The human brain somatostatin interactome: SST binds selectively to P-type family ATPases

Abstract

Somatostatin (SST) is a cyclic peptide that is understood to inhibit the release of hormones and neurotransmitters from a variety of cells by binding to one of five canonical G protein-coupled SST receptors (SSTR1 to SSTR5). Recently, SST was also observed to interact with the amyloid beta (Aβ) peptide and affect its aggregation kinetics, raising the possibility that it may bind other brain proteins. Here we report on an SST interactome analysis that made use of human brain extracts as biological source material and incorporated advanced mass spectrometry workflows for the relative quantitation of SST binding proteins. The analysis revealed SST to predominantly bind several members of the P-type family of ATPases. Subsequent validation experiments confirmed an interaction between SST and the sodium-potassium pump (Na+/K+-ATPase) and identified a tryptophan residue within SST as critical for binding. Functional analyses in three different cell lines indicated that SST might negatively modulate the K+ uptake rate of the Na+/K+-ATPase.

Fig 1. Design of SST interactome experiment.

(A) Schematic representation of the bait peptides used in affinity capture experiments. (B) Workflow of SST interactome analysis designed to compare binders of N-terminally biotinylated-SST28 (x) and SST14 (y) to a truncated mutant peptide, SST14Δ7–10 (z).

Fig 2. Benchmarks of SST interactome analysis and preliminary data.

(A) The number PSMs versus the false discovery rate (FDR) in the interactome dataset. (B) Relative quantitation of propionyl-CoA carboxylase, which displayed similar levels of enrichment across all samples, and Na⁺/K⁺-transporting ATPase subunit alpha-1, which was highly enriched in biotin-SST28 affinity capture eluates. The box plot depicts the enrichment ratios (R) of individual propionyl-CoA carboxylase peptides used for quantitation in log2 space, in addition to the median peptide ratio and Inter Quartile Range (IQR). A subset of peptides (red circles) was eliminated from the quantitation due to redundancy or failure to pass stringency thresholds. Relative protein levels are depicted as ratios, with the ion intensities corresponding to the heaviest isobaric labels acting as the reference. (C) ‘Cellular Component’ and ‘Molecular Function’ Gene Ontology analyses of the shortlisted proteins that displayed the highest enrichment in biotin-SST28 affinity capture eluates.

Fig 3. Identification of ATP1A1 as a candidate interactor of SST28/14.

(A) Workflow of competitive binding experiments undertaken to validate the SST-Na⁺/K⁺-transporting ATPase subunit alpha-1 interaction. (B) SDS-PAGE analysis of SST28 affinity capture eluates visualized by immunoblot and silver stain. Immunoblot against ATP1A1 reveals that binding of ATP1A1 to biotin-SST28 can be blocked by pre-incubation of the brain lysate with free SST14 (50 μM; lane 4). The prominent bands in the eluates corresponding to ATP1A1 are absent when the same samples are analyzed by silver stain (lanes 7, 8). Note the presence of intense bands in the input and unbound fractions (lanes 5, 6) that are absent from SST28 affinity capture eluates, indicating that SST is not inclined to interact with other proteins simply based on their abundance. The bands at approximately 50, 14, and 4 kDa in the eluate samples likely correspond to streptavidin subunits and biotin-SST28 peptides that leached off the affinity capture resin during the elution step, given their absence from the input and unbound fractions.

Fig 4. Validation of SST binding to Na⁺/K⁺-ATPase alpha and beta subunits.

(A) Immunoblot analysis of competitive binding experiment with SST14 and the mutant peptide SST14-W8P (left) and quantification of western blot data (right). Capture of ATP1A1 and ATP1B1 by biotin-SST28 can be blocked by pre-incubation of the brain lysate with free SST14 (50 μM; lanes 5–7), but not with a mutant SST14 with an amino acid substitution (W8P) in the receptor-binding site (50 μM; lanes 8–10). Note that the negative control bait peptide, biotin-SST14Δ7–10, failed to capture any detectable ATP1A1 (lanes 11–13). Coomassie staining of the same blot confirms that an equal amount of protein was loaded in each well, using the streptavidin subunits released from the affinity matrix as a loading control (lower panel). National Institutes of Health (NIH) Image J densitometry analyses of the anti-ATP1A1- and anti-ATP1B1-reactive bands appearing near the 98 kDa and 49 kDa molecular weight markers are shown in the panel to the right (B) Capture of ATP1A1 by biotin-SST28 can be blocked by pre-incubation of the brain lysate with free SST28 (50 μM; lane 6) or SST14 (50 μM; lane 7), but not with SST14-W8P (50 μM; lane 8). Truncated versions of SST14 also fail to block capture of ATP1A1 (50 μM; lanes 9, 10), despite containing the receptor binding sequence (FWKT). Coomassie staining confirms that protein loading was consistent across the gel (middle panel). NIH ImageJ densitometry of the anti-ATP1A1-reactive band appearing is shown in the lower panel (C) Capture of ATP1A1 by biotin SST-28 can be blocked by pre-incubation of the brain lysate with free SST28 (25 μM or 50 μM; lanes 3, 5) but not the similar neuropeptide, VIP (25 μM or 50 μM; lanes 4,6) as shown by western blot (upper panel) and densitometry of the anti-ATP1A1-reactive band (lower panel).

Fig 5. SST14 regulates ⁸⁶Rb⁺ uptake by the Na⁺/K⁺-ATPase in ReN, N2a and HEK293 cells.

The addition of SST14 (50 μM) to ReN, N2a and HEK293 cells reduces the uptake of ⁸⁶Rb⁺ to 91.5%, 53.6% and 74.4%, respectively, compared to controls. Following treatment with ouabain (2 mM) the uptake of ⁸⁶Rb⁺ is reduced to 50.7% in ReN cells, 8.9% in N2a cells and 54.3% in HEK293 cells, respectively. Note that the addition of SST14 to ouabain treated cells does not further reduce the uptake of ⁸⁶Rb⁺ compared to cells treated with ouabain alone.