Messenger RNA in the Pathogenesis of - NCBI

4 downloads 263959 Views 2MB Size Report
role of protease inhibitors in the pathogenesis of AD.8 ... system (Amersham Inc., Arlington Hts., IL). ..... cells among glia that express ATIII mRNA are astro- cytes.
American Journal of Pathology, Vol. 143, No. 3, September 1993 Copynight © American Society for Investigative Pathology

Serine Protease Inhibitor Antithrombin III and Its Messenger RNA in the Pathogenesis of Alzheimer's Disease

Rajesh N. Kalaria, Todd Golde,* Stephanie N. Kroon, and George Perry* From the Departments of Neurology and of Neurosciences and Pathology, * Case Western Reserve University School of Medicine, Cleveland, Ohio

The classical plasma protein antithrombin III (ATIII), an inhibitor ofthe blood coagulation cascade, is a member of the serpins that are gaining import in the nervous system. In this study, we examined the presence of ATIII in the pathological lesions ofAlzheimer's disease (AD). Antibodies to ATIII consistently detected -58-kd protein(s) on immunoblots of cerebral cortex and brain microvessels. Immunocytochemical studies showed ATIII reactivity within amyloid deposits, neurites associated with plaques, and neurofibriUlary tangles in neocortex and hippocampus of virtually aU the AD cases examined. In some cases, astrocytes were also stained, suggesting ATIII in these cells. ATIII immunoreactivity in neurofibriUary tangles was further defined by electron microscopy, which showed it to be associated with paired helicalfilaments. Using the polymerase chain reaction technique to amplify ATIII complementary DNA, wefound low levels of messenger RNA expression, relative to liver, in control human brain samples, and these were increased in AD samples, particularly in the white matter. Our results suggest the increased presence ofATIII commensurate with astrogliosis and association with the neurofibrillarypathology of AD. We conclude that in concert with other amyloid-associated serine protease inhibitors, ATIII may play a role in the pathogenesis of cerebral amyloidosis. (AmJ Pathol 1993, 143.886-

893) Recent studies1-4 on the localization of a1antichymotrypsin in cerebral amyloid deposits of Alzheimer's disease (AD) and on one of the forms of f3-amyloid precursor protein of AD containing a se-

886

quence homologous to the Kunitz family of serine protease inhibitors5-7 have drawn much attention to the role of protease inhibitors in the pathogenesis of AD.8 These observations imply that proteases and protease inhibitors play important complementary roles in the nervous system9'10 and that imbalances in proteolysis might be involved in cerebral amyloidosis. It is now known that the brain expresses an antithrombinlike molecule termed protease nexin (PNI) that exhibits potent thrombin inhibitory activity' 1112 and binds to extracellular amyloid deposits.13 Related studies suggest that PNI is produced by glia and possibly brain capillaries14 and is reduced in AD.15 Whereas the precise role of PNI is unclear, it is thought that PNI around blood vessels may play a protective role against extravasation of thrombin and possibly other plasma proteases into human brain.14 In light of this, we undertook studies to examine antithrombin Ill (ATI 11), a serpin that is a member of the superfamily of proteins that contain a1-antitrypsin,16 a1-antichymotrypsin ovalbumin, and angiotensinogen, all with considerable amino acid homology.17'18 ATIII and thrombin are serum proteins normally produced in the liver and known to be involved in the regulation of blood coagulation.19 ATIII or antithrombin-heparin cofactor has a wide spectrum of activity and inhibits thrombin as well as other activated coagulation factors by forming a stable bimolecular complex with the enzyme. Interestingly, thrombin has been reported to have pronounced mitogenic effects on astrocytes and to inhibit neurite outgrowth in vitro.20 Here, we examined the immunolocalization of ATIII in amyloid deposits and the neurofibrillary pathology of AD. We also used the polymerase chain reaction (PCR) technique to assess the expression of

Supported by the University Hospitals Alzheimer Center and the USPHS for grants AG08012 and AG 10030 (RNK), AG07552, AG09287, and KO4-AG00415 (GP). Accepted for publication April 19, 1993. Address reprint requests to Dr. R. N. Kalaria, Department of Neurology, University Hospitals of Cleveland, 2074 Abington Road, Cleveland, Ohio 44106.

