JOURNAL OF CELLULAR PHYSIOLOGY 172:63–68 (1997)

Cysteine Proteinases Are Responsible forCharacteristic Transketolase Alterations in

Alzheimer FibroblastsFRANCESCO PAOLETTI,* ALESSANDRA MOCALI, AND DONATELLA TOMBACCINI

Istituto di Patologia Generale, Universita’ di Firenze, Firenze, Italy

Cultured fibroblasts from patients affected by Alzheimer’s disease (AD) exhibitedpeculiar alterations of the enzyme transketolase (TK). Abnormalities (dubbed alka-line bands, ab) consisted of enzyme forms having unusually high pl and wereproposed as a marker of the disease in living patients. The mechanisms of TK-ab expression were investigated with the use of cysteine proteinase inhibitorsand purified preparations of either rat liver or human cysteine proteinases. Thecysteine proteinase inhibitors N-acetyl-leu-leu-norleucinal (ALLN), L-trans-Ep-oxy-succinyl-leucylamido(4-guanidino)butane (E-64), and egg white cystatinadded to AD cells just prior to extraction abolished TK abnormalities. Moreover,1 day incubation of AD cultures with either ALLN (10 mg/ml), NH4Cl (10 mM),or KCl (30 mM) prevented TK-ab generation, due, presumably, to an impairmentof lysosomal functions. Isolated rat liver cysteine proteinases were able to degradeTK in normal extracts and reproduce the characteristic TK-ab of AD fibroblasts.Moreover, pure human cathepsin H was also shown to partially induce an Alzhei-mer-like TK pattern and cleave normal TK to a 35 kDa fragment as spontaneouslyoccurring in AD fibroblasts. The explanation of mechanisms of TK-ab formationprovided evidence for an underlying imbalance of proteolysis in AD fibroblastsdue to a relative increase/derangement of the cysteine proteinases cathepsinswhich might be also involved in the reported abnormal processing of multiplecellular components. J. Cell. Physiol. 172:63–68, 1997. q 1997 Wiley-Liss, Inc.

Alzheimer’s disease (AD) is a progressive neurode- (Horecker and Smyrniotis, 1953) that catalyzes twosteps of the nonoxidative pentose cycle (Horecker et al.,generative disorder characterized neuropathologically1954). Enzyme alterations, consisting of TK forms withby the formation throughout the cerebral cortex of in-unusually high pI values and therefore called alkalinetraneuronal neurofibrillary tangles and extracellularbands (TK-ab), did not represent physiological prod-amyloid neuritic plaques (Terry, 1963; Price, 1986).ucts of intact cells. Rather, TK-ab were found only inBrain alterations seem to correlate with the degree ofcell lysates as the result of abnormal enzyme degrada-either sporadic or familial dementia (Blessed et al.,tion (Paoletti et al., 1990). Interestingly, however, these1968) which is affecting an increasing number of people‘‘artifacts’’ have never been detected in fibroblasts fromin middle and late life (Kay, 1991).either healthy subjects or patients with other neurolog-In search of pathogenetic mechanims of AD, muchical diseases (Paoletti et al., 1990), thus proving to beinterest was addressed to the hypothesis that an under-a specific marker of AD fibroblasts.lying abnormality of protein catabolism may be present

The expression of TK-ab in AD fibroblasts was modu-in this disorder (for review see contributions to the vol-lated by culture conditions (Tombaccini et al., 1994)ume edited by Banner and Nixon, 1992). To supportand tentatively ascribed to a relative enhancement ofthis view, there were reports on the aberrant pro-endogenous Ca//-independent cysteine proteinases,cessing of amyloid precursor protein (APP) to amyloido-exhibiting chymotrypsin-like activity, which were re-genic peptides (Selkoe, 1989b; Sisodia et al., 1990; Eschleased and/or activated at the moment of cell extractionet al., 1990) by abnormal proteolytic cleavage (Selkoe,(Paoletti et al., 1990; Paoletti and Mocali, 1991).1989a), altered protein processing (Zhang et al., 1989),

To understand further the mechanism(s) of genera-and increases in lysosomal proteases in the brain (Ca-tion of altered TK forms in AD fibroblasts, we at-taldo and Nixon, 1990). Moreover, a number of alter-tempted to inhibit cysteine proteinases and to identifyations were reported to occur also in several nonner- the enzymes which might be responsible for alterations.vous tissues (Scott, 1993) to suggest the systemic na-

ture of AD.Previuos work from this laboratory has shown that *Correspondence to: Dr. Francesco Paoletti, Istituto di Patologia

Generale, Universita’ di Firenze, Viale G.B. Morgagni 50, 50134dermal AD fibroblasts exhibit peculiar alterationsFirenze, Italy.(Paoletti et al., 1990) of transketolase (TK) (EC 2.2.1.1),

a cytosolic thiamine pyrophosphate-dependent enzyme Revision received 23 January 1997; Accepted 27 January 1997

q 1997 WILEY-LISS, INC.

