Inhibition of Aβ42 oligomerization in yeast by a PICALM ortholog and certain FDA approved drugs

Abstract

The formation of small Aβ₄₂ oligomers has been implicated as a toxic species in Alzheimer disease (AD). In strong support of this hypothesis we found that overexpression of Yap1802, the yeast ortholog of the human AD risk factor, phosphatidylinositol binding clathrin assembly protein (PICALM), reduced oligomerization of Aβ₄₂ fused to a reporter in yeast. Thus we used the Aβ₄₂-reporter system to identify drugs that could be developed into therapies that prevent or arrest AD. From a screen of 1,200 FDA approved drugs and drug-like small compounds we identified 7 drugs that reduce Aβ₄₂ oligomerization in yeast: 3 antipsychotics (bromperidol, haloperidol and azaperone), 2 anesthetics (pramoxine HCl and dyclonine HCl), tamoxifen citrate, and minocycline HCl. Also, all 7 drugs caused Aβ₄₂ to be less toxic to PC12 cells and to relieve toxicity of another yeast AD model in which Aβ₄₂ aggregates targeted to the secretory pathway are toxic. Our results identify drugs that inhibit Aβ₄₂ oligomers from forming in yeast. It remains to be determined if these drugs inhibit Aβ₄₂ oligomerization in mammals and could be developed as a therapeutic treatment for AD.

Keywords: Alzheimer; Aβ42 oligomerization; HTS; PICALM; yeast.

Figure 1. FIGURE 1: Effect of genetic modifiers of toxic HDEL-Aβ₄₂ on Aβ₄₂-RF oligomer formation and Aβ₄₂-MRF aggregates into toxic SDS-stable oligomers in yeast.

(A) Overexpressed Yap1802 but not other previously described HDEL-Aβ₄₂ modifiers dramatically enhanced the translational termination factor activity of Aβ₄₂-RF. Aβ₄₂-RF expressing cells (CUP1::Aβ₄₂-RF) lacking chromosomal SUP35 were transformed with a plasmid carrying the indicated gene under control of the GAL1 promoter. Ten-fold serial dilutions of transformants are shown on plasmid selective non-inducing media (SD-Ura); galactose plasmid selective media (SGal-Ura) to turn on each gene (↑ORF); +Ade medium containing copper to overexpress Aβ₄₂-RF (↑ORF ↑Aβ₄₂-RF); the identical -Ade medium SGal-Ura-Ade+Cu²⁺ (-Ade ↑ORF ↑Aβ₄₂-RF) to determine translational termination factor activity as read-through of the ade1-14 nonsense mutant. The Aβ₄₂-RF overexpressed in cells was shown to have reduced translational termination factor activity as cells grew on -Ade due to aggregation of the fusion protein into small oligomers (Aβ₄₂-RF-e. v.). However, the translation termination factor activity was retained in yeast cells overexpressing Aβ₄₂m2-RF (Aβ₄₂ aggregation-deficient mutant) or YAP1802, due to the absence of oligomer formation, resulting in reduced growth on -Ade (Aβ₄₂m2-RF-e. v.). (B) Yap1802 suppression of Aβ₄₂-RF oligomerization by immunoblot analysis. Total cell lysates were prepared from 2 independent transformants of an Aβ₄₂-RF strain carrying YAP1802 (+) or an empty vector (-) plasmid. Both Aβ₄₂-RF and YAP1802 were overexpressed (SGal-Ura-Ade+Cu²⁺) prior to lysis. Immunoblots were probed with anti-Sup35 RF to evaluate the level of oligomers and monomers. ↑Yap1802 indicates overexpression of YAP1802. The identification of bands as Aβ₄₂-RF dimers and monomers was determined by the estimated sizes of the bands in the immunoblot.

Figure 2. FIGURE 2: Dose-dependent effects of 8 drugs on Aβ₄₂-RF activity as measured by cell growth.

Aβ₄₂-RF expressing cells were treated with each drug at the indicated concentrations. Their effects on cell growth were measured as OD₆₀₀ in +Ade and -Ade media after 4 and 5 days respectively (see Materials and Methods for details). Growth in -Ade is dependent upon Aβ₄₂-RF release factor activity. This activity is expected to be reduced by oligomerizartion. Growth on +Ade does not require Aβ₄₂-RF release factor activity and is used as a control in the presence of each drug. Shown is the relative growth in the presence vs. absence of each drug. Error bars show the standard deviation from three replicates. The asterisks show significance levels of **P < 0.01 or *P < 0.05 according to the Student’s t test between DMSO (0) control and drug treatment at marked concentration for growth in -Ade.

