligand to mutation sites. (D–F) Representative distributions of mutants that exhibited single peaks shifted from the AP-wt (D), additional peaks partially overlapping with the AP-wt (E), and shifted additional peaks (F). The population of double occupancy of enzyme molecule (peak 2) was shown with cyan (AP-wt) and red (mutants) asterisks. The total number of enzyme molecules was higher than 1000.

Figure 3

Alteration of catalytic activity and functional substates in highly active mutants. Cyan and red bins show the distribution of the AP-wt (first y-axis) and mutants (second y-axis), respectively. Population of double occupancy of enzyme molecule (peak 2) was shown with cyan (AP-wt) and red (mutants) asterisks. The total number of enzyme molecules was higher than 1000.

Figure 4

Correlation between functional substates and promiscuous activity. (A) Chemical structures of phosphate monoesters (p-nitrophenyl phosphate, pNPP), phosphate diesters (bis-p-nitrophenyl phosphate, bpNPP), phosphonate monoesters (p-nitrophenyl phenyl phosphonate, pNPPP), and 4-methylumbelliferyl sulfate substrates. (B–E) Correlation between activities against those substrates and CV_n. The activity of mutant AP was comparable to that of AP-wt. The CV_n of the mutants was measured using the FDP shown in Figure 2. The green circle shows AP-wt, and red and blue plots exhibit that the distance of the mutation site from the active center is less or greater than 10 Å. Different forms of the plots indicate mutants; red circle (D101S), triangle (G118D), rectangular (D153S), diamond (K328R), and light blue circle (A119T), triangle (L196V), and rectangular (N263F). For the mutants with distinct bimodal distribution (D153S, K328R, L196V, and N263F), the data were also fitted with a double Gaussian function to determine CV_n represented in filled symbols. The linear approximation of data points in red symbols or blue ones provided fitting slopes; 0.06 (R² = 0.24) and −0.01 (0.003) in (B), 1.13 (0.86) and 0.23 (0.34) in (C), 1.93 (0.86) and 0.07 (0.40) in (D), and 0.06 (0.04) and 0.06 (0.59) in (E), respectively. T59R, G118D, A119T, or L196V was excluded from (B–E), (E), (D), or (E) because the mutants did not exhibit measurable activity against the substrates (see Table S1).