Quantitative analysis of autophagy-related protein stoichiometry by fluorescence microscopy

Abstract

In yeast, approximately 31 autophagy-related (Atg) proteins have been identified. Most of them reside at the phagophore assembly site (PAS), although the function of the PAS mostly remains unclear. One reason for the latter is the lack of stoichiometric information regarding the Atg proteins at this site. We report the application of fluorescence microscopy to study the amount of Atg proteins at the PAS. We find that an increase in the amount of Atg11 at the PAS enhances the recruitment of Atg8 and Atg9 to this site and facilitates the formation of more cytoplasm-to-vacuole targeting vesicles. In response to autophagy induction, the amount of most Atg proteins remains unchanged at the PAS, whereas we see an enhanced recruitment of Atg8 and 9 at this site. During autophagy, the amount of Atg8 at the PAS showed a periodic change, indicating the formation of autophagosomes. Application of this method and further analysis will provide more insight into the functions of Atg proteins.

Figure 1.

The amount of Atg11 affects the capacity of the Cvt pathway. Strains were cultured in SMD and then starved for 2 h. Cells were harvested before or after starvation and analyzed by Western blotting with anti-Ape1 antiserum. (A) Overexpression of prApe1 overwhelms the Cvt pathway. The wild-type (SEY6210) or prApe1 overexpression (JGY069) strain was analyzed. (B) Overexpression of Atg8 and 19 could not restore the maturation of overexpressed prApe1. The prApe1 overexpression strain harboring pCuAtg19, pCuAtg8, or both was analyzed. (C) A higher level of Atg11 enhanced the processing of overexpressed prApe1. Cells expressing a combination of pHA-Atg11, pCuAtg8, and pCuAtg19 were examined. (D) The CC domains (CC2, CC3, and CC4 but not CC1) were required for the effect of Atg11 on Cvt pathway capacity. Besides Atg19 and 8, full-length (FL) or truncations of Atg11 were expressed in the prApe1 overexpressing strain and the maturation of prApe1 was examined.

Figure 2.

More Cvt vesicles were formed in Atg11-overexpressing cells. Cells (pep4Δ vps4Δ CuApe1; JGY089) expressing either pAtg11, pCuAtg8, and pCuAtg19 or empty vectors were cultured in SMD medium to mid-log phase. Cells were harvested and prepared for EM as described in Materials and methods. Representative images of TEM and IEM are shown in A and B, respectively. Areas with typical Cvt vesicles are shown with higher magnification at the right. In A, the Cvt vesicles were drawn in black in the far right. Bars: (A) 500 nm; (B) 200 nm. (C and D) Quantification of Cvt body accumulation. The number was determined from cells with clearly defined Cvt bodies inside the vacuole.

Figure 3.

Linear relationship between GFP-tagged protein amount and fluorescence intensity. The standard curve showed the relationship between protein amount and fluorescence intensity. Fluorescence intensity and GFP-tagged protein amount were quantified in the following strains: (a) P_ATG27GFP-Atg8; (b) Mlc2-GFP; (c) P_ATG27GFP-Atg8 pep4Δ; (d) P_VPS30-ATG18GFP-Atg8; (e) P_VPS30-ATG18GFP-Atg8 pep4Δ; (f) P_ATG3GFP-Atg8; (g) P_ATG3GFP-Atg8 pep4Δ; (h) P_ATG8GFP-Atg8; (i) Cdc12-GFP; and (j) P_ATG8GFP-Atg8 pep4Δ. Error bars indicate the SD of three independent experiments. For fluorescence intensity, >200 cells of each strain were analyzed by microscopy as described in Materials and methods. Linear regression was performed to generate the formula y = 11.029x.

Figure 4.