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP

PAOLETTI ET AL.64

Results presented herein describe how TK-ab were (0.4 mM final concentration) as basically reported byBajkoswki and Frankfater (1975). Volume in the cu-downmodulated by the action of cysteine proteinase in-

hibitors, NH4Cl, and KCl. Moreover, cysteine protein- vette was 0.97 ml, and enzyme activity was calculatedassuming for CLN an extinction coefficient of 3.79 1ases isolated from rat liver and pure human cathepsins

were shown to degrade human TK of normal fibroblasts 103 M01 cm01 at 405 nm. The substrate Na-benzoyl-L-arginine ethyl ester (BAEE) (Sigma) was also employedand yield TK-ab as typically encountered in extracts

of AD fibroblasts. These results proved that cysteine for the determination of papain activity (Anastasi etal., 1983) as monitored at 253 nm.proteinases were the class responsible for abnormal TK

degradation and pointed to cathepsin H as one of theIsoelectric focusing (IEF), SDS-PAGE, andenzymes potentially involved in the process.

immunostaining of TKMATERIALS AND METHODS The isoelectrophoretic separation of the enzyme was

Cell cultures carried out in a Multiphor LKB apparatus (Uppsala,Sweden) equipped with a constant power supply. GelPrimary cultures of fibroblasts were derived from the

dermal biopsy of either patients affected by Alzheimer’s slabs were prepared with 1% Agarose IEF (Pharmacia,Bromma, Sweden) containing 5.4% ampholine solutiondisease or age-matched healthy subjects (three cell

lines for each group). The clinical diagnosis of AD ful- (Pharmalyte 3-10; Pharmacia) and were run for about1 h at 1 W/cm, using 0.05 M sulfuric acid and 1 Mfilled NINCDS-ADRDA criteria as reported by

McKhann et al. (1984). AD cell lines were from primary sodium hydroxide as the anode and cathode buffers,respectively. Separated proteins were transferred ontocultures obtained in our laboratory as part of a local

research programme (Regione Toscana: III Programma nitrocellulose by Southern capillary blotting, and theenzyme was revealed by incubating the membrane withdi Ricerca Finalizzata 384/C) and provided also by CNR

(Consiglio Nazionale delle Ricerche) Cell Tissue Bank anti-TK rabbit IgG preparations followed by biotinyl-ated anti-rabbit antibodies and by the avidin/biotin-of Florence (Italy). Fibroblast cell lines from three cen-

tenerian females (kindly provided by Prof. C. Fran- immunoperoxidase system (Vectastain ABC kit; VectorLabs, Burlingame, CA) as reported previously (Mocaliceschi, Istituto di Patologia Generale, Universita di Mo-

dena, Italy) have been additionally employed as over- and Paoletti, 1989). The formation of the pH gradienton the gel was assessed by running a mixture of pIaged controls.

Fibroblasts were grown in Dulbecco’s modified Ea- standards (pH range 3–10) (Pharmacia).SDS-PAGE analyses were performed according togle’s medium (DMEM) 4500 (GIBCO, Paisley, England)

supplemented with 10% fetal calf serum (FCS) (Boeh- Laemmli (1970) using a mixture of biotinylated SDSmolecular weight standards (range 14–100 kDa) (Phar-ringer-Mannheim, Mannheim, Germany) at 377C in a

5% CO2 humidified atmosphere. Each culture was usu- macia).ally split in three parts every 15 days, using a trypsin

Inhibitors of cysteine proteinasessolution (0.5 mg/ml) for detachment. Confluence wasreached within 7–10 days from inoculation, and final Cystatin was purified from egg white as basically

reported in the literature (Anastasi et al., 1983). Thecell density was around 0.5 1 106 cells/cm2.homogeneity as well as the inhibitory activity of puri-

