Phorbol Esters Affect Multiple Steps in β-Amyloid Precursor Protein Trafficking and Amyloid β-Protein Production

Phorbol

Esters

Affect

Multiple

Steps

in

3,-Amyloid

Precursor

Protein

Trafficking

and

Amyloid

0-Protein

Production

Edward H. Koo

Department of

Pathology and Center

for

Neurologic Diseases,

Brigham and Women's

Hospital,

and

Department

of

Pathology,

Harvard Medical

School, Boston,

Massachusetts,

U.S.A.

ABSTRACT

Background: Amyloid

,3-protein

(A,B), the major con-stituent of amyloid deposits found in Alzheimer's dis-ease, is derived from the (3-amyloid precursor protein

(O3PP).

Constitutiveproteolysisby a-secretase and secre-tion of soluble ,BPP

(f3PPJ)

are stimulated by protein

kinase C (PKC) activation, whereas

AP3

productionand release are inhibited. The cellular mechanism that un-derlies the PKC-mediateddown-regulationof

AP3

gener-ation is unclear. Because endocytic processing of ,BPP from the cell surface is amajorpathway of

AP3

produc-tion, theeffect ofPKCactivationby phorbol 12,1 3-dibu-tyrate(PDBu)onendocytictraffickingof/3PPwas exam-ined.

Materials and Methods: In this study, trafficking of

O3PP

wasassayedinChinese hamsterovary cells (CHO) cellsstablytransfected withfull-length1PP751.

Results: Treatment with PDBu resultedin arapid and

strikingreduction of upto80% in the amount of

O3PP

at thecellsurface.This lossofcell-surfacemoleculescould not be accounted for by changes in the trafficking of cell-surface

O3PP

molecules, as determined by a radiola-beledantibody assay. Rather, the decrease in ,BPP was dueprimarilyto areduction in thesorting of

O3PP

to

the cellsurface. This alteration was correlated with

accelerated intracellular a-secretase-mediated ,BPP

cleavage andaccelerated,3PPtraffickinginthe

exo-cyticpathway.

Conclusions: The data suggest that the displacement of

I3PP

awayfrom thecellsurfaceafter

phorbol

ester

treat-mentreduces the substrateavailable for endocytic pro-cessingand, in turn, results in the inhibition of A3 pro-duction.

INTRODUCTION

Alzheimer's disease is characterized by the dep-osition ofamyloid(3-protein

(AP3)

inparenchyma and blood vessel walls and the intracellular ac-cumulation of neurofibrillary tangles. A,B is de-rived by proteolysis of the larger ,B-amyloid pre-cursor protein (,3PP), a 100-140 kDa integral membrane protein. Processing of,3PPinthe con-stitutive pathway results in the secretion of a large N-terminal soluble product

(1PPj)

and a membrane-retained C-terminal fragment of ap-proximately 10 kDa (1). Release of the N-termi-nal ectodomain of ,BPP results from cleavage by

Addresscorrespondenceand reprintrequests to:EdwardH.

Koo, Departmentof Neurosciences 0691, University of

Cal-ifornia, SanDiego, 9500Gilman Drive,LaJolla,CA

92093-0691, U.S.A.Phone: 619-822-1024;Fax:619-822-1021.

204

"a-secretase"

an as yet unidentified protease that cleaves ,BPP within the

AP3

sequence (2). Thus, formation of a full-length, 40-43 residue

AP3

peptide isprecludedby a-secretase proteoly-sis. Both intracellular and cell-surface ,3PP mol-ecules are substrates fora-secretase; indeed, ac-cumulating evidence suggests proteolysis of

O3PP

occurs in both cellularcompartments (3).

Generation and release ofA,B from ,BPP oc-curs constitutively after two proteolytic events, designated ",3- and -y-secretase" cleavages at the

AP3

N- andC-termini, respectively. Inadditionto

I3PPS,

Af3

and a 3-kDa

fragment

(p3)

are

nor-mally produced and released from a variety of cells both in vivo and in vitro (4). The precise cellularcompartments in which

A,(3

isgenerated are unclear, although processing in an acidic

FIG. 1. Schematic of ,BPP showing antibody

epitopes

Theblack box represents AP3

with the amino acid

se-quenceshownbelow. The

verticaldashed boxes and horizontalstippledbox

rep-resenttheplasma membrane andthe Kunitzprotease in-hibitor (KPI) domain,

respec-tively. "a"and "j3" indicate

a- and f-secretase cleavage

sites. Horizontalblack bars

representtheapproximate epitopes ofantibodiesB5, 5A3, IG7, CT15, 1736, and 1282.

compartmentappearstoberequired. Both

secre-tory (5) and endocytic (6) pathways have been

shownto contributeto

AP3

production; however, the major source of A,B derived from wild-type

OPP

in cultured cells appearsto bethe

endocytic

pathway (6,7).

