- Email: [email protected]
Perforant Path Transection Induces Complement C9 Deposition in Hippocampus
EXPERIMENTAL NEUROLOGY ARTICLE NO.
138, 198–205 (1996)
0058
Perforant Path Transection Induces Complement C9 Deposition in Hippocampus STEVEN A. JOHNSON,1 CISSY S. YOUNG-CHAN,2 NICHOLAS J. LAPING,3
AND
CALEB E. FINCH4
Neurogerontology Division, Andrus Gerontology Center and Department of Biological Sciences, University of Southern California, Los Angeles, California 90089-0191
INTRODUCTION The presence of complement system proteins in amyloid plaques and the up-regulation of several complement mRNAs in neurons and glial cells in affected brain regions during Alzheimer disease (AD) provided a basis for further examination of complement protein expression in a rodent lesion model of AD. Perforant path transection in rats was used as a model for the degeneration of entorhinal cortex (EC) layer II neurons and the consequent deafferentation of the hippocampus that occurs during AD. Immunostaining for C9, a key terminal component of the complement cascade membrane attack complex (MAC), showed extracellular C9 deposition in parenchyma around the EC wound and in hippocampus as early as 1 day, and disappeared by 14 days postlesion. Apoptosis of EC layer II neurons was seen and was presumably due to severing of their axonal projections to the hippocampus by the transection lesion. However, apoptotic EC layer II neurons were not immunostained by anti-rat C9 antibody, suggesting complement was not involved in inducing apoptosis. In the deafferented hippocampus, extracellular C9 immunostaining was localized to the dentate gyrus middle molecular layer, a region of synaptic loss, dendritic degeneration, and early synaptogenesis. In addition, intracellular C9 immunostaining was seen only in select hippocampal interneurons. Dentate gyrus granule neurons and pyramidal neurons were not C9 immunostained. Clusterin (SGP-2), a soluble inhibitor of the MAC that is up-regulated in AD, was also detected in the wound area (extracellular), the dentate gyrus middle molecular layer (extracellular), and intracellularly in scattered hippocampal interneurons. The data support the hypothesis that the complement system generally participates in responses to brain injury, as well as in AD. r 1996 Academic Press, Inc.
1 Present address: Cortex Pharmaceuticals, Inc., 15241 Barranca Parkway, Irvine, CA 92718. 2 Present address: Neuroscience program, Columbia University, New York, NY 10032. 3 Present address: Smithkline Beecham Pharmaceuticals, 709 Swedeland Rd., King of Prussia, PA 19406-0939. 4 To whom correspondence should be addressed.
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Activated classical complement (C) cascade components, including the membrane attack complex (MAC), and known classical C cascade inhibitors, such as clusterin and C4 binding protein, are detected by immunocytochemistry in neuritic plaques, dystrophic neurites, and neurofibrillary tangle-bearing neurons in Alzheimer disease (AD) (9, 19, 21, 26, 30). The C system consists of more than 25 proteins that are major inflammatory mediators in nonneural tissues (21, 27). C1q is the initiator of the classical C pathway and, while normally activated by substrate-bound immunoglobulin, can also be activated by aggregated amyloid b (Ab) peptide (26), the major component of amyloid in neuritic plaques in AD. Activation of the C cascade releases chemotactic peptides that attract and activate macrophages for enhanced phagocytosis and the production of proinflammatory cytokines and cytotoxic reactive oxygen species. The C cascade may proceed to the formation of the cytolytic MAC, a pore-forming heteromultimer containing C5b, C6, C7, C8, and multiple C9 monomers. Clusterin, also known as apoJ, CLI, SGP-2, SP-40,40, or TRPM-2 (15, 20), can inhibit formation of the MAC (15, 27) by complexing with C5b-7, thereby preventing addition of C8 or C9 necessary for MAC function. Brain clusterin has a slightly smaller apparent size than serum clusterin on Western blots, but is indistinguishable in activity as a MAC inhibitor or in slowing the aggregation of synthetic Ab (22a). C system proteins are secreted into the blood by liver and circulating cells, but are also made by fibroblasts and macrophages in response to injury. The presence of C system proteins in the AD brain was not anticipated because of the blood–brain barrier. To identify potential sources, we have localized several C system mRNAs in brain cells. By combined immunocytochemistry (ICC)/in situ hybridization, hippocampal pyramidal neurons were shown to contain C4 mRNA, and microglia shown to contain C1qB mRNA (17, 23, 28). Other C components are produced by cultured astrocytes and microglia (12). Furthermore, we and others have detected mRNAs for many classical C
198
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
system components in mammalian brain RNA by either Northern blot or reverse transcriptase–PCR analyses (17, 31). New evidence (28; these studies) that demonstrates intracellular localization of select classical C system proteins (C1q, C9, clusterin) in lesioned brain suggests that brain C system mRNAs are functional. Although early immunochemical studies found the MAC in dystrophic neurons of AD brains (21), this has not been extended with other anti-MAC antibodies (30). No C9 immunostaining is indicated for senile plaques that immunostain for C1q, other early classical pathway proteins (19, 21, 30), or clusterin (22, 30). Transection of the perforant path from the entorhinal cortex to the molecular layer of the dentate gyrus in rodents is a useful model for select features of AD pathology in rodent brain (3, 10, 17). In AD brain, entorhinal cortex layer II neurons form neurofibrillary tangles and degenerate early in the course of the disease (14), removing a significant part of the excitatory cortical input through the perforant path to the hippocampus. Entorhinal cortex layer II neurons send afferents that terminate in the outer two-thirds of the dentate gyrus molecular layer; their neurodegeneration results in a large reduction of synaptic density in the middle and outer dentate molecular layers, which is also a common site of neuritic plaques in AD (3, 4, 10, 14, 29). The dentate gyrus inner molecular layer, which lacks perforant path innervation but receives subcortical and commissural/associational (C/A) projections, expands by sprouting in response to loss of the entorhinal cortex afferents (4, 10). Perforant path transection in rats also results in degeneration of a subset of entorhinal cortex layer II neurons (25) and initiates synaptic reorganization in the middle and outer dentate molecular layers, as shown by electrolytic lesions of the entorhinal cortex (6, 10, 29). Studies have shown that projections from the medial portion of the entorhinal cortex terminate in the middle molecular layer, while projections from the lateral portion terminate in the outer molecular layer (reviewed in 29). By comparison, the inner dentate molecular layer is not deafferented by perforant path transection and, as in AD hippocampus, expands by compensatory sprouting of C/A afferents (3, 6, 10). Perforant path lesions, as well as excitotoxin lesions, induce C1q and C4 mRNAs in rat brain (17, 23, 28). The present studies examine in detail the immunolocalization of C9 after perforant path lesions. We include observations on clusterin immunolocalization. METHODS
Perforant Path Lesion Fischer 344 male rats (3 months) (Charles River) were fed ad libitum and housed (five/cage) with a 12 h day/12 h night cycle. Rats were anesthetized with
199
