ACID-STABLE SERINE PROTEINASE INHIBITORS IN THE URINE OF ALZHEIMER DISEASE SUBJECTS

GIULIA SPARRO, SALVATORE BONAIUTO*, GABRIELLA GALOENZI, ANNA MARIA ELEUTERI, MAURO ANGELETTI, GIULIO LUPIDI, ROSALIA TACCONI, EL VIO GIANNANOREA 0, ANDREA VESPRINI*, EV ANORO FIORETTI Department of Molecular, Cellular and Animal Biology, Post-Graduate School in Biochemistry and Clinical Chemistry, University of'Camerino, 62032-Camerino (MC), Italy. *°ltalian National Research Centers on Axing (INRCA) of Ferrno*, 63023 Fenno (AP) and AppignanoO, 62010 Appignano (MC), Italy.


INTRODUCTION
Imbalance in the dynamic equilibrium between active proteinases and their specific inhibitors is suspected to be a pathogenic factor for many human diseases (Laskowski and Kato, 1980;Fritz and Wunderer, 1983;Travis and Salvesen, 1983;Jochum et ai., 1986).
A particularly significative association appears to be that between some malignancies and the expression of low molecular weight trypsin inhibitors such as the Pancreatic Secretory Trypsin Inhibitor (PSTI) (Matsuda et aI., 1983;Huhtala et aI., 1982;Ogawa et al., 1987;Tomita et aI., 1990) and the so called Acid Stable Trypsin Inhibitor (ASTI) Okumichi et ai., 1984;Onitsukaetal., 1985;Sumi et aI., 1987;Yoshida etai., 1989).The latter, a multipotent inhibitor found in several biological fluids, including urine, corresponds to the Urinary Trypsin Inhibitor (UTI) (Shulman, 1955;Gebhard and Hochstrasser, 1986) and can be obtained as a product of acid, or enzymatic, hydrolysis from the large plasma protein Inter-a-Trypsin Inhibitor (ITI) (Gebhard and Hochstrasser, 1986).Owing to its antifibrinolytic activity, UTI appears to play an important role in the metabolism of tumour cells and levels are increased in the urine of cancer patients (Chawla et aI., 1978;Onitsuka et ai., 1985).
Recently a role has been proposed for abnormalities of proteolysis in the aetiology of Alzheimer's disease, a degenerative condition, mainly characterized by the presence of neuritic plaques in several brain regions (Abraham and Potter, 1989).It has been suggested that these proteinaceous deposits could originate from an aberration of the proteolysis in the brain (Masters et ai., 1985).Supportive evidence has been deduced from the finding, in the neuritic plaque, of the serine proteinase inhibitor a1-antichymotrypsin (Abraham et aI., 1988) and a 57 residue domain resembling the kallikrein inhibitor, aprotinin or "Kunitz inhibitor" (Fritz and Wunderer, 1983) within the precursor of the ~-amyloid protein (Ponte et ai., 1988, Tanzi et aI., 1988;Kitaguchi et aI. , 1988), the most abundant protein component of the plaque.Recently , it has been reported that, in proteolytic processing of the ~-amyloid precursor, a proteinase very similar to cathepsin G could be involved (Razzaboni et ai., 1992).This enzyme which is expressed in astrocytes (Razzaboni et aI., 1992) in intimate contact with the neuritic plaque, could be regulated by UTI-like inhibitors (Gebhard and Hochstrasser, 1986) and kallikrein inhibitors (Razzaboni et aI., 1992).
We have recently found that acid treated human urine contains antiproteolytic activities against kallikrein (Eleuteri et ai., 1994), a serine proteinase which appears to play an important role also in cerebral function (Kizuki et aI., 1994).We have, therefore, undertaken a comparative study of the levels of acid-stable proteinase inhibitors (kallikrein and trypsin inhibitors) in the urine of healthy and Alzheimer subjects.