Preparation of extracts and protein assay fied cystatin was ascertained by either SDS-PAGEanalysis or titration with known amounts of authenticTo prepare cell extracts, we washed fibroblasts three

times with phosphate buffered saline (PBS) in a bench papain (EC 3.4.22.2; 26.4 mg prot/ml; cat. no. P-3125;Sigma), respectively. The other two cysteine proteinasecentrifuge, and pellets were lysed by ultrasonics (100

W 1 20 sec) in the presence of 10 mM cold Tris-HCl inhibitors, namely L-trans-Epoxy-succinyl-leucylam-ido(4-guanidino)butane (E-64) (cat. no. 1585673) (Bar-buffer at pH 7.2 (approximately 106 cell/0.2 ml buffer).

The lysates were centrifuged at 11,000 rpm for 2 min rett et al., 1982) and N-acetyl-Leu-leu-norleucinal(ALLN) (cat. no. 1086090) (Sasaki, 1990) were from(Beckmann Minifuge, Palo Alto, CA) in Eppendorf

tubes, and the resulting supernatants were used for Boehringer-Mannheim.the analysis.

Cysteine proteinase preparationsProtein concentration was determined according tothe method of Bradford (1976) using bovine serum al- Pure human liver cathepsin H (EC 3.4.22.6; 200–600

mU/mg protein, 10 mU) and cathepsin B (EC 3.4.22.1;bumin as the standard.50–200 U/mg protein, 5 U) were purchased from Calbi-

Determination of enzyme activities ochem (San Diego, CA). Papain, employed for the puri-fication of egg white cystatin, was a product of MerckAssays for TK activity were carried out as previously

reported (Paoletti, 1983) in the presence of a mixture (Darmstadt, Germany).Rat liver cysteine proteinases were isolated from ap-of pentose phosphates (ribose-5P and xylulose-5P) as

the substrates by monitoring at 340 nm for rates of proximately 300 g of fresh tissue as basically reportedby Barrett and Kirschke (1981) for extraction, autoly-NADH oxidation. Aldolase was determined according

to Willnow (1984). sis, and acetone fractionation. Subsequent purificationsteps were changed as follows. The acetone-precipi-Cysteine proteinase activities were routinely deter-

mined at 377C by a spectrophotometric assay at 405 tated material was dissolved with 0.1 M phosphatebuffer, pH 6.0, and activated by the addition of 5 mMnm in the presence of 0.1 M sodium phosphate buffer,

pH 5.5, containing also 1 mM EDTA, 2 mM L-cysteine EDTA and 10 mM L-cysteine for 15 min at room tem-perature. The preparation was applied to a Sephacryland Na-CBZ-L-lysine p-nitrophenyl ester (CLN)

(Sigma Chemical Co., St. Louis, MO) as the substrate S-200 (Pharmacia) column equilibrated with the phos-

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP

ALTERED PROTEOLYSIS IN ALZHEIMER FIBROBLASTS 65

Fig. 2. Modulation of TK abnormalities of AD fibroblasts by treat-ments in culture. AD cultures were incubated for 1 day at 377C inDMEM 4500 plus 10% FCS with or without the addition of eitherALLN (10–40 mg/ml), NH4Cl (10 mM), or KCl (30 mM) in the medium.Cells were then detached, washed with PBS, and harvested to beextracted. IEF analysis and detection of TK was carried out as above.Results shown were from a typical experiment out of three.

Fig. 1. Modulation of TK abnormalities by cysteine proteinase inhib-itors added during AD cell extraction. AD fibroblasts from resting RESULTScultures were harvested and sonicated in 10 mM Tris-HCl buffer, pH Inhibition of TK-ab expression in AD fibroblasts7.2, with or without several cysteine proteinase inhibitors. Extracts

by treatments during cell extractionwere then subjected to IEF analysis, and TK was revealed immunolog-ically following transfer onto a nitrocellulose membrane (see Materials AD fibroblasts have been challenged with cysteineand Methods). Effects of egg white cystatin (0, 0.08, 0.15, 0.8, 1.5 mg/

proteinase inhibitors checking for variations in the ap-ml corresponding to lane 1, 2, 3, 4, 5, respectively), ALLN (10 mg/ml),and E-64 (1 mg/ml) were reported as compared to controls (no addi- pearance of TK-ab. These were revealed immunologi-tion). The formation of pH gradient was assessed by running a mixture cally with anti-TK antibodies (see Materials and Meth-of pI markers (pl 3–10) (Pharmacia). Results shown were from a ods) following isoelectrophoretic (IEF) separation andtypical experiment out of four.

