Cell, Vol. 122, 435–447, August 12, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.cell.2005.05.022
Nicastrin Functions as a
Sanjiv Shah,1Sheu-Fen Lee,1,6Katsuhiko Tabuchi,1,6
Yi-Heng Hao,1,6Cong Yu,1Quincey LaPlant,1
Haydn Ball,2Charles E. Dann III,2
Thomas Südhof,1,4,5and Gang Yu1,3,*
1Center for Basic Neuroscience
2Department of Biochemistry
3Department of Cell Biology
4Department of Molecular Genetics
5Howard Hughes Medical Institute
The University of Texas Southwestern Medical Center
Dallas, Texas 75390
(Lemberg and Martoglio, 2002; Urban and Freeman,
2003; Ye et al., 2000). This, however, does not appear
to be the case for γ-secretase substrates (Lichtenthaler
et al., 1999). Indeed, γ-secretase appears to have a very
broad substrate specificity (reviewed in Kopan and Ila-
gan, 2004). The only known prerequisites for γ-secre-
tase substrates are (1) type I membrane proteins and
(2) shedding the bulk of the extracellular domains from
the full-length precursor proteins (Struhl and Adachi,
2000). Thus, one critical unanswered question concern-
ing γ-secretase-mediated RIP is how γ-secretase speci-
fically recognizes its many substrates that share no
overt sequence similarity.
It is now clear that γ-secretase activity resides in a
multimeric membrane-protein complex with presenilin
(PS1 or PS2), nicastrin (Nct), APH-1 (APH-1aL, APH-1aS,
or APH-1b), and PEN-2 as four essential components.
Presenilin exists in the active γ-secretase complex as a
heterodimer of the N- and C-terminal fragments (NTF
and CTF) resulting from endoproteolysis. Current evi-
dence is compatible with presenilin being the catalytic
subunit of γ-secretase (Wolfe et al., 1999). The precise
biochemical functions of the other γ-secretase sub-
units, particularly nicastrin, remain unclear. Consistent
with their largely membrane-embedded topologies,
APH-1 and PEN-2 may play structural roles in the as-
sembly and maturation of the γ-secretase complex (Lee
et al., 2004). Unlike the multipass membrane proteins
APH-1, PEN-2, and presenilin, nicastrin is a type I mem-
brane protein with a large ectodomain that accounts for
approximately 45% of the calculated protein molecular
mass of the four γ-secretase components combined.
Extensive glycosylation further increases the size of the
nicastrin ectodomain in relation to the other proteins.
Presenilin-dependent nicastrin hyperglycosylation is
important for γ-secretase maturation and trafficking to
the cell surface but is not required for γ-secretase activ-
ity (Herreman et al., 2003). Examination of residues
261–502 in the nicastrin ectodomain reveals sequence
similarity to a peptidase family that includes aminopep-
tidases, carboxypeptidases, and transferrin receptor
proteins (pfam04839; Marchler-Bauer et al., 2005; Fa-
gan et al., 2001), although no peptidase activity has
been detected to date for nicastrin. This region also
spans the most conserved sequence amongst nicastrin
orthologs (near residues 306–360)—the DYIGS motif.
Mutation of the DYIGS motif in previous studies has
shown significant effects on γ-secretase maturation
and/or activity (Chen et al., 2001; Shirotani et al., 2004;
Yu et al., 2000), but the mechanistic role of the DYIGS
motif in γ-secretase-mediated RIP is not clear. For ease
of presentation, we refer to the DYIGS and peptidase
homologous region as the DAP domain.
In the current study, we address two interrelated
questions central to the molecular mechanism of γ-secre-
tase-mediated RIP. First, what is the biochemical func-
tion of nicastrin in the γ-secretase complex? Second,
how is the recognition of a broad range of substrates
determined in the γ-secretase complex? We show that
?-secretase catalyzes the intramembrane cleavage of
amyloid precursor protein (APP) and Notch after their
extracellular domains are shed by site-specific prote-
olysis. Nicastrin is an essential glycoprotein compo-
nent of the ?-secretase complex but has no known
function. We now show that the ectodomain of ni-
castrin binds the new amino terminus that is gener-
ated upon proteolysis of the extracellular APP and
Notch domains, thereby recruiting the APP and Notch
substrates into the ?-secretase complex. Chemical-
or antibody-mediated blocking of the free amino ter-
minus, addition of purified nicastrin ectodomain, or
mutations in the ectodomain markedly reduce the
binding and cleavage of substrate by ?-secretase.
