Dictyostelium possesses highly diverged presenilin/gamma-secretase that regulates growth and cell-fate specification and can accurately process human APP: a system for functional studies of the presenilin/gamma-secretase complex

Abstract

Presenilin (PS) is the catalytic moiety of the gamma-secretase complex. PS and other gamma-secretase components are well conserved among metazoa, but their presence and function in more-distant species are not resolved. Because inappropriate gamma-secretase processing of amyloid precursor protein (APP) in humans is associated with familial Alzheimer's disease, understanding essential elements within each gamma-secretase component is crucial to functional studies. Diverged proteins have been identified in primitive plants but experiments have failed to demonstrate gamma-secretase activity. We have identified highly diverged orthologs for each gamma-secretase component in the ancient eukaryote Dictyostelium, which lacks equivalents of APP, Notch and other characterized PS/gamma-secretase substrates. We show that wild-type (WT) Dictyostelium is capable of amyloidogenic processing of ectopically expressed human APP to generate amyloid-beta peptides Abeta(40) and Abeta(42); strains deficient in gamma-secretase cannot produce Abeta peptides but accumulate processed intermediates of APP that co-migrate with the C-terminal fragments alpha- and beta-CTF of APP that are found in mammalian cells. We further demonstrate that Dictyostelium requires PS for phagocytosis and cell-fate specification in a cell-autonomous manner, and show that regulation of phagocytosis requires an active gamma-secretase, a pathway suggested, but not proven, to occur in mammalian and Drosophila cells. Our results indicate that PS signaling is an ancient process that arose prior to metazoan radiation, perhaps independently of Notch. Dictyostelium might serve to identify novel PS/gamma-secretase signaling targets and provide a unique system for high-throughput screening of small-molecule libraries to select new therapeutic targets for diseases associated with this pathway.

Fig. 1.

Dictyostelium PS proteins align with human PS proteins. (A) Phylogenic comparison of amino acid sequences of Dictyostelium proteins and the human PS and SPP proteins. Sequence alignments excluded the highly variable cytosolic loops of PSs. The human SPPs and PSs are separated on different branches of the tree. The Dictyostelium PS proteins cluster closely with human PSs. The more distantly related Dictyostelium proteins are more closely related to human SPPs; we term them SPP1 and SPP2. (B) Residue alignment surrounding the N-terminal enzymatic aspartate (asterisk); similar amino acid residues are indicated in red. (C) Residue alignment surrounding the C-terminal enzymatic aspartate (asterisk); similar amino acid residues are indicated in red. (D) Amino acid length of the nonconserved cytoplasmic loop. Length is determined between conserved residues PAXIYXS (in B) and VKLGLGD (in C). (E) Residue alignment within the conserved, overlined PALP domain region; similar amino acid residues are indicated in red.

Fig. 2.

Developmental expression of ps1, ps2, nct, aph1 and pen2. RNA was isolated from WT Dictyostelium during growth or at 5-hour increments during development. Gene expression patterns were analyzed by either northern blotting (PS1, Nct, Aph1, Ig7) or semi-quantitative RT-PCR (PS2, Pen2, Ig7). Ig7 is a nonregulated mRNA control.

Fig. 3.

Dictyostelium has PS-dependent γ-secretase activity that processes human APP to release Aβ peptides. (A) Diagrams of human APP α, β and γ proteolytic cleavage sites, and processed fragments; shown are FL (full-length), ΔN (N-terminal deletion), α- and β-CTFs, Aβ₄₀ peptide, and AICD regions. (B) Media conditioned by growing WT and ps1-null cells that express ΔN-APP (a truncated human APP) were analyzed for levels of Aβ₄₀ and Aβ₄₂ peptides by quantitative ELISA. Fresh media and media conditioned by native WT cells were used as negative controls. Bars indicate standard errors that are derived from two independent experiments, each with two replicates. (C)Protein samples were collected from growing native WT Dictyostelium, or WT and ps1-null Dictyostelium that express ΔN-APP. ΔN-APP expression and processed fragments (arrows) were determined by immunoblot assay using α-APP C-terminus. (D) Protein samples were collected from native and APP-expressing CHO cells untreated or incubated with DAPT, and from growing native WT Dictyostelium or WT, aph1-null, ps1;ps2-null and ps1-null Dictyostelium that express ΔN-APP. APP expression and processing was determined by immunoblot assay using α-APP C-terminus. (E) Protein samples were collected from growing or 16-hour-developed WT, aph1-null, nct-null, ps2-null, and aph1;ps2-null Dictyostelium that express ΔN-APP. ΔN-APP expression and processing was determined by immunoblot assay using α-APP C-terminus. (F) Total cell fractions (T), membrane fractions (M) and cytosolic fractions (C) were prepared from growing WT, aph1-null and nct-null Dictyostelium that express ΔN-APP. ΔN-APP and α- and β-CTFs in WT and γ-secretase-null mutants were detected by immunoblotting with α-APP C-terminus. Positions of molecular weight markers, APP, ΔN-APP and α- and β-CTFs are indicated.

