Mohan, Swetha, Sampognaro, Paul J., Argouarch, Andrea R., Maynard, Jason C., Welch, Mackenzie, Patwardhan, Anand, Courtney, Emma C., Zhang, Jiasheng, Mason, Amanda, Li, Kathy H., Huang, Eric J., Seeley, William W., Miller, Bruce L., Burlingame, Alma, Jacobson, Mathew P., and Kao, Aimee W.
Additional file 1 Fig. S1. Enzymes that do not cleave PGRN are biologically active. a, 1 μM of mature CTSD was incubated with 400 ng recombinant human PGRN or 400 ng BSA at the indicated pH. The incubation was stopped after 60 min or 16 h, and reaction contents were subjected to Coomassie staining (60 min) or silver staining (16 h) respectively. Bands correlate to proteins indicated on the right. BSA was used as a control for CTSD activity. At lower pH of 3.4, CTSD self-digest leading to a decreased level of mature CTSD compared to pH 4.5 [52]. b, FRET-based enzyme activity assay at pH 4.5 was performed on all enzymes that do not cleave PGRN. The values obtained were normalized to timepoint t = 0 and plotted relative to activity. All enzymes show activity against the optimized FRET peptides compared to substrate only control. Fig. S2. Antibody validation for individual granulin peptides. a, Dot blots confirming the specificity of each custom antibody towards the granulin peptide. 2.5 μg of each peptide was used with each antibody (11,000 dilution). Paragranulin, p-Gran; granulin G, Gran-G; granulin F, Gran-F; granulin B, Gran-B; granulin A, Gran-A; granulin C, Gran-C; granulin D, Gran-D; granulin E, Gran-E. b, Peptide blocking experiment to demonstrate antibody specificity. 400 ng of rhPGRN was incubated with 1uM CTSL for 20 min. The antibody was incubated with or without the respective blocking peptide (10 or 100-fold molar ratio of peptide to antibody) and then incubated with the membrane overnight. Increase in peptide concentrations show a decrease in the bands of each antibody tested. Fig. S3. Lysosomal proteases unable to digest PGRN in vitro. Full western blot images of the proteases that do not cleave PGRN. 1 μM of each enzyme was incubated with 400 ng of PGRN for 20 min. C-terminal Gran-E antibody or anti-PGRN (Invitrogen) antibody was used to assess the results of the in vitro assay. a, Cysteine proteases Cathepsins C (CTSC), H (CTSH), X (CTSX), O (CTSO) and F (CTSF). b, Aspartyl proteases cathepsin D (CTSD). c, Serine protease Pro-X carboxypeptidase (PRCP) and Cathepsin A (CTSA). The lower molecular weight bands (indicated by *) correspond to the enzyme. Fig. S4. PGRN processing by lysosomal proteases in vitro at pH 6.5. 1 μM of each enzyme was incubated with 400 ng of PGRN for 20 min at pH 6.5. Anti-p-Gran, anti-Gran-F, and anti- Gran-E antibodies were used to assess the results of the in vitro assay. Cathepsin B (CTSV), cathepsin L (CTSL), cathepsin K (CTSK), cathepsin S (CTSS), cathepsin V (CTSV), asparagine endopeptidase (AEP), cathepsin G (CTSG), paragranulin (p-Gran), granulin F (Gran-F), granulin E (Gran-E). Fig. S5. Summary of PGRN processing into granulins by multiple proteases. Summary of the results of the in vitro assays. a, Processing of PGRN to granulins by different proteases is dependent on pH. The range of cleavage is represented in greyscale with no cleavage in white and complete cleavage in black. The range takes into account the amount of full-length PGRN processed into multi-granulin fragments and individual granulins within 20 min. b, Represented is the ability of each enzyme to liberate individual paragranulin (p) in red, granulin F (F) in green, and granulin E (E) in blue. The protease classes are also indicated. CTSE, cathepsin E; CTSV, cathepsin V; CTSL, cathepsin L; CTSB, cathepsin B; CTSK, cathepsin K; AEP, asparagine endopeptidase; CTSG, cathepsin G; CTSS, cathepsin S. Fig. S6. CTSL is highly efficient at liberating paragranulin and granulin E from PGRN. A time course analysis to determine efficiency of cleavage. 