These results indicate that nicastrin is a receptor for
the amino-terminal stubs that are generated by ecto-
domain shedding of type I transmembrane proteins.
Our data are consistent with a model where nicastrin
presents these substrates to ?-secretase and thereby
facilitates their cleavage via intramembrane prote-
Regulated intramembrane proteolysis (RIP) has been
uncovered in diverse biological processes such as cho-
lesterol metabolism, immune surveillance, intercellular
communication, and Alzheimer’s disease. Several classes
of intramembrane proteases have been identified, in-
cluding the site 2 protease (S2P) family of metallopro-
teases, the Rhomboid family of serine proteases, and
the γ-secretase and signal peptide peptidase (SPP)
family of aspartyl proteases. Known substrates of these
unusual enzymes include the sterol regulatory element
binding protein (SREBP) for S2P (reviewed in Brown et
al., 2000), TGFα-like growth factor Spitz for Rhomboid
(Urban et al., 2001), signal peptides for SPP (Weihofen
et al., 2002), and, for γ-secretase, the amyloid precursor
protein APP (De Strooper et al., 1998) and Notch (De
Strooper et al., 1999). In many recognized substrates
for S2P, Rhomboid, and SPP, helix-breaking residues
6These authors contributed equally to this work.
the APP- and Notch-derived γ-secretase substrates
stoichiometrically, directly, and functionally interact
with nicastrin. We demonstrate that this interaction is
mediated by the nicastrin ectodomain and the extracel-
lular N-terminal stub of the substrate. Using both in vi-
tro and in vivo assays, we find that the extracellular
DAP domain of nicastrin is essential for γ-secretase-
substrate recognition but not catalysis. These results
thus define a biochemical function for nicastrin as a
receptor for γ-secretase substrates.
The Nicastrin Ectodomain Interacts
with ?-Secretase Substrates
Previous studies have shown that γ-secretase sub-
strates could be coimmunoprecipitated with both pre-
senilin and nicastrin in total-cell extracts (Xia et al.,
2000; Yu et al., 2000). To study whether nicastrin binds
γ-secretase substrates directly or indirectly via preseni-
lin, we examined whether nicastrin copurifies with the
C-terminal APP fragment C99, a cleavage product of
β-secretase, from baculovirus-infected Sf9 cells (Figure
1A). For ease of detection, nicastrin was fused with a
His6 tag at the C terminus (Nct-His) and C99 was
C-terminally Flag tagged (C99-Flag). Triton X-100 ex-
tracts of membranes from Sf9 cells infected with Nct-
His, Nct-His plus C99-Flag, or C99-Flag baculoviruses
were subjected to immunoprecipitation (IP) with an
anti-Flag antibody. His-tagged nicastrin from Sf9 cells
appears to be present as an w120 kDa full-length pro-
tein, as well as an w10 kDa C-terminal membrane
bound derivative (ctNct-His, Figure 1 and Figure S1A in
the Supplemental Data available with this article on-
line). Analysis of the Flag-peptide-eluted products by
Western blotting or on Coomassie-stained SDS-poly-
acrylamide gels showed that only full-length Nct-His
but not ctNct-His was coimmunoprecipitated with C99-
Flag in high-salt Triton X-100 extracts (Figure 1B). Al-
though the nature of the membrane bound ctNct is not
clear, this result suggests that the cytoplasmic region
and transmembrane region (TMR) of nicastrin do not
bind to C99. Flag-peptide eluates from anti-Flag beads
were also subjected to affinity pull-down using immobi-
lized nickel-nitrilotriacetic acid (Ni-NTA). Retention of
C99-Flag on Ni-NTA beads was dependent on the pres-
ence of Nct-His (Figure 1B). Addition of Flag peptide
or EDTA prevented purification of the Nct-His:C99-Flag
complex on anti-Flag or Ni-NTA beads (Figure S1A).