Nembutal (30 mg/kg, ip) prior to placement in stereotaxic frame for perforant path transection (6). A Scouten retractable wire knife (coordinates: 11 mm anterior, 5.7 mm lateral, 21 to 24 mm from dura, with respect to lambda) was used to cut the left perforant path in two vertical swaths with a 2-mm knife extension, at 135° and 300° (0°, anterior). Sham lesions included anesthetization, placement in stereotaxic frame, and drilling of access hole in the skull, but did not include placement of the retractable knife. Initially, 15 rats were lesioned, allowed to survive 1, 2, or 4 days (n 5 5 per group), and sacrificed by decapitation; the brains were quick-frozen on powdered dry ice and cryosliced; 5 sham-lesioned rats provided controls. Since on-slide immunohistochemistry with frozen sections gave poor-quality immunostaining, we lesioned 5 additional rats and perfusion fixed the brains at 3 days postlesion as described below. For a time-course study 20 lesioned rats were randomly assigned to four groups, 3, 7, 14, and 28 days postlesion, with 5 sham-lesioned rats as controls. At the appropriate postlesion time, rats were given an overdose of Nembutal (60 mg/kg), and brains were perfusion fixed with 3% buffered (0.1 M Na/K phosphate) paraformaldehyde/0.2% glutaraldehyde (after saline perfusion) through the left aorta for immunocytochemistry. Brains were further fixed for 6–12 h, immersed in 15 and 30% sucrose/0.1 M PB at 4°C, and sectioned at 30 µm with a sliding microtome. Immunocytochemistry Rabbit anti-rat C1q (1/500) and anti-rat C9 (1/2000) antisera were gifts of Drs. Sara Piddleston and B. Paul Morgan (University of Wales, Cardiff, UK). The antirat C9 antibody identified the expected single 70-kDa band on Western blots of rat serum (T. Oda and C. Finch, unpublished); this polyclonal antibody does not distinguish the C5b-9 MAC from C9. Rabbit anti-rat clusterin (SGP-2; 1/400) was a gift from Dr. Michael Griswold (Washington State University). Other antisera were mouse anti-human GFAP (1/400) from Boehringer-Mannheim (Indianapolis, IN) and goat antihuman C1q (1/3000) from Sigma (C3900) (St. Louis, MO). Thirty-micrometer free-floating sections were rinsed twice in PBS, blocked for 2 h in 0.1 M lysine, 0.2% Triton X-100, 4% normal serum in PBS, incubated in primary antibody (diluted in PBS, 4% serum) at 4°C overnight, and rinsed in PBS (four changes). Biotinylated secondary antibodies were preadsorbed with sera from rat and several mammals by the manufacturer (Jackson Immunoresearch, West Grove, PA). The remaining steps used the Vectastain ABC kit according to directions by Vector Labs (Burlingame, CA) and used diaminobenzidine (DAB) for detection. In some cases the DAB stain was enhanced by 0.5% nickelous ammonium sulfate, to produce a purple DAB reaction prod-
200
JOHNSON ET AL.
uct. Sections were mounted on slides, dried, and stained with cresyl violet. Photomicrographs were taken with a Zeiss Photomicroscope III. RESULTS
C9 Figure 1 shows the hippocampus at 3 days postlesioning, with reference staining for cell distribution by cresyl violet (Fig. 1A) and immunostaining for glial fibrillary acidic protein (GFAP) (Fig. 1B), synaptophysin (Fig. 1C), clusterin (Fig. 1D), C1q (Fig. 1E), and C9 (Fig. 1F). In the ipsilateral hippocampus, extracellular C9 staining was more intense in the middle third of the dentate gyrus molecular layer than in the inner or outer thirds (Fig. 1F): contralateral hippocampus or entorhinal cortex were not immunostained with this anti-rat C9 polyclonal antibody. C9 immunostaining of the middle molecular layer and stratum lacunosum was detected as early as 1 day, was strongest at 3 and 7 days, and essentially disappeared by 14 days. The temporal rise and fall of C9 immunostaining was similar in hippocampus and subiculum, but persisted through the 14-day time point in the wound area of the entorhinal cortex. Initial C9 ICC studies with frozen sections included 1-, 2-, and 4-day time points, but gave poor immunostaining and are not shown. Subsequent studies with perfusion-fixed brains included 3, 7, 14, and 28 days postlesioning; only data at 3 days are shown because all time points examined were qualitatively similar and the C9 immunostaining was strongest at 3 days. The middle molecular layer (and stratum lacunosum) had a notable loss of synaptophysin immunoreactivity at 3 days (Fig. 1C), consistent with synapse loss and dendritic regression in this layer (3). These same areas also showed notable enhancement of GFAP immunoreactivity (Fig. 1B), showing astrocyte hypertrophy in the deafferented areas. Controls for specificity of all antibodies, including anti-rat C9, included the use of secondary antibodies that were preadsorbed with rat IgG by the manufacturer (Jackson Immunoresearch), as well as omission of primary and secondary antibodies; no immunostaining was observed in any instance. Besides the hippocampus, extracellular C9 immunostaining was present throughout the entorhinal cortex and subiculum and was particularly intense around the transection wound site (not shown). The contralateral entorhinal cortex and hippocampus were not stained. Adjacent to the wound, we observed numerous intensely stained cells with morphology similar to that of macrophage/activated microglia (not shown). C9 immunostaining was also seen at intracellular locations. In the ipsilateral hippocampal formation, scattered C9-immunopositive cells with neuronal morphology were located outside of the dentate granule or
pyramidal cell layers (Fig. 2A). C9-immunopositive neurons were never seen in the densely packed dentate granule cell or pyramidal cell layers ipsilateral to the lesion. Figure 2B shows the absence of C9 immunostaining in the area around (or within) the pyramidal neuron layer in field CA1 contralateral to the lesion. In the scattered C9-immunopositive cells, immunostaining was prominent in the perikarya, but was also present in extended neuronal processes (Fig. 2C). Overall, C9-immunopositive neurons were located at the hilar edge of the dentate granule layer, in the hilus, and adjacent to the pyramidal cell fields from CA3 to subiculum in the ipsilateral hippocampal formation (not shown). This broad nonpyramidal localization is consistent with that of various subclasses of hippocampal interneurons (5, 7, 8) that generally serve to modulate the excitatory tone of the hippocampus. At 3 days postlesion, degenerating neuronal nuclei were seen in superficial layers of ipsilateral entorhinal cortex (Figs. 3A and 3B), consistent with retrograde degeneration of layer II neurons after perforant path lesions (25). These nuclei had the morphological characteristics of apoptotic nuclei, including nuclear compaction and fragmentation, as seen at high power in Fig. 3B. However, these degenerating neurons were clearly not immunostained for C9 (Fig. 3B), suggesting that C9 is not involved in initiating apoptosis of layer II neurons after perforant path transection. Clusterin At 3 days postlesioning, extracellular clusterin immunostaining was also prominent in the ipsilateral dentate middle molecular layer and the stratum lacunosum (Fig. 1D). GFAP was present in a dense network of hypertrophic astrocytes in these same areas (Fig. 1B). In addition, intracellular clusterin immunostaining was detected in hippocampal interneurons in the same general anatomical locations as the C9-immunopositive interneurons; e.g., stratum radiatum or stratum lacunosum (Fig. 2D), dentate molecular layer, and/or hilus of the dentate gyrus. Like C9, clusterin was not detected in hippocampal pyramidal or dentate granule neurons. Unlike C9, clusterin was seen in astrocytes, and clusterin-containing astrocytes were found throughout the ipsilateral entorhinal cortex and hippocampus, irrespective of the location of extracellular clusterin (not shown). C1q Weak C1q immunostaining was detected in the hippocampal formation, including presubicular areas contiguous to the entorhinal cortex. C1q immunostaining was detected in the ipsilateral CA1 stratum lacunosum
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
201
FIG. 1. Immunocytochemistry for C9, clusterin (SGP-2), synaptophysin, C1q, and GFAP in dentate gyrus molecular layer 3d after perforant path transection. Each panel shows a portion of the ipsilateral hippocampus in the horizontal plane. All sections, except that in C, were counterstained with cresyl violet, and all represent semiadjacent sections. A shows cresyl violet stain of hippocampus ipsilateral to lesion. GFAP ICC (B) shows increased astrocyte reactivity in middle molecular layer (arrow, m) and stratum lacunosum (sl). C shows a reduction of synaptophysin immunoreactivity in the ipsilateral middle molecular layer (arrow, m) and stratum lacunosum, where in D and F, there is increased clusterin and C9 immunoreactivity, respectively. Clusterin (SGP-2) is an inhibitor of the formation of the complement C5b-9 MAC, of which C9 is a key component necessary for MAC function. Immunocytochemistry for C1q, the initiating component of the complement cascade pathway, showed weak, but detectable staining primarily in the distal stratum lacunosum. The arrow in B, C, D, and F points out the dentate gyrus middle molecular layer. Abbreviations: DG, dentate gyrus; i, dentate inner molecular layer; m, dentate middle molecular layer; o, dentate outer molecular layer; sl, stratum lacunosum. Size bar, 250 µm.
