Non-catalytic participation of the Pin1 peptidyl-prolyl isomerase domain in target binding

Abstract

Pin1 is a phosphorylation-dependent peptidyl-prolyl isomerase (PPIase) that has the potential to add an additional level of regulation within protein kinase mediated signaling pathways. Furthermore, there is a mounting body of evidence implicating Pin1 in the emergence of pathological phenotypes in neurodegeneration and cancer through the isomerization of a wide variety of substrates at peptidyl-prolyl bonds where the residue preceding proline is a phosphorylated serine or threonine residue (i.e., pS/T-P motifs). A key step in this regulatory process is the interaction of Pin-1 with its substrates. This is a complex process since Pin1 is composed of two domains, the catalytic PPIase domain, and a type IV WW domain, both of which recognize pS/T-P motifs. The observation that the WW domain exhibits considerably higher binding affinity for pS/T-P motifs has led to predictions that the two domains may have distinct roles in mediating the actions of Pin1 on its substrates. To evaluate the participation of its individual domains in target binding, we performed GST pulldowns to monitor interactions between various forms of Pin1 and mitotic phospho-proteins that revealed two classes of Pin-1 interacting proteins, differing in their requirement for residues within the PPIase domain. From these observations, we consider models for Pin1-substrate interactions and the potential functions of the different classes of Pin1 interacting proteins. We also compare sequences that are recognized by Pin1 within its individual interaction partners to investigate the underlying basis for its different types of interactions.

Keywords: Pin1; WW domain; peptidyl-prolyl isomerase; peptidyl-prolyl isomerization; phosphorylation-dependent interactions.

Figure 1

(A) The sequential binding model proposes that the WW domain binds first, bringing the PPIase domain proximal to its targets (Zhou et al., ; Lu et al., 2002). (B) The multimeric binding model proposes that the WW domain anchors Pin1 in multimeric complexes, allowing the PPIase domain to target other substrates in the complex (Jacobs et al., 2003). (C) The catalysis-first binding model proposes that the PPIase domain catalyzes the cis to trans isomerization of the target site to allow trans-isomer-specific WW domain binding (Wintjens et al., 2001). (D) The simultaneous binding model proposes that the WW and PPIase domains bind simultaneously with low-affinity to multiply phosphorylated targets.

Figure 2

GST pull-downs with Pin1 mutants reveal two different types of binding proteins. (A) Large-scale GST pull-downs. One to two milligrams of nocodazole-treated HeLa lysates were incubated with GST fusion proteins bound to glutathione beads. Interacting proteins were run on a 5–12% gradient gel and stained with Coomassie Blue. (B) GST pull-downs were performed as in (A) with 1 mg of HeLa lysate. Interacting proteins were run on 10% SDS-PAGE, transferred to PVDF and blotted with the indicated antibody. To better compare the amount of fusion protein on the beads, 1/200 of each pull-down was run on a gel and stained with Coomassie Blue (PD). (C) Quantification of blots shown in (B). Results are the mean of three independent pull-down experiments ±S.E.M.

Figure 3

p54nrb/NonO binding to R68A/R69A is not due to protein abundance. (A) One to two milligrams of nocodazole-treated HeLa lysates were incubated with GST fusion proteins bound to glutathione beads. Serial dilutions of the pull-down were run on a 5–12% gradient gel, and immunoblotted with anti-p54nrb/NonO. (B) Hundred micrograms of nocodazole-treated HeLa lysates (N/C) were incubated with GST fusion proteins bound to glutathione beads. Interacting proteins were run on a 5–12% gradient gel and stained with Coomassie Blue. (C) HeLa cells were either transfected with HA-NonO or left untransfected (NT). Lysates were used in GST pull-downs (PD's) as above and immunoblotted with the indicated antibody. Endogenous NonO is indicated with End. (D) HeLa cells were transfected with HA-NonO or HA-Cdc25C or left untransfected (NT). Lysates were used in pull-downs (PD) as above and transfected proteins were detected with anti-HA. For the combined sample, 1 mg of HA-NonO lysate was mixed with 1 mg of HA-Cdc25C lysate.

Figure 4

(A) Functional categorization of Pin1 interactors. (B) Subcellular location of Pin1 interactors, from UniProtKB. (C) Frequency chart of amino acids in the alignment of Pin1 binding sites oriented at the S/T-P motif, using the WebLogo service (Schneider and Stephens, ; Crooks et al., 2004).

Figure 5

Phosphopeptide binding by the WW and PPIase domains. (A) Structures of the WW domain (left, 1F8A) and PPIase domain (right, 2ITK) of Pin1 bound to peptides or peptide inhibitors. A general trace of the backbone of the peptide is shown in blue. (B) Sequences of sites in the two Pin1 binding proteins used in this study, Cdc25C and NonO, with pT-P and +1 residues highlighted.