Densitometry analyses of the Coomassie-blue-stained
bands corresponding to Nct-His and C99-Flag showed
an approximately equal molar ratio (Figure S1B). Under
conditions that preserve the association of nicastrin
and C99, neither nicastrin nor C99 binds to control
membrane proteins (Figure S1C). Taken together, these
observations indicate that the association of nicastrin
with C99 is stoichiometric, specific, and direct.
In light of the absence of detectable association of
C99 and the membrane bound C-terminal derivate of
nicastrin, we decided to examine whether γ-secretase
substrate directly binds to the large ectodomain of ni-
castrin using highly purified proteins (Figures 2A and
2B).A mixtureof purified
Figure 1. Nicastrin Stoichiometrically Copurifies with APP-Derived
(A) Schematic representation of anti-Flag immunoprecipitation (IP)
and Ni-NTA pull-down experiments. Dashed lines represent inter-
(B) Triton X-100 extracts of Sf9 cell membranes expressing Nct-
His, Nct-His plus C99-Flag, or C99-Flag were incubated with anti-
Flag beads. A portion of the input (lanes 1–3), flowthrough (lanes
4–6), and Flag-peptide-eluted products (lanes 7–9) were analyzed
on SDS-polyacrylamide gels. The remaining Flag-peptide eluates
were subjected to Ni-NTA pull-down (lanes 10–12). The SDS-poly-
acrylamide gels were either probed with anti-His or anti-Flag (top
panels) or stained with Coomassie blue (bottom panel). Mass-
spectrometry analysis identified the 25 kDa band (*) as the mi-
tochondrion protein ADP/ATP translocase, which likely represents
a nonspecific contaminant as it inconsistently coelutes with C99-
Flag (also see Figure 2B).
2002, 2004; Yu et al., 1998, 2000). All experiments in this paper
were performed at least four times with multiple replications. The
GraphPad Prism software was used for statistical analysis and
graphing. More details of reagents and methods are described in
the Supplemental Experimental Procedures.
tion of an aminopeptidase was united with the catalytic
function of an intramembrane aspartyl protease. This
enables the γ-secretase complex to execute the un-
usual events associated with RIP of type I membrane
proteins in the lipid bilayers.
The effects of mutations of the DAP domain on the
binding of nicastrin to APP-derived substrates gen-
erally correlate with the inhibitory effects of these muta-
tions on intramembrane proteolysis. The observation
that functional effects are greater than the structural
effects agrees with earlier studies that found that sub-
strate binding was not grossly disrupted but might be
subtly altered by specific point mutations of conserved
glutamates in aminopeptidases (Thompson et al., 2003;
Vazeux et al., 1998). The high sensitivity of the APP-GV
and C99-GV transactivation assays also contributes to
the ease of detecting smaller functional differences for
γ-secretase catalysis. On the other hand, the observa-
tion that, under the stringent conditions used in this
study, the E333A mutation partially reduces the ni-
castrin:substrate interaction while the ?312–340 or
?312–369 mutation abolished the interaction (Figures
3B and 4E) implies that other residues of the DAP do-
main may cooperate with Glu333 in substrate binding.
Under gentle conditions (e.g., low-salt digitonin buff-
ers), association of C99/C83 and Nct ?312–340 has
also been observed (Yu et al., 2000). This observation
may reflect the role of the TMRs of the γ-secretase sub-
units in docking of substrates before their catalysis (see
Figure 7D for model), which could be captured under
the conditions that are compatible with γ-secretase as-
sembly and activity. Although it is conceivable that se-
quences outside residues 312–340 of the DAP domain
may contribute to substrate binding, our data clearly
establish that a major substrate-recognition site con-
sists of Glu333 and its nearby residues.
Our studies indicate that nicastrin is a receptor with
broad specificity in recognizing short peptide-like ex-
tracellular domains of type I membrane proteins. How-
ever, the binding properties of nicastrin to different sub-
strates could be differentially regulated by the number
and composition of amino acids of the extracellular
portion of the substrates. This view is supported by the
finding that alteration of the N terminus of C99 subtly
affects its binding to nicastrin (Figure 6C) and its cleav-
age by γ-secretase (data not shown) and that Notch
and APP processing is differentially affected by muta-
tions of the DYIGS motif of nicastrin (Chen et al., 2001).
Designing or screening for compounds or antibody de-
rivatives that specifically or preferentially block the
binding of the N terminus of β-secretase-cleaved APP
(i.e., C99) to the nicastrin ectodomain may prove to be a
tractable strategy as a therapy for Alzheimer’s disease.
