Target identification by chromatographic co-elution: monitoring of drug-protein interactions without immobilization or chemical derivatization

Abstract

Bioactive molecules typically mediate their biological effects through direct physical association with one or more cellular proteins. The detection of drug-target interactions is therefore essential for the characterization of compound mechanism of action and off-target effects, but generic label-free approaches for detecting binding events in biological mixtures have remained elusive. Here, we report a method termed target identification by chromatographic co-elution (TICC) for routinely monitoring the interaction of drugs with cellular proteins under nearly physiological conditions in vitro based on simple liquid chromatographic separations of cell-free lysates. Correlative proteomic analysis of drug-bound protein fractions by shotgun sequencing is then performed to identify candidate target(s). The method is highly reproducible, does not require immobilization or derivatization of drug or protein, and is applicable to diverse natural products and synthetic compounds. The capability of TICC to detect known drug-protein target physical interactions (K(d) range: micromolar to nanomolar) is demonstrated both qualitatively and quantitatively. We subsequently used TICC to uncover the sterol biosynthetic enzyme Erg6p as a novel putative anti-fungal target. Furthermore, TICC identified Asc1 and Dak1, a core 40 S ribosomal protein that represses gene expression, and dihydroxyacetone kinase involved in stress adaptation, respectively, as novel yeast targets of a dopamine receptor agonist.

Fig. 1.

TICC method for monitoring ligand-protein interactions. a, schematic illustrating representative characteristic elution profiles of drug (panel i) or target alone (panel ii) during nondenaturing IEX-HPLC. Once bound to its direct target (panel iii), the retention time of the compound “shifts” to that of its interacting protein partner (i.e. ligand-protein complex), even in the presence (panel iv) of many irrelevant competitor proteins. b, typical TICC workflow: drug-treated (in vivo) or untreated cells (in vitro dosing) are lysed, and soluble protein extracts containing the ligand of interest are fractionated using dual IEX-HPLC. All of the fractions are collected and analyzed for the presence of drug by LC-MS/MS using SRM to characterize the profiles of free (unbound) and bound (shifted) ligand. Proteins present in the bound drug-containing (shifted) fractions are then analyzed by nanoflow LC-MS/MS to identify candidate co-eluting targets.

Fig. 2.

Stable interaction of methotrexate with its primary target dihydrofolate reductase. a, TICC of recombinant DHFR (100 μm) with various amounts (as indicated) of MTX by dual IEX-HPLC. The resulting UV absorbance (280 nm) traces for DHFR alone (top panel) or DHFR mixed with 50 (second panel), 100 (third panel), 125 (fourth panel), 150 (fifth panel), or 200 μm of MTX (bottom panel) are shown. A drug-induced conformational change is evident from the preferential detection of one of two prominent DHFR peaks as the MTX concentration is increased. Target saturation occurred at 125 μm MTX, whereas the peak intensity of free drug further increased proportionally to the amount of excess compound. b, selective DHFR-MTX co-elution in a complex mixture. The top three panels show dual IEX-HPLC elution profiles of E. coli cell lysate (650 μg of total protein) or extract dosed with either recombinant DHFR (25 μm) or MTX (50 μm) only. The presence of DHFR is indicated by black shading, whereas the black circle denotes DHFR-bound MTX. The bottom three panels show the corresponding elution profiles of extract dosed with a saturating amount (50 μm) of MTX and 5, 0.5, or 0.05 μm DHFR. Despite the presence of many other proteins, the specific interaction (co-elution) of MTX with DHFR was evident in each case by TICC, although the proportion of DHFR in the protein mixture was only 1, 0.1, or 0.01%, respectively.

Fig. 3.

Target detection and identification for radicicol and sordarin using TICC. a, in vitro drug dosing experiments using HeLa cell cytosolic protein extract. The top four chromatograms show dual IEX-HPLC elution profiles of radicicol (20 μm) mixed with 50, 200, 400, or 800 μg of lysate. A single peak representing protein-bound drug (black circle) was detected with increasing intensity by SRM in proportion to total protein load, whereas no column retention was observed for free drug. The heat map (bottom panel) shows the spectral counts of high confidence proteins identified by LC-MS/MS, including the known target Hsp90 (arrow) whose proteomic pattern correlated most closely with the radicicol profile. b and c, dual IEX-HPLC fractionation of radicicol (20 μm) dosing to a yeast whole cell extract (b) or after in vivo treatment of yeast for 20 min prior to cell lysis (c). Protein-bound radicicol detected by SRM (black circle), and the spectral counts obtained for Hsp90 in the same fractions by LC-MS/MS are reported. N.D., not determined. d, quantitative comparison of sordarin-binding in protein lysate prepared from wild type (WT) or sordarin-resistant (Sor^R) yeast strains. A marked reduction in protein-bound drug, denoted with a stippled box, was observed in the resistant strain, reflecting the lower affinity of mutant elf2 for sordarin. Excess free drug is indicated with a bracket.

Fig. 4.

Target identification for trichostatin A. a, The amount of TSA in each fraction is quantified after tandem heparin dual IEX-HPLC fractionation (method 2) of 2.4 mg of HeLa lysate dosed with 1 μm of TSA (top panel). The overlaid profiles of TSA with histone deacetylases HDAC1 and HDAC2 are shown in bottom panel, and HDAC1 and HDAC2 are known interactors of TSA. b, Drug elution profiles overlaid with two example histone deacetylase complexes shows good agreement of drug with nucleosome remodelling and histone deacetylation (NURD) complex (top panel) and poor agreement with BHC histone deacetylase complex (bottom panel) using the protein data collected in this study.

Fig. 5.

Target identification for the anti-fungal compound 4513-0042. Dual IEX-HPLC elution profile of the antifungal compound 4513-0042 alone (a) or after in vitro dosing to a yeast whole cell extract (b). A new peak representing protein-bound drug (black circle) is evident at 64–66 min. c, plot showing the relative binding of 4513-0042 to purified GST-Erg6p, GST alone, or a BSA control. After incubation and separation by spin column gel filtration, the amount of compound in the protein-bound flow-through fraction (void volume) was quantified by SRM. d, plasmid-based overexpression of GST-Erg6p confers significant resistance to 4513-0042 as compared with a parental wild type (WT) strain or cells expressing an unrelated yeast factor (GST-Fra1p). Growth was recorded using triplicate cell culture readings at A₆₀₀, and the ratio of drug treatment to no drug control was plotted. The error bars represent the coefficient of variation; a two-tailed Student's t test (equal variance) was applied. *, p < 0.01 compared with wild type 300 μm sample.

Fig. 6.

Target confirmation for the psychoactive drug A77636 using TAP immunoprecipitation. Amount of drug bound (pmol) in fractions 17–31 (a) and fractions 45–47 (b) after subjecting individual IP lysates of each putative protein target dosed with 2 μm A 77636 to weak cation exchange fractionation pH 6.0 method. Using this method, A77636 interacted nonspecifically with the column, and a low leakage background level (around limit of quantitation, 1 pmol) was observed in all fractions. The results for drug control (DC) fractionation and control wild type (WT) fractionations are also included as references. Dashed lines illustrate significance thresholds calculated as three times the average amount of drug found in drug control and wild type. ALD6 exceeds significance threshold but is not considered to bind A77636 because it does not exhibit the expected peak shape and has poor correlation to bound drug profile. ASC1 and DAK1 are validated as bona fide interacting proteins for A77636.