202
JOHNSON ET AL.
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
(Fig. 1E) compared to the contralateral side or cresyl violet-stained control sections (Fig. 1A). The dentate molecular layer lacked immunodetectable C1q. At 3 days postlesion, extracellular C1q was strongest directly adjacent to the wound in the entorhinal cortex, but declined with increasing distance from the wound. Intracellular C1q was confined to the wound site and adjacent areas, where we observed putative infiltrating blood cells and scattered macrophage/microglia (not shown). The contralateral side did not stain for C1q. DISCUSSION
The terminal components of the C cascade require C9 to form a fully functional MAC. The robust increases of C9 in the lesioned EC and deafferented hippocampus seen in this study may allow formation of the MAC. Neurons of three morphotypes showed C9 immunostaining: large process-bearing cells with pyramidal neuron morphology in ipsilateral EC layers II–IV and subiculum (not shown), cells on the inner border of the deafferented zone of the dentate gyrus (not shown), and multipolar cells in the CA1 stratum lacunosum moleculare and stratum radiatum (Fig. 2A). Near the EC wound cavity, cells with macrophage/amoeboid microglial morphology were also densely immunostained for C9. These findings are consistent with immunocytochemical observations of the hippocampus in AD, in which McGeer et al. (21) reported that anti-human C9 stained plaques ‘‘. . . along with tangles, neuropil threads and dystrophic neurites.’’ That photomicrograph could also be interpreted to show immunostaining of select neuronal perikarya, as observed here in a rat model of AD. The same report shows similar staining, except for plaques, by an anti-C5b-9 (anti-MAC) antibody (21), suggesting to those authors the presence of the lytic MAC on membranes of tangled neurons and dystrophic neurites in AD hippocampus. We do not know the identity or function of the C9-immunopositive neurons in the deafferented rat hippocampus shown in Figs. 2A and 2C. Since there is no C9 immunolabeling of CA field pyramidal or dentate gyrus granule neurons, we suggest that the C9immunopositive neurons represent one or more classes of interneurons. AD brains show a loss of neuropeptide
203
Y/somatostatin (NPY/SS) interneurons in the facia dentata of the hippocampus (5), i.e., the same region that shows C9 and clusterin-immunopositive cells with neuronal morphology in the present studies. The same NPY/SS interneurons might be compromised after perforant path transection (8) and subject to complement attack. Further evidence supports the concept that the C system has an active role during neurodegeneration. The phagocytosis of myelin by macrophages during responses to peripheral neuronal lesions is impaired when various C components are lacking or deficient (2). Because the C cascade is subject to regulation at many stages by C inhibitors, including clusterin (15, 27), it is pertinent to the argument that brain cells also make clusterin. Clusterin mRNA is induced in these same neuron types, as well as in the numerous astrocytes in the deafferented dentate middle molecular layer and stratum lacunosum (17, 24, 28). In AD, clusterin protein occurs in neuritic plaques (22). The secretion of clusterin by primary astrocyte cultures (24) suggests that the extracellular clusterin in the middle molecular layer of this study and in neuritic plaques of AD was secreted by local astrocytes. Strong, diffuse extracellular C9 immunostaining was found in the CA1 stratum lacunosum and the dentate gyrus middle molecular layer, while there was only weak staining in the outer or inner dentate molecular layers. This laminar restriction of C9 immunostaining correlates well with reactive synaptogenesis and increased dendritic arborization that begins first in the middle molecular layer between 2 and 4 days after lesioning (4, 13). It also correlates with the loss of synaptophysin immunoreactivity in the middle molecular layer. It is not entirely clear why synaptophysin was not reduced in the outer molecular layer; incomplete transection of the perforant path, especially the portion of the pathway from the lateral entorhinal cortex that terminates in the outer molecular layer, may be an explanation. We suggest that, along with a possible role in neurodegeneration, as was suggested previously for AD (21), the complement system may also play a role in repair mechanisms after brain injury.
FIG. 2. Intracellular immunolocalization of complement C9 and clusterin in hippocampal interneurons after perforant path transection. Three days after unilateral perforant path transection scattered interneurons that do not reside within the tightly packed pyramidal or granule cell layers are immunopositive for rat C9 ipsilateral (A), but not contralateral (B) to the transection lesion. Arrowheads identify several C9-immunopositive neurons in A that are located in stratum radiatum or stratum oriens. At higher magnification it is clear that both soma and proximal dendrites are labeled by the anti-rat C9 polyclonal antibody (C). Anti-rat clusterin polyclonal antibody also immunostains scattered interneurons located outside the pyramidal or granule cell layers of the hippocampal formation. Several clusterin-immunopositive neurons, located in the ipsilateral stratum lacunosum, are shown in D. The underlying parenchymal staining by anti-clusterin in D, which can be seen at lower power in Fig. 1D, is present only on the ipsilateral side. Abbreviations: sp, stratum pyramidale (pyramidal cell layer); sr, stratum radiatum. A and B are at same magnification; size bar, 50 µm. C and D are at same magnification; size bar, 20 µm.
204
JOHNSON ET AL.
FIG. 3. Apoptotic morphology of nuclei in degenerating entorhinal cortex layer II neurons. Pyknotic nuclei were present in entorhinal cortex layers II and III after perforant path transection. Such nuclei were fragmented and more intensely stained than normal nuclei and generally resembled the morphology of nuclei from cells undergoing apoptosis. A shows a portion of medial entorhinal cortex 3 days after the transection lesion; apoptotic nuclei are indicated by arrowheads. The area in the upper right corner of A enclosed by the dashed lines is shown at higher magnification in B. Note the densely stained, fragmented nuclear bodies that appear to be contained within the nuclear membrane. Also note that although the parenchyma was stained with anti-rat C9 polyclonal antibody, suggesting extracellular deposition of C9, apoptotic layer II neurons were not C9 immunopositive. V, large vessel. Size bars equal 20 and 10 µm in A and B, respectively.
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
ACKNOWLEDGMENTS The authors thank Mr. Chris Anderson for excellent technical assistance. We also thank Mr. Hank Hogan for excellent technical and photographic assistance. This work was supported by AG10673 (S.A.J.) and AG07909 (C.E.F.).