Activation ofprotein kinase C (PKC) by a

variety of agents stimulates

a-secretase-medi-ated 1PPs secretion (8). This alteration,

how-ever, does not require direct phosphorylation

of

OPP,

which suggests an indirect mode of

action of PKC on regulating

OPP

cleavage

(9,10). Acceleration of ,BPP metabolism through the a-secretory pathway by PKC

acti-vation results in a corresponding decrease in

,B-secretory processing and A,B production (11,12). Thecellular andmolecular basis ofthe PKC-mediated reduction of

AP3

production

re-mains to be clearly defined. In view of our

results indicatinga major role ofthe endocytic

pathway in AP production, this study was

de-signed to determine how processing of

O3PP

in this pathway is affected by PKC activation. To

date, cell-surface receptors have shown

heter-ogeneousandunpredictableresponsesto

phor-bol esters with respect to trafficking in the

receptor-mediated pathway (13). The results showed that PKC activation by phorbol esters

leads to a rapid and major reduction in

O3PP

sorting to the cell surface. In turn, this loss in cell-surface J3PP results in a corresponding

re-duction in available substrate for endocytic processing and A,B production.

MATERIALS AND METHODS

Cell Cultureand Metabolic Labeling

Chinese hamster ovary (CHO) cells were grown

inDulbecco's modified Eagle's medium contain-ing 10% fetal calfserum. Stablytransfected CHO

cell line expressing wild-type ,3PP751 has been establishedpreviously (6). In pulse-chase

exper-iments, confluent ,BPP-transfected CHO cells

were labeled with 35S-methionine for 10 min

and chased for 20 min or 2 hr with or without

1 ,uMphorbol-12,13-dibutyrate (PDBu). This

la-beling paradigm examines only the effect of

PDBu on post-translational events.

Immunopre-cipitations of

AP3

(media) and full-length ,BPP (cell lysates) were carried outwith 1282, an

an-tibody recognizing

AP3,

and a C-terminal

anti-body, CT15, respectively (14,15) (Fig. 1).

O3PP,

frommediaandsaponin buffer (see below) were

immunoprecipitated by antibodies B5 (16) and

1736 (12),which recognize the extracellular do-main of

OPP

or a-secretase cleaved

.PPs,

respec-tively (Fig. 1).

To detect intracellularly cleaved IPPS, CHO cells pulse-labeled for 10 min were quickly

chilled with ice-cold Dulbecco's phosphate buff-ered saline (DPBS) after a 10- or 20-min chase

period. The cells were then incubatedat 4°C for 40 min with 0.1% saponin in DPBS supple-mented with protease inhibitors as described

(17). Saponin wasusedbecause this mild

deter-gent permeabilizes cells without solubilizing cell

membranes, thereby allowing soluble

intracellu-B5 I7

1.G75^.-W-,:... ..

-#d AB

CT15

a

206 MolecularMedicine, Volume 3,Number 3,March 1997

lar molecules to diffuse out of the cells.

Immu-noprecipitations using antibodies B5 and 1736

were then carried out from the saponin buffer supplemented with 10% calf serum, to recover

OPPS

released from the

permeabilized

cells. In surface iodinationexperiments, transfected CHO cells were radioiodinated with Na'25I as

de-scribed previously (18). After labeling, the cells

were chased withnormal CHO medium, withor

without PDBU, for 2 hr. A,3 wasthen

immuno-precipitated from the chase media with antibody 1282.Theimmunoprecipitatedmaterialwas

sep-arated by SDS-PAGE (6% tris-glycine or 16.5%

tris-tricine gels for

jPPs

or

AP3,

respectively).

Dried gels were either quantitated by

densitom-etryafter fluorographicenhancement orexposed

directly to phosphorimager screen. All labeling

experiments were repeated two to four times.

