Allograft-Inflammatory-factor-1 is upregulated in microglial cells in human cerebral infarctions

Allograft-Inflammatory-factor-1 is upregulated in microglial cells in human cerebral infarctions

Journal of Neuroimmunology 104 (2000) 85–91 www.elsevier.com / locate / jneuroin Allograft-Inflammatory-factor-1 is upregulated in microglial cells i...

712KB Sizes 11 Downloads 53 Views

Journal of Neuroimmunology 104 (2000) 85–91 www.elsevier.com / locate / jneuroin

Allograft-Inflammatory-factor-1 is upregulated in microglial cells in human cerebral infarctions E. Postler, A. Rimner, R. Beschorner, H.J. Schluesener, R. Meyermann* ¨ ¨ , Calwer Strosse 3, D-72076 Tubingen , Germany Institute for Brain Research, University of Tubingen Received 18 June 1999; received in revised form 27 September 1999; accepted 29 September 1999

Abstract Allograft inflammatory factor-1 (AIF-1) is a 17-kDa-peptide identified in rat cardiac allografts undergoing chronic rejection and in activated microglial cells in inflammatory autoimune disease of the CNS. We have investigated the expression of AIF-1 in 18 autopsy cases of human focal cerebral infarction. AIF-1-positive cells show the morphology of microglia and are CD68- but not GFAP-positive. The peptide is expressed at a low level in normal brain. In infarctions, activated microglial cells in the area of glial reaction show strongly enhanced cytoplasmic immunoreactivity. The density of AIF-1-expressing cells increases during the first three days post infarction and remains elevated until chronic cystic stages. The upregulation of AIF-1-immunoreactivity precedes the rise in expression of the S-100-protein MRP-8. We conclude that AIF-1 is a sensitive marker of human microglial activation not only in inflammation but also in non-inflammatory lesions of the CNS.  2000 Elsevier Science B.V. All rights reserved. Keywords: Allograft-inflammatory-factor-1; Microglia; Cerebral infarction; Human

1. Introduction Allograft-inflammatory-factor (AIF-1), a 17-kDa-polypeptide, was originally identified in macrophages in transplanted rat and human hearts subject to chronic rejection. Human AIF-1 shows 90% amino acid sequence homology with the rat protein (Utans et al., 1995, 1996). It is identical to two recently characterized peptides, the ionized calcium binding adapter molecule-1 (Imai et al., 1996) and microglia response factor-1 (Tanaka et al., 1998). AIF-1 has also been detected in cultured rat microglia cells (Imai et al., 1996) and is constitutively expressed by dendritic cells of lymphatic organs, tissue macrophages in intestine, liver, pancreas and in vivo in rat brain microglia cells (Chen et al., 1997). Strong constitutive AIF-1-immunoreactivity has also been found in rat peripheral nerve and testis. (Utans et al., 1995; Imai et al., 1996, Schluesener et al., 1998). The AIF-1-molecule contains a calcium-binding-EFhand-motif, indicating a function in cellular calcium homeostasis. Calcium (Ca 21 ) functions as a cytoplasmatic *Corresponding author. Tel.: 149-7071-298-2293; fax: 149-7071-294846. E-mail address: [email protected] (R. Meyermann)

second messenger and is associated with activation of myeloid cells, monocytes (Gilchrist et al., 1994) and microglia (McLarnon et al., 1997, Moller et al., 1997). In these cell types, calcium responses precede phagocytosis and oxidative burst and influence the transcription of a number of eukaryotic genes via transcription factors interacting with calcium response elements (Roche and Prentki, 1994). Other activation-related EF-hand-proteins of myeloid specifity belong to the S-100-family: S100A8 (macrophage-inhibiting-factor-related-protein-8; MRP-8), S-100A9 (MRP-14) and S100A12 (Weissman and Lagasse, 1992; Roth et al., 1994; Mahnke et al., 1995; for review ¨ see Schafer and Heizman, 1996). In a previous immunohistochemical study, we have described the upregulation of MRP-8 and -14 by activated microglial cells early in human focal cerebral infarction (Postler et al., 1997). In contrast to the S-100-group, the gene encoding AIF-1 is localized to the the MHC-III-region (Imai et al., 1996), together with other inflammatory proteins as tumor-necrosis-factor-a and -b, complement- (C2; C4 and factor B) and heat-shock-proteins (HSP70-1H and HSP-2). For TNF-a and HSP-70, an early upregulation within the first hours after cerebral infarction has been proven in rats (Buttini et al., 1996; Kawagoe et al., 1992). In the present study we have investigated the time course of AIF-1-