ATIII in the Pathology of AD

887

AJP September 1993, Vol. 143, No. 3

its messenger (m)RNA in brain tissue obtained at autopsy from subjects with AD and normal aging controls.

Materials and Methods Tissue Brain samples from the frontal (Brodmann 10), temporal (Brodmann 22), parietal (Brodmann 1 and 2) and occipital cortex (Brodmann 17), cerebellum, hippocampus, and parahippocampal gyrus, and the corpus callosum were obtained at autopsy from a total of 22 subjects with an age range of 55 to 93 years. Fifteen of these were from subjects with histopathologically confirmed AD, and the rest were non-AD or nonneurological controls.21 In addition, samples from liver, kidney, and spleen were obtained from two control subjects without evidence of systemic disease. The postmortem interval ranged between 2 and 8 hours for all samples.

Immunoblotting and Immunoprecipitation Immunoblots of solubilized proteins or immunoprecipitates from the cerebral cortex and cerebral microvessels were essentially prepared (Figure 1), as

AL

described previously.21'22 The ATIII immunoprecipitates were prepared by extracting tissue in 50 mM Tris-Hcl buffer pH 7.5, containing 150 mM NaCI, 2 mM EDTA, 0.2% Nonidet-P40 and 0.2% Triton X-10022 and incubation of the high-speed supernatant with ATIII antibodies followed by precipitation with protein A-agarose to obtain insoluble complexes. For immunodetection, solubilized proteins or the immunocomplexes were suspended in Laemmli sample buffer, boiled and electrophoresed on 10% acrylamide gels. The electrophoresed samples were then transferred onto Immobilon-P membranes

(Millipore, Bedford, MA) and immunodetected using the electrochemical luminescence Western blotting system (Amersham Inc., Arlington Hts., IL). To evaluate specificity of the antibodies, the blots were probed with antibodies absorbed with normal human serum or ATIII (Sigma Chemical Co., St. Louis, MO). The diluted antibodies were incubated (4 C for 11 to 16 hours) in a ratio of 1:10 with human serum and used as primary antibody. The electrochemical luminescence method also gave staining in lower molecular weight (MW) bands that were shown to be nonspecific based on the absorption experiments.

I

4',,

SI.,, -p

;.... ...

. .i or..

Figure 1. ATIMI immunostaining of NFT, amyloid deposits, cortical capillaries, and astrocytes in AD. A and B: intense staining of both extracellular and intracellular NFT recognized by a MAb to ATII( Chemicon, Temecula, CA). C: Both amyloid plaques (arrowhead) and intracortical capillaries were stained by an antiserum to ATII (DAKO, Carpinteria, CA). D: staining of astrocytes in temporal cortex from an AD subject. Immunoperoxidase viewed under differential interference (Nomarski) optics. Scale bar = 50 y.

888

Kalaria et al

AJP September 1993, Vol. 143, No. 3

Immunocytochemistry Except where noted, fresh tissue blocks of about 0.5-cm thickness were fixed by immersion in 3 to 4% formaldehyde in sodium phosphate buffer, pH 7.4, for 12 to 24 hours at 4 C. After fixation, blocks were cryoprotected in sucrose and cut into 20- to 25-p sections with a freezing microtome. Freefloating sections were immunostained with various antibodies to ATIII (Table 1) as described previously.21 Although Triton X-100 was used throughout to permeabilize the tissue, preliminary experiments showed that protease treatments4 were not generally necessary for detecting ATIII immunoreactivity. In some cases, sections were cut from unfixed frozen tissue, which following sectioning was postfixed in ice-cold acetone.4 Immunostained or unstained adjacent tissue sections from each case were also stained with thioflavin S. To verify staining in various lesions and cellular structures, adjacent sections were immunostained with antibodies to A4/f3protein, T, ubiquitin, glial acidic fibrillary protein (GFAP), or collagen IV (Table 1). To test specificity of ATIII immunostaining, control sections were incubated with either irrelevant antisera, ATIII antiserum preabsorbed with the ATIII protein, preimmune serum, or ascites, and in the absence of primary antibody.

