Heterogeneous expression of CD59 on human Purkinje cells

Neuroscience Letters 362 (2004) 21–25 www.elsevier.com/locate/neulet

Heterogeneous expression of CD59 on human Purkinje cells Anette Storsteina,*, Anette Knudsena, Line Bjørgeb,c, Seppo Merid, Christian Vedelera a Department of Neurology, Haukeland University Hospital, N-5021 Bergen, Norway Department of Gynecology and Obstetrics, Haukeland University Hospital, Bergen, Norway c Department of Microbiology and Immunology, The Gade Institute, University of Bergen, Bergen, Norway d Department of Bacteriology and Immunology, University of Helsinki, Helsinki, Finland b

Received 23 September 2003; received in revised form 26 January 2004; accepted 29 January 2004

Abstract The expression of CD59 and other complement regulators was studied in human cerebellum from 14 individuals with no cerebellar pathology, from one patient with multiple sclerosis (MS) and from two patients with paraneoplastic cerebellar degeneration (PCD). CD59 was present on the Purkinje cells at various levels in eight of the 14 cases with no cerebellar pathology. CD59 was also present on the Purkinje cells of the patient with MS, but not on the scarce remaining Purkinje cells of the two patients with PCD. Other complement regulators (CD35, CD46 and CD55) were not expressed on the Purkinje cells, whereas CD59, CD46 and CD55 were present on the molecular, granulosa and endothelial cells. The results suggest that Purkinje cells not expressing CD59 could be especially prone to complement-mediated damage. q 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Purkinje cells; Cerebellum; Complement; CD59; Multiple sclerosis; Paraneoplastic disease

Complement regulators are proteins that are either expressed on the cell surface or present in body fluids. The membrane-bound proteins are divided into two groups: CD35 (complement receptor 1), CD46 (membrane cofactor protein) and CD55 (decay accelerating factor) which inhibit the key enzymes in the early stages of complement activation, C3/C5 convertases, and CD59 (protectin) which inhibits the formation of the membrane attack complex (MAC) [8]. In immunohistochemistry and in situ hybridization studies CD46, CD55 and CD59 expression has been demonstrated on cells from numerous human tissues, whereas the expression of CD35 is more limited, as this regulator is mainly present on circulating blood cells and on glomerular podocytes [8]. The complement regulators are crucial in protecting autologous cells from complement-mediated damage. This has been demonstrated in mice, where the knockout of one of the regulators, Crry, resulted in complement-mediated rejection of the fetoplacental unit in utero [7]. The presence of soluble Crry in cerebrospinal fluid, however, protects * Corresponding author. Tel.: þ 47-55-975044; fax: þ 47-55-975164. E-mail address: [email protected] (A. Storstein).

transgenic mice from developing neurological deficits in experimental allergic encephalomyelitis [7]. Neurons express low levels of CD46 and CD59, while microglia express CD55 and CD59 [11,15]. Whether neurons in vivo can be induced to express increased levels of complement regulators is not known, although this has been demonstrated in vitro [19]. The limited expression of complement regulators in the central nervous system and the mounting evidence of local synthesis of complement components by activated glial cells are likely to be important factors for the risk of complement-mediated damage to neuronal cells [14]. Deficiency of CD59 has been found to contribute to neurodegeneration in Alzheimer’s disease [18]. So far little attention has been paid to complement regulators in the cerebellum. However, mRNA of the soluble MAC inhibitor vitronectin has been demonstrated in Purkinje cells from both healthy individuals and from patients with Alzheimer’s and Parkinson’s disease [17]. To evaluate the possible vulnerability of Purkinje cells to complement-mediated damage, we studied the presence of CD59 and other complement-regulatory proteins on cells of the cerebellum from 14 patients without any disease of the nervous system, one patient with multiple sclerosis (MS)

0304-3940/03/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2004.01.078

22

A. Storstein et al. / Neuroscience Letters 362 (2004) 21–25

and two patients with paraneoplastic cerebellar degeneration (PCD). Cerebellar tissue from 17 patients was obtained at autopsies performed at the Department of Pathology, Haukeland University Hospital, Bergen, Norway (Table 1). The tissue samples were obtained within 28 – 100 h after death (median 40 h). Autopsies revealed no cerebellar pathology in patients 1 – 14. Patient 15 had extensive cerebellar demyelination consistent with secondary progressive MS, and patients 16 and 17 had PCD with the presence of Purkinje cell antibodies (Yo antibodies) and widespread loss of Purkinje cells. Mouse monoclonal antibodies (mAbs) to human CD59 (BRIC 229) and CD55 (BRIC 216 and BRIC 230), to human CD46 (J4-48), and to human CD35 (M07510) were purchased from Bioproducts Laboratories (Elstree, UK), Serotec (Oxford, UK) and Dako (Glostrup, Denmark), respectively. Mouse anti-human CD59 mAbs HELC-1, HELC-2 and HELC-3 as well as rat anti-human CD59 mAb YTH53.1 were obtained from Professor S. Meri, University of Helsinki, Finland. Goat anti-human CD59 polyclonal antibody (pAb) was provided by Professor F. Tedesco, University of Trieste, Italy. Peroxidase-conjugated rabbit anti-mouse and rabbit anti-rat antibodies, as well as an EnVisione þ kit were purchased from Dako (Glostrup, Denmark). Peroxidase-conjugated rabbit anti-goat antibody was purchased from Sigma (St. Louis, MO). Tissue samples were obtained from both cerebellar hemispheres and snap-frozen in isopentane precooled in