Kinetics of Surface f8PP Trafficking

To determine the kinetics of ,BPP secretion and internalization from the cell surface, a method

that quantitatively assesses both the release of

I3PPs

into media and the internalization of

OPP

from the cell surface using radiolabeled 1G7

monoclonal antibody was performed as

de-scribed (19). 1G7 is amonoclonal antibody that

recognizes the extracellular domain of ,BPP within the midregion and was added at a

con-centration of -7 nM to confluent CHO cells in 12-well tissue culture plates. Inbrief, after incu-bation at40C, the cells were either lysed

imme-diately (time 0) or placed inprewarmed (370C)

medium with or without PDBu. After 5, 10, 30,

and 60 min, the media were collected and cells

rapidly chilled with ice-cold DPBS at pH 2.8.

Afteranadditional5-minwash with acidic buffer

to detach residual surface-bound antibody, the cellswerelysedin0.2MNaOH.Radioactivitywas

then determined from the resultant three frac-tions: medium, acidwash, and acid resistant

ly-sate, which represent secreted, cell-surface, and intracellular ,BPP pools, respectively.TCA precip-itable counts were used from media and lysate

samples. Specific binding from all the antibody-binding experiments was calculatedby

subtract-ing the radioactivity from untransfected CHO

cellsperformedinparallel. Theresultswerethen

expressed in the three fractions at each time point as a percentage of the total radioactivity obtained at time 0. All experiments were

per-formed intriplicate and theresultsareexpressed as average (± SEM) of three repetitions.

Cell-Surface ,BPP

Three methods were used to determine the amount of

O3PP

onthe cell surface of transfected CHO cells. In the first approach, radioiodinated 1G7 antibody was addedtoconfluent CHO cells precooled to4°C asdescribed (19). Inall exper-iments, parallel sets of transfected and untrans-fected CHO cells were pretreatedwith PDBu for 0, 5, 10, or 30 min priortoantibody binding. In the secondapproach, full-length ,BPPwas immu-noprecipitated with CT1 5 antibodyimmediately after surface radioiodination from control cellsor from cells treated with 1 ,uM PDBu for 15 min beforelabeling.Inthe thirdapproach, the arrival of ,3PPtothecell surface fromnewly synthesized molecules was assayed by pulse labeling trans-fected CHO cells for 10 min, followed by chase periods of 10and20minwithorwithoutPDBu. After rapid coolinginice-cold DPBS, surface

O3PP

molecules were recoved by incubating the cells with 5A3/ 1 G7monoclonal antibodies for 1 hrat 4°C. 5A3/1G7, which recognize nonoverlapping epitopes inthe extracellular domain of

O3PP

(6), were used together to obtain higher signal. The cells were then lysed in the presence of 2-fold excess coldunlabeled transfected CHOcells. An-tibody-bound cell-surface fPPwas then precipi-tated by incubating the lysates with precleared anti-mouse agarose beads (1). Immunoprecipi-tated material was fractionated by SDS-PAGE and visualizedbyautoradiography or phosphor-imaging.

RESULTS

,BPP Trafficking afterPDBu Treatment

To examine whether PKC activation by PDBu affects the release orinternalization of ,BPP from the cell surface, the trafficking of ,BPP molecules

A

PDBu

- +

AB*

p3.'

35S

1251

1 di--- -I

----+

0 10 20 30 40 50 60

Time(mins.)

-0- PDBu -°- Control

Internal

-0o

0,S

-I

or 0

5

10 20 30 40 50 60

Time (mins.)

-e- PDBu -0- Control

100t

i

cn

8

a)

a.

80

Surface

60F

40

20

00 10 2I0 50- 60

10 20 30 40 so 60

Time(mins.)

FIG. 2. Time-course study ofcell-surface I3PP trafficking after PDBu

Radioiodinated 1G7IgG wasused todetermine the

rate of ,3PPrelease andinternalization from the cell surface inwild-type CHO cellline. The radioactivity from media (A), acid resistantlysates (B), and acid labile wash (C) wastakentorepresent secreted, in-ternalized, and cell surface ,3PP, respectively.Within eachof the three fractions, the result obtainedat

eachtime pointis expressedas apercent of radioac-tivity ofcell lysatesobtained attime 0. The value from the threefractions is <100% because only TCA-precipitable radioactivity wastaken into

ac-countandbecause of the variation inradioactivity obtained at eachtimepoint. The dataare the

aver-ages ± SEM ofthreeindependent experiments. The

differences inthe media and internalized fractions between control and PDBu treated cells were

statisti-cally significantateach ofthe four timepoints (p < .05 top < .003).

