Figure 1. Overview of autophagy4
To degrade organic particles or organelles, a double membrane structure called phagophore gradually expands and surrounds the target with the help of the LC3 lipidated to the PE (now termed LC3-II) on the membrane.
55
Figure 2. Ubiquitin-like Atg8 (LC3) is conjugated to membrane PE through E1-E3 like
system23
In eukaryotic cells, Atg7 acts as E1-like ligase, which transfer LC3 (LC3-I) to E2-like ligase Atg3. Atg12-Atg5 conjugate serves as E3-ligase that transfer LC3 on Atg3 to the PE on phagophore membrane. The lipidated LC3 (LC3-II) could promote phagophore recruitment and it is then removed from the PE (membrane) by Atg4 after autophagosome completion.
56
Figure 3. Structure of the human LC3
(a) N-terminal helical domain (NHD) and C-terminal ubiquitin-like domain (ULD) in human LC3. The NHD consists of two α-helices.
(b) Key amino acids residues of the LC3 in this study: Thr6 and Thr12 on NHD were mutated to phosphomimetic aspartate (i.e., T6D and T12D); Asp106, Phe108, and Tyr110 on ULD were selected to make fluorescent mutants (i.e., D106W, F108W, and Y110W). PDB ID, 1V49.
57
Figure 4.CD spectra of various secondary structures30
Solid line, α-helix; long dashed line, antiparallel β-sheet; dotted line, type I β-turn; cross dashed line, extended 31-helix or poly (Pro) II helix; short dashed line, irregular structure. α-helix denaturation could lead to drastic change (mostly significant decrease) of the CD signal at wavelength 222 nm.
58
Figure 5. PCR check of the LC3 T6D, T12D, T6A, and T12A mutageneses
(a) Both PCR products of the T6D and the T12D mutageneses were shown at the bands near 7 kb, about the size of our plasmid (6969 bp).
(b) PCR products of the T6A and T12A mutageneses, also about 7 kb from the same template. Note there were byproducts in the PCR of both the T6D and the T6A mutageneses, each showing single bands lower than 0.5 kb.
59
Figure 6. PCR check for the fluorescent LC3s and their phosphomimetic double
mutants
Each mutagenesis reaction produced an expected product around 7 kb, and all were further undergone blunt-end ligation and confirmed for mutations with sequencing. The phosphomimetic-fluorescent double mutants were built on the template pHM T6D MBP-LC3 or pHM T12D MBP-LC3.
Note: in the Y110W and its double mutants, there were notable byproducts showing single bands lower than 0.5 kb; to increase the efficiency in subsequent transformations, the agarose gel showing only the bands of interest (at positions near 7 kb) were sliced and re-dissolved separately before the next step (the DpnI digestion).
60
Figure 7.SDS-PAGE check for the amylose-resin-pulled down fusion protein MBP-LC3 wild type (about 57 kDa)
Each lane had 15 μl loaded. The source of the proteins was the spun cell lysate (80 ml) from 4-L cell culture. Ft: the supernatant from the lysate freely flowing through the column; Wash: 100-ml column buffer washing through the resin to remove any unbound proteins; E1-E4: elution of the MBP-conjugated protein through sequential additions of 20 ml H2O for four times. Note in this gel, the 15-μl E2 was spilled into 2 wells, thus having two lanes. E1-4 were then selected and pooled (80 ml).
61
Figure 8. SDS-PAGE check for the amylose-resin-pulled down fusion protein
MBP-LC3 T6D (about 57 kDa)
Each lane had 15 μl loaded. The source of the proteins was the spun cell lysate (40 ml) from 2-L cell culture. Ft: the supernatant from the lysate freely flowing through the column; W1: 10-ml column buffer washing through the resin to remove any unbound proteins; W2: 10-ml column buffer further washing through the resin to check the clearance of unbound proteins; E1-E4: elution of the MBP-conjugated protein through sequential additions of 10 ml H2O for four times. E2 and E3 were then selected and pooled (20 ml). Note that the protein marker used for all SDS-PAGE in this study was Prestained Protein Ladder 4 (Geneaid), except in this PAGE, the marker Prestained Protein Marker Broad Range (New England BioLabs) was loaded instead, which showed the poorer resolution in my experimental condition.