Immunoelectron Microscopy For electron microscopy, 40- to 50-p-thick Vibratome sections prepared from fixed tissue were incubated free-floating with primary antiserum containing 0.3% Triton X-100. These sections were imTable 1. Immunocytochemical Staining by Antibodies to ATII and Other Proteins in AD Brains

Antibody and source Dilution NFP PLQ Vessels ATIII MAb 1

1:10

+

±

0/±

1:10

0

+

+1±

1:100

+

+

+/+

1:1000

+

+1±

1:100

± ±

+

+/+

1:100

0

+

±/0

1:100 1:100 1:100

+ + 0

0 +

0/0 0/0

(Chemicon)

ATIII MAb 2

(Pierce)*

ATIII rabbit 1

(DAKO)

ATIII rabbit 2

(Sigma)*

ATIII sheep 1

(Serotec)

A4/0 protein

(Perry) T (Perry) Ubiquitin (Perry) Collagen IV (Chemicon)

±

+/+

+: positive; ±: variable; 0: negative. Antibodies to other proteins

were used to identify amyloid lesions. * Positive staining was evident in frozen sections.

munostained using either ATIII-1 or ATIII-2 as primary antibody (Table 1) and were either 1) immunodetected by the peroxidase-antiperoxidase method of Sternberger23 or 2) stained by the immunogold method.24 Ultrathin sections were placed onto copper grids, and in the case of immunogold preparations contrasted with uranyl acetate and lead citrate and examined on a JEOL 100CX microscope at 60 kV.

Reverse Transcribed-Generated Complementary DNA PCR PCR was performed essentially as described previously by Golde et al.25 Total cellular RNA extracted from either individual or pooled samples of cortical grey and white matter and the peripheral tissues was reverse transcribed with the murine leukemia virus reverse transcriptase (Bethesda Research Laboratories, Grand Island, NY). The complementary DNAs thus obtained were used as templates for the PCR. ATIII oligonucleotide pairs were synthesized on an Applied Biosystems Inc (ABI) 380A DNA synthesizer making use of the phosphoramidite method. The oligonucleotides were purified using an OPC cartridge (ABI), dried down, and resuspended in diethylpyrocarbonate water. The sequence of the forward primer, ATIII+1 1 was AAGCCGCGGGACATTCCCATGAATCC, corresponding to bases 11 to 19 of the ATIII sequence. The sequence of the reverse primer, ATIII-161 was GAAAGCGGGAATTGGCCTT and is complementary to bases 161 to 155 of the ATIII sequence.26 To assess relative amounts of amplified products, control ribosomal protein S6 (RPSP6) primers were run in parallel. The sequences for RPS6 were +114 TGACGCTCTGGGTGAAGAATG and -213 GCCATGGGTCAAGACACCCT. Amplification and quantification of the ATIII and RPSPG complementary DNAs were carried out as previously described25 except that 1) amplified products were labeled with 0.1 pCi [32P]aATP, 2) the concentration of aNTPs was 50 pM, and 3) complementary DNAs for ATIII and RPSPG were amplified for 30 and 22 cycles, respectively.

Results Our initial finding27 suggested the localization of ATIII-like immunoreactivity in lesions typically associated with AD. Further intensive immunocytochemical studies revealed that ATIII reactivity was evident in both neurofibrillary tangles (NFT) and amyloid

