Figure 1.
Characterization of cell lines expressing BACE1.
HEK (left column) and HeLa (right column) cell lines were generated that stably express either BACE1 or an empty vector as control. To monitor changes in APP processing cells were treated with the metalloprotease (α-secretase) inhibitor GM6001, the β-secretase inhibitor C3 and the γ-secretase inhibitor DAPT. (A–B) Cell lysates were probed for the presence of myc-tagged BACE1 (top panel); endogenous full-length APP, which is present in a mature, fully glycosylated form (mAPP) and an incompletely glycosylated immature form (iAPP, middle panel); and membrane-bound APP CTFs (bottom panel) produced by ectodomain shedding. (C–D) Abundance of APPs in conditioned medium of the various treatment conditions. (E) Quantification of secreted APPs levels arising from the different treatments. Data from both cell lines were combined, and normalized to the control (DMSO) condition. The fold-accumulation of APPs arising from the various treatment conditions vs. DMSO control is graphed; * p < 0.05. “+” indicates the addition of a drug while “−” indicates the addition of DMSO as a control. The molecular weight in kDa is shown to the left of each Western blot panel.
Figure 2.
Regulation of single-pass and GPI-linked proteins by BACE1.
Putative substrates identified to undergo BACE1 shedding in HEK and HeLa cells were examined for topology and proposed function. (A) Venn diagram indicating the total number of putative BACE1 substrates identified, and the number of these substrates that overlap or were unique to each cell type. (B) Membrane topology of the putative BACE1 substrates. (C) Putative BACE1 substrates were divided into several functional categories based on known protein functions and gene ontology classifications. (D) Sequence of APP, with peptides identified to be elevated by BACE1 expression indicated in red. The APP transmembrane sequence is highlighted in yellow, and arrowheads indicate the β- (major and minor sites), α-, γ-40 and γ-42 secretase cleavage sites (from left to right). APP-770 amino acid numbering is indicated on the left.
Table 1.
Putative β-secretase substrates identified by quantitative proteomics.
Figure 3.
BACE1 shedding of GPI-linked and type II transmembrane proteins.
Identified BACE1 substrates ephrin-A5 and GOLIM4 were cloned, FLAG-tagged, and stably expressed in HEK cells that express BACE1 or empty vector as control. The left column shows Western blots of cell lysates, and the right column shows blots of conditioned medium. Cells were treated with the β-secretase inhibitor C3 to confirm the necessity of BACE1 activity for ectodomain shedding. (A) Ephrin-A5, a GPI-linked protein, was robustly expressed and produced two prominent bands, the lower presumably representing the processed and mature GPI-linked form. BACE1 activity decreased the levels of full-length protein, and the shed product was visible within the cellular lysate (left panel). Conditioned medium revealed one minor (ephrin-A5sα) and one major (ephrin-A5sβ) band indicative of shed ephrin-A5, the major band corresponding to the BACE1 cleavage product (right panel). (B) GOLIM4, a type II transmembrane protein, was poorly expressed in cellular lysates (left panel), but accumulation of the shed ectodomain was found in conditioned medium of BACE1 expressing cells (right panel).
Figure 4.
BACE1 shedding of type I transmembrane proteins.
Identified BACE1 substrates of type I topology were either cloned and stably expressed or the endogenous protein was analyzed in HEK cells by Western blot. Cell lysates are shown in the left column, and conditioned medium shown in the right column. (A) LRIG2 was expressed in cell lysates as several distinct bands, likely owing to differential glycosylation (left panel). LRIG2 shedding by BACE1 was observed in the conditioned medium (right panel). (B) LRIG3 was stably expressed, as shown in cell lysates (left panel). LRIG3 was shed by BACE1 into the conditioned medium (right panel). (C) Endogenous IGF2R was analyzed with an ectodomain directed antibody. β-secretase activity produced a prominent decline in the full-length protein (left panel), and an increase in the shed ectodomain in the conditioned medium (right panel). (D) Endogenous APLP1 was expressed at undetectable levels in the cell lysate (left panel; the asterisk denotes a background band) but accumulated in the conditioned medium due to BACE1-mediated ectodomain shedding (right panel).
Figure 5.
Semaphorin 4C is processed by BACE1 and γ-secretase.
Semaphorin 4C was cloned with an N-terminal FLAG tag and a C-terminal HA tag, and stably expressed in HEK cells overexpressing BACE1 or empty vector as control. (A) Cell lysates show a single prominent band for mature Sema4C (left panel), which is shed by β-secretase activity into the medium (right panel). (B) Cells were treated with the γ-secretase inhibitor DAPT and cell lysates were probed for the presence of CTFs using the C-terminal HA epitope tag.
Figure 6.
Single pass membrane proteins unaltered by β-secretase activity.
Endogenous membrane proteins nicastrin (Nct, A) and Itgβ1 (B) were found not to be shed by BACE1 or another secretase. ACE (C) and BACE1 (D) were both shed by metalloproteases, but not affected by β-secretase activity. Left panels are from cell lysates, and right panels are from conditioned medium.
Figure 7.
Alignment of β-secretase substrates and putative cleavage sites.
The primary amino acid sequence of BACE1 substrates (A) and non-substrates (B) are shown. The first ten amino acids of the transmembrane domain are included, if present. Known cleavage sites are indicated with a black box and arrowhead, and potential cleavage sites are in gray. All sequences are human, except ST6Gal I, whose cleavage site was determined in rat. Sequence from type II proteins is listed from C- to N-terminal to maintain membrane orientation.