were identified as the immunoreactive enzyme forms(usually two or three bands on the gel) migrating withina pH interval of 8.5–9.2. Treatments were carried outon pellets of washed AD fibroblasts by adding the inhib-phate buffer. Cysteine proteinase activities were de-

tected in the fractions corresponding to a molecular itor just prior to sonication. Experiments with cystatin,a physiological cysteine proteinase inhibitor from eggmass of 25–35 kDa, which were collected to be pro-

cessed through a Thiopropyl-Sepharose 4B (Phar- white (Anastasi et al., 1983), have shown that the addi-tion of this compound to the lysis buffer within themacia) column as indicated by purchaser (Pharmacia,

1974). The active cysteine proteinases bound to the ma- range of 0–1.5 mg/ml was effective in decreasing TK-ab appearance in a dose-dependent manner (Fig. 1).trix were then displaced with 20 mM L-cysteine and 2

mM EDTA in 0.1 M phosphate buffer, pH 6.0. The sam- Moreover, the synthetic inhibitor ALLN (Sasaki, 1990)at the concentration of 10 mg/ml was found to abolishple was concentrated by ultrafiltration (YM10 filter

membrane; Amicon Corporation, Danvers, MA), and the formation of TK-ab which, in turn, occurred sponta-neously in untreated AD fibroblasts. A similar decreaseexcess of L-cysteine was removed by desalting on G-25

Sephadex. The Sephacryl S-200 and Thiopropyl-Sepha- in TK-ab was observed also in the presence of E-64(1 mg/ml) that was one of the most specific cysteinerose 4B steps were repeated once again.

The activity of the cysteine proteinase fraction was proteinase inhibitors (Barrett et al., 1982).routinely assayed at pH 5.5. However, experiments to

Inhibition of TK-ab expression in AD fibroblastsassess the stability of enzyme activities at differentby treatments in culturevalues of pH have shown maximum activity within the

pH range of 6.5–7. Activity was strictly dependent on In the attempt to modulate TK-ab formation in ADfibroblasts, we incubated cultures with a panel ofthe presence of 10 mM L-cysteine and 2 mM EDTA in

the assay mixture. Isolated rat liver cysteine protein- agents which might affect intracellular milieu and/orpotentially alter cysteine proteinase activities. Resultsases were capable to inactivate aldolase and were fully

inhibited by either E-64 and egg white cystatin. The in Figure 2 showed that TK-ab were deleted whenALLN was delivered to AD cultures at the final concen-SDS-PAGE profile of isolated proteinase preparation

was heterogeneous and showed three major protein tration of 10–40 mg/ml of medium and incubated for 1day before cell harvesting and extraction. Moreover,bands within a range of approximately 31, 29, and 25

kDa. The specific activity of rat liver proteinase prepa- TK-ab were also absent in AD fibroblasts maintainedwith 10 mM NH4Cl, which was known as an acidotropicration was 1.5 U/mg protein at 257C with Na-CBZ-L-

lysine p-nitrophenyl ester as the substrate under condi- agent that decreases intralysosomal protein processingdue to a disturbance of pH gradients in acidic cell or-tions given above.

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP

PAOLETTI ET AL.66

Fig. 3. Generation of TK abnormalities in control fibroblast extractFig. 4. Generation of TK abnormalities in control fibroblast extractstreated with or without rat liver cysteine ptoteinases. Cultured fibro-by pure human cathepsin B and H. Aliquots (45 ml) of control cellblasts from normal subjects were harvested and sonicated in 10 mMextracts (10 mg/ml) were incubated for 1 h at 377C in 25 mM Tris-Tris-HCl, pH 6.8. Aliquots (25 ml) of the extract (10 mg/ml) wereHCl, 1 mM EDTA, and 5 mM L-cysteine, pH 6.8, without additionsincubated separately with an equal volume of either 0.1 M Tris-HClor with either human cathepsin H (94 mg/ml, corresponding to 47 mU/plus 1 mM EDTA, pH 6.8 (control), or rat liver cysteine proteinasesml) and B (235 mg/ml, corresponding to 20.2 U/ml). Total volume in(7 mg/ml) equilibrated in the same buffer (see Materials and Methodsincubation mixtures was 85 ml. At the end of incubation, a portion offor details) up to 12 h at room temperature. The cysteine proteinaseeach mixture was employed for IEF (35 ml) and SDS-PAGE (15 ml)fraction was tested without any previous activation, following activa-analyses. Results shown were from a typical experiment out of three.tion with 1 mM EDTA and 5 mM L-cysteine or following activation

and subsequent addition of E-64 (0.1 mg/ml, final concentration). Ali-quots (45 ml) of each sample were used for IEF analysis. Results shownwere from a typical experiment out of four.