ATIII in the Pathology of AD 889 AJP September 1993, Vol. 143, No. 3

plaques (Table 1, Figure 1) of virtually all neocortical and hippocampal samples examined. Whereas staining in both the compact plaques and to a lesser extent in the diffuse type was noted, intense staining was particularly evident in neurites and in neuropil threads associated with amyloid plaques (Figure 1). ATIII staining corresponded to the degree of pathology verified by thioflavin S staining or immunostaining with A4/f3 protein antibodies in adjacent sections. In a few AD samples, consistent patterns of staining in astrocytes bordering the glia limitans were also evident (Figure 1). The astrocytic staining was verified by positive GFAP immunoreactivity in adjacent sections (not shown). However, ATIII antibodies invariably stained intracortical vessels and capillaries as readily evident with collagen IV immunoreactivity in adjacent sections. In contrast to the striking staining of amyloid deposits and tangles in AD, brain tissue without detectable pathology from control subjects revealed only diffuse staining of capillaries and occasionally astrocytes. The immunoreactivity in capillaries was presumably due to residual plasma and vessel-bound protein (Figure 1C). The presence of ATIII immunoreaction product in the NFT in tissue from AD subjects was further localized by electron microscopy using the

immunoperoxidase method (Figure 2). At the ultrastructural level, immunoreactive paired helical filaments in tangle-bearing neurons were readily detected (Figure 2). The specificity of the immunostaining was confirmed by use of several antibodies to ATIII (Table 1) and immunoblotting experiments (Figure 3). Whereas all antibodies revealed staining, two of the antibodies gave positive staining only when frozen sections were used. The most useful antibodies were a monoclonal (AT111-1) and a polyclonal antibody (ATIII-2) that consistently identified NFT and amyloid plaques. The polyclonal ATIII-2, however, often exhibited vascular staining and produced a higher background (Figure 1C). Control experiments in the absence of primary antibody, with preimmune sera or with ascites or irrelevant antibodies showed no specific staining. Moreover, staining was markedly diminished upon absorption of the antibody with ATIII (not shown). Immunoblotting experiments showed that ATIIIlike immunoreactive proteins were present in cerebral cortex and cerebral microvessels and in immunoprecipitates of cortical homogenates (Figure 3). Under reducing conditions, ATIII antibodies recognized protein(s) of approximately 55 to 58 kd in sol-

Figure 2. Immunocytochemical localization of ATIJI in NFT by electron microscopy. A: Intraneuronal (note nucleus and cytoplasm) filaments stained by MAb ATmI (immunoperoxidase). B: Higher power vieuw of the NFT shoun in A in which the paired helical structure (arrowhead) is clearly evident. Scale bars, A = 2.5,u, B = 0.25 y.

890

Kalaria et al

AJP September 1993, Vol. 143, No. 3

A mwa

06. 1..06. *80. *-

B.

C D E

F

s: :r l:

.. .....fi

'.....

."....

fs

*_ ......

--

j-.

-

....

,

50 33.: Figure 3. Immunoblots of proteins extracted or immunoprecipitated from temporal cortex and cerebral microvessels. All samples analyzed were from AD subjects. Antibodies to ATII detected -58-kd proteins (arrow) in a purified ATIIIpreparation (lane A), solubilized cerebral microvessels (lane B), cerebral cortex (lane C), and ATIJI immunoprecipitate from a cortical bomogenate (lane D). Specificity of the antibody (ATIII-3 rabbit polyclonal) was sbown in blots immunoreacted with absorbed antibody (lane E) and irrelevant antibody, to a actin (lane F). MW markers shown in kd.

ubilized fractions of the cortex and microvessels. This was consistent with the electrophoretic mobility of authentic ATIII run in parallel (Figure 3). In a few cortical samples, we also observed variable immunoreactivity in higher mass proteins, which presumably reflects ATIII-protease complexes (Figure 3, lane C). Further characterization of these complexes was not undertaken. These immunoblotting results were in agreement with previous reports on ATIII purified from human plasma.26 We also immunoprecipitated ATIII from a small number of AD and non-AD control samples chosen at random (Figure 4). These results showed sample-to-sample variation in ATIII protein; however, specimens from AD

