Processing of Fluorescent Proteins May Prevent Detection of Prion Particles in [ PSI+] Cells

Abstract

Yeast is a convenient model for studying protein aggregation as it is known to propagate amyloid prions. [PSI⁺] is the prion form of the release factor eRF3 (Sup35). Aggregated Sup35 causes defects in termination of translation, which results in nonsense suppression in strains carrying premature stop codons. N-terminal and middle (M) domains of Sup35 are necessary and sufficient for maintaining [PSI⁺] in cells while preserving the prion strain's properties. For this reason, Sup35NM fused to fluorescent proteins is often used for [PSI⁺] detection and investigation. However, we found that in such chimeric constructs, not all fluorescent proteins allow the reliable detection of Sup35 aggregates. Particularly, transient overproduction of Sup35NM-mCherry resulted in a diffuse fluorescent pattern in the [PSI⁺] cells, while no loss of prions and no effect on the Sup35NM prion properties could be observed. This effect was reproduced in various unrelated strain backgrounds and prion variants. In contrast, Sup35NM fused to another red fluorescent protein, TagRFP-T, allowed the detection of [PSI⁺] aggregates. Analysis of protein lysates showed that Sup35NM-mCherry is actively degraded in the cell. This degradation was not caused by vacuolar proteases and the ubiquitin-proteasomal system implicated in the Sup35 processing. Even though the intensity of this proteolysis was higher than that of Sup35NM-GFP, it was roughly the same as in the case of Sup35NM-TagRFP-T. Thus, it is possible that, in contrast to TagRFP-T, degradation products of Sup35NM-mCherry still preserve their fluorescent properties while losing the ability to decorate pre-existing Sup35 aggregates. This results in diffuse fluorescence despite the presence of the prion aggregates in the cell. Thus, tagging with fluorescent proteins should be used with caution, as such proteolysis may increase the rate of false-negative results when detecting prion-bearing cells.

Keywords: GFP; RFP; Sup35; [PSI+]; mCherry; prion; yeast.

Figure 1

Fusion of Sup35NM to mCherry does not allow the detection of [PSI⁺] aggregates. Clones expressing copper-inducible constructs in OT56 ([PSI⁺][PIN⁺]), 1-OT56 ([psi^-][PIN⁺]), 2-OT56 ([psi^-][pin^-]) (A), and other [PSI⁺] strains (B) were analyzed with fluorescence microscopy. Shown are representative images of either green (GFP) or red (RFP) fluorescence acquired using respective filter sets. Scale bar equals 10 $\mathsf{\mu }$M. Cells producing Sup35NM-mCherry and Sup35NM-TagRFP-T are henceforth denoted as Sup35NM-mC and Sup35NM-TR-T, respectively. (C). Quantification of the frequencies of the fluorescent foci detection in cells from B (see Supplementary Table S2 for the detailed cell count data). NM stands for Sup35NM. **, p < 0.01; ***, p < 0.001 in Fisher’s exact test. Error bars indicate percentage error.

Figure 2

Sup35NM-mCherry protein retains the prion properties and does not cause the loss of [PSI⁺]. (A). Phenotypic assay performed by replica plating of OT56 [PSI⁺][PIN⁺], 1-OT56 ([psi^-][PIN⁺]), and 2-OT56 ([psi^-][pin^-]) strains, transformed with vectors pRS315CmC, pRS315CNMmC, and pRS315CNMG for copper-inducible expression of mCherry, SUP35NM-mCherry, and SUP35NM-GFP, respectively. Similar to Sup35NM-GFP, Sup35NM-mCherry enhances nonsense-suppression and prion toxicity in the [PSI⁺][PIN⁺] strain and promotes de novo [PSI⁺] formation in the [psi^-][PIN⁺] strain. (B). Tenfold serial dilutions of the OT56 transformed with vectors from A, as well as pRS315CG, pR15CUP-NM-yTagRFP-T, and pR15CUP-yTagRFP-T for the inducible expression of GFP, SUP35NM-yTagRFP-T, and yTagRFP-T, respectively.

Figure 3

Both Sup35NM-TagRFP-T and Sup35NM-mCherry, but not Sup35NM-GFP, are subjected to proteolysis in yeast cells. (A,B) Western blot analysis of protein lysates from clones shown in Figure 2A and Figure 1A, respectively. Arrowheads indicate major additional product detected by anti-Sup35NM antibodies, which is similar in weight to the untagged Sup35NM. FP—fluorescent protein (GFP, or mCherry, or TagRFP-T).

Figure 4

Sup35NM-mCherry degradation leads to decrease in relative amount of aggregates detectable by fluorescence. Proteins from OT56 ([PSI⁺] [PIN⁺]) or 2-OT56 ([psi^-] [pin^-]) cells expressing indicated constructs were extracted in non-denaturing conditions. After addition of the SDS-containing buffer, samples were either boiled for 5 min (+) or incubated at room temperature (−). SDS-PAGE was visualized using a gel documentation system with filter sets for either green (GFP) or red (RFP) fluorescence.

Figure 5

Vacuolar proteases do not influence Sup35NM-mCherry properties. (A). Fluorescent images of the prb1Δ or pep4Δ cells overexpressing SUP35NM fused to different fluorescent protein genes. (B). Western blot analysis of protein lysates from prb1Δ strains overproducing the indicated proteins. Arrowheads indicate major additional products detected by anti-Sup35NM antibodies. (C). Tenfold serial dilutions of the prb1$\Delta$0-P-74-D694 strain co-transformed either with pRS315 together with pIM35 (TagRFP-T) or pR16CUP-NM-yTagRFP-T (Sup35NM-TR-T), or with pRS316 together with pRS315CmC (mCherry) or pRS315CNMmC (Sup35NM-mC). The panels for each medium are taken from the same plate. (D). Tenfold serial dilutions of the pep4Δ strains transformed with plasmids for expression of the indicated constructs.

Figure 6

Proteasomal degradation is not responsible for the Sup35NM-mCherry processing and diffuse distribution in [PSI⁺] cells. OT56 cells expressing SUP35NM-mCherry under control of CUP1 promoter were treated with MG132 for 4 hours; DMSO treatment was used as a control. MG132 was either added simultaneously with CuSO₄ (1) or one hour prior to CuSO₄ addition (2). Cells were analyzed with fluorescence microscopy (A) or were subjected to protein extraction and Western blotting (B). Anti-Sup35NM antibodies were used.