Protein Purification and Modification
Purification of recombinant γ-secretase and its substrates (N100-
Flag/His, C99-Flag, and nFlag-C99) from insect cell membranes
and the nicastrin ectodomain from conditioned media of mamma-
lian cells is described in the Supplemental Experimental Pro-
cedures. N-formylated C99-Flag (fC99) was purified on anti-Flag
beads from detergent extracts of ?acrAB E. coli strain AG100A(DE3)
(a gift from Dr. R.T. Sauer) transformed with pET21b/C99-Flag and
induced by 1 mM IPTG plus 2 ?g/ml actinonin for 5 hr at room
temperature. Cys-C99 was expressed in E. coli as a fusion protein
with an N-terminal His6tag and a recognition site for tobacco etch
virus (TEV) protease as well as a C-terminal Flag tag (Figure S7B).
The fusion protein purified on Ni-NTA beads was cleaved with His-
tagged TEV protease. Purified Cys-C99 was obtained after Ni-NTA
depletion of the undigested fusion protein and TEV protease.
Chemical synthesis of fluorescein- and biotin-LC-Gly-Gly-thioester
and their ligation to the α-amino group of Cys-C99 are described
in detail in the Supplemental Experimental Procedures.
N100-Flag/His, nFlag-C99, C99-Flag, Cys-C99, and their N-ter-
minally modified variants were incubated at 37°C with recombinant
γ-secretase purified from Sf9 cells or with cell-free γ-secretase from
HeLa cells in a reaction buffer containing 0.25% CHAPSO, 50 mM
PIPES (pH 7.0), 5 mM MgCl2, 5 mM CaCl2, 0.0125% phosphatidyl-
ethanolamine, and 0.1% phosphatidylcholine. A typical 50 ?l reac-
tion uses 25 ?g γ-secretase preparation and 0.1 ?g substrate.
AICD-Flag or NICD#-Flag/His was analyzed on Tris-Tricine SDS-
polyacrylamide gels using anti-APP-CTD, anti-Flag, or anti-cleaved
Notch-1 (Val1744) antibody. γ-secretase cleavage of APP or C99
was also analyzed in HEK293 or fibroblast cells using the luciferase
reporter gene assay system as described (Cao and Sudhof, 2001;
Lee et al., 2004). The fluorogenic reporter peptide mimicking the
γ-secretase-cleavage sites of APP was assayed as reported (Farm-
ery et al., 2003).
cedures, Supplemental References, and seven figures and can be
found with this article online at http://www.cell.com/cgi/content/
include SupplementalExperimental Pro-
We are indebted to Drs. M.S. Brown and J.L. Goldstein for discus-
sions and critical reading of the manuscript. C.E.D. is supported by
the Sara and Frank McKnight Fund for Biochemical Research. G.Y.
is the Thomas O. Hicks Scholar in Medical Research at the Univer-
sity of Texas Southwestern Medical Center. This work was sup-
ported by the National Institutes of Health grant R01 AG023104,
the Welch Foundation, the American Health Assistance Founda-
tion, the American Federation for Aging Research, and the Alzhei-
Received: September 20, 2004
Revised: February 14, 2005
Accepted: May 18, 2005
Published: August 11, 2005
Reagents and General Methods
Wild-type and mutant cDNAs for γ-secretase subunits and sub-
strates were cloned into baculoviral, bacterial, and mammalian ex-
pression vectors. nicastrin-deficient embryonic fibroblasts were
obtained from nicastrin knockout mice, which will be reported else-
where (K.T., G.Y., and T.S., unpublished data). Standard methods
such as site-directed mutagenesis, transfection, Western blotting,
and affinity precipitation were performed as described (Lee et al.,
Brown, M.S., Ye, J., Rawson, R.B., and Goldstein, J.L. (2000). Regu-
lated intramembrane proteolysis: a control mechanism conserved
from bacteria to humans. Cell 100, 391–398.
Cao, X., and Sudhof, T.C. (2001). A transcriptionally [correction of
Nicastrin Is a γ-Secretase-Substrate Receptor
transcriptively] active complex of APP with Fe65 and histone ace-
tyltransferase Tip60. Science 293, 115–120.