400 ng of recombinant human PGRN was incubated with 50 nM of enzyme for the time points indicated. P-Gran, Gran-F and Gran-E antibodies were used to assess the results of the assay. PGRN, progranulin; CTSL, cathepsin L; CTSB, cathepsin B. Fig. S7. Antibody specificity to detect both PGRN and granulin sized bands in wild-type and PGRN knock out iPSC cell lysates. a, Lysates from isogenic WTC11 and Pgrn KO iPSCs were probed with custom anti-granulin antibodies. b, Commercial antibodies from Invitrogen and Sigma were tested on the same iPSC cell lines. Fig. S8. Expression profile of the PGRN proteases in differentiated SH-SY5Y cells. a, qPCR analysis to assess the expression of candidate PGRN proteases in differentiated SH-SY5Y cells. The graph represents the mean expression value of each enzyme normalized to GAPDH. The absolute mean values are noted for each sample. All enzymes were run in triplicates with n = 4 biological replicates. b, Differentiated SH-SY5Y cells were subjected to treatment with AEP siRNA or scramble control. A separate quantification and statistical analysis of PGRN, Gran-F and mature AEP was performed, similar to that seen in Fig. 3 of the manuscript. c, PGRN protein is assessed with an Anti-Gran E antibody. PGRN KO SH-SY5Y cells are used to determine specificity of PGRN band. d, Quantification of c. Fig. S9. PGRN is processed by AEP to liberate individual granulins F and B. Recombinant human PGRN was incubated with and without AEP for 1 h at pH 4.5. The reaction was stopped, and the cleavage bands were separated by SDS-PAGE. a, Multiple cleavage bands and individual granulin sized bands can be visualized upon silver stain. b-e, Bands were cut and digested for mass spectrometry to identify the cleavage sites (also see Fig. 4c). Identified peptides are annotated with observed fragment ions. Previous and next amino acids are in () and carbamidomethylated Cys are colored red. The precursor mass of each peptide is as follows: b, 428.5229 m/z; c, 847.8607 m/z; d, 639.3205 m/z; e, 478.9133 m/z. Fig. S10. Progranulin and granulin F levels in the human brain. a, Western blot analysis of the levels of PGRN and Gran-F in the non-degenerating IOC regions of the control brain compared to the same region in FTLD-TDP-Pgrn subjects. b, quantification of PGRN and Gran-F level, normalized to actin. c, Western blot analysis of the levels of PGRN and Gran-F in the degenerating MFG regions of the control brain compared to the same region in FTLD-TDP-Pgrn subjects. d, quantification of PGRN and Gran-F level, normalized to actin. (**, p = 0.008). Unpaired, two-tailed student t-test was performed between the pairs. Error bars represent mean with standard deviation. e-g, Levels of PGRN and Gran-F in a and c are plotted against age. There is no correlation between age and level of PGRN and Gran-F. Fig. S11. AEP levels are significantly increased in a degenerating brain region of FTLD-TDP-Pgrn subjects. a, Western blot analysis of endogenous immature (pro AEP) and mature AEP levels in control brain. b, Quantification of pro and mature AEP levels between IOC and MFG regions in control subjects. Paired two-tailed student t-test was performed between the different groups. No significant (ns) difference was observed. c, Western blots of endogenous Pro and mature AEP in IOC and MFG brain regions of FTLD-TDP-Pgrn subjects. d, Levels of mature AEP in the degenerating MFG region are increased compared to the non-degenerating IOC region from the same FTLD-TDP-Pgrn samples (*, p = 0.01, n = 6). Paired two-tailed student t-test was performed between the groups. IOC, inferior occipital cortex; MFG, middle frontal gyrus; AEP, asparagine endopeptidase. Sample numbers listed below each Western blot correspond to subject number in Table 2.