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
AUGER, M. J., AND J. A. ROSS. 1992. The biology of the macrophage. Pages 1–74 in C. E. Lewis and J. O. McGee, Eds., The Macrophage. IRL Press, Oxford. BRUCK, W., AND R. L. FRIEDE. 1991. The role of complement in myelin phagocytosis during PNS wallerian degeneration. J. Neurol. Sci. 103: 182–187. CABALKA, L. M., B. T. HYMAN, C. R. GOODLETT, T. C. RITCHIE, AND G. W. VAN HOESEN. 1992. Alteration in the pattern of nerve terminal protein immunoreactivity in the perforant pathway in Alzheimer’s disease and in rats after entorhinal lesions. Neurobiol. Aging 13: 283–291. CACERES, A., AND O. STEWARD. 1983. Dendritic reorganization in the denervated dentate gyrus of the rat following entorhinal cortical lesions: A golgi and electron microscopic analysis. J. Comp. Neurol. 214: 387–403. CHAN-PALAY, V. 1987. Somatostatin immunoreactive neurons in the human hippocampus and cortex shown by immunogold/ silver intensification on vibrotome sections: Coexistence with neuropeptide Y neurons, and effects in Alzheimer-type dementia. J. Comp. Neurol. 260: 201–223. CUMMINGS, B. J., G. J. YEE, AND C. W. COTMAN. 1992. bFGF promotes the survival of entorhinal layer II neurons after perforant path axotomy. Brain Res. 591: 271–276. DE LANEROLLE, N. C., J. H. KIM, R. J. ROBBINS, AND D. D. SPENCER. 1989. Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy. Brain Res. 495: 387–395. DELLER, T., AND C. LERANTH. 1990. Synaptic connections of neuropeptide Y (NPY) immunoreactive neurons in the hilar area of the rat hippocampus. J. Comp. Neurol. 300: 433–447. EIKELENBOOM, P., S-S. ZHAN, W. A. VAN GOOL, AND D. ALLSOP. 1994. Inflammatory mechanisms in Alzheimer’s disease. Trends Pharmacol. Sci. 15: 447–450. GEDDES, J. W., D. T. MONAGHAN, C. W. COTMAN, I. T. LOTT, R. C. KIM, AND H. C. CHUI. 1985. Plasticity of hippocampal circuitry in Alzheimer’s disease. Science 230: 1179–1181. GEHRMAN, J., S. W. SCHOEN, AND G. W. KREUTZBERG. 1991. Lesion of the rat entorhinal cortex leads to a rapid microglial reaction in the dentate gyrus. A light and electron microscopical study. Acta Neuropathol. 82: 442–455. HAGA, S., K. IKEDA, M. SATO, AND T. ISHII. 1993. Synthetic Alzheimer amyloid b/A4 peptides enhance production of complement C3 component by cultured microglial cells. Brain Res. 601: 88–94. HOFF, S. F., S. W. SCHEFF, AND C. W. COTMAN. 1982. Lesioninduced synaptogenesis in the dentate gyrus of aged rats: II. Demonstration of an impaired degeneration clearing response. J. Comp. Neurol. 205: 253–259. HYMAN, B. T., G. W. VAN HOESEN, A. R. DAMASIO, AND C. L. BARNES. 1984. Alzheimer’s disease: Cell-specific pathology isolates the hippocampal formation. Science 225: 1168–1170. JENNE, D. E., AND J. TSCHOPP. 1993. Clusterin: The intriguing guises of a widely expressed glycoprotein. Trends Biochem. Sci. 17: 154–159. JOHNSON, S. A., T. MCNEILL, B. CORDELL, AND C. E. FINCH. 1990.
17.
18.
19.
20.
21.
22.
22A.
23.
24.
25.
26.
27. 28.
29.
30.
31.
205
Relation of neuronal APP-751/APP-695 mRNA ratio and neuritic plaque density in Alzheimer’s disease. Science 248: 854– 857. JOHNSON, S. A., M. LAMPERT-ETCHELLS, G. M. PASINETTI, I. ROZOVSKY, AND C. E. FINCH. 1992. Complement mRNA in the mammalian brain: Responses to Alzheimer’s disease and experimental brain lesioning. Neurobiol. Aging 13: 641–648. JOHNSON, S. A., G. M. PASINETTI, AND C. E. FINCH. 1994. Expression of complement C1qB and C4 mRNAs during rat brain development. Dev. Brain Res. 80: 163–174. LAMPERT-ETCHELLS, M., G. PASINETTI, C. FINCH, AND S. JOHNSON. 1993. Regional localization of cells containing C1q and C4 mRNAs in the frontal cortex during Alzheimer’s disease. Neurodegeneration 2: 113–121. MAY, P. C., AND FINCH, C. E. 1992. Sulfated glycoprotein-2: New relationships of this multifunctional protein to neurodegeneration. Trends Neurosci. 15: 391–396. MCGEER, P. L., H. AKIYAMA, S. ITAGAKI, AND E. G. MCGEER. 1989. Immune system response in Alzheimer’s disease. Can. J. Neurol. Sci. 16: 516–527. MCGEER, P. L., T. KAWAMATA, AND D. G. WALKER. 1992. Distribution of clusterin in Alzheimer brain tissue. Brain Res. 579: 337–341. ODA, T., G. M. PASINETTI, H. H. OSTERBURG, C. ANDERSON, S. A. JOHNSON, AND C. E. FINCH. 1994. Purification and characterization of brain clusterin. Biochem. Biophys. Res. Commun. 204: 1131–1136. PASINETTI, G. M., S. A. JOHNSON, I. ROZOVSKY, M. LAMPERTETCHELLS, D. G. MORGAN, M. N. GORDON, T. E. MORGAN, D. WILLOUGHBY, AND C. E. FINCH. 1992. Complement C1qB and C4 mRNAs responses to lesioning in rat brain. Exp. Neurol. 118: 117–125. PASINETTI, G. M., S. A. JOHNSON, T. ODA, I. ROZOVSKY, AND C. E. FINCH. 1994. Clusterin (SGP-2): A multifunctional glycoprotein with regional expression in astrocytes and neurons of the adult rat brain. J. Comp. Neurol. 339: 387–400. PETERSON, D. A., C. A. LUCIDI-PHILLIPI, K. L. EAGLE, AND F. H. GAGE. 1994. Perforant path damage results in progressive neuronal death and somal atrophy in layer II of entorhinal cortex and functional impairment with increasing postdamage age. J. Neurosci. 14: 6872–6885. ROGERS, J., N. R. COOPER, S. WEBSTER, J. SCHULTZ, P. L. MCGEER, S. D. STYREN, W. H. CIVIN, L. BRACHOVA, B. BRADT, P. WARD, AND I. LIEBERBURG. 1992. Complement activation by beta-amyloid in Alzheimer disease. Proc. Natl. Acad. Sci. USA 89: 10016–10020. ROTHER, K., AND G. O. TILL, EDS. 1988. Page 535 in The Complement System. Springer-Verlag, Berlin. ROZOVSKY, I., T. E. MORGAN, D. A. WILOUGHBY, M. M. DUGICHDJORDEVICH, G. M. PASINETTI, S. A. JOHNSON, AND C. E. FINCH. 1994. Selective expression of clusterin (SGP-2) and complement C1qB and C4 during responses to neurotoxins in vivo and in vitro. Neuroscience 62: 741–758. STEWARD, O., AND S. L. VINSANT. 1983. The process of reinnervation in the dentate gyrus of the adult rat: A quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis. J. Comp. Neurol. 214: 370–386. VEERHUIS, R., P. VAN DER VALK, I. JANSSEN, S. S. ZHAN, W. E. VAN NOSTRAND, AND P. EIKELENBOOM. 1995. Complement activation in amyloid plaques in Alzheimer’s disease brains does not proceed further than C3. Virchows Arch. 426: 603–610. WALKER, D. G., AND P. L. MCGEER. 1992. Complement gene expression in human brain: Comparison between normal and Alzheimer disease cases. Mol. Brain Res. 14: 109–116.