62
Figure 9. SDS-PAGE check for the amylose-resin-pulled down fusion protein
MBP-LC3 T12D (about 57 kDa)
Each lane had 15 μl loaded. The source of the proteins was the spun cell lysate (about 180 ml) from 6-L cell culture. Ft: the supernatant from the lysate freely flowing through the column; Wash: 210-ml column buffer washing through the resin to remove any unbound proteins; E1-E4: elution of the MBP-conjugated protein through sequential additions of 22 ml H2O for four times; E5: adding 10 ml H2O to the column after E4 to complete the elution. E1-5 were then selected and pooled (about 100 ml).
63
Figure 10. SDS-PAGE check for the amylose-resin-pulled down fusion protein
MBP-ULD (about 53 kDa)
Each lane had 15 μl loaded. The source of the proteins was the spun cell lysate (about 50 ml) from 4-L cell culture. Ft: the supernatant from the lysate freely flowing through the column; Wash: 42-ml column buffer washing through the resin to remove any unbound proteins; E1-E4: elution of the MBP-conjugated protein through sequential additions of 10 ml H2O for four times. E1-4 were then selected and pooled (about 40 ml).
64
Figure 11.SDS-PAGE check for MBP-removal from the Factor Xa-treated MBP-LC3 wild type by the amylose resin
Each lane had 15 μl loaded. The source of the proteins was the pulled-down MBP-LC3 wild-type (80 ml, but concentrated and dialyzed with the Factor Xa reaction buffer to 20 ml before Factor Xa treatment) from 4-L cell culture, treated with the Factor Xa at 4°C for one week. Ft: the 20-ml reaction product (after dialysis with the column buffer) freely flowing through the column; E1-E4: elution of the MBP-free LC3 through sequential additions of 10-ml column buffer for four times. Wash: 10-ml H2O washing through the resin to check any residual proteins. Ft, E1, and E2 were then selected and pooled (about 40 ml). The upper bands might be the tag MBP (about 42 kDa) whereas the lower bands were the MBP-free LC3 wild type (about 14 kDa).
65
Figure 12. SDS-PAGE check for MBP-removal from the Factor Xa-treated MBP-LC3
T6D by the amylose resin
Each lane had 15 μl loaded. The source of the proteins was the pulled-down MBP-LC3 T6D (20 ml, but concentrated and dialyzed with the Factor Xa reaction buffer to 4 ml before Factor Xa treatment) from 2-L cell culture, treated with the Factor Xa at 4°C for 10 days. Ft: the 4-ml reaction product (after dialysis with the column buffer) freely flowing through the column; E: elution of the MBP-free LC3 T6D by 10-ml column buffer. Ft and E were then selected and pooled (about 14 ml). The dim upper band indicated the tag MBP (about 42 kDa) whereas the lower bands were the MBP-free LC3 T6D (about 14 kDa). Note: in this SDS-PAGE, the samples were run in a 15% gel under constant 120 V for only 65 min, not the typical 90 min.
66
Figure 13. SDS-PAGE check for MBP-removal from the Factor Xa-treated MBP-LC3
T12D by the amylose resin
Each lane had 15 μl loaded. The source of the proteins was the pulled-down MBP-LC3 T12D (100 ml) from 6-L cell culture, treated with the Factor Xa (after dialysis with the Factor Xa reaction buffer) at 4°C for 3 days. Ft: the reaction product (approximately 90 ml, after dialysis with the column buffer) freely flowing through the column; E: elution of the MBP-free LC3 T12D by 35 ml column buffer. Wash: 40 ml H2O washing through the resin to check any residual proteins (e.g., free MBP or MBP-LC3 T12D). Ft and E were then selected and pooled (about 125 ml). The upper bands indicated the MBP (about 42 kDa) and the fusion protein remaining intact (about 57 kDa), and the lower bands were the MBP-free LC3 T12D (about 14 kDa). Note the fraction “E” still contained many unwanted proteins (e.g., MBP and the fusion protein), and the affinity chromatography was repeated for the pool (see Fig. 15).