liquid nitrogen. Cryostat sections, 8 mm thick, were fixed in ice-cold acetone for 5 min and then washed in phosphatebuffered saline (PBS, pH 7.2) for 5 min. The cerebellar sections were incubated with 1 – 10 mg/ml of the antibodies against the various complement regulators. The sections were then washed in PBS and incubated with peroxidaseconjugated secondary antibodies for 1 h at room temperature. The EnVisione þ kit was used according to the instructions of the manufacturer. Sections were then washed in PBS, mounted and examined by light microscopy. Cytospins of human leukocytes were used as positive controls for all the primary antibodies, except for anti-CD35 which stained glomeruli in human renal sections. All experiments included negative controls without primary antibodies. The anti-CD59 mAbs BRIC 229, HELC-2, HELC-3 and YTH53.1, but not HELC-1, stained the surface membrane of all the Purkinje cells in four of the 14 control cerebelli. The pAb also stained the cytoplasm of all the Purkinje cells in these specimens (Fig. 1b). In a further four cerebelli, CD59 was only detected on the Purkinje cells using the pAb with no staining of the anti-CD59 mAbs. In the remaining six cerebelli, CD59 staining was not found on the Purkinje cells using any of the antibodies (Fig. 2). Similar results were obtained using enhanced secondary expression with the EnVisione þ kit. The CD59 expression was unrelated to sex, age, pre-existing diagnosis, cause of death or postmortem time before autopsy (Table 1). The various mAbs against CD59 (except for HELC-1)

Table 1 Patient characteristics and CD59 expression on Purkinje cells Patient no.

Sex/age

Cause of death

Pre-existing illness

1 2

M/73 M/83

Pneumonia Pontine infarction

3 4 5 6 7 8 9

F/71 F/71 M/51 M/81 F/70 M/52 F/71

Pneumonia Coronary infarction Endotracheal hemorrhage Renal insufficiency Frontal cerebral hemorrhage Perforation of small intestine Pulmonary embolus

10 11

M/79 M/55

Coronary insufficiency Pneumonia

12

M/67

Coronary infarction

13 14 15

F/56 M/70 F/46

Pneumonia Coronary infarction Pneumonia

16

F/84

Pneumonia

17

F/55

Pneumonia

Pulmonary cancer Atherosclerotic coronary disease Metastatic pulmonary cancer Diabetes mellitus Metastatic pulmonary cancer Diabetes, colon cancer Pericarditis, rheumatoid arthritis Metastatic malignant melanoma Chronic obstructive lung disease, aortic valve stenosis Atherosclerotic coronary disease Chronic obstructive lung disease, coronary insufficiency Metastatic small cell lung cancer, renal insufficiency Metastatic cancer of the urinary bladder Atherosclerotic coronary disease Multiple sclerosis, cerebellar demyelination Ovarian cancer, paraneoplastic cerebellar degeneration Ovarian cancer, paraneoplastic cerebellar degeneration

Hours postmortem

Monoclonal anti-CD59

Polyclonal anti-CD59

28 37

Positive Positive

Positive Positive

70 31 28 100 52 29 46

Positive Positive Negative Negative Negative Negative Negative

Positive Positive Positive Positive Positive Positive Negative

40 55

Negative Negative

Negative Negative

40

Negative

Negative

30 48 33

Negative Negative Positive

Negative Negative Positive

100

Negative

Negative

72

Negative

Negative

A. Storstein et al. / Neuroscience Letters 362 (2004) 21–25

23

Fig. 1. (a,b) Membrane and cytoplasmic staining of Purkinje cells (arrows) using a polyclonal antibody against CD59. Bar, 30 mm.

