The Journal of Comparative Neurology 513:1–20 (2009)
Autoantibodies in Autoimmune Polyglandular Syndrome Type I Patients React with Major Brain Neurotransmitter Systems SERGUEI¨ O. FETISSOV,1* SOPHIE BENSING,2 JAN MULDER,1 ERWAN LE MAITRE,1 ANNA-LENA HULTING,2 ¨ LDBERG,4 EYSTEIN S. HUSEBYE,6 TIBOR HARKANY,3 OLOV EKWALL,4,5 FILIP SKO ¨ MPE,4 AND TOMAS HO ¨ KFELT1* JAAKKO PERHEENTUPA,7 FREDRIK RORSMAN,4 OLLE KA 1 Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden 2 Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden 3 Department of Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden 4 Department of Medical Sciences, University Hospital, Uppsala University, Uppsala, Sweden 5 Department of Paediatrics, The Sahlgrenska Institute at Go¨teborg University, Go¨teborg, Sweden 6 Institute of Medicine, University of Bergen, and Department of Medicine, Haukeland University Hospital, Bergen, Norway 7 The Hospital for Children and Adolescents, Helsinki University Hospital, Helsinki, Finland
ABSTRACT Patients with autoimmune polyglandular syndrome type I (APS1) often display high titers of autoantibodies (autoAbs) directed against aromatic L-amino acid decarboxylase (AADC), tyrosine hydroxylase (TH), tryptophan hydroxylase (TPH), and glutamic acid decarboxylase (GAD). Neurological symptoms, including stiff-man syndrome and cerebellar ataxia, can occur in subjects with high levels of GAD autoAbs, particularly when patient sera can immunohistochemically stain ␥-aminobutyric acid (GABA) neurons. However, it was not known if APS1 sera can also stain major monoamine systems in the brain. Therefore, in this work we applied sera from 17 APS1 patients known to contain autoAbs against AADC, TH, TPH, and/or GAD to rat brain sections and processed the sections according to the sensitive immunohistochemical tyramide signal amplification method. We found that autoAbs
in sera from 11 patients were able to stain AADC-containing dopaminergic, serotonergic, and noradrenergic as well as AADC only (D-group) neurons and fibers in the rat brain, in several cases with a remarkably high quality and sensitivity (dilution up to 1:1,000,000); and, since they are human antibodies, they offer a good opportunity for performing multiplelabeling experiments using antibodies from other species. Six APS1 sera also stained GABAergic neuronal circuitries. Similar results were obtained in the mouse and primate brain. Our data demonstrate that many APS1 sera can immunostain the major monoamine and GABA systems in the brain. Only in a few cases, however, there was evidence that these autoAbs can be associated with neurological manifestations in APS1 patients, as, e.g., shown in previous studies in stiff-man syndrome. J. Comp. Neurol. 513:1–20, 2009. © 2008 Wiley-Liss, Inc.
Indexing terms: dopamine; serotonin; noradrenaline; enzymes; monoamines; autoimmunity
Autoimmune polyglandular syndrome type I (APS1), also called autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED), is an autosomal, recessively inherited disease. It is now known that the underlying cause is mutations in a gene on human chromosome 21 (Ahonen et al., 1990) and in the mouse on chromosome 10 (Mittaz et al., 1999). This gene encodes a transcription factor named AIRE (autoimmune regulator) (Aaltonen et al., 1997; Nagamine et al., 1997). The syndrome is associated with several clinical manifestations including chronic mucocutaneous candidiasis, hypoparathyroidism, and Addison’s disease (Neufeld et al., 1980). It has been recognized that APS1 patients have high titer of autoantibodies (autoAbs) against a number of important enzymes involved in the monoamine biosynthesis (Winqvist et al., 1992; Rorsman et al., 1995; Tuomi et al., 1996; Gebre-Medhin et al., 1997; Ekwall et al., 1998, 1999; Hedstrand et al., 2000; Peterson et al., 2000; So¨derbergh et al., 2004), including aromatic L-amino acid decarboxylase (AADC), a pyridoxal-dependent decarboxylase (Husebye et al., 1997; So¨derbergh et al., 2000) and tyrosine hydroxylase (TH)
© 2008 Wiley-Liss, Inc.