Figure 3. FIGURE 3: Seven drugs suppress Aβ₄₂-RF oligomerization in yeast in a dose-dependent manner.

SDS-resistant Aβ₄₂-RF oligomers were detected by immunoblot analysis (upper). The assay strain expressing Aβ₄₂-RF was grown in complex medium in the presence of each compound at the indicated concentrations. Equal amounts of lysate proteins were treated with 1% SDS for 7 mins at room temperature and analyzed by SDS PAGE followed by immunoblotting with anti-Sup35 RF antibodies. Aβ₄₂-RF cells grown in DMSO (0) were used as controls. PGK, yeast 3-Phosphoglycerate Kinase, detected with anti PGK antibodies was an internal control to show the effect of the drugs on protein synthesis in general. Immunoblot signals for Aβ₄₂-RF monomers and oligomers were quantified and converted into % inhibition of oligomer formation from the ratios of oligomers to monomers compared to DMSO controls (0 means no inhibition) (lower). Error bars represent standard deviation from 3 independent immunoblots except for ifenprodil tartrate which shows data from 1 of 2 immunoblots that showed no drug effects. The asterisks show significant levels of **P < 0.01 or *P < 0.05 according to Student’s t test.

Figure 4. FIGURE 4: The protective effect of drugs on Aβ₄₂ induced cytotoxicity in PC12 cells.

(A) Toxicity of Aβ₄₂ assembled under conditions favoring oligomer formation was reduced when assembly was done in the presence of drugs. Aβ₄₂ at 200 μM assembled under conditions favoring oligomer formation in the presence of 100 μM drug or DMSO was diluted and added to PC12 cells at a concentration of 20 μM Aβ₄₂ and 10 μM drug. Cell viability was assessed after 24 hrs growth at 37°C using the MTT assay. The percentage of viable cells is shown. Error bars are the standard deviation of triplicate experiments. (B) Dose-dependent effects of minocycline HCl on Aβ₄₂ associated cytotoxicity. Methods were as in (A). PC12 cells were treated with final concentrations of 20 μM Aβ₄₂ with 5, 10, and 15 µM minocycline HCl or DMSO only (blue bars). Cells were also treated with the indicated amount of minocycline HCl alone (red bars). (C) Effect of drugs on in vitro Aβ₄₂ high molecular fibril assembly in fibril favorable conditions. Aβ₄₂ was assembled into high molecular fibrils at 37°C in the presence of 50 µM drug or control DMSO according to . Thioflavine T (ThT) was added to aliquots at the indicated times and fluorescence, indicative of fiber formation, was measured. (D) Effect of drugs on in vitro labeled Aβ₄₂ (TMR-K-Aβ₄₂) oligomerization. Reduced TMR fluorescence, which measures oligomerization, was recorded as a function of time immediately after dilution of the TMR-K-Aβ₄₂ with addition of 10 µM of each drug or DMSO for a negative control. The asterisks show significance levels of **P < 0.01 or *P < 0.05 according to Student’s t test between DMSO control and drug treatment (A) and DMSO control and drug treatment at 5 hrs (C).

Figure 5. FIGURE 5: The effect of drugs that inhibit Aβ₄₂-RF from forming oligomers on toxicity caused by overexpression of HDEL-Aβ₄₂ in yeast.

Effects of 7 drugs on suppression of HDEL-Aβ₄₂ toxicity and general cell growth are shown. Yeast cells carrying either HDEL-Aβ₄₂ (HDEL-Aβ₄₂ strain) or an empty vector (control) grown overnight in non-inducing plasmid selective media (2% raffinose, SR-Leu) were normalized, diluted to 1 x 10⁵ cells/100 µl of media in each of 96 wells and further grown for 3 days in 2% galactose inducing media (SRGAL-Leu) or 2% raffinose noninducing media (SR-Leu) in the presence of each drug at the indicated concentrations. Growth was measured as OD₆₀₀. Error bars show the standard deviation from three trials. The asterisks show increased relative growth with drug treatment compared to DMSO control (black bar in left panel) at significance levels of **P < 0.01 or *P < 0.05 according to Student’s t test.