FIG. 3. Reduction inA8production and release afterPDBu

(A) Representative autoradiograms after the two dif-ferentlabeling methods. Inboth approaches, PDBu

wasadded towild-type CHO cellsduring the 2-hr chaseperiodaftereithera 10-minpulselabelingby

35S-methionine or 125I surface iodination. The

de-creaseinAPrelease in medium from 35S-methio-nine-labeled cells is mirrored byan increaseinp3

fragment. Recovery ofp3 after surface iodination is

inconsistent, as wasshownpreviously (6). (B) The

average percentreduction (± SEM) of A,B release after 35S-methionine (n = 4) and 1251 surface

label-ing (n = 3).The difference between thetwo labeling

methods wasstatistically significant (p < .005).

Release ofAj8 afterPDBuTreatment

BecauseA,3canbederived fromcell-surface

O3PP

processed in the endocytic pathway, one would

predict that the changes in

OPP

internalization demonstrated above should result inareduction

inA/3 release from surfaceprecursors.Indeed,by

using selective cell-surface iodination, addition of PDBu at the beginning of the chase period decreased recovery of labeled A,3 from medium

by -30% (Fig. 3). The magnitude of this

de-crease is consistent with the reduction of ,3PP

A

_{-i- PDBu}

lOOr

-°- Control

80K

Media

'a)a)

1--U) 0)

8

a)

a.

60 40 20

0 0

a_

B

.N

C a)

8

C

a-B

100

80 60 AAL

1001

20

/

goF

V0

-7.

60F

40F

C

0

-h c

8

IE

20

F

o

Effect ofPDBuonABrelease

(m Iw| Ad.

now

208 MolecularMedicine,Volume3, Number3, March 1997

A

PDBu-+

: 200

8...

997

.. ...

...

.:... :. ....

..:. 68

B

C

lOOr

0

a.

d*

60.

a-w

CO 40

t 20

o

0 5 10 30

Time after PDBu(mins)

FIG. 4. Levels of cell-surface f3PP after PDBu

(A) Immunoprecipitation of full-length

O3PP

with CT1 5 antibodyimmediatelyafter surface radioiodination. The

amount ofradiolabeled ,3PPwas reduced 15 min after PDBu pretreatment. Molecularweights determinedfrom prestained markers areshown on the right. (B) Levels of cell-surface ,BPP weremeasuredbyradioiodinated 1G7

monoclonal antibodybinding. ,3PP-transfected CHO cellswerepretreated withPDBu for 5, 10, or 30min before incubating with radiolabeled 1G7antibody. The results areexpressedas percentradioactivityofuntreated control cells measuredattime 0 (average ± SEM). (C) Levels ofnewly synthesized,BPP arrivingto the cell surface were determinedby selectively immunoprecipitating cell-surface

O3PP

with 5A3/1G7 monoclonal antibodies addedto

in-tactcells. Transfectedwild-typeCHO cells werepulselabeled with 35S-methionine for 10 min, followedbychase periodsof 10 or20 min withor without PDBu. Treatment with PDBu (+) showed marked reductions inthe

amount ofnewly synthesized cell surface ,BPP (bracket) ascomparedwithuntreated control cells (-). The

back-ground is higherinthis assaybecause the labeled lysates cannotbe thoroughly precleared.

internalization after PKC activation (Fig. 2b). However, by metabolic labeling with 35S-methi-onine, A,B release was inhibited by -80%

(Fig. 3), a result comparable to that reported by others (3).Therefore, when the production of

AP3

is measured from total cellular

O3PP

labeled by 35S-methionine, PDBu inhibitionof

AP3

releaseis substantially greater than when only the

cell-surface,BPP poolisexaminedby selective surface radioiodination.

Cell Surface fPP afterPDBu Treatment

The above results suggestthatthe effect ofPDBu

ontraffickingofcell-surface,BPP moleculesalone

isinsufficient to accountforthe overall decrease

in

AP3

production. Therefore, the amountof /3PP

on the cell surface after PDBu treatment was

examined. As detected by surface

radioiodina-tion and immunoprecipitation, the level of

cell-surface,3PPwasmarkedly decreased15 min after

the addition of PDBu (Fig. 4A). To more

accu-rately determinethe time-dependent changesin surface ,BPP expression, transfected CHO cells

werepretreatedwith PDBufrom 5 to 30minand then assayed using radioiodinated monoclonal

antibody 1G7 binding. Surprisingly, an 80%

de-cline incell-surface 13PPwas seenwithin5minof

addingPDBu and remainedatapproximately the

same level up to 30 min following treatment

(Fig. 4B). Since changes in endocytic trafficking

are minor (Fig. 2), in the range of 30%, this reduction in cell-surface

O3PP

must be due to otherfactors, inparticular, to diminishedsorting

of moleculesto the cell surface.

