Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 10 (1), 1691

Deep Analysis of Residue Constraints (DARC): Identifying Determinants of Protein Functional Specificity

Affiliations

Deep Analysis of Residue Constraints (DARC): Identifying Determinants of Protein Functional Specificity

Farzaneh Tondnevis et al. Sci Rep.

Abstract

Protein functional constraints are manifest as superfamily and functional-subgroup conserved residues, and as pairwise correlations. Deep Analysis of Residue Constraints (DARC) aids the visualization of these constraints, characterizes how they correlate with each other and with structure, and estimates statistical significance. This can identify determinants of protein functional specificity, as we illustrate for bacterial DNA clamp loader ATPases. These load ring-shaped sliding clamps onto DNA to keep polymerase attached during replication and contain one δ, three γ, and one δ' AAA+ subunits semi-circularly arranged in the order δ-γ123-δ'. Only γ is active, though both γ and δ' functionally influence an adjacent γ subunit. DARC identifies, as functionally-congruent features linking allosterically the ATP, DNA, and clamp binding sites: residues distinctive of γ and of γ/δ' that mutually interact in trans, centered on the catalytic base; several γ/δ'-residues and six γ/δ'-covariant residue pairs within the DNA binding N-termini of helices α2 and α3; and γ/δ'-residues associated with the α2 C-terminus and the clamp-binding loop. Most notable is a trans-acting γ/δ' hydroxyl group that 99% of other AAA+ proteins lack. Mutation of this hydroxyl to a methyl group impedes clamp binding and opening, DNA binding, and ATP hydrolysis-implying a remarkably clamp-loader-specific function.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
DARC workflow diagram. Algorithmic steps are illustrated schematically using two distinct and independent analyses: on the left, an analysis of P-loop GTPases that shows the steps to estimate CDC, 3DCDC and DCCP with Ras GTPase (pdbid: 1ctq) as the query; and, on the right, an analysis of GCN5-related N-acetyltransferases (GNAT) that shows the steps to estimate CLCP with Gna1 (pdbid: 4ag9) as the query; both analyses show the steps to estimate CP and 3DCP. The input, computational operations, and output are indicated by green, yellow, and white text, respectively. First, DARC applies BPPS to define the query protein lineage (within an implicit superfamily hierarchy) based on discriminating pattern residues, which correspond to CP constraints and which are visualized both in available 3D structures and in a contrast alignment. Note that, for clarity, each lineage in this diagram only extends to the query family level, even though DARC can extend the lineage to the subfamily and deeper levels. Second, DARC performs DCA on the family sub-MSA (as shown for the GTPases); the highest DC-scoring residue pairs (i.e., CDC constraints) are displayed in available 3D-structures. Third, using Initial Cluster Analysis (ICA)(as described in Supplementary Methods), DARC identifies statistically surprising 3D-clusters of pattern residues (i.e., CLCP constraints); this is shown for the Gna1 family, which exhibits a very high degree of clustering. Finally, using ICA, DARC identifies any significant correspondence among 3D contacts, DC-pairs and BPPS patterns (i.e., 3DCDC, DCCP, and 3DCP constraints).
Figure 2
Figure 2
Constraints shown within the E. coli DNA clamp loader complex bound to primer DNA and an ATP analog (pdb_id: 3glf). (a) The γ-subunit. An AAA+ module, which consists of an ATPase domain (I) and a three-helix bundle (II), is followed by a collar domain (III) that conjoins the five subunits. The γ2 subunit is shown; δ and δ’ are inactive but have similar architectures. (b) Domain I of γ2 showing Walker A (K51), Walker B (E127), and sensor 1 (T157; red) residues associated with ATP binding and hydrolysis, the trans-acting “R-finger” (R169) (also present in δ’ and most AAA + proteins) and T165 (blue), which was mutated to valine. (cg) DARC-defined pattern residues and top DC-scoring pairs. (c) Interface between the γ1 ATP binding site and the γ2 helices α3 and