Additional supporting information may be found in the online version of this article. Grant sponsor: Swedish Research Council; Grant sponsor: Torsten and Ragnar So¨derberg Foundation; Grant sponsor: Marianne and Marcus Wallenberg Foundation; Grant sponsor: Knut and Alice Wallenberg Foundation; Grant sponsor: European Union (EU); Grant number: NEWMOOD, LSHMCT-2004-503474; Grant sponsor: EU FP6 Program on rare diseases, project entitled EurAPS; Grant sponsor: Alzheimer’s Association (to T. Harkany); Grant sponsor: European Union; Grant number: MEMOLOAD, FP2-2007-201159 (to T. Harkany); Grant sponsor: Alzheimer’s Research Trust UK (to J.M.). *Correspondence to: Sergueı¨ O. Fetissov, ADEN Laboratory, Faculte´ de Me´decine-Pharmacie, 22, Bld Gambetta, Rouen, 76183 Cedex 1, France. E-mail:
[email protected] or Tomas Ho¨kfelt, Department of Neuroscience, Karolinska Institutet, Retzius va¨g. 8, 17177, Stockholm, Sweden. E-mail:
[email protected] Received 31 December 2007; Revised 22 May 2008; Accepted 15 October 2008 DOI 10.1002/cne.21913 Published online in Wiley InterScience (www.interscience.wiley.com).
The Journal of Comparative Neurology 2
S.O. FETISSOV ET AL.
(Hedstrand et al., 2000) and tryptophan hydroxylase (TPH) (Ekwall et al., 1998, 1999), two tetrahydropterin-dependent hydroxylases. AutoAbs against these enzymes can be used as markers for APS1 and their presence is predictable for different manifestations of this disease. For example, autoAbs directed against TPH have been associated with gastrointestinal dysfunction (Ekwall et al., 1998, 1999), autoAbs to TH correlate with Alopecia areata (Hedstrand et al., 2000), while AADC autoAbs are strongly associated with autoimmune hepatitis (So¨derbergh et al., 2004). Another type of autoAbs frequently found in APS1 patients is directed against glutamic acid decarboxylase (GAD) (Christie et al., 1994; Lernmark, 1996), a pyridoxal-dependent decarboxylase, the enzyme synthesizing the inhibitory neurotransmitter ␥-aminobutyric acid (GABA). Increased levels of GAD autoAbs were found in insulin-dependent diabetes mellitus (IDDM) (Baekkeskov et al., 1990) and some neurological diseases such as stiff-man syndrome (SMS) (Solimena et al., 1988), epilepsy, and cerebellar ataxia (Honnorat et al., 1995). Interestingly, most of the patients with cerebellar ataxia are females with IDDM and polyendocrine autoimmunity having a high titer of GAD autoAbs (Honnorat et al., 2001). Such autoAbs have been used in immunohistochemical studies to stain GABA neurons in sections of rat brain (Solimena et al., 1988). In fact, this staining can only be obtained when using sera from subjects presenting neurological symptoms (Vianello et al., 2005). Thus, it appears that there is a link between GAD autoAbs and neurological manifestations relevant to the functioning of GABA neurons in the central nervous system. Fur-
thermore, using animal models, a recent study showed that certain GAD autoAbs can be causative for neuronal dysfunctions similar to cerebellar ataxia and SMS (Manto et al., 2007). The immunohistochemical studies of Solimena et al. (1988) raise the possibility that sera from APS1 patients could also be of interest for immunohistochemical analysis of peripheral and central monoamine and GABA neuron systems. In fact, Ekwall et al. (1998) showed that serum from an APS1 patient stained 5-hydroxytryptamine (5-HT)-positive enterochromaffin cells in the small intestine epithelium, and Cocco et al. (2005) reported that an APS1 serum stains presumable dopamine (DA) terminals in the median eminence of the rat hypothalamus. In the present study we therefore analyzed to what extent APS1 sera can be used to detect major transmitter system in the rat brain. We applied a number of APS1 patient sera onto formalin/picric acid-fixed rat brain sections and compared these staining patterns with, in particular, the well-established distribution of DA, noradrenaline (NA), 5-HT (5-hydroxytryptamine or serotonin), and adrenaline neurons (Dahlstro¨m and Fuxe, 1964; Bjo¨rklund and Lindvall, 1984; Ho¨kfelt et al., 1984a,b; Moore and Card, 1984; Steinbusch, 1984) as well as GABA neurons (Ho¨kfelt and Ljungdahl, 1975; Ottersen and Storm-Mathisen, 1984; Mugnaini and Oertel, 1985). It was shown early with histochemical approaches that, as expected, both AADC protein (Ho¨kfelt et al., 1973) and transcript (Tison et al., 1991) are present both in catecholamine (CA) and 5-HT neurons.