and the pAb also stained the Purkinje cells of the patient with MS. None of the CD59 antibodies stained the Purkinje cells of the two patients with PCD (Table 1). The mAbs against the other complement-regulatory proteins CD35, CD46 and CD55 did not stain the Purkinje cells in any of the 17 specimens, not even when the enhanced secondary expression method was used. CD59 was expressed to an equal extent on the molecular, granulosa and endothelial cells in the cerebellar cortex from the 17 patients using all of the mAbs and the pAb. CD46 and CD55 were weakly expressed in the molecular layer, but abundantly expressed in the granular layer and on the endothelium in both layers of the cerebellar cortex in all

specimens tested. CD35 was not expressed in any of the cerebellar specimens except for intravascular staining. The present study demonstrates that CD59 is expressed on some human Purkinje cells, but not all. This difference was evident using both mAbs and a pAb together with a sensitive immunocytochemical method [10]. In some specimens, CD59 was only detected using the pAb. Such heterogeneity could result from masking of epitopes detected by the various mAbs, but more likely results from variable expression of CD59. In contrast, we found no evidence of variable expression of CD59 on the granulosa, molecular or endothelial cells. Heterogeneous expression of CD59 has been described

24

A. Storstein et al. / Neuroscience Letters 362 (2004) 21–25

Fig. 2. Purkinje cells not stained by a polyclonal antibody against CD59. Bar, 30 mm.

in tumor cell lines [3], but not in single, normal cell types. Thus, whether genetic regulation exists for CD59 expression on the Purkinje cells is unknown. The expression of CD35 on human glomerular podocytes and erythrocytes is, however, genetically determined [5,6]. Genetic regulation of CD59 expression could not be studied further in our specimens due to the long interval between death and tissue removal and instability of mRNA in postmortem samples. Interestingly, we found a variation in the staining pattern of the different anti-CD59 mAbs. The HELC-1 mAb stained the granulosa, molecular and endothelial cells, but not the Purkinje cells. This mAb has no CD59-blocking activity, and therefore the epitope probably lies in a region other than the active site. Further, the reactivity of the HELC-1 mAb may be affected by differences in the glycosylation of CD59 [2]. The BRIC 229, HELC-2, HELC-3 and YTH53.1 mAbs are blocking antibodies and therefore likely to overlap with the active site of CD59. These mAbs gave a similar staining pattern of the Purkinje cells. The results suggest that polymorphic variants of the CD59, such as variation in glycosylation, may be expressed on different cells in the human cerebellum. Polymorphisms in the CD59 expression have been described in mice, where there is a CD59 gene duplication. The two forms of murine CD59, CD59a and CD59b, are simultaneously expressed in multiple tissues, but in the testes, CD59a is expressed exclusively in spermatids, whereas CD59b is expressed in more mature sperm cells [9]. This suggests that the two gene products may have different functions. Although the number of samples studied was limited, the varying CD59 expression did not correlate with sex, age, pre-existing diagnosis, cause of death or the postmortem time to autopsy. Nevertheless, the expression of CD59 on the Purkinje cells could be influenced by cytokines or other regulating factors. Tumor necrosis factor-a, interleukin-1b

and interleukin-6 upregulate the expression of CD59 on cell surfaces, whereas others, such as interferon-g, may antagonize this effect [13]. We have shown that CD59 is upregulated on Schwann cells in experimental allergic neuritis, an effect that may be induced by proinflammatory cytokines [16]. In the present study we found CD59 on Purkinje cells in the cerebellum from a patient with MS, but not in the cerebelli from two patients with PCD. Whilst this may simply reflect the variation we see in non-disease tissue samples, it may also suggest that cytokines released in some cancers and autoimmune diseases may cross the blood – brain barrier and influence the expression of CD59 on the Purkinje cells. Whether complement activation, cytotoxic T lymphocytes or other factors are responsible for the loss of Purkinje cells in PCD is still not known. Anti-Purkinje cell antibodies (anti-Yo antibodies) are primarily of the IgG1 subclass and able to fixate complement [1]. We have found deposits of MAC in the cerebellum of PCD patients (unpublished data), indicating that complement activation has taken place. Deficiency of CD59 could, therefore, be one explanation for the loss of Purkinje cells. Alternatively, the CD59 deficiency is a secondary effect of other factors mediating the degeneration of the Purkinje cells in PCD. We found that Purkinje cells did not express the complement regulators CD35, CD46 or CD55, whilst CD46 and CD55 were present on other cells in the cerebellar cortex, such as the molecular, granulosa and endothelial cells. The findings are in line with other reports showing that the expression of complement regulators on neurons in other parts of the brain is sparse [11], but that microglia and astrocytes express CD59, CD55 and CD46 as well as soluble complement inhibitors [12]. This difference may, in part, explain why the area around the Purkinje cells is particularly susceptible to ischemia-induced MAC attack [4]. In conclusion, the expression of complement regulators on human Purkinje cells is very limited, rendering these neurons highly vulnerable to complement-mediated killing. Our results suggest that this susceptibility may be heterogeneous and that Purkinje cells with no CD59 expression may be especially prone to complement attack. Whether this is a predisposing factor for PCD or other inflammatory or degenerative diseases of the cerebellum remains to be determined.