near to the cathode. Confirming the specificity of TKdegradation by cysteine proteinases was the fact thatTK-ab were 1) not formed when the inhibitor E-64 wasganelles (Hasilik and von Figura, 1984). Also very effec-added from the beginning to the incubation mixturetive was the incubation with 30 mM KCl, although itscontaining the extract plus active cysteine proteinasesinhibition of TK-ab expression could not be readily ex-and 2) only partially produced in the presence of pro-plained as in the case of NH4Cl. In turn, the additionteinases which were not previously activated by L-cys-of E-64 to AD cultures did not exert any inhibitoryteine and EDTA.effect on TK-ab appearance (data not shown) to confirm

However, it was noteworthy that TK activity in allprevious observations on the poor ability of this com-samples was maintained unchanged (approximatelypound to enter living cells (Wilcox and Mason, 1992;0.009 mU enzyme/ml) irrespective of the presence ofTombaccini et al., 1994).active proteinases. Therefore, cysteine proteinase-me-

Generation of TK-ab in extracts of normal diated degradation of TK to yield TK-ab must be nothuman fibroblasts by a preparation of isolated so extensive or at least a limited proteolytic process

cysteine proteinases from rat liver since these fragments were still catalytically and im-munologically active.The role of cysteine proteinases in the generation of

TK-ab was assessed further by effects of direct interac- Effects of commercially pure cathepsin B and Htion between extracts of normal human fibroblasts and on the generation of TK-ab in extracts ofrat liver cysteine poteinases. These have been isolated normal human fibroblastsfollowing the procedure reported in Materials andMethods and were 1) activated by EDTA and L-cys- The identification of the proteolytic enzyme responsi-

ble for TK alteration was achieved by the use of pureteine, 2) fully inhibited by cystatin and E-64, and 3)exhibiting a molecular mass within a range of approxi- commercial human cathepsin B and H. These enzymes

were activated and incubated with fresh extracts ofmately 20–35 kDa. Results of a typical experiment(Fig. 3) have showed that the normal IEF profile of TK normal human fibroblasts for 1 h at 377C. Results of

IEF analysis and TK immunodetection (Fig. 4, IEF)in control extracts (no proteinase added) was signifi-cantly modified by 1 h incubation with proteinases acti- have shown that cathepsin H was able to induce TK

degradation and formation of two alkaline enzymevated with 1 mM EDTA and 5 mM L-cysteine. Particu-larly, there was a large shift of TK immunoreactive forms which resembled, at least qualitatively, the TK-

ab occurring in AD cell extracts. Cathepsin B was onlymaterial towards the alkaline region of gel to recalltypical TK-ab. Even more marked was the picture of partially effective on TK by producing a single ab, and

the combination of both cathepsin H and B yielded athe extract incubated for 12 h with the activated pro-teinase fraction, as denoted by a decrease in normal TK profile quite comparable to that of addition of cathepsin

H alone (data not shown). The effect of cathepsin H onforms and concurrent increase in the dark background

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP

ALTERED PROTEOLYSIS IN ALZHEIMER FIBROBLASTS 67

TK was even more convincing considering that, in in particular showed a significant activity against TKbeing able 1) to induce the characteristic appearanceterms of activity, this proteinase was approximately