138, 198–205 (1996)
0058
Perforant Path Transection Induces Complement C9 Deposition in Hippocampus STEVEN A. JOHNSON,1 CISSY S. YOUNG-CHAN,2 NICHOLAS J. LAPING,3
AND
CALEB E. FINCH4
Neurogerontology Division, Andrus Gerontology Center and Department of Biological Sciences, University of Southern California, Los Angeles, California 90089-0191
INTRODUCTION The presence of complement system proteins in amyloid plaques and the up-regulation of several complement mRNAs in neurons and glial cells in affected brain regions during Alzheimer disease (AD) provided a basis for further examination of complement protein expression in a rodent lesion model of AD. Perforant path transection in rats was used as a model for the degeneration of entorhinal cortex (EC) layer II neurons and the consequent deafferentation of the hippocampus that occurs during AD. Immunostaining for C9, a key terminal component of the complement cascade membrane attack complex (MAC), showed extracellular C9 deposition in parenchyma around the EC wound and in hippocampus as early as 1 day, and disappeared by 14 days postlesion. Apoptosis of EC layer II neurons was seen and was presumably due to severing of their axonal projections to the hippocampus by the transection lesion. However, apoptotic EC layer II neurons were not immunostained by anti-rat C9 antibody, suggesting complement was not involved in inducing apoptosis. In the deafferented hippocampus, extracellular C9 immunostaining was localized to the dentate gyrus middle molecular layer, a region of synaptic loss, dendritic degeneration, and early synaptogenesis. In addition, intracellular C9 immunostaining was seen only in select hippocampal interneurons. Dentate gyrus granule neurons and pyramidal neurons were not C9 immunostained. Clusterin (SGP-2), a soluble inhibitor of the MAC that is up-regulated in AD, was also detected in the wound area (extracellular), the dentate gyrus middle molecular layer (extracellular), and intracellularly in scattered hippocampal interneurons. The data support the hypothesis that the complement system generally participates in responses to brain injury, as well as in AD. r 1996 Academic Press, Inc.
1 Present address: Cortex Pharmaceuticals, Inc., 15241 Barranca Parkway, Irvine, CA 92718. 2 Present address: Neuroscience program, Columbia University, New York, NY 10032. 3 Present address: Smithkline Beecham Pharmaceuticals, 709 Swedeland Rd., King of Prussia, PA 19406-0939. 4 To whom correspondence should be addressed.
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Activated classical complement (C) cascade components, including the membrane attack complex (MAC), and known classical C cascade inhibitors, such as clusterin and C4 binding protein, are detected by immunocytochemistry in neuritic plaques, dystrophic neurites, and neurofibrillary tangle-bearing neurons in Alzheimer disease (AD) (9, 19, 21, 26, 30). The C system consists of more than 25 proteins that are major inflammatory mediators in nonneural tissues (21, 27). C1q is the initiator of the classical C pathway and, while normally activated by substrate-bound immunoglobulin, can also be activated by aggregated amyloid b (Ab) peptide (26), the major component of amyloid in neuritic plaques in AD. Activation of the C cascade releases chemotactic peptides that attract and activate macrophages for enhanced phagocytosis and the production of proinflammatory cytokines and cytotoxic reactive oxygen species. The C cascade may proceed to the formation of the cytolytic MAC, a pore-forming heteromultimer containing C5b, C6, C7, C8, and multiple C9 monomers. Clusterin, also known as apoJ, CLI, SGP-2, SP-40,40, or TRPM-2 (15, 20), can inhibit formation of the MAC (15, 27) by complexing with C5b-7, thereby preventing addition of C8 or C9 necessary for MAC function. Brain clusterin has a slightly smaller apparent size than serum clusterin on Western blots, but is indistinguishable in activity as a MAC inhibitor or in slowing the aggregation of synthetic Ab (22a). C system proteins are secreted into the blood by liver and circulating cells, but are also made by fibroblasts and macrophages in response to injury. The presence of C system proteins in the AD brain was not anticipated because of the blood–brain barrier. To identify potential sources, we have localized several C system mRNAs in brain cells. By combined immunocytochemistry (ICC)/in situ hybridization, hippocampal pyramidal neurons were shown to contain C4 mRNA, and microglia shown to contain C1qB mRNA (17, 23, 28). Other C components are produced by cultured astrocytes and microglia (12). Furthermore, we and others have detected mRNAs for many classical C
198
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
system components in mammalian brain RNA by either Northern blot or reverse transcriptase–PCR analyses (17, 31). New evidence (28; these studies) that demonstrates intracellular localization of select classical C system proteins (C1q, C9, clusterin) in lesioned brain suggests that brain C system mRNAs are functional. Although early immunochemical studies found the MAC in dystrophic neurons of AD brains (21), this has not been extended with other anti-MAC antibodies (30). No C9 immunostaining is indicated for senile plaques that immunostain for C1q, other early classical pathway proteins (19, 21, 30), or clusterin (22, 30). Transection of the perforant path from the entorhinal cortex to the molecular layer of the dentate gyrus in rodents is a useful model for select features of AD pathology in rodent brain (3, 10, 17). In AD brain, entorhinal cortex layer II neurons form neurofibrillary tangles and degenerate early in the course of the disease (14), removing a significant part of the excitatory cortical input through the perforant path to the hippocampus. Entorhinal cortex layer II neurons send afferents that terminate in the outer two-thirds of the dentate gyrus molecular layer; their neurodegeneration results in a large reduction of synaptic density in the middle and outer dentate molecular layers, which is also a common site of neuritic plaques in AD (3, 4, 10, 14, 29). The dentate gyrus inner molecular layer, which lacks perforant path innervation but receives subcortical and commissural/associational (C/A) projections, expands by sprouting in response to loss of the entorhinal cortex afferents (4, 10). Perforant path transection in rats also results in degeneration of a subset of entorhinal cortex layer II neurons (25) and initiates synaptic reorganization in the middle and outer dentate molecular layers, as shown by electrolytic lesions of the entorhinal cortex (6, 10, 29). Studies have shown that projections from the medial portion of the entorhinal cortex terminate in the middle molecular layer, while projections from the lateral portion terminate in the outer molecular layer (reviewed in 29). By comparison, the inner dentate molecular layer is not deafferented by perforant path transection and, as in AD hippocampus, expands by compensatory sprouting of C/A afferents (3, 6, 10). Perforant path lesions, as well as excitotoxin lesions, induce C1q and C4 mRNAs in rat brain (17, 23, 28). The present studies examine in detail the immunolocalization of C9 after perforant path lesions. We include observations on clusterin immunolocalization. METHODS
Perforant Path Lesion Fischer 344 male rats (3 months) (Charles River) were fed ad libitum and housed (five/cage) with a 12 h day/12 h night cycle. Rats were anesthetized with
199
Nembutal (30 mg/kg, ip) prior to placement in stereotaxic frame for perforant path transection (6). A Scouten retractable wire knife (coordinates: 11 mm anterior, 5.7 mm lateral, 21 to 24 mm from dura, with respect to lambda) was used to cut the left perforant path in two vertical swaths with a 2-mm knife extension, at 135° and 300° (0°, anterior). Sham lesions included anesthetization, placement in stereotaxic frame, and drilling of access hole in the skull, but did not include placement of the retractable knife. Initially, 15 rats were lesioned, allowed to survive 1, 2, or 4 days (n 5 5 per group), and sacrificed by decapitation; the brains were quick-frozen on powdered dry ice and cryosliced; 5 sham-lesioned rats provided controls. Since on-slide immunohistochemistry with frozen sections gave poor-quality immunostaining, we lesioned 5 additional rats and perfusion fixed the brains at 3 days postlesion as described below. For a time-course study 20 lesioned rats were randomly assigned to four groups, 3, 7, 14, and 28 days postlesion, with 5 sham-lesioned rats as controls. At the appropriate postlesion time, rats were given an overdose of Nembutal (60 mg/kg), and brains were perfusion fixed with 3% buffered (0.1 M Na/K phosphate) paraformaldehyde/0.2% glutaraldehyde (after saline perfusion) through the left aorta for immunocytochemistry. Brains were further fixed for 6–12 h, immersed in 15 and 30% sucrose/0.1 M PB at 4°C, and sectioned at 30 µm with a sliding microtome. Immunocytochemistry Rabbit anti-rat C1q (1/500) and anti-rat C9 (1/2000) antisera were gifts of Drs. Sara Piddleston and B. Paul Morgan (University of Wales, Cardiff, UK). The antirat C9 antibody identified the expected single 70-kDa band on Western blots of rat serum (T. Oda and C. Finch, unpublished); this polyclonal antibody does not distinguish the C5b-9 MAC from C9. Rabbit anti-rat clusterin (SGP-2; 1/400) was a gift from Dr. Michael Griswold (Washington State University). Other antisera were mouse anti-human GFAP (1/400) from Boehringer-Mannheim (Indianapolis, IN) and goat antihuman C1q (1/3000) from Sigma (C3900) (St. Louis, MO). Thirty-micrometer free-floating sections were rinsed twice in PBS, blocked for 2 h in 0.1 M lysine, 0.2% Triton X-100, 4% normal serum in PBS, incubated in primary antibody (diluted in PBS, 4% serum) at 4°C overnight, and rinsed in PBS (four changes). Biotinylated secondary antibodies were preadsorbed with sera from rat and several mammals by the manufacturer (Jackson Immunoresearch, West Grove, PA). The remaining steps used the Vectastain ABC kit according to directions by Vector Labs (Burlingame, CA) and used diaminobenzidine (DAB) for detection. In some cases the DAB stain was enhanced by 0.5% nickelous ammonium sulfate, to produce a purple DAB reaction prod-
200
JOHNSON ET AL.