Chemicals
Bovine trypsin, and the synthetic substrate N-a-Benzoyl-DL-Arginine-p-Nitroanilide (BzArgNan) were obtained from Sigma Chemical Co (USA).Porcine pancreas kallikrein and aprotinin were a kind gift of Bayer AG (Germany).The synthetic substrate S-2266 used for measuring the kallikrein activity was obtained from Kabi Vitrum (Sweden).Antibodies anti-human Inter-a-Trypsin Inhibitor were obtained from Dakopatts (Denmark) and utilized without further purification.Polyclonal antibodies against aprotinin, raised in rabbits, were kindly supplied by Bayer AG (Germany).Sep-Pak Cl ~ cartridges were obtained from Millipore (USA), and prepacked HiTrap ™ affinity columns (I ml volume) from Pharmacia (Sweden).All other chemicals were of analytical grade.

Urine samples
Urine samples, collected in the morning, were obtained from healthy or Alzheimer subjects of both sexes (fi ve males and five females) of comparable age, and were treated as soon as possible after collection.Quantitation of the inhibitors was performed in triplicate on three different samples for each subject considered.Dementia was diagnosed according to the criteria of the Diagnostical Statistical Manual of mental disorders (DSM-III-Revised) (American Psychiatric Association, 1980).The subjects were classified as affected by Alzheimer's di sease when the criteria reported by the NINCDS-ADRDA work group were satisfied (McKhann et ai., 1984).

Affinity chromatography
Porcine pancreas kallikrein, or bovine trypsin were immobilised on HiTrapTM-NHS activated columns according to the suppliers' instructions.

Isolation of the inhibitors
Isolation of the acid-stable urinary inhibitors was carried out taking advantage of a procedure which allows the quantitative recovery of small amounts of active inhibitors in biological fluids essentially based on affinity-and reversed phase-HPLC chromatography as major purification tools (Fioretti et ai., 1993) (see Scheme 1).

HPLC
The HPLC analyses were performed with a System Gold apparatus (Beckman, USA) using a reversed phase (RP) column Supelcosil LC 18-DB, equipped with a guard column, both obtained from Supelco (USA).The column was equilibrated in 0.1 % trifluoroacetic acid (TFA) in water (solvent A) and eluted with a gradient obtained by mixing solvent A with a solution of 4 : I (v/v) acetonitrile-2-propanol in 0.1 % TFA (solvent B) at a flow-rate of I mllmin.The inhibitor samples were prepared by dissolution with solvent A just before use .Absorbance of the eluted fractions was recorded at 254 nm.Chromatographic data were analyzed with the Gold software.

Concentration of the inhibitors
The concentration of the urinary inhibitors was determined by titration with an aprotinin standardized trypsin solution, as previously reported (Fioretti et aI., 1993).

Antiproteolytic activity
The antiproteolytic activity of the urinary inhibitors, in the HPLC eluates, was determined by measuring their antitrypic orland antikallikrein activity as previously reported (Fioretti et al., 1993).The presence of organic solvents, which affects the proteolytic activity, was taken into account adding, in the reference test, an eluate volume from a run performed without the inhibitor.

Immunochemical characterization of the inhibitors
Cross-reactivity between polyclonal antibodies raised against bovine aprotinin, or human ITl, and the urinary inhibitors was assayed as previously reported (Fioretti et ai., 1984;Fioretti et ai., 1985) measuring the ability of the antibodies to remove the antitryptic activity of the inhibitors.

RESULTS
The experimental procedure utilized for the isolation of the acid-stable urinary proteinase inhibitors is summarized in Scheme 1. Usually, 200-250 ml of freshly collected urine was acidified by slow addition, under stirring, of a concentrated solution (50%) of trichloroacetic acid to a final TCA concentration of 2.5%.The supernatant obtained after centrifugation, was brought to pH 8.6 with concentrated NaOH solution and applied to a I ml prepacked HiTrapTM-kallikrein column, for isolation of the kallikrein inhibitors, or, alternatively, to a I ml HiTrap TM-trypsin column forthe isolation of the trypsin inhibitors.Both columns were washed with the equilibrium buffer (200 mM Tris-HCl pH 8.6 containing 20 mM CaCI 2 ) and eluted with a solution of HCI at pH 2.0.In both cases, the fractions showing anti proteolytic activity (measured as antitryptic activity) were pooled, neutralized and loaded on a Sep-Pak C 18 cartridge.The cartridge was eluted with a mixture obtained by mixing 60% of Solvent A and 40% of Solvent B Scheme 1. Procedure adopted for the isolation of the urinary acid-stable proteinase inhibitors.(see Materials and Methods).This fraction (about 3.0 ml) was utilized for the quantitative determination of the inhibitors in the urine samples (see Table 1).The determination was carried out using a titration with an aprotinin-standardized trypsin solution (see Material and Methods) taking advantage of the high affinity (Kd <lO•IIM) shown by both kinds of urinary inhibitors for the enzyme (see also below).Further information on the nature of the inhibitors recovered from the affinity columns was obtained through an HPLCreversed phase (RP) analysis and a preliminary molecular characterisation.The fraction obtained from the Sep-Pak C I8 cartridge was concentrated to dryness, dissolved in 50 f.l.I