67
Figure 14. SDS-PAGE check for MBP-removal from the Factor Xa-treated MBP-ULD
by the amylose resin
Each lane had 15 μl loaded. The source of the proteins was the pulled-down MBP-ULD (40 ml) from 4-L cell culture, treated with the Factor Xa (after dialysis with the Factor Xa reaction buffer) at room temperature for 3 days. Ft1: the 40-ml reaction product (after dialysis with the column buffer) freely flowing through the column; E1: elution of the MBP-free ULD with 40 ml column buffer. W1: 40 ml H2O washing through the resin to remove any residual proteins. Then “Ft1” and “E1” went through the regenerated column again, to remove the unwanted free MBP or the intact fusion protein more thoroughly. Ft2, Ft3, E2, and W2 were from the repeated chromatography.
Ft2 and Ft3: the 40-ml “Ft1” and 40-ml “E1” freely flowing through the column, respectively; E2: elution of any free LC3 ULD by additional 10 ml column buffer. W2:
10 ml H2O washing through the resin to check any residual proteins. Ft2, Ft3, E2, and
68
W2 were then selected and pooled (about 100 ml). Note there were many unknown bands in the gel, probably from nonspecific over-digestion of the MBP at room temperature. The upper bands showed the tag MBP (about 42 kDa) whereas the lower bands were the MBP-free ULD (about 11 kDa).
69
Figure 15. Repeated MBP-capture (cleaning) from the LC3 T12D “pool” (see Fig. 13)
with the affitnity chromatography
Ft: the 125-ml LC3 T12D “pool” freely flowing through the amylose resin (the same bed volume as the first time); E: elution of the MBP-free LC3 T12D by 10 ml column buffer; Wash: 20 ml ddH2O washing off any bound proteins (e.g., free MBP or MBP-LC3 T12D); Ft and E were then selected and pooled again (about 135 ml). Note the unwanted proteins (e.g., MBP and the fusion protein) did not show significant bands since the previous fraction containing more impurities (i.e., “E” in Fig. 13, 35 ml) was diluted in the earlier pooling before repeating the chromatography.
70
Figure 16.Typical LC3 elution monitor in the FPLC (the wild type)
Injected about 100 μl, flow rate 0.75 ml/min. The arrow indicates the effluent (the 43rd-49th fractions, each 300 μl) with high UV 280 nm absorbance from the LC3 wild type (about 14 kDa). Note: x-axis indicates how much buffer had been eluted at different fractions; y-axis shows the absorbance intensity at λ280 of each fraction.
71
Figure 17.Typical LC3 elution monitor in the FPLC (the T6D)
This was the 5th inject for purifying the sample from 2-L cell culture (the final volume concentrated to approximately 1000 μl); injected about 100 μl, flow rate 0.75 ml/min.
The arrow indicates the effluent (the 41st-45th fractions, each 300 μl) with high UV 280 nm absorbance from the LC3 T6D (about 14 kDa).
72
Figure 18.Typical LC3 elution monitor in the FPLC (the T12D)
This was the first inject for purifying the final 1050 μl-sample (from 2-L cell culture);
injected about 150 μl, flow rate around 0.6 ml/min. The arrow indicates the effluent (the 41st-45th fractions, each 300 μl) with high UV 280 nm absorbance from the LC3 T12D (about 14 kDa).
73
Figure 19.The ULD elution monitor in the FPLC
This was the 4th inject for purifying the final 700 μl-sample (from 4-L cell culture);
injected about 100 μl, flow rate 0.64 ml/min. The arrow indicates the effluent (the 41st-45th fractions, each 300 μl) with high UV 280 nm absorbance from the ULD (about 11 kDa). Note that the injection syringe carrying 700 μl-sample was kept at the injection hole of the machine at 4°C, but the monitoring chromatograms changed significantly after several injections (see Fig. 20).
