The PDZ domain as a complex adaptive system

Abstract

Specific protein associations define the wiring of protein interaction networks and thus control the organization and functioning of the cell as a whole. Peptide recognition by PDZ and other protein interaction domains represents one of the best-studied classes of specific protein associations. However, a mechanistic understanding of the relationship between selectivity and promiscuity commonly observed in the interactions mediated by peptide recognition modules as well as its functional meaning remain elusive. To address these questions in a comprehensive manner, two large populations of artificial and natural peptide ligands of six archetypal PDZ domains from the synaptic proteins PSD95 and SAP97 were generated by target-assisted iterative screening (TAIS) of combinatorial peptide libraries and by synthesis of proteomic fragments, correspondingly. A comparative statistical analysis of affinity-ranked artificial and natural ligands yielded a comprehensive picture of known and novel PDZ ligand specificity determinants, revealing a hitherto unappreciated combination of specificity and adaptive plasticity inherent to PDZ domain recognition. We propose a reconceptualization of the PDZ domain in terms of a complex adaptive system representing a flexible compromise between the rigid order of exquisite specificity and the chaos of unselective promiscuity, which has evolved to mediate two mutually contradictory properties required of such higher order sub-cellular organizations as synapses, cell junctions, and others--organizational structure and organizational plasticity/adaptability. The generalization of this reconceptualization in regard to other protein interaction modules and specific protein associations is consistent with the image of the cell as a complex adaptive macromolecular system as opposed to clockwork.

Figure 1. Binding of 95 artificial phage-displayed ligands to six PDZ domains of PSD95 and SAP97.

Binding histograms were obtained by individual phage ELISA performed on purified GST fusions of the indicated domains immobilized in micro-titer plate wells . For accurate relative affinity evaluations and cross-domain comparison, the slopes of individual ELISA kinetics were determined and normalized by the highest slope value in each of the six sets shown. The 96^th well in each set was loaded with a library aliquot to indicate background. Axis X indicates identification (i.d.) numbers of individual artificial ligands (see individual ligand sequences together with their i.d. numbers in Table S1). Axis Y indicates the normalized relative affinity. The domain organization of the PSD95 and SAP97 proteins is shown above histograms.

Figure 2. RFP (residue frequency patterning) analysis of artificial ligands.

A, The observed-to-expected ratios of individual amino acid frequencies within the whole set of artificial peptide ligands isolated from cDNA and random peptide libraries by TAIS using SAP PDZ domains as targets. B, The frequency distributions of the indicated amino acids within the last sixteen C-terminal positions of aligned artificial peptide ligands. For both A and B: axis Y indicates the observed-to-expected frequency ratio values. The dotted line corresponds to the expected frequency value. Statistically significant overrepresentation is indicated by star symbols. n is sample size, p indicates the chi-square (A) or binomial (B) tests P-values. C, The aligned sequences of artificial peptide ligands are arranged in four groups based on their relative affinities to the PSD95-PDZ1. The numbers in parentheses indicate the range of normalized phage ELISA values within a given affinity group. The ligand positions from “−4” to “−7” are boxed to indicate the area of relative concentration of positively charged residues. Arginines and lysines are highlighted green, while aspartic and glutamic acids are red. The unique i.d. numbers of artificial ligands are indicated on the left from their sequences. The digits above columns indicate the C-terminal position numbering of ligand residues. Analogous arrangements of ligands for other five PDZ domains are shown in Fig. S1A, Fig. S1B and Fig. S1C.

Figure 3. Binding of 126 proteomic fragments (natural ligands) to six PDZ domains of PSD95 and SAP97.

Binding histograms were obtained by peptide ELISA performed on purified GST fusions of the indicated domains immobilized in micro-titer plate wells as described previously . The X axis indicates the i.d. numbers of natural ligands (see Table S2). The Y axis indicates the peptide ELISA kinetics slope value in arbitrary units (a.u.). The 127^th and 128^th wells in each of the six sets were loaded with irrelevant peptides to indicate background signal. To illustrate internal consistency of the affinity evaluations obtained by peptide ELISA and their external consistency with the previously published affinity measurements, the reported affinities of the five PDZ domain-ligand pairs obtained by three different research groups using fluorescence polarization , , are shown. The previously reported affinities provide calibration, suggesting that the individual signals that are higher than 200 a.u. roughly correspond to the interaction affinities of 15 µM K_D or stronger.

Figure 4. Analysis of amino acid frequency biases within the positional window “−4 to −7” in artificial and natural binders of the PSD95-PDZ1 domain.

A, The observed-to-expected ratios of individual (top) and grouped (bottom) amino acid frequencies within the positional window “−4 to −7” in the 32 best artificial (open bars) and 32 best natural (filled bars) peptide ligands. B, The observed-to-expected ratios of individual (top) and grouped (bottom) amino acid frequencies within the positional window “−4 to −7” in the 32 worst natural peptide ligands. Grouping of twenty natural amino acids into eleven conserved physicochemical classes is shown below histograms. Statistically significant over-(under)representation is indicated by star symbols. n is sample size, p indicates the binomial test P-values. C, Comparison of the best (on the left) and worst (on the right) natural binders of the PSD95-PDZ1 domain. The i.d. numbers of natural ligands are indicated on the left from their sequences. Arginines and lysines are highlighted green, while aspartates and glutamates are red. Notice how the relative abundance of negatively charged residues at the “−1” position in the best binders mirrors the relative abundance of positively charged residues at the “−1” position in the worst binders.

Figure 5. Pattern recognition differences between individual PDZ domains of PSD95.

A, Comparison of the sequences that exhibited the highest differential ratios in their relative affinities to PDZ2 domain versus PDZ3 domain (on the left) to the sequences of the best PDZ3 binders (on the right). B, Comparison of the sequences that showed the highest differential ratios in their relative affinities to PDZ2 domain versus PDZ1 domain (on the left) to the sequences of best PDZ1 binders (on the right). Arginines and lysines are shown green, aspartates and glutamates are red.