Chen, F., Yu, G., Arawaka, S., Nishimura, M., Kawarai, T., Yu, H.,
Tandon, A., Supala, A., Song, Y.Q., Rogaeva, E., et al. (2001). Ni-
castrin binds to membrane-tethered Notch. Nat. Cell Biol. 3, 751–
Dawson, P.E., and Kent, S.B. (2000). Synthesis of native proteins
by chemical ligation. Annu. Rev. Biochem. 69, 923–960.
Dawson, P.E., Muir, T.W., Clark-Lewis, I., and Kent, S.B. (1994). Syn-
thesis of proteins by native chemical ligation. Science 266, 776–
De Strooper, B., Annaert, W., Cupers, P., Saftig, P., Craessaerts, K.,
Mumm, J.S., Schroeter, E.H., Schrijvers, V., Wolfe, M.S., Ray, W.J.,
et al. (1999). A presenilin-1-dependent gamma-secretase-like pro-
tease mediates release of Notch intracellular domain. Nature 398,
De Strooper, B., Saftig, P., Craessaerts, K., Vanderstichele, H.,
Guhde, G., Annaert, W., Von Figura, K., and Van Leuven, F. (1998).
Deficiency of presenilin-1 inhibits the normal cleavage of amyloid
precursor protein. Nature 391, 387–390.
Fagan, R., Swindells, M., Overington, J., and Weir, M. (2001). Nic-
astrin, a presenilin-interacting protein, contains an aminopepti-
dase/transferrin receptor superfamily domain. Trends Biochem.
Sci. 26, 213–214.
Farmery, M.R., Tjernberg, L.O., Pursglove, S.E., Bergman, A., Win-
blad, B., and Naslund, J. (2003). Partial purification and character-
ization of gamma-secretase from post-mortem human brain. J.
Biol. Chem. 278, 24277–24284.
Herreman, A., Van Gassen, G., Bentahir, M., Nyabi, O., Craessaerts,
K., Mueller, U., Annaert, W., and De Strooper, B. (2003). gamma-
Secretase activity requires the presenilin-dependent trafficking of
nicastrin through the Golgi apparatus but not its complex glycosy-
lation. J. Cell Sci. 116, 1127–1136.
Kopan, R., and Ilagan, M.X. (2004). Gamma-secretase: proteasome
of the membrane? Nat. Rev. Mol. Cell Biol. 5, 499–504.
Kornilova, A.Y., Bihel, F., Das, C., and Wolfe, M.S. (2005). The initial
substrate-binding site of gamma-secretase is located on presenilin
near the active site. Proc. Natl. Acad. Sci. USA 102, 3230–3235.
Lee, S.F., Shah, S., Li, H., Yu, C., Han, W., and Yu, G. (2002). Mam-
malian APH-1 interacts with presenilin and nicastrin and is required
for intramembrane proteolysis of amyloid-beta precursor protein
and Notch. J. Biol. Chem. 277, 45013–45019.
Lee, S.F., Shah, S., Yu, C., Wigley, W.C., Li, H., Lim, M., Pedersen,
K., Han, W., Thomas, P., Lundkvist, J., et al. (2004). A conserved
GXXXG motif in APH-1 is critical for assembly and activity of the
gamma-secretase complex. J. Biol. Chem. 279, 4144–4152.
Lemberg, M.K., and Martoglio, B. (2002). Requirements for signal
peptide peptidase-catalyzed intramembrane proteolysis. Mol. Cell
Lichtenthaler, S.F., Wang, R., Grimm, H., Uljon, S.N., Masters, C.L.,
and Beyreuther, K. (1999). Mechanism of the cleavage specificity of
Alzheimer’s disease gamma-secretase identified by phenylalanine-
scanning mutagenesis of the transmembrane domain of the amy-
loid precursor protein. Proc. Natl. Acad. Sci. USA 96, 3053–3058.
Luciani, N., Marie-Claire, C., Ruffet, E., Beaumont, A., Roques, B.P.,
and Fournie-Zaluski, M.C. (1998). Characterization of Glu350 as a
critical residue involved in the N-terminal amine binding site of
aminopeptidase N (EC 18.104.22.168): insights into its mechanism of ac-
tion. Biochemistry 37, 686–692.