74
Figure 20.Changing chromatogram monitoring the ULD elution with time in the FPLC This was the 7th inject for purifying the final 700 μl-sample (from 4-L cell culture);
injected about 100 μl, flow rate 0.51-0.58 ml/min. The right arrow indicates the decreased UV absorbance around the 41st-45th fractions whereas the absorbance from the fractions containing solutes with greater mass (the left arrow) became relatively stronger, which might result from ULD aggregation when awaiting incoming injections at 4°C.
75
Figure 21.MS confirmation of the gel-filtrated LC3 wild type (average MW 14129.29 Da)
Figure 22. MS confirmation of the gel-filtrated LC3 T6D (average MW 14143.27 Da)
76
Figure 23. MS confirmation of the gel-filtrated LC3 T12D (average MW 14143.27 Da)
Figure 24. MS confirmation of the gel-filtrated ULD (average MW 10839.53 Da)
77
Figure 25.How to determine the protein concentration for the CD using the spectrophotometry
Measure the absorbances of the protein sample at 280, 300, and 400 nm. Sketch a straight line determined by (300, Abs300) and (400, Abs400) and obtain its linear equation y = ax + b, where x is the wavelength (nm) and when x = 280, y is considered the background value at 280 nm; subtract this value from the Abs280 readout and the difference is then put into the Beer-Lambert law with the estimated molar extinction coefficient (e.g., for LC3 wild type, ε = 5960 M-1cm-1) to calculate the protein’s molar concentration. Measure the absorbances of each sample in triplicate, and take the average.
78
Figure 26.CD spectra of the LC3s
All signals were collected at room temperature in triplicate, averaged, and normalized to 50 μM, pathlength 1 mm.
79
Figure 27.CD spectra of the LC3s in the 8 M urea
All signals were collected at room temperature in triplicate, averaged, and normalized to 50 μM, pathlength 1 mm.
Figure 28.CD spectra of the LC3s in the 6 M GuHCl
All signals were collected at room temperature in triplicate, averaged, and normalized to 50 μM, pathlength 1 mm.
80
Figure 29. CD spectra of the LC3 T12D under different pH conditions
All signals were collected at room temperature in triplicate, averaged, and normalized to 50 μM, pathlength 1 mm.
81
Figure 30.How to prepare the 50 μM LC3 under the different urea concentrations
82
Figure 31.Urea denaturation curves of the LC3 variants based on the CD signal at 222 nm. All signals were collected at room temperature, and normalized to 50 μM,
pathlength 1 mm.
Figure 32.GuHCl denaturation curves of the LC3 variants based on the CD signal at 222 nm. All signals were collected at room temperature, and normalized to 50 μM, pathlength 1 mm.
83
Figure 33.Examination of pH dependence of the LC3 T12D by urea denaturation curves. All signals were collected at room temperature, and normalized to 50 μM, pathlength 1 mm.
84
a
b
Figure 34. PKA activity assay using the standard substrate Kemptide
(a) The control without PKA addition showed only the mass peak 772.5 corresponding to the Kemptide without any phosphorylation. (b) The Kemptide incubated with PKA in the presence of ATP and Mg2+ showed an additional mass peak 852.4.
85
a
b
Figure 35. PKA activity assay on the human LC3 wild type
(a) The control without PKA addition shows only mass peak 14129 Da corresponding to the LC3 without any phosphorylation. (b) The LC3 wild type incubated with PKA in the presence of ATP and Mg2+ did not show the expected mass peak 14209 Da
corresponding to the LC3 with phosphorylation.
86
Figure 36. Hypothesis of LC3 denaturation based on the “open-closed” idea
When challenged with increasing denaturant (e.g., urea), the loose bound NHD would first dissociate from the ULD, which is thought relatively compact; simultaneously the NHD denatures under thorough exposure to the surrounding urea; then the ULD begins massive denaturation or unfolding when the NHD has (almost) assumed
“random-coiled.”
87