Localization patterns do not affect the linearity of the fluorescent signal. (A) Representative images of GFP-Atg8 localization in pep4Δ, atg2Δ, and atg5Δ strains, which result in autophagic body, PAS, and diffuse cytosolic staining patterns, respectively. GFP-Atg8 pep4Δ (YZX254), GFP-Atg8 atg2Δ (JGY054), or GFP-Atg8 atg5Δ (JGY053) cells were starved for 2 h and examined by fluorescence microscopy. Bar, 2 μm. (B) The protein amount of GFP-Atg8 was the same in the pep4Δ, atg2Δ, or atg5Δ backgrounds. The amount of Pgk1 protein in each strain was used as a loading control. (C) Relative fluorescence intensity of GFP-Atg8 per cell in pep4Δ, atg2Δ, and atg5Δ strains (n = 60). Images of all of these strains were quantified as described in Materials and methods. The intensity of GFP-Atg8 in the atg5Δ strain was normalized to 1 as a reference. Error bars indicate the SEM of three independent experiments.

Figure 5.

Effect of Atg11 overexpression on the PAS intensity of Atg8 and 9. (A) On the 2D projection image, circles were drawn around the PAS (arrow) and the adjacent cytosolic region (arrowhead) as background. The difference between these two regions was recorded as the PAS intensity. Bar, 2 μm. (B) The CuApe1 Atg9-GFP strain (JGY072) expressing pHA-Atg11, pCuAtg8, or pCuAtg19 was cultured in SMD medium and examined by microscopy (n = 80). (C) CuApe1 atg8Δ cells (JGY071) harboring pGFP-Atg8 together with pHA-Atg11, pCuAtg19, or empty vector were cultured in SMD medium and examined by microscopy (n = 80). (D) Increase of Atg11-GFP intensity at the PAS in Atg11 overexpression cells. The CuApe1 Atg11-GFP strain (JGY087) transformed with empty vector or pAtg11-GFP was grown in SMD media and the PAS intensity of Atg11-GFP was analyzed by microscopy (n = 80). (E) Overexpression of Ape1 or Atg11 did not induce lipidation of Atg8. Wild-type (SEY6210) cells or the CuApe1 strain (JGY069) expressing pHA-Atg11 or pCuAtg19 were cultured in nutrient-rich medium, and a protein extract was resolved by SDS-PAGE and examined by Western blot with anti-Atg8 antiserum. The asterisk marks a nonspecific band. (F) The total GFP-Atg8 signal was not affected by the increase of Atg11. The same strains as in C were used, but instead of measuring the signal at the PAS, the GFP-Atg8 intensity of the whole cell was measured (n = 60). The relative intensity of cells expressing empty vector was set to 1 as reference, and error bars indicate the SEM of three independent experiments.

Figure 6.

Amount of Atg proteins at the PAS during either the Cvt pathway or autophagy. (A) Boxplot of the amount of Atg proteins at the PAS in growing conditions. The local protein amount was calculated as described in Materials and methods and graphed into a boxplot (n = 100). (B) Difference in the PAS intensity between the Cvt pathway and autophagy. GFP-tagged strains were cultured in nutrient-rich conditions (YPD medium; n = 100) and then shifted to starvation conditions (SD-N medium; n = 60) for 2 h. Microscopy was done before and after starvation. Error bars indicate the SEM of three independent experiments.

Figure 7.

Kinetics of Atg proteins at the PAS during autophagy. GFP-tagged strains were starved for 2 h and analyzed by microscopy. (A) For Atg9-GFP, the PAS intensity remained as a relative constant for 30 min, but only nine time points are shown. Fluorescence intensity at time 0 was normalized to 1 (n = 9). Essentially the same results were seen for Atg1-, Atg11-, Atg16-, Atg17-, and Atg19-GFP. (B) For GFP-Atg8, the PAS intensity changed over time. The time point at which the punctate structure could first be seen was set as time 0 and fluorescence intensity at this time point was set to 1 (n = 11). In A and B, all data points came from three independent experiments and error bars indicate the SEM. (C) Representative images of GFP-Atg8 showing the change of PAS (arrow) intensity. Bar, 2 μm.