uct. Sections were mounted on slides, dried, and stained with cresyl violet. Photomicrographs were taken with a Zeiss Photomicroscope III. RESULTS
C9 Figure 1 shows the hippocampus at 3 days postlesioning, with reference staining for cell distribution by cresyl violet (Fig. 1A) and immunostaining for glial fibrillary acidic protein (GFAP) (Fig. 1B), synaptophysin (Fig. 1C), clusterin (Fig. 1D), C1q (Fig. 1E), and C9 (Fig. 1F). In the ipsilateral hippocampus, extracellular C9 staining was more intense in the middle third of the dentate gyrus molecular layer than in the inner or outer thirds (Fig. 1F): contralateral hippocampus or entorhinal cortex were not immunostained with this anti-rat C9 polyclonal antibody. C9 immunostaining of the middle molecular layer and stratum lacunosum was detected as early as 1 day, was strongest at 3 and 7 days, and essentially disappeared by 14 days. The temporal rise and fall of C9 immunostaining was similar in hippocampus and subiculum, but persisted through the 14-day time point in the wound area of the entorhinal cortex. Initial C9 ICC studies with frozen sections included 1-, 2-, and 4-day time points, but gave poor immunostaining and are not shown. Subsequent studies with perfusion-fixed brains included 3, 7, 14, and 28 days postlesioning; only data at 3 days are shown because all time points examined were qualitatively similar and the C9 immunostaining was strongest at 3 days. The middle molecular layer (and stratum lacunosum) had a notable loss of synaptophysin immunoreactivity at 3 days (Fig. 1C), consistent with synapse loss and dendritic regression in this layer (3). These same areas also showed notable enhancement of GFAP immunoreactivity (Fig. 1B), showing astrocyte hypertrophy in the deafferented areas. Controls for specificity of all antibodies, including anti-rat C9, included the use of secondary antibodies that were preadsorbed with rat IgG by the manufacturer (Jackson Immunoresearch), as well as omission of primary and secondary antibodies; no immunostaining was observed in any instance. Besides the hippocampus, extracellular C9 immunostaining was present throughout the entorhinal cortex and subiculum and was particularly intense around the transection wound site (not shown). The contralateral entorhinal cortex and hippocampus were not stained. Adjacent to the wound, we observed numerous intensely stained cells with morphology similar to that of macrophage/activated microglia (not shown). C9 immunostaining was also seen at intracellular locations. In the ipsilateral hippocampal formation, scattered C9-immunopositive cells with neuronal morphology were located outside of the dentate granule or
pyramidal cell layers (Fig. 2A). C9-immunopositive neurons were never seen in the densely packed dentate granule cell or pyramidal cell layers ipsilateral to the lesion. Figure 2B shows the absence of C9 immunostaining in the area around (or within) the pyramidal neuron layer in field CA1 contralateral to the lesion. In the scattered C9-immunopositive cells, immunostaining was prominent in the perikarya, but was also present in extended neuronal processes (Fig. 2C). Overall, C9-immunopositive neurons were located at the hilar edge of the dentate granule layer, in the hilus, and adjacent to the pyramidal cell fields from CA3 to subiculum in the ipsilateral hippocampal formation (not shown). This broad nonpyramidal localization is consistent with that of various subclasses of hippocampal interneurons (5, 7, 8) that generally serve to modulate the excitatory tone of the hippocampus. At 3 days postlesion, degenerating neuronal nuclei were seen in superficial layers of ipsilateral entorhinal cortex (Figs. 3A and 3B), consistent with retrograde degeneration of layer II neurons after perforant path lesions (25). These nuclei had the morphological characteristics of apoptotic nuclei, including nuclear compaction and fragmentation, as seen at high power in Fig. 3B. However, these degenerating neurons were clearly not immunostained for C9 (Fig. 3B), suggesting that C9 is not involved in initiating apoptosis of layer II neurons after perforant path transection. Clusterin At 3 days postlesioning, extracellular clusterin immunostaining was also prominent in the ipsilateral dentate middle molecular layer and the stratum lacunosum (Fig. 1D). GFAP was present in a dense network of hypertrophic astrocytes in these same areas (Fig. 1B). In addition, intracellular clusterin immunostaining was detected in hippocampal interneurons in the same general anatomical locations as the C9-immunopositive interneurons; e.g., stratum radiatum or stratum lacunosum (Fig. 2D), dentate molecular layer, and/or hilus of the dentate gyrus. Like C9, clusterin was not detected in hippocampal pyramidal or dentate granule neurons. Unlike C9, clusterin was seen in astrocytes, and clusterin-containing astrocytes were found throughout the ipsilateral entorhinal cortex and hippocampus, irrespective of the location of extracellular clusterin (not shown). C1q Weak C1q immunostaining was detected in the hippocampal formation, including presubicular areas contiguous to the entorhinal cortex. C1q immunostaining was detected in the ipsilateral CA1 stratum lacunosum
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
201
FIG. 1. Immunocytochemistry for C9, clusterin (SGP-2), synaptophysin, C1q, and GFAP in dentate gyrus molecular layer 3d after perforant path transection. Each panel shows a portion of the ipsilateral hippocampus in the horizontal plane. All sections, except that in C, were counterstained with cresyl violet, and all represent semiadjacent sections. A shows cresyl violet stain of hippocampus ipsilateral to lesion. GFAP ICC (B) shows increased astrocyte reactivity in middle molecular layer (arrow, m) and stratum lacunosum (sl). C shows a reduction of synaptophysin immunoreactivity in the ipsilateral middle molecular layer (arrow, m) and stratum lacunosum, where in D and F, there is increased clusterin and C9 immunoreactivity, respectively. Clusterin (SGP-2) is an inhibitor of the formation of the complement C5b-9 MAC, of which C9 is a key component necessary for MAC function. Immunocytochemistry for C1q, the initiating component of the complement cascade pathway, showed weak, but detectable staining primarily in the distal stratum lacunosum. The arrow in B, C, D, and F points out the dentate gyrus middle molecular layer. Abbreviations: DG, dentate gyrus; i, dentate inner molecular layer; m, dentate middle molecular layer; o, dentate outer molecular layer; sl, stratum lacunosum. Size bar, 250 µm.