subjects generally showed more immunoprecipitable ATIII compared to the controls (Figure 4). Moreover, the intensity of immunoreactivity seemed to be related to presence of astrogliosis as evident from GFAP-stained sections (not shown). In view of the possibility that ATIII present in residual plasma within vessels of the tissue could contribute to the sample variability, further analysis of the ATIII immunoprecipitates was not performed. To determine whether ATIII might be derived from a local source in brain or the circulation, we searched for its mRNA using PCR to amplify reverse transcribed RNA.25 As expected, results revealed that the liver exhibited the highest amount of ATIII mRNA of all tissues examined. Low-level expression was evident in the kidney, but there was no detectable expression in spleen (Figure 5) or the cerebral meninges (not shown). However, more interestingly, ATIII mRNA was expressed in low levels in both grey and white matter of normal human brain (Figure 5). Using pooled complementary DNAs from 10 to 12 AD and 10 to 12 age-matched control samples, we detected increased amounts of ATIII mRNA in both grey and white matter (Figure 5). Similar results were evident in two other experiments, with more consistent change in the AD white matter compared to the grey. We also attempted quantitation of the PCR-amplified ATIII mRNA relative to RPS6 mRNA, used as a control index, in individual brain samples from AD and controls. These results showed that although the mean ratio of ATIII/ RPS6 for AD samples was about 2.5 times higher compared to that of controls (Figure 6) the difference between mean ratios did not reach statistical significance at the 5% level (P = 0.058, Student's two-tailed t-test). This was presumably due to large

Figure 4. ATIII immunoprecipitates from bomogenates of occipital cortex from AD and non-AD control subjects. ATII-reactive proteins (arrow) were immunoprecipitated with the polyclonal ATII rabbit 1 antibody ( Table 1) and detected with ATII MAb 2. Other ATII antibodies gave similar results (not shown). Samples in lanes 1, 2, and 4 to 6 uere from AD subjects and those in 3 and 7 to 9 were non-AD controls. Sample 7 was from

a subject diagnosed as non-AD with some astrogliosis. The intensity of ATII immunoreactivity in e-ach sample was qualitative/v related to the degree of Alzbeimer pathology and GFAP immunostaining in sections from the same cases. MW markers shown in kd.

ATIII in the Pathology of AD

891

AJP September 1993, Vol. 143, No. 3

K. S

L

WM GREY C AD: C AD

0.60 0

a. _

0.40 0.20

ATI:II 0.00 CtrI AD Ctrl AD Individual Pooled of ATIJI/PS6 mR/VA expression in individual and

Figure 6. Ratios pooled brain samples. For individual analysis, results re-present mean ± SEMforfive controls and five AD samples. The mean ratios were compared by Student's two tailed t-test, which showed differences u'ere not signifi cant at P < 0.05, giving a P value of 0.058. For pooled analysis run in parallel, ratios represent values obtained for 10 to 12 pooled samples of white matter from controls and AD. The re-lative ratios for samples of liver, kidney,~and spleen from controls w'ere 36.31, 0.24, and 0.0, respectively.

RPS6 Figure 5. Expression of ATII and RPS6 mRNAs in human brain and

peripheral organs. A77I mRNA was readily evident in liver (L) and kidney (K) but tIot itn spleen (5). In brain, ATII mRNA expression was detected in both uhite ( WM) and grey matter. In pooled samples from 11 to 12 controls and AD, ATIJI mRNA was increased in both white and grey matter in AD compared to controls. However, relative to RPS6, although the uhite matter showed increased ATIII/RPS6 ratio in AD compared to controls, this uas not evident in the grey matter.

sample-to-sample variation in the control group. In sum, the PCR experiments showed ATIII mRNA expression to be increased in samples from AD subjects compared to age-matched controls, but the change was most evident in the white matter.