General abbreviations AIRE APECED autoAbs APS1 AADC BSA BT-FITC CA CSF CR DA GABA DIC GAD GFP HRP IDDM ir IR ITT NCBI NA PFA PB PBS PV RRX RT SMS TH TPH TSA 5-HT
Autoimmune regulator Autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy Autoantibodies Autoimmune polyglandular syndrome type I Aromatic L-amino acid decarboxylase Bovine serum albumin Biotinyl tyramide-fluorescein Catecholamine Cerebro-spinal fluid Calretinin Dopamine ␥-aminobutyric acid Differential interference contrast Glutamic acid decarboxylase Green fluorescent protein Horseradish peroxidase Insulin-dependent diabetes mellitus Immunoreactive Immunoreactivity In vitro transcription and translation National Center for Biotechnology Information Noradrenaline Paraformaldehyde Phosphate buffer Phosphate-buffered saline Parvalbumin Rhodamine Red X Room temperature Stiff-man syndrome Tyrosine hydroxylase Tryptophan hydroxylase Tyramide signal amplification 5-hydroxytryptamine or serotonin
Anatomical abbreviations AHN Arc Aq BLA cp CA3 CPu Crb Cx DG DRN f GP Gr hbc IP IPL LC ME MFB MPT ot Po PVN py Pyr RM SNC SNR VTA 3V 4V
Anterior hypothalamic nucleus Arcuate nucleus Cerebral aqueduct Basolateral nucleus of amygdala Cerebral peduncle Ammon’s horn field CA3 Caudate putamen Cerebellum Cerebral cortex Dentate gyrus Dorsal raphe nucleus Fornix Globus pallidus Granular layer Habenular commissure Interpeduncular nucleus Intermediate pituitary lobe Locus coeruleus Median eminence Medial forebrain bundle Medial pretectal area Optic tract Polymorph layer Paraventricular nucleus Pyramidal tract Pyramidal cell layer Nucleus raphe magnus Substantia nigra pars compacta Substantia nigra pars reticulata Ventral tegmental area 3rd ventricle 4th ventricle
The Journal of Comparative Neurology APS1 AUTOANTIBODIES STAIN MAJOR BRAIN SYSTEMS TABLE 1. According to Immunostaining, Sera of 17 APS1 Patients Were Divided Into Four Groups (See Results)
APS1 patient no. Group 1 2 3 4 5 6 Group 7 8 9 Group 10 11 12 Group 13 14 15 16 17
In vitro detected autoantibodies against:
Immunostaining in the rat brain
AADC
TH
TPH
GAD
AADC/TH/ TPH
GAD
ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ
ⴙ ⴙ ⴚ ⴚ ⴚ ⴙ
ⴙ ⴙ ⴙ ⴙ ⴚ ⴙ
ⴙ ⴙ ⴙ ⴙ ⴚ ⴙ
ⴙⴙⴙ ⴙⴙⴙ ⴙⴙⴙ ⴙⴙⴙ ⴙⴙⴙ ⴙⴙ
ⴚ ⴚ ⴚ ⴚ ⴚ ⴚ
ⴙ ⴙ ⴙ
ⴙ ⴙ ⴚ
ⴙ ⴙ ⴙ
ⴙ ⴙ ⴙ
ⴙⴙⴙ ⴙⴙⴙ ⴙⴙⴙ
ⴙⴙⴙ ⴙⴙⴙ ⴙⴙ
ⴙ ⴙ ⴙ
ⴙ ⴚ ⴚ
ⴙ ⴙ ⴙ
ⴙ ⴙ ⴙ
ⴙ/ⴚ ⴚ ⴙ/ⴚ
ⴙⴙⴙ ⴙⴙ ⴙⴙ
ⴚ ⴚ ⴚ ⴙ ⴙ
ⴚ ⴙ ⴙ ⴚ ⴚ
ⴙ ⴚ ⴚ ⴚ ⴙ
ⴚ ⴚ ⴚ ⴚ ⴙ
ⴚ ⴚ ⴚ ⴚ ⴚ
ⴚ ⴚ ⴚ ⴚ ⴚ
I
II
III
IV
Sign (ⴙ) or (ⴚ) for in vitro detection reflect presence or absence, respectively, of autoantibody against corresponding enzyme. In the last two columns subjective estimation of intensity of immunostaining was made in the microscope using a rating scale from not detectable (ⴚ), very low (ⴙ/ⴚ), low (ⴙ), medium (ⴙⴙ), or high (ⴙⴙⴙ) levels.
Brains from mouse and Microcebus murinus (gray mouse lemur), a prosimian primate, were also examined in a preliminary way.