Further evidence for this mechanism was

provided by assaying for thearrival of ,BPPtothe

cell surface from newly synthesized molecules. Usingthisapproach,thelevels ofcell-surface 3PP

derivedfrom newly synthesized molecules were

dramatically decreased after PDBu treatment

(Fig. 4C). Specifically, treatment with PDBu for

10to20min afterthe initialpulse labeling period

resulted in -80-90% reduction in the amount

of labeled cell-surface ,BPP. This experimental paradigm was such that the labeled molecules

must havebeen newly synthesizedand recently transported to the cell surface, in contrast to earlier experiments (Fig. 2, 4A and B) that

ex-amined the fate of the steady-state pool of

O3PP

already present at the cell surface when PDBu

treatment wasinstituted. Thus, thelatterpoolis the summation of the decrease in nascent ,BPP arriving to the cell surface, as well as to the increase in I3PPS being released from the cell surface.

10mm

20mm

PDBu - +

OPP

E

_200

-97

Intracellular Cleavage ofj3PP

To confirm the above postulate that targeting of

O3PP

to the cell surface in the secretory/exocytic pathway is altered, intracellular

3PPP

was immu-noprecipitated from saponin-treated CHO cells after treatment with PDBu using a pulse-chase paradigm. Treatment with low concentrations of saponin (0.1%) permeabilized but did not solu-bilize cellmembranes because incontrol experi-ments, full-length ,BPP was virtually undetect-able from the saponin buffer (Fig. 5A). Therefore, predominantly soluble nonmem-brane-bound 3PP molecules, i.e.,

,PPs,

were an-alyzed in this assay. As expected, secretion of

.PPs

into the medium was

markedly

increased

byPDBu (Fig. 5B). The levels of I3PPs immuno-precipitated by both B5 and 1736 antibodies were increased by 2- to 3-fold, which is consis-tentwith the generationof increased amounts of a-secretase cleaved species. Concomitantly, as determined by saponin permeabilization, the amountof intracellular

I3PPS

detectedby bothB5 and 1736 immunoprecipitations was markedly decreased, by -70%. This finding suggests that there was a redistribution of the intracellular pool ofIBPPSmoleculesintothemedium,inother words,anaccelerationof both a-secretory cleav-age and release of

I3PPS,

bypassing the cell sur-face.

DISCUSSION

The data presented above provide compelling evidence that PKC activation by phorbol esters resultsinmajorandrapid alterationsin,BPP traf-ficking in the endocytic pathway. In particular, an80-90% decreaseinthe targeting ofIPP mol-ecules to the cell surface was seen within min-utesofPDBu treatment. Surprisingly, the reduc-tion intheamountofcell-surface ,BPP afterPDBu treatment was caused by a major decrement in sorting of ,BPP to the cell surface and only sec-ondarily from an increase in secretion of cell-surface molecules. This loss of cell-cell-surface ,BPP molecules derived from the exocytic pathway would lead to a corresponding decrease in sub-strate available for subsequent endocytic pro-cessing. Thus, the effect of phorbol esters on trafficking of ,BPP molecules at the cell surface appears tobe relatively minor. Finally, the data showed that the perturbations of ,BPP sorting to the cell surface are caused principally by an in-crease in intracellular a-secretase cleavage and

A

464

-zoo

-97

-68

B

PIE'

Thi - 43. +

-.97

FIG. 5. Immunoprecipitation ofI8PPSafter saponintreatment

(A) Transfectedwild-type CHOcells werepulse la-beled with 35S-methionine for 10minwithout a chase periodandtreated either with 0.1% or 0.2% saponinat4°Cfor40 min. Immunoprecipitation of

full-length

O3PP

was carriedoutwith CT1 5 antibody

from saponin buffer (saponin) orcelllysate (lysate), the latter after solubilization with 1% NP40.