Discussion Our findings indicate the presence of ATIII in normal human brain and in NFT and amyloid deposits that typically characterize AD pathology. These observations would suggest the release of ATIII, along with other serine protease inhibitors,4 presumably in response to proteolytic enzymes liberated from degenerating elements during the pathogenesis of

AD. Previous attempts to demonstrate ATIII in lesions associated with AD were unsuccessful. 1-3,28 The disparity in our findings and those of previous studies is possibly due to differences in tissue processing and use of different antibodies. We also found that some of the antibodies used in our studies were not readily reactive and only weakly active in acetone-fixed frozen sections. Here, we used a number of antibodies with immunocytochemistry on free-floating tissue sections and avoided the use of conventionally fixed paraffin-embedded tissue. In an attempt to localize the source of ATIII in brain, we used PCR to examine mRNA expression. We showed that the human brain not only expresses ATIII mRNA, but the expression levels are also apparently increased in AD, particularly in the white matter, as evident from the pooled samples. It is reasonable to suggest that the source of ATIII mRNA in the white matter, which contains essentially no neuronal RNA, is glial cells. The possibility

Table 2. Plasma Serine Protease and Other Protease Inhibitors in AD

Inhibitor protein al -antichymotrypsin ca l-antitrypsin ATIII Protease nexin Inter al-trypsin inhibitor a2-macroglobulin Protease nexin 11 (PN II) Factor Xla-inhibitor (same as PN II) Serum amyloid P Plasma concentrations nd: not determined.

Plasma

Mr

concn.

(kd)

0.25 1.3 0.2

68 53 58 43 180 725 110 112 230

0.50 2.5 0.04

Target protein(s)

Chymase

Elastase Thrombin Thrombin Trypsin Various Various Factor Xla Elastase

mRNA in brain

References 1,4

nd

4

This study nd

nd

(g/l) and molecular mass (Mr) where known are provided. Detection of mRNA in brain:

14 36 37 38 39 27 +: indicates yes; -: no; and

892 Kalaria et al AJP September 1993, Vol. 143, No. 3

that ATIII mRNA found in the brain could have originated from blood cells remnant in RNA-extracted tissue is also negated by the observation that there was no detectable expression of ATIII mRNA in the spleen or the meninges. The most likely candidate cells among glia that express ATIII mRNA are astrocytes. This is consistent with recent studies showing the expression of ATIII mRNA in astroglial cultures29 and our observations of ATIII reactivity in astrocytes and immunoprecipitated proteins. It is likely that increased ATIII mRNA is commensurate with the astrocytic response that is prominent in the ensuing pathology of AD.3>32 Fundamentally, if ATIII is produced by astrocytes, then our studies implicate that these cells contain a large repertoire of inhibitor molecules (see ref. 33 and Table 2). The inhibitors are secreted presumably to neutralize the actions of proteases that might be released as a result of neuronal or vascular injury. It is further possible that the protease inhibitors are taken up retrogradely by neurons,4 which may explain ATIII immunoreactivity in NFT within neurons. Nevertheless, these considerations on origin of ATIII do not preclude the possibility that at least some of the protein within the parenchyma could be derived from the circulation34 or produced by neurons.12 Our results are of relevance to the localization of PNI, an antithrombin-like molecule,14 a1-antichymotrypsin, 1-4 and more recently a1antitrypsin4 in AD brain, which all have serpin sequences. It is intriguing that recent studies not only demonstrate their localization in normal brain but also a number of the plasma protease inhibitors, including the serpins8 and nexins, seem to be involved in the pathogenesis of AD (Table 2). However, the exact role(s) of these protease inhibitors,3S39 including ATIII, in the neurofibrillary pathology of AD largely remain unclear. It is also noteworthy that the mRNAs of many of these inhibitors, albeit at low levels (Golde et al, unpublished observations), should be found to be expressed in brain tissue (Table 2). This indicates that, as in the periphery, the brain has a remarkable capacity to synthesize molecules associated with the acute phase response35 or processes analogous to chronic inflammation in brain.

Acknowledgments We thank Peggy Richey and Andrea Pax for help with the photography and technical assistance. We also wish to thank Drs. Kim and Wisniewski (IBR, New York) for the provision of MAb 4G8.

