Substitutions at a rheostat position in human aldolase A cause a shift in the conformational population

Abstract

Some protein positions play special roles in determining the magnitude of protein function: at such "rheostat" positions, varied amino acid substitutions give rise to a continuum of functional outcomes, from wild type (or enhanced), to intermediate, to loss of function. This observed range raises interesting questions about the biophysical bases by which changes at single positions have such varied outcomes. Here, we assessed variants at position 98 in human aldolase A ("I98X"). Despite being ~17 Å from the active site and far from subunit interfaces, substitutions at position 98 have rheostatic contributions to the apparent cooperativity (n_H ) associated with fructose-1,6-bisphosphate substrate binding and moderately affected binding affinity. Next, we crystallized representative I98X variants to assess structural consequences. Residues smaller than the native isoleucine (cysteine and serine) were readily accommodated, and the larger phenylalanine caused only a slight separation of the two parallel helixes. However, the diffraction quality was reduced for I98F, and further reduced for I98Y. Intriguingly, the resolutions of the I98X structures correlated with their n_H values. We propose that substitution effects on both n_H and crystal lattice disruption arise from changes in the population of aldolase A conformations in solution. In combination with results computed for rheostat positions in other proteins, the results from this study suggest that rheostat positions accommodate a wide range of side chains and that structural consequences manifest as shifted ensemble populations and/or dynamics changes.

Keywords: aldolase; mutation; rheostatic positions; structural repacking; substitution.

FIGURE 1

Activity response curves aldolase A to varying Fru‐1,6‐BP concentration. (a) Evaluating reproducibility of K _app determination for the wild type enzyme. For each titration, each point represents an average of three sample replicates. The titration curves included in this graph show 28 biological replicates, collected by four different individuals. As an example, technical replicates collected on 1 day are in black and biological replicates collected for different protein preparation on different days are in gray. Lines are best fits to Equation (1). (b) The response of human aldolase A I98X substitutions. Although not always apparent due to overlap in points, at each substrate concentration, data were collected in triplicate. The I98S position had low activity (possibly low expression, instability during protein preparation or low V _max activity), thus when converted to relative activity that data sets show considerably more scatter than other data sets. Lines are best fits to Equation (1). When position 98 was substituted with glycine or tyrosine, no activity was detected

FIGURE 2

Comparison of K _app and n _H fit values for WT replicates and I98X variants. The wild‐type values (black dots) represent the >84 individual measurements obtained (see text for more detail). The errors of the wild type replicates are illustrated by their ranges of measurements; the averages and standard deviations for the wild type values are in Table 1. The values for the I98X variants (gray dots) represent the averages of three replicates; most error bars (gray lines) represent errors of the fits (standard deviations). The exception in the I98X column is the “I” data point (i.e., wild type protein), which represents the average and standard deviation of the 27 replicates that were collected on the same days as the other I98X variants; this difference is denoted with a black error bar. The “Y” and “G” variants are shown as high values on the K _app plot (open gray dots); these variants lacked activity, which could arise from either a catastrophic effect on k _cat or on K _app and/or the absence of folded protein. For I98Y unfolding was ruled out by CD (Figure S3). Both the wild type and I98X data sets are displayed with horizontal offset to aid visual inspection of the data. The ranges of outcomes for I98X variants, as compared to the wild type ranges, confirm that position 98 is a rheostat for the control of n _H and at least a moderate rheostat for the tuning of K _app

FIGURE 3

Structure of wild type human aldolase A. (Left) The homotetramer. (Right) A Zoomed in image of position I98 and a phosphate molecule that occupies the active site. The I98 position is distant from the active site and from subunit interfaces

FIGURE 4

Crystallized structures of proteins substituted at position 98. The space between the two helices where the wild type I98 resides easily accommodates smaller amino acids. Substituting with phenylalanine (a larger amino acid) causes the two helices to separate, breaking hydrogen bonds between K101 and N71 and D68. This shift also seems a likely cause of stress in the crystal lattice that results in lower resolution of the I98F structure. Insertion of even larger amino acids like tyrosine into position 98 may cause sufficient lattice strain to prevent successful structure determination

FIGURE 5

The correlation between cooperativity (i.e., n _H ) and crystallographic resolution. Values and errors from Table 1 were used in an error‐weighted linear fit to obtain the line and R value