202
JOHNSON ET AL.
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
(Fig. 1E) compared to the contralateral side or cresyl violet-stained control sections (Fig. 1A). The dentate molecular layer lacked immunodetectable C1q. At 3 days postlesion, extracellular C1q was strongest directly adjacent to the wound in the entorhinal cortex, but declined with increasing distance from the wound. Intracellular C1q was confined to the wound site and adjacent areas, where we observed putative infiltrating blood cells and scattered macrophage/microglia (not shown). The contralateral side did not stain for C1q. DISCUSSION
The terminal components of the C cascade require C9 to form a fully functional MAC. The robust increases of C9 in the lesioned EC and deafferented hippocampus seen in this study may allow formation of the MAC. Neurons of three morphotypes showed C9 immunostaining: large process-bearing cells with pyramidal neuron morphology in ipsilateral EC layers II–IV and subiculum (not shown), cells on the inner border of the deafferented zone of the dentate gyrus (not shown), and multipolar cells in the CA1 stratum lacunosum moleculare and stratum radiatum (Fig. 2A). Near the EC wound cavity, cells with macrophage/amoeboid microglial morphology were also densely immunostained for C9. These findings are consistent with immunocytochemical observations of the hippocampus in AD, in which McGeer et al. (21) reported that anti-human C9 stained plaques ‘‘. . . along with tangles, neuropil threads and dystrophic neurites.’’ That photomicrograph could also be interpreted to show immunostaining of select neuronal perikarya, as observed here in a rat model of AD. The same report shows similar staining, except for plaques, by an anti-C5b-9 (anti-MAC) antibody (21), suggesting to those authors the presence of the lytic MAC on membranes of tangled neurons and dystrophic neurites in AD hippocampus. We do not know the identity or function of the C9-immunopositive neurons in the deafferented rat hippocampus shown in Figs. 2A and 2C. Since there is no C9 immunolabeling of CA field pyramidal or dentate gyrus granule neurons, we suggest that the C9immunopositive neurons represent one or more classes of interneurons. AD brains show a loss of neuropeptide
203
Y/somatostatin (NPY/SS) interneurons in the facia dentata of the hippocampus (5), i.e., the same region that shows C9 and clusterin-immunopositive cells with neuronal morphology in the present studies. The same NPY/SS interneurons might be compromised after perforant path transection (8) and subject to complement attack. Further evidence supports the concept that the C system has an active role during neurodegeneration. The phagocytosis of myelin by macrophages during responses to peripheral neuronal lesions is impaired when various C components are lacking or deficient (2). Because the C cascade is subject to regulation at many stages by C inhibitors, including clusterin (15, 27), it is pertinent to the argument that brain cells also make clusterin. Clusterin mRNA is induced in these same neuron types, as well as in the numerous astrocytes in the deafferented dentate middle molecular layer and stratum lacunosum (17, 24, 28). In AD, clusterin protein occurs in neuritic plaques (22). The secretion of clusterin by primary astrocyte cultures (24) suggests that the extracellular clusterin in the middle molecular layer of this study and in neuritic plaques of AD was secreted by local astrocytes. Strong, diffuse extracellular C9 immunostaining was found in the CA1 stratum lacunosum and the dentate gyrus middle molecular layer, while there was only weak staining in the outer or inner dentate molecular layers. This laminar restriction of C9 immunostaining correlates well with reactive synaptogenesis and increased dendritic arborization that begins first in the middle molecular layer between 2 and 4 days after lesioning (4, 13). It also correlates with the loss of synaptophysin immunoreactivity in the middle molecular layer. It is not entirely clear why synaptophysin was not reduced in the outer molecular layer; incomplete transection of the perforant path, especially the portion of the pathway from the lateral entorhinal cortex that terminates in the outer molecular layer, may be an explanation. We suggest that, along with a possible role in neurodegeneration, as was suggested previously for AD (21), the complement system may also play a role in repair mechanisms after brain injury.
FIG. 2. Intracellular immunolocalization of complement C9 and clusterin in hippocampal interneurons after perforant path transection. Three days after unilateral perforant path transection scattered interneurons that do not reside within the tightly packed pyramidal or granule cell layers are immunopositive for rat C9 ipsilateral (A), but not contralateral (B) to the transection lesion. Arrowheads identify several C9-immunopositive neurons in A that are located in stratum radiatum or stratum oriens. At higher magnification it is clear that both soma and proximal dendrites are labeled by the anti-rat C9 polyclonal antibody (C). Anti-rat clusterin polyclonal antibody also immunostains scattered interneurons located outside the pyramidal or granule cell layers of the hippocampal formation. Several clusterin-immunopositive neurons, located in the ipsilateral stratum lacunosum, are shown in D. The underlying parenchymal staining by anti-clusterin in D, which can be seen at lower power in Fig. 1D, is present only on the ipsilateral side. Abbreviations: sp, stratum pyramidale (pyramidal cell layer); sr, stratum radiatum. A and B are at same magnification; size bar, 50 µm. C and D are at same magnification; size bar, 20 µm.
204
JOHNSON ET AL.
FIG. 3. Apoptotic morphology of nuclei in degenerating entorhinal cortex layer II neurons. Pyknotic nuclei were present in entorhinal cortex layers II and III after perforant path transection. Such nuclei were fragmented and more intensely stained than normal nuclei and generally resembled the morphology of nuclei from cells undergoing apoptosis. A shows a portion of medial entorhinal cortex 3 days after the transection lesion; apoptotic nuclei are indicated by arrowheads. The area in the upper right corner of A enclosed by the dashed lines is shown at higher magnification in B. Note the densely stained, fragmented nuclear bodies that appear to be contained within the nuclear membrane. Also note that although the parenchyma was stained with anti-rat C9 polyclonal antibody, suggesting extracellular deposition of C9, apoptotic layer II neurons were not C9 immunopositive. V, large vessel. Size bars equal 20 and 10 µm in A and B, respectively.
C9 EXPRESSION IN DEAFFERENTED HIPPOCAMPUS
ACKNOWLEDGMENTS The authors thank Mr. Chris Anderson for excellent technical assistance. We also thank Mr. Hank Hogan for excellent technical and photographic assistance. This work was supported by AG10673 (S.A.J.) and AG07909 (C.E.F.).