References 1. Abraham CR, Selkoe DJ, Potter H: Immunocytochemical identification of the serine protease inhibitor a1antichymotrypsin in the brain amyloid deposits of Alzheimer's disease. Cell 1988, 52:487-501 2. Pasternack JM, Abraham CR, Van Dyke BJ, Potter H, Younkin SG: Astrocytes in Alzheimer's disease grey matter express alphal-antichymotrypsin. Am J Pathol 1989, 135:827-834 3. Picken MM, Larrando-Lillo M, Coria F, Gallo GR, Shelanski ML, and Frangione B: Distribution of the protease inhibitor a1-antichymotrypsin in cerebral and systemic amyloid. J Neuropathol Exp Neurol 1990, 49: 41-48 4. Gollin PA, Kalaria RN, Eikelenboom P, Rozemuller A, and Perry G: al-antitrypsin and al-antichymotrypsin are in the lesions of Alzheimer disease. Neuroreport 1992, 3:201-203 5. Kitaguchi N, Takahashi Y, Tokushima Y, Shiojiri S, and Ito H: Novel precursor of Alzheimer's disease amyloid protein shows protease inhibitory activity. Nature 1988, 331:530-532 6. Ponte P, Gonzalez-DeWhitt P, Schilling J, Miller J, Hsu D, Greenberg B, Davis K, Wallace W, Lieberburg I, Fuller F, and Cordell B: A new A4 amyloid mRNA contains a domain homologous to serine proteinase inhibitors. Nature 1988, 331:525-527 7. Tanzi RE, McClatchey Al, Lamperti ED, Villa-Komaroff L, Gusella JF, and Neve RL: Protease inhibitor domain encoded by an amyloid protein precursor mRNA associated with Alzheimer's disease. Nature 1988, 331: 528-530 8. Carell RW: Enter a protease inhibitor. Nature 1988, 331:478-479 9. Dihanich M, Kaser M, Reinhard E, Cunningham D, Monard D: Prothrombin mRNA is expressed by cells of the nervous system. Neuron 1991, 6:575-581 10. Monard D: Cell-derived proteases and protease inhibitors as regulators of neurite outgrowth. Trends Neurosci 1988, 11:541-544 11. Farrell DH, Cunningham DD: Human fibroblasts accelerate the inhibition of thrombin by protease nexin. Proc Natl Acad Sci USA 1986, 83:6858-6862 12. Wagner SL, Lau AL, Nguyen A, Mimuro J, Loskutoff DJ, Isackson PJ, Cunningham D: Inhibitors of urokinase and thrombin in cultured neural cells. J Neurochem 1991, 56:234-242 13. Rosenblatt DE, Geula C, Mesulam M-M: Protease nexin immunostaining in Alzheimer's disease. Ann Neurol 1989, 26:628-634 14. Choi BH, Suzuki M, Taiseung K, Wagner SL, Cunningham DD: Protease nexin-1. Localization in the human brain suggests a protective role against extravasated serine proteases. Am J Pathol 1990, 137:741-747 15. Wagner SL, Geddes JW, Cotman CW, Lau AL, Gurwitz D, Isackson PJ, Cunningham D: Protease nexin-1, an