α4. Likely hydrogen or ionic bond forming oxygen and nitrogen atoms are shown in red and blue, respectively. (d) Top-scoring clamp-loader-specific DC-pairs and γ- and γ/δ’-pattern residues. The γ-residues cluster around the catalytic base E127-γ1 (circled in red) (CLSP: p = 1.8 × 10−9). The γ/δ’-residues cluster around L140-γ2 (circled in blue) (CLSP: p = 6.7 × 10−10). Color scheme: Backbones of γ1, γ2 and γ3: blue, yellow, and red, respectively. Putative β-clamp binding loops C1 and C2 in γ1, γ2, and γ3, marine blue, orange, and pink, respectively. Helices α2, α3, and α4 of γ2, orange. Walker A and B catalytic residues in γ1 and the trans-acting R-fingers in γ2 and γ3, yellow; γ/δ’-residues, blue; γ-residues, red. Rods denoting high DC-scoring, clamp-loader-specific residue pairs: between the α2 and α3 N-termini, magenta; linked to the C1 or C2 loops, purple; between the α1 helix and the β4 strand, green; DNA and ATP analog ADP•BeF3, cyan; Zn++, gray; Mg++, green. (e) γ-residues near the α1 helix, the N-terminus of which coordinates with ATP. (f) K121-γ and the C1 and C2 loops proposed to bind to the β-clamp. (g) Region in δ’ corresponding to that shown for γ2 in panel (f); orange residues are distinctive of δ’. For clarity, DC-pairs between the α2 and α3 N-termini are not shown.
Figure 3
Figure 3
Bacterial DNA clamp loader contrast alignments. BPPS-generated alignments highlighting residues distinctive of the AAA+ superfamily, of the γ/δ’ subgroup, and of γ but not δ’. Residues are highlighted to indicate amino acid biochemical properties based on the following color code: red font with yellow highlight, non-polar (AVILMWFY); blue font with yellow highlight, cysteine (C); red, acidic (DE); cyan, basic (KR); magenta, polar (STNQ); green, glycine (G); blue, histidine (H); black, proline (P). Non-conserved positions and non-pattern residues are shown in gray font. The leftmost columns are colored the same as the residue sidechains in Fig. 2 and give the phylum for each sequence except for the two (proteobacterial) E. coli proteins used as queries, which are denoted by their pdb identifiers. (For more extensive alignments and for NCBI sequence identifiers, see Fig. S1.) The heights of the red bars above each highlighted column estimate the selective pressure imposed on pattern residues at that position using a semi-logarithmic scale. Directly below the representative aligned sequences, the characteristic residues at each position in the full alignment are shown and, directly below these, corresponding frequencies (after weighting for sequence redundancy) are given in integer tenths. A ‘7’, for example, indicates that the corresponding residue occurs in 70–80% of the sequences in the alignment. (a) Contrast alignment highlighting residues distinguishing γ from δ’. Below this the residue positions for the E. coli γ subunit are given and below these are predicted secondary structure elements (symbol: H, helix; E, strand) and their designations. Secondary structure assignments were calculated for the γ subunit using DSSP. (b) Contrast alignment highlighting residues distinguishing γ and δ’ (top and bottom five sequences, respectively) from other AAA+ modules. The positions listed at the bottom again correspond to the E. coli γ subunit. Below these are indicated the putative clamp binding C1 and C2 loops and the taThrγδ’, which was mutated to valine. (c) Contrast alignment highlighting those residues most distinctive of the AAA+ superfamily. Below this, the locations of motifs characteristic of AAA+ ATPases are indicated.
Figure 4
Figure 4