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
AUGER, M. J., AND J. A. ROSS. 1992. The biology of the macrophage. Pages 1–74 in C. E. Lewis and J. O. McGee, Eds., The Macrophage. IRL Press, Oxford. BRUCK, W., AND R. L. FRIEDE. 1991. The role of complement in myelin phagocytosis during PNS wallerian degeneration. J. Neurol. Sci. 103: 182–187. CABALKA, L. M., B. T. HYMAN, C. R. GOODLETT, T. C. RITCHIE, AND G. W. VAN HOESEN. 1992. Alteration in the pattern of nerve terminal protein immunoreactivity in the perforant pathway in Alzheimer’s disease and in rats after entorhinal lesions. Neurobiol. Aging 13: 283–291. CACERES, A., AND O. STEWARD. 1983. Dendritic reorganization in the denervated dentate gyrus of the rat following entorhinal cortical lesions: A golgi and electron microscopic analysis. J. Comp. Neurol. 214: 387–403. CHAN-PALAY, V. 1987. Somatostatin immunoreactive neurons in the human hippocampus and cortex shown by immunogold/ silver intensification on vibrotome sections: Coexistence with neuropeptide Y neurons, and effects in Alzheimer-type dementia. J. Comp. Neurol. 260: 201–223. CUMMINGS, B. J., G. J. YEE, AND C. W. COTMAN. 1992. bFGF promotes the survival of entorhinal layer II neurons after perforant path axotomy. Brain Res. 591: 271–276. DE LANEROLLE, N. C., J. H. KIM, R. J. ROBBINS, AND D. D. SPENCER. 1989. Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy. Brain Res. 495: 387–395. DELLER, T., AND C. LERANTH. 1990. Synaptic connections of neuropeptide Y (NPY) immunoreactive neurons in the hilar area of the rat hippocampus. J. Comp. Neurol. 300: 433–447. EIKELENBOOM, P., S-S. ZHAN, W. A. VAN GOOL, AND D. ALLSOP. 1994. Inflammatory mechanisms in Alzheimer’s disease. Trends Pharmacol. Sci. 15: 447–450. GEDDES, J. W., D. T. MONAGHAN, C. W. COTMAN, I. T. LOTT, R. C. KIM, AND H. C. CHUI. 1985. Plasticity of hippocampal circuitry in Alzheimer’s disease. Science 230: 1179–1181. GEHRMAN, J., S. W. SCHOEN, AND G. W. KREUTZBERG. 1991. Lesion of the rat entorhinal cortex leads to a rapid microglial reaction in the dentate gyrus. A light and electron microscopical study. Acta Neuropathol. 82: 442–455. HAGA, S., K. IKEDA, M. SATO, AND T. ISHII. 1993. Synthetic Alzheimer amyloid b/A4 peptides enhance production of complement C3 component by cultured microglial cells. Brain Res. 601: 88–94. HOFF, S. F., S. W. SCHEFF, AND C. W. COTMAN. 1982. Lesioninduced synaptogenesis in the dentate gyrus of aged rats: II. Demonstration of an impaired degeneration clearing response. J. Comp. Neurol. 205: 253–259. HYMAN, B. T., G. W. VAN HOESEN, A. R. DAMASIO, AND C. L. BARNES. 1984. Alzheimer’s disease: Cell-specific pathology isolates the hippocampal formation. Science 225: 1168–1170. JENNE, D. E., AND J. TSCHOPP. 1993. Clusterin: The intriguing guises of a widely expressed glycoprotein. Trends Biochem. Sci. 17: 154–159. JOHNSON, S. A., T. MCNEILL, B. CORDELL, AND C. E. FINCH. 1990.
17.
18.
19.
20.
21.
22.
22A.
23.
24.
25.
26.
27. 28.
29.
30.
31.
205
Relation of neuronal APP-751/APP-695 mRNA ratio and neuritic plaque density in Alzheimer’s disease. Science 248: 854– 857. JOHNSON, S. A., M. LAMPERT-ETCHELLS, G. M. PASINETTI, I. ROZOVSKY, AND C. E. FINCH. 1992. Complement mRNA in the mammalian brain: Responses to Alzheimer’s disease and experimental brain lesioning. Neurobiol. Aging 13: 641–648. JOHNSON, S. A., G. M. PASINETTI, AND C. E. FINCH. 1994. Expression of complement C1qB and C4 mRNAs during rat brain development. Dev. Brain Res. 80: 163–174. LAMPERT-ETCHELLS, M., G. PASINETTI, C. FINCH, AND S. JOHNSON. 1993. Regional localization of cells containing C1q and C4 mRNAs in the frontal cortex during Alzheimer’s disease. Neurodegeneration 2: 113–121. MAY, P. C., AND FINCH, C. E. 1992. Sulfated glycoprotein-2: New relationships of this multifunctional protein to neurodegeneration. Trends Neurosci. 15: 391–396. MCGEER, P. L., H. AKIYAMA, S. ITAGAKI, AND E. G. MCGEER. 1989. Immune system response in Alzheimer’s disease. Can. J. Neurol. Sci. 16: 516–527. MCGEER, P. L., T. KAWAMATA, AND D. G. WALKER. 1992. Distribution of clusterin in Alzheimer brain tissue. Brain Res. 579: 337–341. ODA, T., G. M. PASINETTI, H. H. OSTERBURG, C. ANDERSON, S. A. JOHNSON, AND C. E. FINCH. 1994. Purification and characterization of brain clusterin. Biochem. Biophys. Res. Commun. 204: 1131–1136. PASINETTI, G. M., S. A. JOHNSON, I. ROZOVSKY, M. LAMPERTETCHELLS, D. G. MORGAN, M. N. GORDON, T. E. MORGAN, D. WILLOUGHBY, AND C. E. FINCH. 1992. Complement C1qB and C4 mRNAs responses to lesioning in rat brain. Exp. Neurol. 118: 117–125. PASINETTI, G. M., S. A. JOHNSON, T. ODA, I. ROZOVSKY, AND C. E. FINCH. 1994. Clusterin (SGP-2): A multifunctional glycoprotein with regional expression in astrocytes and neurons of the adult rat brain. J. Comp. Neurol. 339: 387–400. PETERSON, D. A., C. A. LUCIDI-PHILLIPI, K. L. EAGLE, AND F. H. GAGE. 1994. Perforant path damage results in progressive neuronal death and somal atrophy in layer II of entorhinal cortex and functional impairment with increasing postdamage age. J. Neurosci. 14: 6872–6885. ROGERS, J., N. R. COOPER, S. WEBSTER, J. SCHULTZ, P. L. MCGEER, S. D. STYREN, W. H. CIVIN, L. BRACHOVA, B. BRADT, P. WARD, AND I. LIEBERBURG. 1992. Complement activation by beta-amyloid in Alzheimer disease. Proc. Natl. Acad. Sci. USA 89: 10016–10020. ROTHER, K., AND G. O. TILL, EDS. 1988. Page 535 in The Complement System. Springer-Verlag, Berlin. ROZOVSKY, I., T. E. MORGAN, D. A. WILOUGHBY, M. M. DUGICHDJORDEVICH, G. M. PASINETTI, S. A. JOHNSON, AND C. E. FINCH. 1994. Selective expression of clusterin (SGP-2) and complement C1qB and C4 during responses to neurotoxins in vivo and in vitro. Neuroscience 62: 741–758. STEWARD, O., AND S. L. VINSANT. 1983. The process of reinnervation in the dentate gyrus of the adult rat: A quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis. J. Comp. Neurol. 214: 370–386. VEERHUIS, R., P. VAN DER VALK, I. JANSSEN, S. S. ZHAN, W. E. VAN NOSTRAND, AND P. EIKELENBOOM. 1995. Complement activation in amyloid plaques in Alzheimer’s disease brains does not proceed further than C3. Virchows Arch. 426: 603–610. WALKER, D. G., AND P. L. MCGEER. 1992. Complement gene expression in human brain: Comparison between normal and Alzheimer disease cases. Mol. Brain Res. 14: 109–116.