400-fold lower than cathepsin B (see legend to Fig. 4). of TK-ab, although less efficiently than the rat liverpreparation, and 2) to fish out TK in the crude extractIt is also worth noting that control extracts (without

any addition) showed no TK-ab despite the long incuba- and typically degrade it to a 35 kDa fragment as occursspontaneously in AD fibroblasts (Paoletti et al., 1990).tion at 377C to confirm the stability of TK in the absence

of exogenous proteinases. Taken together, these data suggest that, beyond thepotential role of cathepsin H, additional cysteine pro-Aliquots of samples used for the above IEF experi-

ments were also employed for SDS-PAGE analysis (Fig. teinases (most unlikely cathepsin B) might be involvedin the typical TK degradation by acting either sepa-4). Control TK subunits, represented by a single band

of approximately 68 kDa, were partially converted by rately or in a cooperative manner.The proteolytic machinery of AD fibroblasts could becathepsin H to a lower molecular mass immunoreactive

protein which migrated around 35 kDa. We cannot say altered due to a relative increase in cysteine protein-ases because of augmented expression of these enzymesthat this fragment corresponded to the TK-ab revealed

on IEF; however, it is remarkable that a cleavage prod- or decreased levels of their endogenous inhibitors.However, this assumption seemed against preliminaryuct of the same size was reported to occur spontane-

ously in AD-fibroblast extracts (Paoletti et al., 1990). results showing that AD fibroblasts contained signifi-cantly less cysteine proteinase activity than normal fi-

DISCUSSION broblasts. An alternative hypothesis we are presentlyevaluating is that cysteine proteinases, irrespective ofInhibitory effects of ALLN, E-64, and egg white cys-

tatin on TK-ab formation in AD fibroblasts allowed us their levels, might be unevenly distributed in AD fi-broblasts, and this would allow an abnormally closeto focus on cysteine proteinases as the class of enzymes

involved and address more specifically the cellular com- interaction between cysteine proteinases and TKwithin the short time of cell extraction.partment in which they operate. A role of calpains in

TK degradation seemed unlikely since these cytosolic TK-ab can be regarded as a useful yet nonphysiologi-cal marker to disclose a constitutive alteration of prote-cysteine proteinases were 1) strictly Ca//-dependent

for activity (Murachi, 1983), while TK degradation was olysis in AD vs. age-matched controls, overaged con-trols from three centenarians (data not shown), andnot mediated by calcium ions (Paoletti and Mocali,

1991), and 2) poorly inhibited by egg cystatin, which, other neurological diseases (Paoletti et al., 1990). How-ever, rather than TK, the major target of this alterationon the contrary, was a powerful inhibitor of lysosomal

cysteine proteinases of the cathepsin family (Barrett, should be those proteins which were currently chan-nelled through the lysosomal/endosomal compartment1987). The involvement of lysosomal enzymes was also

denoted by the following observations. First, abnormal- and therefore might be abnormally processed. Thisview was in keeping with the occurrence of a varietyities of TK (which is a cytosolic enzyme) in AD fibro-

blasts were produced only at the moment of cell extrac- of altered products in AD cells (Selkoe, 1989) and withthe fact that the amyloidogenic peptide b-AP was re-tion (Paoletti et al., 1990), suggesting the release of

segregated enzymes. Second, the inhibition of TK-ab ported to derive from incorrect processing of amyloidprecursor protein (APP) in the lysosome (Sisodia et al.,formation in AD cells by NH4Cl in culture was strongly

in favor of an impairment of lysosomal functions due 1990), possibly through an alternative pathway (Haasset al., 1992). We do not know yet whether the mecha-to a disturbance of pH gradients in acidic cell organelles

(Hasilik and von Figura, 1984). This finding, together nisms reported for TK-ab generation could also applyto the formation of b-AP. It is intriguing, however, thatwith results of experiments with ALLN in culture, has

proved that abnormal degradative processes occurring the cysteine proteinases cathepsin S (Munger et al.,1995; Lemere et al., 1995) and B (Cataldo and Nixon,in AD fibroblasts possibly might be corrected in intact

cells by the use of agents capable of crossing intracellu- 1990) have been inferred to play a role in APP abnormalprocessing.lar membranes and inhibiting lysosomal cysteine pro-

teinase activities. With regard to the inhibitory effect ACKNOWLEDGMENTSof KCl on TK-ab formation, this might correlate withThis work was supported by grants from MURSTthe increase in free intracellular Ca// occurring in AD

(60% and 40%) Progetto: Patologia e Fisiopatologiafibroblasts by the activation of the 166-pS K/ channelGenerale dell’Invecchiamento and from Regione Tos-in response to elevated external K/ (Etcheberrigaraycana: III Programma di Ricerca Finalizzata 384/C.et al., 1993). In fact, it has been shown previously that

TK abnormalities were reduced in the presence of Ca// LITERATURE CITEDand augmented by the addition of chelators (Paoletti

Anastasi, A., Brown, M.A., Kembhavi, A.A., Nicklin, M.J.H., Sayers,and Mocali, 1991). C.A., Sunter, D.C., and Barrett, A.J. (1983) Cystatin, a protein in-Direct evidence for the involvement of lysosomal cys- hibitor of cysteine proteinases. Biochem. J., 211:129–138.