0165-5728 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0165-5728( 99 )00222-2

86

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91

expression after human focal cerebral infarctions by performing immunohistochemistry on human autopsy brain tissue. The results were compared to well-characterized markers of microglial activation such as macrophage antigen-387 (MAC 387), MRP-8, CD68 and MHC-classII-antigen (HLA-DR).

2. Materials and methods We investigated 18 human brains with focal cerebral infarction at different stages (mean age: 66.4 years, SD615,1 Table 1). The post-mortem-interval amounted to

12–48 hours. Other medical conditions leading to an altered state of actitvity of the immune system as sepsis, meningitis / encephalitis or immunosuppressive medication were excluded as well as infarctions with prominent hemorrhage. This group was compared to 10 control patients who died from non-neurological diseases (e.g. leukemia, peritonitis, myocardial infarction, myocarditis, ARDS, gastrointestinal bleeding, aortal dissection) and, considering clinical history and results of microscopic investigation, did not suffer from prolonged cerebral hypoxia (mean age: 49.4 years, SD611.9). The tissue was fixed, dehydrated and embedded in paraffin following routine histological protocols. For estimation of infarct

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91

age, hematoxyline-eosine-stained slides were evaluated for the histological features defined as standard indicators of infarct age (Graham, 1992). In particular, spongiform changes of the neuropil, pallor of the core of infarct, endothelial swelling and ischemic neuronal damage without detection of granulocytes were regarded as indicators of an early stage (,24 h). In case 1, clinical files documented that the patient died less than 24 hours after a combined dissection of aorta and carotid arteries which caused unilateral cerebral infarction. The age of processes with additional sharp demarcation of the infarct from the surrounding parenchyma, presence of granulocytes and circumscript hemorrhages was estimated to be more than

87

24 hours. Infarctions with pronounced infiltration of foam cells removing necrotic tissue debris were classified as older than three days while the age of cases with prominent cystic change with or without macrophages and astrogliosis was regarded to be at least one week. The control cases did not have any history of neurological symptoms and their brains did not show hypoxic changes. Immunhistochemistry for AIF-1 was performed using a recently described monoclonal antibody (Schluesener et al., 1998). In brief, the slides were rehydrated and pretreated for 735 minutes in a microwave oven (600 Watts) in citrate buffer (pH 6,0). Unspecific binding sites were blocked by normal swine serum in 1:10 dilution. Sections were incubated with AIF-1-antibody-containing hybridoma supernatant for 1 h at room temperature. Antibody binding was visualized with biotinylated secondary antibody, the avidin-peroxidase-complex (DAKO, Hamburg, FRG) and diaminobenzidine as chromogen. For double labeling, the slides were pretreated with microwaves for 735 min a second time after AIF-1-had been stained and incubated with one of the following antibodies: anti-HLA-DR, -MAC387, -GFAP, -CD68 (DAKO, Hamburg. FRG) and -MRP-8 (Bachem, Heidelberg, FRG). The second antigen was visualized by the alkaline-phosphatase-anti-alkalinephosphatase-complex (APAAP, DAKO) with Fast-BlueBB-salt (Sigma, Deisenhofen, FRG) as chromogen.