ATIII in the Pathology of AD

893

AJP September 1993, Vol. 143, No. 3

16. 17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

antithrombin with neurite outgrowth activity, is reduced in Alzheimer disease. Proc Natl Acad Sci USA 1989, 86:8284-8288 Carrell R, Travis J: a1-antitrypsin and the serpins: variation and countervariation. TIBS 1985, 7:20-24 Chandra T, Stackhouse R, Kidd VJ, Robson JH, Woo SLC: Sequence homology between human alantichymotrypsin, al-antitrypsin, and antithrombin ll. Biochemistry 1983, 22:5055-5061 Hill RE, Shaw PH, Boyd PA, Baumann H, Hastie ND: Plasma protease inhibitors in mouse and man: divergence within the reactive centre regions. Nature 1984, 311:175-177 Harpel PC: Blood proteolytic enzyme inhibitors: their role in modulating blood coagulation and fibrinolytic pathways. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Edited by Colman RW, Hirch J, Marder VJ, Salzman EW. New York, Lippincott, 1987, pp 219-267 Gurwitz D, Cunningham DD: Thrombin modulates and reverses neuroblastoma neurite outgrowth. Proc Natl Acad Sci USA 1988, 85:3440-3444 Kalaria RN, Galloway PG, Perry G: Widespread amyloid P component immunoreactivity in cortical amyloid deposits of Alzheimer's disease and other degenerative disorders. Neuropathol Appl Neurobiol 1991, 17: 189-201 Weidemann A, Konig G, Bunke D, Fischer P, Salbaum ML, Masters CL, Beyreuther K: Identification, biogenesis, and localization of precursors of Alzheimer's disease A4 amyloid protein. Cell 1989, 57:115-126 Perry G, Rizzuto N, Autilio-Gambetti L, Gambetti P: Paired helical filaments from Alzheimer disease patients contain cytoskeletal components. Proc NatI Acad Sci USA 1985, 82:3916-3920 Perry G, Mulvihill P, Manetto V, Autilio-Gambetti L, Gambetti P: Immunocytochemical properties of Alzheimer straight filaments. J Neurosci 1987, 7:3736-3742 Golde TE, Estus S, Usiak M, Younkin LH, and Younkin SG: Expression of f amyloid protein precursor mRNAs: recognition of a novel alternatively spliced form and quantitation in Alzheimer's disease using PCR. Neuron 1990, 4:253-267 Prochownik EV, Markham AF, and Orkin SH: Isolation of a cDNA clone for human antithrombin ll. J Biol Chem 1983, 258:8389-8394

27. Kalaria RN, and Kroon SN: Serum proteins in neurofibrillary pathology of Alzheimer's disease: antithrombin III-like immunoreactivity. Soc Neurosci Abstr 1991, 17:725 28. Rozemuller JM, Abbink JJ, Kamp AM, Stam FC, Hack CE, and Eikelenboom P: Distribution pattern and functional state of a1-antichymotrypsin in plaques and vascular amyloid deposits in Alzheimer's disease. Acta Neuropathol 1991, 82:200-207 29. Deschepper CF, Bigornia V, Berens ME, and Lapointe MC: Production of thrombin and antithrombin IlIl by brain and astroglial cell cultures. Mol Brain Res 1991, 11:355-358 30. Duffy PE, Rapport M, and Graf L: Glial fibrillary acidic protein and Alzheimer-type senile dementia. Neurology 1990, 30:778-780 31. Delacourte A: General and dramatic glial reaction in Alzheimer brains. Neurology 1990, 40:33-37 32. Frederickson RA: Astrocytes in Alzheimer's disease. Neurobiol Aging 1992, 14:25-39 33. Stornetta KL, Hawelu-Johnson CL, Guyenet PG, Lynch KR: Astrocytes synthesize angiotensinogen in brain. Science 1988, 242:1444-1446 34. Kalaria RN, Grahovac I: Serum amyloid P immunoreactivity in hippocampal tangles, plaques and vessels: implications for leakage across the blood-brain barrier in Alzheimer's disease. Brain Res 1990, 516:349-353 35. Kalaria RN: Serum amyloid P and related molecules associated with the acute-phase response in Alzheimer's disease. Res Immunol 1992, 143:637-641 36. Yoshida E, Yoshimura M, Ito Y, Mihara H: Demonstration of an active component of Inter-a-trypsin inhibitor in the brains of Alzheimer type dementia. Biochem Biophys Res Commun 1991, 174:1015-1021 37. Bauer J, Strauss S, Schreiter-Gasser U, Ganter U, Schlegel P, Witt I, Yolk B, Berger M: Interleukin-6 and a-2-macroglobulin indicate an acute-phase state in Alzheimer's disease cortices. FEBS Letts 1990, 285: 111-114 38. Van Nostrand WE, Wagner SL, Suzuki M, Choi BH, Farrow JS, Geddes JW, Cotman CW, Cunningham DD: Protease nexin-l1, a potent antichymotrypsin, shows identity to amyloid 13-protein precursor. Nature 1989, 341:546-549 39. Smith RP, Higuchi DA, Broze GJ: Platelet coagulation factor Xla-inhibitor, a form of Alzheimer amyloid precursor protein. Science 1990, 248:1126-1128