Bajkoswki, A.S., and Frankfater, A. (1975) Specific spectrophotomet-teine proteinases in TK degradation was provided byric assays for cathepsin B1. Anal. Biochem., 68:119–127.results of the incubation of normal cell extracts with

Banner, C.D.B., and Nixon, R.A., eds. (1992) Proteases and Proteasethe rat liver enzyme preparation. This, which was os- Inhibitors in Alzheimer’s Disease Pathogenesis. The New Yorktensibly a mixture of cathepsin B, H, L, and S and Academy of Sciences, New York, Vol. 674, pp. 1–252.

Barrett, A.J. (1987) The cystatins: A new class of peptidase inhibitors.dipeptidyl peptidase I, was capable of turning a normalTrends Biochem. Sci., 12:193–196.TK pattern into an AD pattern with the typical appear-

Barrett, A.J., and Kirschke, H. (1981) Cathepsin B, cathepsin H, andance of very intense TK-ab. More specific identification cathepsin L. Proteolytic enzymes, part C. In: Methods in Enzymol-of the proteinase involved was attempted by using com- ogy. L. Lorand, ed. Academic Press Inc., New York, Vol. 80, pp.

535–561.mercially pure human cathepsin B and H. The latter

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP

PAOLETTI ET AL.68

Barrett, A.J., Kembhavi, A.A., Brown, M.A., Kirschke, H., Knight, Department of Health and Human Services Task Force on Alzhei-mer’s disease. Neurology, 34:939–944.C.G., Tamai, M., and Hanada, K. (1982) L-trans-epoxy-succinyl-

leucylamido(4-guanidino)butane (E-64) and its analogues as inhibi- Mocali, A., and Paoletti F. (1989) Transketolase from human leuko-cytes. Isolation, properties and induction of polyclonal antibodies.tors of cysteine proteinases including cathepsins B, H and L. Bio-

chem. J., 201:189–198. Eur. J. Biochem., 180:213–219.Munger, J.S., Haass, C., Lemere, C.A., Shi, G.-P., Wong, W.S.F.,Blessed, G., Tomlinson, B.E., and Roth, M. (1968) The association

between quantitative measures of dementia and senile change in Teplow, D.B., Selkoe, D.J., and Chapman, H.A. (1995) Lysosomalprocessing of amyloid precursor protein to b-peptides: A distinctthe cerebral grey matter of elderly subjects. Br. J. Psychiatry,

114:797–811. role for cathepsin S. Biochem. J., 311:299–305.Murachi, T. (1983) Calpain and calpastatin. Trends Biochem. Sci.,Bradford, M. (1976) A rapid and sensitive method for the quantitation

8:167–169.of microgram quantities of protein utilizing the principle of protein-Paoletti, F. (1983) Purification and properties of transketolase fromdye binding. Anal. Biochem., 72:248–254.

fresh rat liver. Arch. Biochem. Biophys., 222:489–496.Cataldo, A.N., and Nixon, R.A. (1990) Enzymatically active lysosomalPaoletti, F., and Mocali, A. (1991) Enhanced proteolytic activities inproteases are associated with amyloid deposits in Alzheimer brain.

cultured fibroblasts of Alzheimer patients are revealed by peculiarProc. Natl. Acad. Sci. U. S. A., 87:3861–3865.transketolase alterations. J. Neurol. Sci., 105:211–216.Esch, F.S., Keim, E.C., Beattie, R.W., Blacher, A.R., Culwell, T., Olter-

Paoletti, F., Mocali, A., Marchi, M., Sorbi, S., and Piacentini, S. (1990)sdorf, D., McClure, D., and Ward, P. (1990) Cleavage of amyloidOccurrence of transketolase abnormalities in extracts of foreskinb-peptide during constitutive processing of its precursor. Science,fibroblasts from patients with Alzheimer’s disease. Biochem. Bio-248:1122–1124.phys. Res. Commun., 172:396–401.Etcheberrigaray, R., Ito, E., Oka, K., Tofrl-Grehl, B., Gibson, G.E.,

Pharmacia Fine Chemicals (1974) Activated Thiol-Sepharose 4B.and Alkon, D.L. (1993) Potassium channel dysfunction in fibroblastsBromma, Sweden.identifies patients with Alzheimer disease. Proc. Natl. Acad. Sci.