Marchler-Bauer, A., Anderson, J.B., Cherukuri, P.F., DeWeese-
Scott, C., Geer, L.Y., Gwadz, M., He, S., Hurwitz, D.I., Jackson, J.D.,
Ke, Z., et al. (2005). CDD: a Conserved Domain Database for pro-
tein classification. Nucleic Acids Res. 33 (Database issue), D192–
Shirotani, K., Edbauer, D., Kostka, M., Steiner, H., and Haass, C.
(2004). Immature nicastrin stabilizes APH-1 independent of PEN-2
and presenilin: identification of nicastrin mutants that selectively
interact with APH-1. J. Neurochem. 89, 1520–1527.
Spector, S., Flynn, J.M., Tidor, B., Baker, T.A., and Sauer, R.T.
(2003). Expression of N-formylated proteins in Escherichia coli.
Protein Expr. Purif. 32, 317–322.
Steiner, H., Kostka, M., Romig, H., Basset, G., Pesold, B., Hardy, J.,
Capell, A., Meyn, L., Grim, M.L., Baumeister, R., et al. (2000). Gly-
cine 384 is required for presenilin-1 function and is conserved in
bacterial polytopic aspartyl proteases. Nat. Cell Biol. 2, 848–851.
Struhl, G., and Adachi, A. (2000). Requirements for presenilin-
dependent cleavage of notch and other transmembrane proteins.
Mol. Cell 6, 625–636.
Thompson, M.W., Govindaswami, M., and Hersh, L.B. (2003). Muta-
tion of active site residues of the puromycin-sensitive aminopepti-
dase: conversion of the enzyme into a catalytically inactive binding
protein. Arch. Biochem. Biophys. 413, 236–242.
Urban, S., and Freeman, M. (2003). Substrate specificity of rhom-
boid intramembrane proteases is governed by helix-breaking resi-
dues in the substrate transmembrane domain. Mol. Cell 11, 1425–
Urban, S., Lee, J.R., and Freeman, M. (2001). Drosophila rhomboid-1
defines a family of putative intramembrane serine proteases. Cell
Vazeux, G., Iturrioz, X., Corvol, P., and Llorens-Cortes, C. (1998).
A glutamate residue contributes to the exopeptidase specificity in
aminopeptidase A. Biochem. J. 334, 407–413.
Weihofen, A., Binns, K., Lemberg, M.K., Ashman, K., and Martoglio,
B. (2002). Identification of signal peptide peptidase, a presenilin-
type aspartic protease. Science 296, 2215–2218.
Wolfe, M.S., Xia, W., Ostaszewski, B.L., Diehl, T.S., Kimberly, W.T.,
and Selkoe, D.J. (1999). Two transmembrane aspartates in preseni-
lin-1 required for presenilin endoproteolysis and gamma-secretase
activity. Nature 398, 513–517.
Xia, W., Ray, W.J., Ostaszewski, B.L., Rahmati, T., Kimberly, W.T.,
Wolfe, M.S., Zhang, J., Goate, A.M., and Selkoe, D.J. (2000). Pre-
senilin complexes with the C-terminal fragments of amyloid precur-
sor protein at the sites of amyloid beta-protein generation. Proc.
Natl. Acad. Sci. USA 97, 9299–9304.
Ye, J., Dave, U.P., Grishin, N.V., Goldstein, J.L., and Brown, M.S.
(2000). Asparagine-proline sequence within membrane-spanning
segment of SREBP triggers intramembrane cleavage by site-2 pro-
tease. Proc. Natl. Acad. Sci. USA 97, 5123–5128.
Yu, G., Chen, F., Levesque, G., Nishimura, M., Zhang, D.M., Le-
vesque, L., Rogaeva, E., Xu, D., Liang, Y., Duthie, M., et al. (1998).
The presenilin 1 protein is a component of a high molecular weight
intracellular complex that contains beta-catenin. J. Biol. Chem.
Yu, G., Nishimura, M., Arawaka, S., Levitan, D., Zhang, L., Tandon,
A., Song, Y.Q., Rogaeva, E., Chen, F., Kawarai, T., et al. (2000). Ni-
castrin modulates presenilin-mediated notch/glp-1 signal trans-
duction and betaAPP processing. Nature 407, 48–54.