2.1. Data analysis For evaluation of glial reactions, the white matter immediately adjacent to the completely necrotic infarct core was examined. For every macrophage marker, the positive cells in 20 high power fields (HPF) and cells double-positive for AIF-1 and the respective marker were counted and mean values were calculated for each data set. Fig. 1. Length of scale bar is 100 mm in (A), (B), (D), (E), (G), (H); 50 mm in (C) and (F). (A): CD 68-positive ramified microglia cells (blue) express AIF-1 (brown; case 2). (B): GFAP-positive astrocytes (brown) do not express AIF-1, which is found on GFAP-negative cells (blue) between positive astrocytes (Case 2). (C): Weak expression of AIF-1 by rare ramified microglia cells in the white matter of a 30-year-old patient who died from sudden cardiac failure. (D): Expression of AIF-1 on microglial cells (*) in the white matter adjacent to a cortical laminar infarction (3). In the preserved neighbouring white matter, less immunoreactivity is seen (+; case 2). (E): AIF-1-immunoreactivity is strongly enhanced in microglial cells in the area of glial reaction surrounding necrosis of infarctions with an age of ,3 days (case 9). (F): Coexpression of MRP-8 (blue) and AIF-1 (brown) in microglia (case 2; age ,3days). (G): Upregulation of AIF-1 in microglial cells precedes the expression of other macrophage activation antigens as MRP-8: In case 1 (age less than 24 hours, see Section 2), numerous ramified microglia cells strongly express AIF-1 (brown), while MRP-8 can only be found as a constitutive antigen in intravascular monocytes (blue). Blue immunoreactivity on ramified microglia cells is hardly detectable at all. (H): Neighbourhood of infarct necrosis of case 14 (age of infarction 1–3 weeks). A number of small cells within a spongy glial matrix show AIF-1-positive ramifications. Although the immunostain is less intense, the density of these microglial cells is still significantly elevated above baseline.

88

Table 1 Case data Age

Sex

Time Post infarction

Localization of infarction

History of infarction focus / of remote area

Additional medical conditions

1

67

F

,24 h

left temporal lobe

hypoxic neuronal damage, spongiosis / few eosinophilic neurons

2 3

87 80

F M

1d 1d

left occipital cortex right postcentral gyrus

4

62

F

1d

5

52

F

1–2d

right parietooccipital cortex right occipitobasal cortex

6 7

52 78

F F

2–3d 2–3d

left thalamus right frontal cortex

8

70

F

2–3d

left precentral gyrus

9

75

F

3d

left occipitobasal cortex

10

61

M

3–4d

left occipital cortex

11

64

M

3-4d

12 13

86 75

F M

¯7d .1 week

parietooccipital cortex / white matter left parietal cortex, insula left parietal white matter

14

80

M

.3 weeks

periventricular, pallidum

Endothelial sweeling, hemorrhage / no vacuoles Demarcation, no granulocytes / sparse hypoxic neurons in hippocampus, no vacuolation marked spongiosis no granulocytes, hypoxic neurons in basal ganglia, beginning vacuolation marked spongiosis no granulocytes / no hypoxic cortex neurons beginning beginning vacuuolation in white matter Macrophages capillary proliferation / hypoxic damage of neurons Spongiosis, endothelial swelling rare granulocytes / marked hypoxic damage of neurons in cerebellar cortex and nuclei Demarcation endothelial swelling rare granulocytes / marked hypoxic neuronal changes brain edema Sponiosis, myelin pallor granulocytes / condensed neurons with pycnotic nuclei, vacuolation of myelin myelin containing macrophages / marked hypoxic damage of cortex neurons Colliquation necrosis macrophages capillary proliferation / rare hypoxic cortex neurons beginning edema dense macrophages formation of glial scar / capillary proliferation dense macrophages, beginning of glial reaction / rare hypoxic neurons in cerebellar nuclei rests of necrotic tissue formation of glial scar

15

78

F

.3 weeks

left occipital cortex

16

70

M

left occipital white matter

17

69

F

18

52

F

several weeks several weeks months

cystic media necrosis, dissection of aorta, died less than 24 h post-OP hyperthyreosis, ulcus ventriculi severe arteriosclerosis, coronary heart disease depression malignant neuroleptic syndrome Aneurysm of A. carot. int., arteriosclerosis COPD, hemorrhagic diathesis peripheral vascular occlusive disease, embolism of left carotid artery coronary heart disease valvular insufficiency infantile cerebral damage of uncertain origin infantile cerebral damage severe imbecility Parkinson’s disease bronchial carinoma carcinoma of hypopharynx and esophagus arteriosclerosis, arterial hypertension diabetes mellitus peripheral arterial occlusive disease diabetes COPD coronary heart disease coronary heart disease intestinal carcinoma cor pulmonale diabetes coronary heart disease bronchial carinoma severe arteriosclerosis diabetic polyneuropathy

caput nuclei caudati caudal left cerebellar lobe

rests of necrotic tissue glial raction / marked eosinophilic change of neurons in cerebellar nuclei Macrophages in moderate density moderate astrogliosis fibrotic vessels / small fraction of eosinophilic neurons in cerebellar nuclei chronical cystic lesion with foam cells / scarce eosinophilic neurons

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91

Case no.