Fig. 4.

DAPT inhibits Dictyostelium processing of APP. Growing native and ΔN-APP-expressing WT Dictyostelium were untreated or incubated with DMSO ±50 μM DAPT for 18 hours. Protein samples were collected and analyzed by immunoblot using α-APP C-terminus. Positions of molecular weight markers, ΔN-APP and α- and β-CTFs are indicated.

Fig. 5.

PS/γ-secretase mutants exhibit defects in growth on bacteria. (A) Single cells of WT and mutant strains of Dictyostelium were grown in the presence of bacteria for identical times. Growth zones were illuminated from below and photographed. Bars: 53cm. (B) Single cells of WT and mutant strains of Dictyostelium were grown in the presence of bacteria for identical times. Growth zones were illuminated from above and photographed. Bar: 5 cm. (C) Protein samples were collected from growing WT controls (–) or ps2-null cells expressing either PS2^WT or PS2^DD/AA carrying His-epitope tags. PS2 expression was determined by immunoblot assay using α-His. Similar mobility differences between PS^WT and PS^DD/AA variants have been observed previously. (D) Single cells of ps2-null cells expressing either PS2^WT or PS2^DD/AA were grown in the presence of bacteria for identical times. Growth zones were illuminated from above and photographed. Bars: 5 cm.

Fig. 6.

PS/γ-secretase mutants exhibit defects in growth on bacteria. WT and mutant strains of Dictyostelium were mixed with TRITC-labeled, heat-killed yeast particles; samples were removed at the times indicated and monitored by fluorimetric analyses. Arbitrary fluorescence units were used to normalize each strain relative to the maximum obtained for WT within the same experiment. Each of the cell lines was examined at least three times. ps1-, ps1;ps2- and aph1:nct-null differences to WT were statistically significant (P<0.05); differences of PS2^DD/AA-expressing cells to PS2^WT-expressing cells were statistically significant (P<0.05).

Fig. 7.

PS/γ-secretase mutants exhibit developmental delays and morphological defects. WT and γ-secretase mutant strains were developed for 24 hours on filter matrices. Mutant strains fail to form terminal developmental structures.

Fig. 8.

PS/γ-secretase mutants have cell-autonomous defects in prespore-spore cell differentiation. (A) WT and γ-secretase-mutant strains were starved in monolayer culture in the presence of 8Br-cAMP to induce sporulation. The percentage of differentiated spore cells was calculated from total cells after three days. Bars indicate standard errors that are derived from three independent experiments, each with three replicates. (B) Real-time quantitative RT-PCR was used to determine relative gene expression levels of the prespore genes cotB (black) and pspA (gray) at 15 hours of development in γ-secretase-mutant strains compared with WT. Expression levels were normalized to H3a. Bars indicate standard errors that are derived from two independent experiments, each with two replicates. (C) PS/γ-secretase signaling regulates prespore differentiation in a cell-autonomous manner. WT and ps2-null cells that express lacZ were mixed with a ninefold excess of unmarked WT, ps2-null or ps1;ps2-null cells as indicated. Cell mixes were developed to the slug stage and stained by β-galactosidase activity. Slugs are positioned with their indicated anterior prestalk (pst) zones to the left and posterior prespore (psp) zones to the right.