Equilibrium clamp binding and clamp opening for wild type versus mutant clamp loaders. The binding or opening activity of a clamp loader mutant was measured side-by-side with the wild-type (wt) clamp loader in triplicate. The average values and standard deviations are plotted. These titration data were fit to Equation 3 in Supplementary Methods to calculate apparent dissociation constants (Kd,app) in binding assays and apparent opening constants (Kop,app) in opening assays. The apparent equilibrium constants are a function of both the equilibrium constants for the initial binding step (Kd) and the subsequent opening step (Kop) for the two-step reaction. Because of this, the value of Kd,app will be smaller than Kop,app unless the value for Kop is much less than 1 as for the wt clamp loader. (a-c) Clamp binding assays contained 0.5 mM ATP, 10 nM β-PY, and wt (black circles) or mutant clamp loaders (grey squares). PY fluorescence increases when a clamp loader binds the clamp. (a) Plot for γ-mutant. Kd,app was 2.1 ± 0.4 nM for wt and 2.0 ± 0.6 nM for mutant. (b) Plot for δ’-mutant. Kd,app was 3.8 ± 1.2 nM for wt and 8.4 ± 1.2 nM for mutant. (c) Plot for γδ’-mutant. Kd,app was 2.7 ± 0.4 nM for wt to 19.3 ± 1.5 nM for mutant. (d–f) In clamp opening assays, AF488 fluorescence was measured when γ-clamp loader was added to β-AF4882. Clamp opening separates the two fluorophores and increases fluorescence. Assays contained 0.5 mM ATP, 10 nM β-AF4882, and wt (black circles) or mutant γ complexes (grey squares). (d) Plot for γ-mutant. Kop,app was 3.0 ± 0.6 nM for wt and 10.3 ± 2.9 nM for mutant. (e) Plot for δ’-mutant. Kop,app was 2.7 ± 0.08 nM for wt and 10.4 ± 2.1 nM for mutant. (f) Plot for γ/δ’ -mutant. Kop,app was 3.3 ± 0.6 nM for wt and 52.0 ± 6.8 nM for mutant.
Figure 5
Figure 5
DNA binding, ATP hydrolysis and ATP binding affinities of mutant clamp loaders. Assays were performed as described in Supplementary Methods. For each panel, individual data points from three independent experiments are shown with the horizontal line representing the mean value. (a) Equilibrium binding to X-rhodamine-labeled primed template DNA by wild type (wt) and γ/δ’-mutant complexes (black and grey points, respectively) was measured by fluorescence anisotropy. Assays contained 50 nM DNA and 0.5 mM ATPγS. ATPγS was included instead of ATP to inhibit the clamp loader’s DNA-dependent ATPase activity and thereby facilitate equilibrium DNA binding. (b) Rates of ATP hydrolysis by wt and mutant clamp loaders were measured under Vmax conditions (Km for wt is 9.3 µM ATP) using a saturating concentration of 1 mM ATP in the absence of the β-clamp. The concentration of the wt clamp loader (50 nM) was 4-fold lower than the mutants (200 nM) so that measured rates would be the same order of magnitude. Concentration-adjusted rates for the wt clamp loader are shown above the data measured at 50 nM. Values of kcat were 0.111 ± 0.003 for wt, and 0.011 ± 0.002, 0.024 ± 0.001, and 0.0062 ± 0.0008 s−1 for γ-, δ’-, and γ/δ’-mutants, respectively. (c) Rates of DNA-dependent ATP hydrolysis were measured for the wt (black points) and the γ/δ’-mutant complex (grey points) in assays containing 0.5 mM ATP, 250 nM clamp loader, and primed template DNA. (d) ATP and MgCl2 binding to wt (black points) and γ/δ’-mutant clamp loader (grey points) was measured by differential scanning fluorimetry. Thermal stability (Tm values) were measured in assays containing the clamp loader only, the clamp loader and MgCl2, the clamp loader and ATP, or the clamp loader, ATP and MgCl2. (e) ATP hydrolysis was measured in steady-state clamp loading assays for the wt and γ/δ’-mutant clamp loaders. Assays contained 50 nM wt or mutant clamp loader, 200 nM β-clamp, 500 nM DNA, and 1 mM ATP.

Similar articles

See all similar articles

References

    1. Neuwald AF, Aravind L, Spouge JL, Koonin EV. AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 1999;9:27–43. - PubMed
    1. Tucker PA, Sallai L. The AAA+ superfamily–a myriad of motions. Curr Opin Struct Biol. 2007;17:641–652. doi: 10.1016/j.sbi.2007.09.012. - DOI - PubMed
    1. Capra JA, Singh M. Characterization and prediction of residues determining protein functional specificity. Bioinformatics. 2008;24:1473–1480. doi: 10.1093/bioinformatics/btn214. - DOI - PMC - PubMed
    1. Fischer JD, Mayer CE, Soding J. Prediction of protein functional residues from sequence by probability density estimation. Bioinformatics. 2008;24:613–620. doi: 10.1093/bioinformatics/btm626. - DOI - PubMed
    1. Kalinina OV, Gelfand MS, Russell RB. Combining specificity determining and conserved residues improves functional site prediction. BMC Bioinformatics. 2009;10:174. doi: 10.1186/1471-2105-10-174. - DOI - PMC - PubMed
Feedback