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91

89

These results were compared to data obtained from an unaffected brain region remote from the focal infarction in the same patient and to values obtained from comparable white-matter-sites of the control group. For AIF-1, MRP-8 and CD68, the mean values in early infarction cases (,3 days) were compared to the means of the older processes (.3 days) using the Mann-Whitney-U-test.

3. Results In brain tissue, AIF-1-expression was found exclusively on CD68- positive but not on GFAP-positive cells. These findings prove that AIF-1-positive cells belong to the microglia-macrophage-population and that they are not astrocytes (Fig. 1A and B). In control brains without prolonged hypoxia, immunoreactivity for AIF-1 was restricted to delicate processes of single microglial cells (Fig. 1C), their mean density being low (3.1 cells / high power field (HPF50.25 mm 2 ); SD52.9). Microglial cells positive for the constitutive marker CD68 were found at a density of 15.5 cells / HPF (SD55.2), which was regarded as baseline. In the unaffected (‘remote’) areas of the infarction cases, the expression level of AIF-1 varied from case to case but was clearly higher than in control brains (mean56.1 cells / HPF; SD55.5). In both control cases and remote areas of infarction patients, the microglia activation markers MRP-8 and MHC-II-antigen (HLA-DR) were not detected. In the area of glial reaction around infarctions with an age of less than 3 days, (Fig. 1D) immunoreactivity for AIF-1 was more pronounced in the perinuclear cytoplasm and delicate processes of ramified microglial cells but also in amoeboid microglia cells with short stout processes. The density of positive cells also had increased markedly (Fig. 1E; mean57.2 cells / HPF; SD52.9). Colocalisation of AIF-1-immunoreactivity was seen in MRP-8-positive(Fig. 1F) and HLA-DR-positive cells (not shown), indicating an upregulation of AIF-1-expression during microglial activation. In the earliest infarction case with an age of less than 24 hours (case 1), strong reactivity for AIF-1 was detectable in microglial cells which were still negative for MRP-8 (Fig. 1G). Perivascular cells also showed strong immunoreactivity, in contrast to endothelial cells. After the third day post infarction (p.i.), macrophages with foamy cytoplasm which had invaded the infarct necrosis showed membraneous immunoreactivity for AIF1. At this stage, the density of AIF-1-expressing cells in the area of glial reaction had increased statistically in comparison to the first three days (mean516 cells / HPF; SD58.8 vs. mean57.2; SD52.9; U-test, p,0.05,). In parallel, the number of CD68-positive cells had also increased significantly (mean523.1 cells / HPF; SD56.8 vs. mean513.5; SD57.2; p,0.05) and both antigens remained at a high level during removal of necrosis. MRP-8 was entirely downregulated after 7 days p.i. (Fig.

Fig. 2. The density of AIF-1-expressing cells in the area of glial reaction surrounding necrosis is upregulated during the first 3 weeks after infarction and remains at a high level until the chronic stage. The density of CD68-positive cells also increases during the first days after infarction and remains elevated afterwards. MRP-8 is upregulated on ramified and round cells during the first 4 days after infarction and is not found any more after 1 week.

2). Fig. 1H shows numerous ramified cells exhibiting AIF-1-immunoreactivity in the area of glial reaction of an older infarction (age 1–3 weeks).