MATERIALS AND METHODS APS1 sera Serum samples from 10 Swedish, six Norwegian, and one Finnish patient with APS1 were analyzed by immunohistochemistry (Table 1). Data on antibody presence in these patients as revealed by immunoprecipitation with recombinant GAD, AADC, TH, and TPH were available from previous studies (Ekwall et al., 1998; Hedstrand et al., 2000; So¨derbergh et al., 2004). Additionally, a sample of the cerebrospinal fluid (CSF) was obtained from Patient #8. All serum or CSF samples were collected with the informed consent from the patients and the study was approved by the local ethical committees at Uppsala University and Karolinska Institutet.
Immunohistochemistry Animals and tissue preparation. Rats. The experiments were performed on 10 male Sprague–Dawley rats (body weight 250 –350 g) (Scanbur BK, Stockholm, Sweden). The animals were maintained under standard conditions on a 12-hour day/night cycle (lights on 07:00) with water and food available ad libitum. The rats were deeply anesthetized using sodium pentobarbital (60 mg/kg i.p.; Apoteket, Stockholm, Sweden) and perfused via the ascending aorta with 60 mL of Tyrode’s buffer (37°C), followed by 60 mL of a mixture of 4% paraformaldehyde (PFA) and 0.2% picric acid diluted in 0.16 M phosphate buffer (PB) (pH 6.9) (Pease, 1962; Zamboni and De Martino, 1967) and 300 mL of the same fixative at 4°C, the latter for ⬇5–7 minutes. The brain was dissected out and postfixed in the same fixative for 90 minutes at 4°C, and finally immersed in 10% sucrose diluted in phosphate-buffered saline (PBS) (pH 7.4) containing
3 0.01% sodium azide (Sigma-Aldrich, St. Louis, MO) and 0.02% Bacitracin (Sigma-Aldrich) (4°C) for 48 hours. The brains were snap-frozen with CO2 and sectioned at 14 m in a cryostat (Microm, Heidelberg, Germany). The sections were then mounted on aluminum gelatin-coated slides. Mice. GAD-green fluorescent protein (GFP) heterozygous (⌬neo) knockin (GAD67gfp/ⴙ) mice were provided by Dr. Yuchio Yanagawa (Department of Morphological Brain Science, Kyoto University, Japan). Mice were maintained under standard conditions on a 12-hour day/night cycle (lights on 07:00) with water and food available ad libitum. The mice (n ⴝ 2) were deeply anesthetized in isoflurane in 30% N2O/70% O2 (5%, v/v%, at 1 L/min flow), and transcardially perfused with 15 mL of ice-cold physiological saline followed by 100 mL of 4% PFA in 0.1 M PB (pH 7.4). Whole brains were postfixed in 4% PFA overnight, cryoprotected in 30% sucrose in 0.9% NaCl, and serial 50-m coronal sections were prepared on a cryostat microtome. The generation and genotyping of heterozygous GAD67gfp/ⴙ mice was performed as described (Tamamaki et al., 2003). Briefly, Tamamaki et al. (2003) generated these mice by inserting GFP cDNA in the GAD67 promoter between the GAD67 5ⴕ flanking region and the GAD67 codon. In this way, transgenic mice had a single copy gene for GAD67 expression and a single-copy gene for GFP expression, and both of them were connected to the same promoter, enhancer, and suppressor in the introns or in the 5ⴕ and 3ⴕ flanking region in the GAD67 allele, assuring the parallel pattern of expression of both GFP and GAD67 proteins (Tamamaki et al., 2003). Indeed, as verified in the original (Tamamaki et al., 2003) and subsequent neuroanatomical studies both in embryonic and adult mice (Tanaka et al., 2006; Manent et al., 2006; Berghuis et al., 2007), the vast majority of GFP-positive cells had GAD67 immunoreactivity in their perikarya, while no ectopic GFP expression was found and no significant changes in the distribution, cellular morphology, or migratory behavior of GABAergic neurons were noticed. Furthermore, recent data by Esumi et al. (2008) showed that GFP-positive neurons in these mice express GAD67 mRNA. Overall, the above results validate this animal model to assess the functional and morphological complexity of GABAergic circuitries in the mouse brain. Gray mouse lemurs (Microcebus murinus). These primates were born in a laboratory breeding colony at Brunoy (France, ethical approval: #962773) (Harkany et al., 2005) from a stock originally caught on the southwest coast of Madagascar 35 years ago. The animals were maintained under ambient temperature (24 –26°C), 55% relative humidity, while food was available ad libitum. Seasonal variations of physiological functions were routinely entrained by alternating periods of long and short days with light/dark cycles of 14/10 hours and 10/14 hours, respectively. We used young female animals (n ⴝ 2) with ages of 1.77 and 3.87 years. All animals were anesthetized during the long daylight period with an i.p. injection of sodium pentobarbital (100 mg/kg) and transcardially perfused with 50 mL of ice-cold physiological saline followed by 80 mL of 4% PFA in 0.1 M PB (pH 7.4). Whole brains were postfixed in 4% PFA overnight, cryoprotected in 30% sucrose in physiological saline, and series of 50-m coronal sections were prepared on a freezing microtome. Single staining (TSA). The rat brain sections were washed in PBS and incubated overnight at 4°C with one of the 17 APS1 sera diluted 1:2,000 –1:100,000, or with CSF (Patient #8)
The Journal of Comparative Neurology 4 diluted 1:100. Some sera were diluted up to 1:1,000,000 to test sensitivity. To visualize the immunoreactivity the sections were processed using a commercial kit (PerkinElmer Life Science, Boston, MA) based on tyramide signal amplification (TSA) (Adams, 1992). Briefly, the sections were washed in TNT buffer (0.1 M Tris-HCl, pH 7.5; 0.15 M NaCl; 0.05% Tween 20) for 15 minutes, incubated with TNB buffer (0.1 M Tris-HCl, pH 7.5; 0.15 M NaCl; 0.5% Dupont Blocking Reagent, PerkinElmer) for 30 minutes at room temperature (RT), and incubated with a rabbit antihuman or donkey antihuman IgG, all coupled to horseradish peroxidase (HRP) (Dako, Copenhagen, Denmark) diluted 1:200 in TNB buffer for 30 minutes. The sections were washed in TNT buffer and incubated in a biotinyl tyramide-fluorescein (BT-FITC) conjugate (PerkinElmer) diluted 1:100 in amplification diluent for 10 minutes at RT. The specificity of the binding was tested by preadsorption of the sera (1:2000 –1:50,000) with 120 000 cpm [35S]radiolabeled human GAD, AADC, TH, or TPH expressed in vitro as described below. Double staining. After single-staining with APS1 patient sera using the TSA kit, the rat brain sections were washed in PBS and incubated overnight (or over 2 nights) at 4°C with a rabbit TH antibody (1:400; Chemicon International, Temecula, CA) or a guinea pig 5-HT antibody (1:400; Steinbusch et al., 1983), a mouse monoclonal GAD67 antibody (1:500; Chemicon, Temecula, CA), a rabbit parvalbumin antibody (1:1,000, Swant, Bellinzona, Switzerland), or a rabbit calretinin antibody (1:1,000; Atlas Antibodies, AlbaNova University Center, Stockholm, Sweden) for the conventional immunohistochemistry procedure (Coons, 1958). After washes in PBS the sections were further incubated using a Rhodamine Red X (RRX)conjugated donkey antirabbit, antimouse or antiguinea pig antibody (1:200; Jackson ImmunoResearch, West Grove, PA). Multiple staining in mice and Microcebus. Free-floating sections were preincubated with 5% normal donkey serum (Jackson ImmunoResearch), 2% bovine serum albumin (BSA; Sigma-Aldrich), and 0.3% Triton X-100 in PB for 1 hour at RT. Sections were then exposed to a cocktail of primary antibodies including serum from APS1 Patient #1 (1:2,000) in combination with rabbit anti-TH (1:4,000; Markey et al., 1980), goat anti-parvalbumin (1:1,000, Swant), rabbit anti-parvalbumin (1: 1,000, Swant), mouse anti-calbindin (1:1,000, Swant) diluted in PB containing 0.3% Triton X-100, 0.1% BSA, and 1% normal donkey serum for 48 hours at 4°C. After extensive rinsing in PB, sections were incubated with a mixture of carbocyanine conjugated (Cy2, Cy3, Cy5) antibodies against human, rabbit, goat, and mouse (all raised in donkey; 9 g/mL; Jackson ImmunoResearch) that were diluted in PB containing 2% BSA for 2 hours at RT. Because extensive myelination and lipofuscin accumulation may cause staining artifacts, we used Sudan Black B after the immunostaining to quench tissue autofluorescence (Schnell et al., 1999). Microscopy. The rat brain sections were coverslipped using 2.5% DABCO in glycerol (Sigma); brain sections from mouse and lemur were mounted, air-dried, and coverslipped with Entellan (in toluene; Merck, Darmstadt, Germany). The rat sections were examined using a Nikon Eclipse E600 fluorescence microscope equipped with appropriate filters and ORCA-ER, C4742-80 digital camera (Hamamatsu Photonics K.K., System Division, Hamamatsu City, Japan), using Hamamatsu Photonics Wasabi 150 software. For colocaliza-