Full-length

O3PP

(arrowhead) wasreadilyprecipitated

from celllysatebutwas minimallydetectable from the bufferfraction treated with 0.2% saponin and

essentially undetectableinthe 0.1% saponinbuffer. (B) Immunoprecipitations ofintracellular and secreted

13PPsafterPDButreatment.TransfectedCHO cellswere

labeledby 35S-methioninefor 10 minand chased for 20 minwith (+) orwithout (-) PDBu.

(3PP,

was

im-munoprecipitatedfrom medium (media) and from sa-poninbuffer (saponin) with antibodiesB5 and 1736. Levels ofimmunoprecipitable

OPP,

werehigherin me-diawith both 1736andB5 antibodies but lower

intra-cellularlywhenrecovered from the saponinbuffer,

afterPDButreatment.Antibody 1736consistently pre-cipitates less

f3PP,

thanantibody B5, accounting for the lowersignalinthe 1736lanes. Inlow-percentagegels, I3PP,fromCHO cellsmigratesas adoublet.

an increase in the trafficking of 3PPS out of the cell, and to amuch lesserdegree, from increased a-secretase activityat the cellsurface.

O%AF&^

210 MolecularMedicine,Volume 3, Number 3, March 1997

The results also showed that PKC activation by PDBu altersa number of steps in

AP3

genera-tion. For wild-type ,BPP, where the endocytic pathway ishypothesized to be the major contrib-utor to A,Bproduction, the net decrease in traf-fickingof ,BPP to cell surface is likely tounderlie the inhibition of

AP3

production by markedly re-ducing the substrate available for endocytic pro-cessing. Therefore, reduced internalization of cell-surface JPP only plays a minor role in this decrement in

AP3

release. Moreover, the results are consistent with the hypothesis that process-ing of ,BPP in the endocytic pathway is a major route for

AP3

production and release. In the "Swedish" APP double missense mutation, where

AP3

production appearstobe shiftedtothe secretory pathway, similar mechanisms occur-ring within the secretory pathway may underlie the reduction in

AP3

production after treatment by phorbol esters (20).

The data presented here are consistent with the report that upon PKC activation, ,3PP is re-distributed from the trans Golgi network (TGN) toother cellular locations through increased for-mation of secretory vesicles from the TGN (21). The results reported here complement this recent findingindemonstratingalargedisplacement of

,3PP

awayfromthe cellsurface andanincreasein the trafficking of fPP out of the cell. However, the data also demonstrated that the acceleration in a-secretase cleavage that accompanies this perturbation in

O3PP

sorting is only minimally increased at the plasma membrane. Therefore, a-secretase activity enhanced by PKC activation must occurinan intracellular compartment that remains tobe defined (22). Two potential mech-anisms mediating thisPKC effect come to mind: a directPKC-mediated activation of intracellular a-secretase activity, or a displacement of fPP to the compartmentwhere a-secretase is located.

Finally,inaddition to phorbol esters, a num-ber of other agents have been shown to both increase

I3PPs

secretionand decreaseAf3 produc-tion (8). These include treatment by cytokines, muscarinic activation, and inhibition of protein phosphatases. In view of the multiple effects of PKC activation by PDBu on

IPP

trafficking and processing, one would hypothesize that the other agents will show a range of cellular alter-ations affecting ,3PP and A,Bproduction and re-lease. Understanding the multitude of cellular perturbations that alter I3PP trafficking should generate additional potential targets for inhibit-ing A,B generation.

ACKNOWLEDGMENTS

Ithank Sharon Squazzo and Leslie Caromile for invaluable technical assistance, Drs. Christian Haass, Albert Hung, Sam Sisodia, and David Teplow for helpful discussions, and Drs. Dale Schenk and Dennis Selkoe for use of various antisera.Thisworkwassupported bygrantsfrom the Alzheimer's Association (ZEN 94-011) and the National Institutes ofHealth (AG12376 and NSO18 12).

REFERENCES

1. Weidemann A, Konig G, Bunke D, et al. (1989) Identification, biogenesis, and local-ization of precursor of Alzheimer's disease A4amyloid protein. Cell 57: 115-126. 2. Esch FS, Keim PS, Beattie EC, et al. (1990)

Cleavage of amyloid beta peptide during constitutive processing of its precursor. Sci-ence 248: 1122-1124.

3. Checler F. (1995) Processing of the ,3-amy-loid precursor protein and its regulation in Alzheimer's disease. J. Neurochem. 65: 1431-1444.