4. Discussion The findings of an upregulation of AIF-1 in chronic rejection of rat and human allografts (Utans et al., 1995; 1996), in autoimmune lesions (Chen et al., 1997) and of constitutive AIF-1-expression in lymphatic organs (Autieri, 1996) indicate that AIF-1 probably has a function in the regulation of immunological processes. This assumption was also supported by detection of enhanced microglial AIF-1-immunoreactivity in or near to the lesions of rat experimental autoimmune encephalomyelitis at 10–12 days after immunisation (Schluesener et al., 1998). In the present study, an earlier increase of AIF-1-immunoreactivity in activated microglia cells was found at 2–3 days after human focal cerebral infarction in the zone of glial reaction neighbouring the necrotic area. The cause of the delayed upregulation in EAE probably is the slow activation of autoimmune reaction during the first 14 days

90

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91

after immunisation while tissue necrosis and its subsequent resorption strongly activate microglia cells during the first 3 days post infarction. Microglial upregulation of AIF-1, however, is not unique to inflammatory and tissue-necrotizing processes in the brain: the microglia-responsefactor-1 (mrf-1), a homologue to AIF-1, was recently identified in microglial cells during apoptotic granular cell degeneration in the developing rat cerebellar cortex (Tanaka et al., 1998). Its expression therefore seems to be common to conditions which lead to microglial activation and phagocytosis in the brain. The density of AIF-1-expressing ramified cells in the area next to necrosis after infarction remained elevated during the resorption phase until the chronic stage. This is in contrast to microglial expression of MRP-8, which is found on ramified microglia only during the first three days post infarction (Postler et al., 1997). The prolonged upregulation of AIF-1 after infarction again parallels the findings seen in the inflammatory model of EAE: here, enhanced AIF-1-expression was detectable until 40 days post immunisation, long after dissolution of inflammatory infiltrates and regression of clinical signs (Schluesener et al., 1998). Together with the notion that in heart transplants committed to late-onset chronical rejection, AIF is expressed during the early postoperative stage in spite of absence of acute inflammatory changes (Utans et al., 1996), this speaks for a relation of AIF-1 to chronic (‘smouldering‘) inflammatory processes. Therefore, its functions seem to be different from early-response proinflammatory cytokines as interleukin-1b or tumor-necrosisfactor-a, which are also promptly upregulated by microglia in inflammatory and degenerative brain diseases (Benecsik et al., 1996; Renno et al., 1995; Buttini et al., 1996). Additionally, the fact that aside the CNS, strong AIF-1expression can be found in peripheral nerve and testis, two other organs with a blood-tissue-barrier and a down-regulated immunological status, may prompt the speculation that the protein might be important for immune-regulation in these tissues. The density of AIF-1-positive cells are highest in the area of glial reaction surrounding infarction, lower in the non-affected remote brain areas of infarction patients and lowest in healthy control brains. These results correspond to the different degrees of hypoxia and resulting tissue damage in these three groups. In the remote brain areas of infarction patients, the moderate enhancement of microglial AIF-1-expression suggests an intermediate state of activation compared to infarction area and healthy control brain. This indicates that these ‘remote’ areas are also damaged to a certain extent, probably by global brain hypoxia or secondarily by brain edema and swelling. In general, AIF-1 is a sensitive marker for microglial activation correlating well to the intensity of tissue damage not only in inflammation of the CNS but also in cerebral hypoxia.