Price, D.L. (1986) New perspectives on Alzheimer’s disease. Annu.U. S. A., 90:8209–8213.Rev. Neurosci., 9:489–512.Haass, C., Koo, E.H., Mellon, A., Hung, A.Y., and Selkoe, D.J. (1992)

Sasaki T. (1990) Inhibitory effect of di- and tripeptidyl aldehydes onTargeting of cell-surface b-amyloid precursor protein to lysosomes:calpains and cathepsins. J. Enzym. Inhib., 3:195–201.Alternative processing into amyloid-bearing fragments. Nature,

Scott, R.B. (1993) Extraneuronal manifestations of Alzheimer’s dis-357:500–503. ease. J. Am. Geriatr. Soc., 41:268–276.Hasilik, A., and von Figura, K. (1984) Processing of lysosomal enzymes Selkoe, D.J. (1989a) Biochemistry of altered brain protein in Alzhei-in fibroblasts. In: Lysosomes in Biology and Pathology. J.T. Dingle, mer’s disease. Annu. Rev. Neurosci., 12:463–490.R.T. Dean, and W.S. Sly, eds. Elsevier Scientific Publishing Co., Selkoe, D.J. (1989b) Amyloid b-protein precursor in the pathogenesisAmsterdam, Vol. 7, pp. 3–16. of Alzheimer’s disease. Cell, 58:611–612.Horecker, B.L., and Smyrniotis, P.Z. (1953) The coenzyme function of Sisodia, S.S., Koo, E.H., Beyreuther, K., Unterbeck, A., and Price,thiamine pyrophosphate in pentose phosphate metabolism. J. Am. D.L. (1990) Evidence that b-amyloid protein in Alzheimer’s diseaseChem. Soc., 75:1009–1010. is not derived by normal processing. Science, 248:492–495.

Horecker, B.L., Gibbs, M., Klenow, H., and Smyrniotis, P.Z. (1954) Terry, R.D. (1963) The fine structure of neurofibrillary tangles inThe mechanism of pentose phosphate conversion to hexose mono Alzheimer’s disease. J. Neuropathol. Exp. Neurol., 22:629–641.phosphate. I. With a rat liver enzyme preparation. J. Biol. Chem., Tombaccini, D., Mocali, A., and Paoletti, F. (1994) Characteristic207:393–403. transketolase alteration in dermal fibroblatsts of Alzheimer pa-

Kay, D.W.K. (1991) The epidemiology of dementia: A review of recent tients are modulated by culture conditions. Exp. Mol. Pathol.,work. Rev. Clin. Geront., 1:55–66. 60:140–146.

Laemmli, U.K. (1970) Cleavage of structural proteins during the as- Wilcox, D., and Mason, R.W. (1992) Inhibition of cysteine proteinasessembly of the head of bacteriphage T4. Nature, 227:680–685. in lysosomes and whole cells. Biochem. J., 285:495–502.

Lemere, C.A., Munger, J.S., Shi, G.-P., Natkin, L., Haass, C., Chap- Willnow, P. (1984) Fructose-1,6-bisphosphate aldolase. In: Methodsman, H.A., and Selkoe, D.J. (1995) The lysosomal cystein protease, in Enzymatic Analysis. H.U. Bergmeyer, ed. VCH Verlag Chemiecathepsin S, is increased in Alzheimer’s disease and Down syn- GmbH, Weinheim, Vol. 4, 3rd ed., pp. 346–353.drome brain. Am. J. Pathol., 146:848–860. Zhang, H., Sternberger, N., Rubinstein, L.J., Herman, M.M., Binder,

McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., L.I., and Sternberger, L.A. (1989) Abnormal processing of multipleand Stadlan, E.M. (1984) Clinical diagnosis of Alzheimer’s disease: proteins in Alzheimer disease. Proc. Natl. Acad. Sci. U. S. A.,

86:8045–8049.Report of the NINCDS-ADRDA Work Group under the auspices of

/ 8918$$0485 08-13-97 11:06:12 wlcpa W Liss: JCP