S.O. FETISSOV ET AL. tion analysis, a Bio-Rad Radiance Plus confocal scanning microscope (Bio-Rad, Hemel Hempstead, UK) installed on a Nikon Eclipse E600 fluorescence microscope, equipped with 10ⴛ (0.45 N.A.), 20ⴛ (0.75 N.A.), and 60ⴛ oil (1.40 N.A.) objectives was used. The FITC was excited using the 488 nm argon laser and its signal detected using the HQ 530/60 emission filter (Bio-Rad). For the detection of RRX, the 543 nm HeNe laser in combination with HQ 590/70 emission (Bio-Rad) was used. GAD-GFP mouse and gray mouse lemur sections were inspected using a confocal laser-scanning microscope (Model 510, Zeiss, Jena, Germany) equipped with appropriate excitation and emission filters for maximum separation of Cy2, Cy3, and Cy5 signals. Emission wavelengths for Cy2, Cy3, and Cy5 were limited to 505–530 nm (bandpass filter), 560 – 610 nm (bandpass filter), and >650 nm (longpass filter), respectively. Underlying structures and outlines of cells were visualized using differential interference contrast (DIC) objectives. Digital images resulting from the microscopy were optimized for image resolution, and images with doublelabeling were merged in Adobe PhotoShop 6.0 (Adobe Systems, San Jose, CA). Colors of particular emission wavelengths were user-defined offline in Zeiss LSM viewer software to enhance the visual clarity of photomicrographs. Cy5 signals were encoded in blue. High-resolution TIFF images were exported and processed to reach optimal brightness and contrast levels using Corel Photo-Paint X3 (Corel, Ottawa, ON, Canada) and image plates were assembled in Corel Draw X3 (Corel).
In vitro transcription and translation of enzymes cDNA clones corresponding to GAD, AADC, TPH, and TH were subcloned into a pSP6 poly A vector as previously described (Falorni et al., 1994; Husebye et al., 1997; Ekwall et al., 1998; Hedstrand et al., 2000). Recombinant [35S]-radiolabeled enzymes were produced by in vitro transcription and translation (ITT) in a TnT SP6 Quick coupled reticulocyte lysate system (Promega, Madison, WI). The correct size of the radioactive product was verified by SDS-polyacrylamide gel electrophoresis (Fig. 1), and [35S]-methionine incorporation was measured by trichloroacetic acid precipitation, followed by scintillation counting.
Antisera The rabbit anti-TH polyclonal antiserum (AB152; Chemicon International) was produced using denatured fulllength TH from rat pheochromocytoma to yield an antibody that gives a 62 kDa band by Western blotting. In addition, selectively labeled dopaminergic neurons in the substantia nigra pars compacta (Martin-Ibanez et al., 2006) revealed a cellular staining pattern identical to the rabbit polyclonal antibody provided by the late Dr. Menek Goldstein (1:4,000; produced against rat pheochromocytoma PC-12 cells) (Markey et al., 1980). Both antibodies have been shown to reveal a single immunoreactive protein band on denaturing Western blots at the expected molecular weight (Markey et al., 1980; Rommelfanger et al., 2004). We used goat antiparvalbumin (PV; PVG-214) (Dayer et al., 2008) and rabbit anti-parvalbumin (PV-28) antisera (Swant) raised against rat muscle PV. These antibodies gave identical staining pattern in all applications. In addition, absolute specificity of the rabbit anti-PV antibody has been demonstrated by the absence of any immunoreactivity in PVⴚ/ⴚ mice (Caillard
The Journal of Comparative Neurology APS1 AUTOANTIBODIES STAIN MAJOR BRAIN SYSTEMS
5 adsorption with the 5-HT-BSA conjugate (no staining) (Steinbusch et al., 1983).