References Autieri, M.V., 1996. cDNA-cloning of human allograft inflammatory factor-1: tissue distribution, cytokine induction and m-RNA-expression in injured rat carotid arteries. Biochem. and Biophys. Res. Commun. 228, 29–37. Benecsik, A., Malcus, C., Akoaka, H., Giraudon, P., Belin, M.F., Bernard, A., 1996. Selective induction of cytokines in mouse brain infected with canine distemper virus: structural, cellular and temporal expression. J. Neuroimmunol. 65, 1–9. Buttini, M., Appel, K., Sauter, A., Gebicke-Haerter, P.J., Boddecke, H.W.G.M., 1996. Expression of tumor necrosis factor alpha after focal cerebral ischemia in the rat. Neurosci. 71, 1–16. ¨ ¨ Chen, Z.W., Ahren, B., Ostenson, C.G., Cintra, A., Bergman, T., Moller, C., Fuxe, K., Mutt, V., Efendic, S., 1997. Identification, isolation and characterization of daintain (allograft inflammatory factor-1), a macrophage polypeptide with effects on insulin secretion and abundantly present in the pancreas of prediabetic BB rats. Proc. Natl. Acad. Sci. USA 94, 13879–13884. Gilchrist, J.S., Czubryt, M.P., Pierce, G.N., 1994. Calcium and calciumbinding proteins in the nucleus. Mol. Cell. Biochem. 135, 79–88. Graham, D.I., 1992. Hypoxia and vascular disorders. In: Adams, H.J., and Duchen, L.W., (Eds.), Greenfield‘s neuropathology, Edward Arnold, London, pp. 196-200 Imai, Y., Ibata, I., Ito, D., Ohsawa, K., Kohsaka, S., 1996. A novel gene iba-1 in the major histocompatibility complex calss III region encoding an EF hand protein expressed in a monocytic lineage. Biochem. Biophys. Res. Commun. 224, 855–862. Kawagoe, J., Abe, K., Sato, S., Nagano, I., Nakamura, S., Kogure, K., 1992. Distributions of heat shock protein (HSP)70 and heat shock cognate protein (HDC) 70 mRNAs after transient focal ischemia in rat brain. Brain Res. 587, 195–202. Mahnke, K., Bhardwaj, R., Sorg, C., 1995. Heterodimers of the surfacebinding proteins MRP-8 and MRP14 are expressed on the surface of human monocytes upon adherence to fibronectin and collagen: Relation to TNF-alpha, IL-6 and superoxide production. J. Leukoc. Biol. 57, 63–71. McLarnon, J.G., Xu, R., Lee, Y.B., Kim, S.U., 1997. Ion channels in human microglia in culture. Neuroscience 78, 1217–1228. Moller, T., Kann, O., Prinz, M., Kirchhoff, F., Verkhratsky, A., Kettenmann, H., 1997. Endothelin-induced calcium signaling in cultured mouse microglial cells is mediated through ETB-receptors. Neuroreport 8, 2127–2131. Postler, E., Schluesener, H.J., Meyermann, R., 1997. Expression of the S-100 proteins MRP-8 and -14 in ischemic brain lesions. Glia 19, 27–34. Renno, T., Krakowski, M., Piccirillo, C., Lin, J., Owens, T., 1995. TNF-a expression by resident microglia and infiltrating leucocytes in the central nervous system of mice with experimental allergic encephalomyelitis. J. Immunol. 154, 944–953. Roche, E., Prentki, M., 1994. Calcium regulation of immediate early response genes. Cell Calcium 16, 331–338. Roth, J., Goebeler, M., Wrocklage, V., Van den Bos, C., Sorg, C., 1994. Expression of the calcium-binding proteins MRP-8 and MRP-14 in monocytes is regulated by a calcium-induced suppressor mechanism. Biochem. J. 301, 655–660. ¨ Schafer, B.W., Heizman, C.W., 1996. The S-100-family of EF-hand calcium-binding proteins: functions and pathology. Trends Biochem. Sci. 21, 134–140. Schluesener, H.J., Seid, K., Kretschmar, J., Meyermann, R., 1998. Allograft-inflammatory-factor in rat experimental autoimmune encephalomyelitis, neuritis and uveitis: expression by activated macrophages and microglial cells. Glia 24, 244–251. Tanaka, S., Suzuki, K., Watanabe, M., Matsuda, A., Tone, S., Koike, T., 1998. Upregulation of a new microglial gene, mrf-1, in response to

E. Postler et al. / Journal of Neuroimmunology 104 (2000) 85 – 91 programmed neuronal cell death and degeneration. J. Neurosci. 18, 6358–6369. Utans, U., Areci, R.J., Yamashita, Y., Rusell, M.E., 1995. Cloning and characterisation of allograft inflammatory factor-1: a novel macrophage factor identified in rat cardiac allografts with chronic rejection. J. Clin. Invest. 95, 2954–2962. Utans, U., Quist, W.C., McManus, B.M., Wilson, J.E., Areci, R.J.,

91

Wallace, A.F., Rusell, M.E., 1996. Allograft inflammatory factor: A cytokine-responsive macrophage molecule expressed in transplanted human hearts. Transplantation 61, 1387–1392. Weissman, I.L., Lagasse, E., 1992. Mouse MRP-8 and MRP-14, two intracellular calcium-binding proteins associated with the development of the myeloid lineage. Blood 8, 1907–1915.