References [1] E. Amyes, J. Curnow, Z. Stark, L. Corlett, I. Sutton, A. Vincent, Restricted IgG1 subclass of anti-Yo antibodies in paraneoplastic cerebellar degeneration, J. Neuroimmunol. 114 (2001) 259–264. [2] D.L. Bodian, S.J. Davis, B.P. Morgan, N.K. Rushmere, Mutational analysis of the active site and antibody epitopes of the complementinhibitory glycoprotein CD59, J. Exp. Med. 185 (1997) 507–516. [3] A. Gorter, S. Meri, Immune evasion of tumor cells using membrane-

A. Storstein et al. / Neuroscience Letters 362 (2004) 21–25

[4]

[5]

[6]

[7]

[8]

[9]

[10]

bound complement regulatory proteins, Immunol. Today 20 (1999) 576–582. P.J. Lindsberg, J. Ohman, T. Lehto, M.L. Karjalainen-Lindsberg, T. Paeta Wuorimaa, O. Carpen, M. Kaste, S. Meri, Complement activation in the central nervous system following blood-brain barrier damage in man, Ann. Neurol. 40 (1996) 587–596. R. Matre, C.A. Vedeler, O.J. Halvorsen, Expression of glomerular C3b receptor activity in normal human kidneys, Immunol. Lett. 14 (1986) 3231–3232. R. Matre, C.A. Vedeler, Demonstration of human erythrocyte C3b receptor (CR1) by haemadsorption and indirect haemagglutination techniques, J. Immunol. Methods 96 (1987) 139–144. H. Molina, The murine complement regulator Crry: new insights into the immunobiology of complement regulation, Cell. Mol. Life Sci. 59 (2002) 220–229. B.P. Morgan, S. Meri, Membrane proteins that protect against complement lysis, Springer Semin. Immunopathol. 15 (1994) 369–396. X. Qin, T. Miwa, H. Aktas, M. Gao, C. Lee, Y.M. Qian, C.C. Morton, A. Shahsafaei, W.C. Song, J.A. Halperin, Genomic structure, functional comparison, and tissue distribution of mouse Cd59a and Cd59b, Mamm. Genome 12 (2001) 582 –789. E. Sabattini, K. Bisgaard, S. Ascani, S. Poggi, M. Piccioli, C. Ceccarelli, F. Pieri, G. Fraternali-Orcioni, S.A. Pileri, The EnVisione þ system: a new immunohistochemical method for diagnostics and research. Critical comparison with the APAAP, ChemMatee, CSA, LABC, and SABC techniques, J. Clin. Pathol. 51 (1998) 506–511.

25

[11] S.K. Singhrao, J.W. Neal, N.K. Rushmere, B.P. Morgan, P. Gasque, Differential expression of individual complement regulators in the brain and choroid plexus, Lab. Invest. 79 (1999) 1247–1259. [12] C. Speth, M.P. Dierich, P. Gasque, Neuroinvasion by pathogens: a key role of the complement system, Mol. Immunol. 38 (2001) 669 –679. [13] O.B. Spiller, O. Criado-Garcia, S.R. de Cordoba, B.P. Morgan, Cytokine-mediated up-regulation of CD55 and CD59 protects human hepatoma cells from complement attack, Clin. Exp. Immunol. 121 (2000) 234–241. [14] A. Thomas, P. Gasque, D. Vaudry, B. Gonzales, M. Fontaine, Expression of a complete and functional complement system by human neuronal cells in vitro, Int. Immunol. 12 (2000) 1015–1023. [15] C. Vedeler, E. Ulvestad, L. Bjørge, G. Conti, K. Williams, S. Mørk, R. Matre, The expression of CD59 in normal human nervous tissue, Immunology 82 (1994) 542 –547. [16] C.A. Vedeler, G. Conti, T. Fujioka, E. Scarpini, A. Rostami, The expression of CD59 in experimental allergic neuritis, J. Neurol. Sci. 165 (1999) 154 –159. [17] D.G. Walker, P.L. McGeer, Vitronectin expression in Purkinje cells in the human cerebellum, Neurosci. Lett. 251 (1998) 109 –112. [18] L.-B. Yang, L. Rena, S. Meri, J. Rogers, Y. Shen, Deficiency of complement defense protein CD59 may contribute to neurodegeneration in Alzheimer’s disease, J. Neurosci. 20 (2000) 7505–7509. [19] K.-Z. Zhang, S. Junnikkala, M.G. Erlander, G. Hongqing, J.A. Westberg, S. Meri, L.C. Andersson, Up-regulated expression of decay-accelerating factor (CD55) confers increased complement resistance to sprouting neural cells, Eur. J. Immunol. 28 (1998) 1189– 1196.