RESULTS General aspects
Figure 1. Coupled in vitro transcription and translation of 35S-labeled enzymes analyzed by SDS-PAGE followed by phosphorimaging. Lanes 1 and 6, positive control in vitro translation product (control template DNA added); Lanes 2 and 7, negative control in vitro translation product (no template DNA added); Lane 3, TPH in vitro translation product; Lane 4, AADC in vitro translation product; Lane 5, GAD in vitro translation product; Lane 8, TH in vitro translation product. MW, molecular weight markers in kDa.
et al., 2000). The anti-calbindin D28k monoclonal antibody (Swant) has been produced by hybridization of mouse myeloma cells with spleen cells from mice immunized with calbindin D-28k purified from chicken gut. Specificity of this antibody has been confirmed by immunocytochemistry and Western blot analyses in transfected N18 –RE 105 neuroblastoma-retina hybrid cells (D’Orlando et al., 2002) and calbindin-D28kⴚ/ⴚ mice (Schwaller et al., 2002). A rabbit anti-calretinin (CR) antibody was raised against the recombinant C-terminal fragment of human CR containing 133 amino acids (Atlas Antibodies AlbaNova University Center, Stockholm, Sweden; HPA0073505). This antibody provided a single immunoreactive band of expected molecular weight on denaturing Western blot in rat tissue (data not shown), and its labeling pattern is identical, although more refined, to antibodies raised against this Ca2ⴙ-binding protein used previously (Martin-Ibanez et al., 2006). GAD immunoreactivity was revealed by using a monoclonal IgG2a antibody raised against the 67-kDa isoform of GAD (GAD67; clone 1G10.2 antibody; Chemicon). This antibody selectively reacts with GAD67 in rat, mouse, and primate as tested by Western blot analysis of brain lysates. It has been shown to selectively reveal GABAergic interneurons and GABA-containing processes in rat telencephalon (Singec et al., 2004). Antibodies against 5-HT were raised in guinea pigs against a 5-HT-BSA conjugate as described by Steinbusch et al. (1983). Briefly, serotonin was coupled to BSA after diazotization of para-aminobenzoic acid and coupling of diazotized para-benzoic acid to serotonin. This procedure is described in more detail in Steinbusch et al. (1983) under the heading “Serotonin-Immunogen B.” Specificity tests included preimmune serum (no staining), crossreactivity with chemically similar compounds (negligible), and
Seventeen APS1 patient sera, which by previous in vitro biochemical analysis were shown to contain various combinations of autoAbs against AADC, TH, TPH, and GAD (Table 1), were studied by immunohistochemistry, mainly on rat brain sections. Several sera were very “powerful” and “sensitive,” and dilution up to 1:1,000,000 resulted in strong staining patterns of cell bodies in some regions, but to distinctly visualize fine nerve terminals in areas such as cortex/hippocampal formation higher concentrations were needed (1:100,000). Depending on the extent to which the autoAbs stained monoamine (DA, NA, 5-HT) and/or GABA systems in coronal sections of rat brain, we divided the sera into four groups (Table 1). Group I harbors six sera which stained monoamine, but not GABA neuron-like systems. The three sera in Group II, and the CSF from Patient #8 stained monoamine- and GABA neuronlike systems. In Group III sera (n ⴝ 3) the staining pattern was GABA neuron-like. In Group IV (n ⴝ 5) no distinct staining was observed with immunohistochemistry, in spite of the presence of autoAbs detectable in vitro. Generally, two major patterns were observed, one reflecting typical staining and distribution of GABA neurons, earlier detected with antibodies against GAD (and against GABA, as well as with autoradiographic visualization of 3H-GABA uptake), the other representing monoamine neuron systems, originally detected with the Falck–Hillarp technique (Falck et al., 1962) and subsequently also with immunohistochemistry (Coons, 1958), using antisera directed against all four enzymes in the catecholamine synthesis, that is TH, AADC, dopamine -hydroxylase, and phenylethanolamine N-methyltransferase (for Refs., see introduction). Moreover, some sera, especially the ones in Group I, labeled widely distributed neurons earlier shown to contain AADC, but none of the other three monoamine systems in the rat brain (Jaeger et al., 1984). These neurons were termed “D” groups by Jaeger et al., following the classical subdivision of CA neurons by Dahlstro¨m and Fuxe (1964), designating DA and NA neurons as “A” groups and 5-HT neurons as “B” groups, and the more restricted adrenaline neurons as “C” groups (see Ho¨kfelt et al., 1984a,c). In addition to the neuronal staining, especially the Group I sera showed a strong AADC staining of capillary walls in certain brain regions, in agreement with the well-known localization of this enzyme to endothelial cells of the blood vessels (Melamed et al., 1980; Kang et al., 1992). Coronal sections from the following rat brain regions were chosen for analysis: 1) anterior telencephalon at the level of the head of the caudate nucleus; 2) mid-posterior diencephalon including cortex and hippocampus; 3) the mesencephalon including the substantia nigra; 4) the pons, including the locus coeruleus, the dorsal raphe, and 5) the medulla oblongata. GABA neurons are found in all these regions, but with particularly distinct nerve endings in the cerebellum and the cochlear nuclei. The AADC-only “D” cell groups are spread out over the brain and can be seen, e.g., in various hypothalamic and thalamic nuclei, and several other regions/nuclei known to contain none of the classical “A,” “B,” or “C” cell groups.