RP-Chromatography
of Solvent A and then loaded on a RP-C I8 column, equilibrated and eluted as reported in the Material and Methods section.Figure I reports the elution profile and the antiproteolytic activity , determined against kallikrein and trypsin, of the fractions eluted from the CI S column when the material recovered from the kallikrein affinity column was chromatographed.In this case, the fractions showing antikallikrein activity (hereafter termed " kallikrein inhibitors") are eluted from the column at a concentration ranging between 30% and 40% of Solvent B, and both the elution profile and the anti kallikrein activity appear very similar when the urine of Alzheimer (Figure I B) or control subjects (Figure lA) was analyzed.The kallikrein inhibitors also show antitryptic activity and their anti proteolytic activity is unaffected either by the presence of polyclonal antibodies raised against the bovine aprotinin or by anti-human ITI antibodies which are immunologically identical to anti-UTI antibodies (Gebhard and Hochstrasser, 1986) (see above) .These results prompted further investigation of the nature of urinary kallikrein inhibitors.The fraction recovered from the CI S column was subjected either to SDSpolyacrylamide gel electrophoresis in non reducing conditions (Laemmli, 1970) or to gel permeation chromatography on a Superdex 75 HR 10/30 column (Pharmacia, Sweden).Inhibitor content was determined on the basis of antitryptic residual activity using an aprotinin standardized trypsin solution.Determinations were performed in triplicate on the fractions recovered from the Sep-Pak CIS cartridge (see Materials and Methods).
Both analyses showed that this fraction contained several proteins and that the antikallikrein activity was associated to a protein showing an apparent molecular weight of about 24 KDa.The complete characterization of this inhibitor is in the progress.
When the material recovered from the immobilized trypsin column was chromatographed on the RP-C I8 HPLC column under the conditions reported, the anti proteolytic activity was found in fractions eluted at a concentration ranging between 30% and 80% of Solvent B (Figure 2).As in the previous case, these fractions were assayed for anti-kallikrein activity and for their cross reactivity towards anti-aprotinin and anti-ITI antibodies.The results obtained clearly indicate that, in this case, two kinds of trypsin inhibitors are present in the CIS elute: i) a first one eluted between 30% and 40% of solvent B, accounting for the presence of the kallikrein inhibitors (see above) which shows also antitryptic activity and a second one (hereafter termed the "trypsin inhibitors") eluted between 40% and 80% of solvent B, which is inactive against kallikrein and recognizes anti-human-ITI antibodies (Figure 3).Also in this case, both the elution profile and the anti proteolytic activity of the fractions eluted when urine of Alzheimer (Figure 2B) or control subjects (Figure 2A) was subjected to HPLC analysis, appear very similar.The urinary antitryptic activity is associated, as expected, with multiple protein peaks showing an apparent molecular weight of 30KDa, 24KDa and 18KDa respectively, most likely arising from a proteolytic fragmentation of the native Inter-a-Trypsin Inhibitor.As can be seen in Table I, where the results obtained from this study are summarized, striking differences in the urinary levels of the kallikrein and trypsin inhibitors were observed.The urine content of kallikrein inhibitors (about 10 /lg/l) was very similar for healthy and Alzheimer subjects of both sexes, while in ill subjects a marked increase in the levels of the trypsin inhibitors was observed.As can be seen in Table I a mean content of 135 /lg of trypsin inhibitors per litre of urine was found in control subjects (150 /lg/litre for males and 120 /lg/litre for females).This value is elevated 2-3 fold in the urine of Alzheimer patients