4. Haass C, Selkoe DJ. (1993) Cellular process-ing of

03-amyloid

precursor protein and the genesis ofamyloid ,B-peptide. Cell 75: 1039-1042.

5. Busciglio J, Gabuzda DH, Matsudaira P, Yankner BA. (1993) Generation of ,B-amy-loid in the secretory pathway in neuronal and nonneuronal cells. Proc. Natl. Acad. Sci. U.S.A. 90: 2092-2096.

6. Koo EH, Squazzo SL. (1994) Evidence that production and release of amyloid ,B-protein involves the endocytic pathway. J. Biol. Chem. 269: 17386-17389.

7. LeBlanc AC, Gambetti P. (1994) Production of Alzheimer 4kDa ,B-amyloid peptide re-quires the C-terminal cytosolic domain of the amyloid precurso protein. Biochem. Bio-phys. Res. Commun. 204: 1371-1380.

8. Gandy S, Greengard P. (1994) Processing of Alzheimer Aj3-amyloid precursor protein: Cellbiology, regulation, and role in Alzhei-mer disease. Int. Rev. Neurobiol. 36: 29-50. 9. da Cruz e Silva OAB, Iverfeldt K, Oltersdorf

T, etal. (1993)Regulated cleavage of Alzhei-mer ,B-amyloid precursor protein in the ab-sence of thecytoplasmic tail.Neuroscience. 57: 873-877.

ectodomain phosphorylation and regulated cleavage of ,B-amyloid precursor protein. EMBO J. 13: 534-542.

11. Buxbaum JD, Koo EH, Greengard P. (1993) Protein phosphorylation inhibits production of Alzheimer amyloid ,B/A4 peptide. Proc. Natl. Acad. Sci. U.S.A. 90: 9195-9198. 12. HungAY, Haass C, Nitsch RM, et al. (1993)

Activation of proteinkinase C inhibits cellu-lar production of the amyloid

P3-protein.

J. Biol. Chem. 268: 22959-22962.

13. Backer JM, King GL. (1991) Regulation of receptor-mediated endocytosis by phorbol esters. Biochem. Pharmacol. 41: 1267-1277. 14. Podlisny MB,Ostaszewski BL, Squazzo SL, et

al. (1995) Aggregation of secreted amyloid /3-proteininto sodium dodecyl sulfate-stable oligomers in cell culture. J. Biol. Chem. 270: 9564-9570.

15. Sisodia SS, Koo EH, Hoffman PN, Perry G, Price DL. (1993) Identification and transport of full-length amyloid precursor proteins in rat peripheral nervous system. J. Neurosci. 13: 3136-3142.

16. Oltersdorf T, Ward PJ, Henriksson T, et al. (1990) The Alzheimer amyloid precursor protein: Identification of a stable intermedi-ate inthebiosynthetic/degradativepathway. J. Biol. Chem. 265: 4492-4497.

17. Simmons CF Jr, Schwartz AL. (1984) Cellu-lar pathways of galactose-terminal ligand movement in aclonedhumanhepatoma cell line. Mol. Pharmacol. 26: 509-519.

18. Perez RG, Squazzo SL, Koo EH. (1996) En-hanced release of the amyloid 3-protein from codon 670/671 "Swedish" mutant 13-amyloid precursor protein occurs in both secretory and endocytic pathways. J. Biol. Chem. 271: 9100-9107.

19. Koo EH, Squazzo SL, Selkoe DJ, Koo CH. (1996) Trafficking of cell surface amyloid

,8-protein

precursor. I. Secretion, endocytso-sis, and recycling detectedby labeled mono-clonal antibody. J. Cell Sci. 109: 991-998. 20. Felsenstein KM, Ingalls KM, Hunihan LW,

Roberts SB. (1994) Reversal of the Swedish familial Alzheimer's disease mutant pheno-type in cultured cells treated with phorbol 12,13-dibutyrate. Neurosci. Lett. 174: 173-176. 21. Xu H, Greengard P, Gandy S. (1995) Regu-lated formation of golgi secretory vesicles containing Alzheimer B-amyloid precursor protein. J. Biol. Chem. 270: 23243-23245. 22. Arribas J, Massague J. (1995) Transforming

growth factor-a and ,B-amyloid precursor proteinshare a secretory mechanism. J. Cell Biol. 128: 433-441.