DISCUSSION
A correct balance between proteolytic and anti proteolytic activities appears important for many physiological processes, those occurring at brain level included.Proteinases and proteinase inhibitors have, in fact, been involved in several aspects of neurogenesis and in glial function (Pittman, 1984;Krystosek and Seeds, 1984;Pittman, 1985;Pittman, 1987).A more direct involvement has recently been reported for the serine proteinases kallikrein (Kizuki et ai., 1994) and cathepsin G (Razzaboni et ai., 1992).Immunoreacti vity for glandular kallikrein was found in both the cerebral cortex and the brain stem of rats and appeared to be concentrated in the neuronal cell bodies, indicating, for this protease, an involvement in neuron function (Kizuki et ai., 1994).In addition, a marked decrease in kallikrein activity has been observed in the brain of Alzheimer subjects ( et al .. 1990) as well as in old senescence-accelerated mice (Kizuki et ai., 1994).
A cathepsin G-like activity has been partially purified from the normal human and monke y brain and from the brain of Alzheimer subjects (Razzaboni et al., 1992); this proteolytic acti\ity appears to be sensitive to inhibition by the kallikrein inhibitor (aprotinin) and is localized in astrocytes, glial cells surrounding the neuritic plaque (Razzaboni et al .. 1992).Owing to this localization it has been hypothesized that this proteinase could be involved in processing the p-amyloid precursor and hence could influence formation and deposition of the p-amyloid.We now show that, in human urine, acid-stable proteinase inhibitors showing antikallikrein or antitrypsin acti vity are present and that their levels, at least for the trypsin inhibitors, are increased in subjects with Alzheimer's disease.Both kallikrein-and trypsin inhibitors show a broad specificity towards proteolytic enzymes and are active against proteinases apparently involved in at a concentration of so lvent B ranging between 30% an d 40% (A) and between 40% and 80% (B).The effect was fol lowed measuring the abi lity of the ant ibodi es to remove the antitryptic activity of the inhibitors as previously reported (Fioretti er al., 1984;1985).
the function of nervous system cells such as kallikrein and cathepsin-G (the UTI-like inhibitors).In addition, the urinary proteinase inhibitors are characterized by a relatively low molecular weight and appear potentiall y able to cross the blood-brain barrier.UTIlike inhibitors, deriving from proteolytic processing of the plasma protein Inter-a-Trypsin Inhibitor, are circulating molecules found in many biological fluids.In conclusion, the finding of increased levels of UTI -like inhibitors in urine of Alzheimer's disease patients seems to support the hypothesis that aberrant proteolytic processes could be involved in the pathogenesis of this neurological disorder.Nevertheless the results presented here are not sufficient for assigning a specific role to acid-stable proteinase inhibitors in degenerative phenomena of the nervous system, but can be seen as a starting point for further studies.In particular, it could be important to examine a larger sample of Alzheimer subjects and to perform a similar analysis on biological fluids (such as the cerebral spinal fluid) in a closer contact with the brain tissue .
Pak CI S + C I8 HPLC column) (Sep-Pak CI S + C I8 HPLC column) Kallikrein inhibitors Trypsin inhibitors Figure I. HPLC elution profile and antiproteolytic activity of the urinary kallikrein inhibitors isolated from the urine of Alzheimer (B) or control subjects (A).Column: Supelcosil LC,,-DB.Flow-rate: 1.0 mUmin.Detection wavelength: 254 nm.The shaded area accounts for the fractions showing antiproteolytic activity determined against kallikrein and trypsin.For other details see Materials and Methods.

Figure 2 .
Figure 2. HPLC el ut io n profile and antiproteol ytic activity of the urinary trypsin inhibitors isolated from the urin e of .'.. ILhe imer r B ) or control subjects (AJ.Column: Supelcosil LC I8 -DB.Flow-rate: 1.0 mllmin.Detec tio n \\ J\ elc n ~th : 254 nm.The shaded area accounts for the fractions showing antitryptic activity.For other detail, ,ee \laterials and Methods.

Figure 3 .
Figure 3.Effect of increasing quantities of anti-aprotin in antibodies (0 ) and anti-Inter-a-Trypsin Inhibitor an tibodies (.) on the antiproteolytic activity of the urinary trypsin inhibitors eluted from the RP-C " column

Table 1 .
Urine content of acid-stable, low molecular weight, proteinase inhibitors in Alzheimer and healthy subjects.