The Journal of Comparative Neurology 6
S.O. FETISSOV ET AL.
Figure 2. Immunostaining of rat brain using APS1 sera from Group I (patients: A #2; B,C #3; D,E #1) reveals labeling of various monoamine neuronal systems containing AADC. A: DA neurons in the arcuate nucleus (Arc) of the hypothalamus projecting to the median eminence (ME) and to the intermediate pituitary lobe (IPL). B: NA neurons of the locus coeruleus (LC). C: 5-HT neurons of the DRN. D: DA neurons of the ventral mesencephalon including the substantia nigra shown in (E) at higher magnification in the SNC. Aq, cerebral aqueduct; cp, cerebral peduncle; IP, interpeduncular nucleus; 3V, 3rd ventricle; 4V, 4th ventricle. Scale bars ⴝ 200 m in A–D; 100 m in E.
Specific staining patterns: rat Sera from Group I, in general, gave extensive staining patterns in all parts of the rat brain. In principle, they labeled the CA and 5-HT cell bodies at various coronal levels, including the A12 group in the arcuate nucleus (Fig. 2A), NA neurons in the locus coeruleus (LC) (A6 cell groups) (Fig. 2B), DA cell bodies in the ventral mesencephalon (A9, A10) (Fig. 2D,E), and adrenaline neurons (C1/2) in the medulla oblongata (data not shown). Moreover, cell bodies in the dorsal raphe (DRN) (B7 cell group) (Figs. 2C, 3C,C’,C“ and 3D,D’,D”) as well as medullary raphe nuclei were positive. The catecholaminergic or serotonergic nature of these neurons was confirmed by
double-staining with antibodies to, respectively, TH (Fig. 3A– A“,B–B”) and 5-HT (Fig. 3C–C“,D–D”). The sensitivity was further tested by diluting serum #1 to 1:1,000,000, resulting in distinct staining of cell bodies in the substantia nigra pars compacta and ventral tegmental area (Fig. 4A–L) and DRN (data not shown). Such sections could then be double-stained with antibodies raised, e.g., in mouse (Fig. 4B–D) or rabbit (Fig. 4E–L) against 1) GAD67, showing close apposition of GABA terminals on DA cell bodies and dendrites (Fig. 4A–D); 2) CR, demonstrating CR expression in some but not all DA cell bodies (Fig. 4F–H); and 3) PV, showing lack of coexistence with, but close apposition of PV-immunoreactive (ir) neurons to DA neurons (Fig. 4J–L). Interestingly, the APS1 immunoreactivity (IR)
Figure 3. Immunostaining of rat brain using APS1 sera (Group I, Patient #1). APS1-positive structures are localized: in DA neurons of the SN (A) double-stained with TH rabbit antibodies (A’,A”) and in 5-HT neurons of the DR (C) double-stained with 5-HT guinea pig antibodies (C’,C”). At higher magnification of the SN (B) and DR (D) it is clear that APS1 sera stain subcellular domains distinctly different from TH (B”) or 5-HT (D”) inside DA or 5-HT neurons, respectively. Note patchy APS1 staining in DA cell bodies (see also Fig. 4). Magenta-green copy is available as Supplementary Figure 1. Scale bars ⴝ 50 m in A,C; 10 m in B,D.
The Journal of Comparative Neurology 8
S.O. FETISSOV ET AL.
Figure 4. Staining of rat ventral mesencephalon after incubation with serum of Patient #1 at a dilution of 1:1,000,000. A: APS1-IR (green) predominates in the ventral tegmental area (VTA) and substantia nigra pars compacta (SNC) and exhibits mutually exclusive localization with (GAD67) (red) (A), known to preferentially label GABAergic perikarya and terminals in the SNR (Oertel et al., 1982). High-resolution laser-scanning microscopy substantiates these findings by showing that GAD67-IR concentrates around, but does not colocalize with, APS1-positive neuronal perikarya (arrows) and dendrites both in the VTA (B) and SNC (C). In fact, GAD67-ir boutons appear in close apposition (arrows) to APS1-labeled neurons and their proximal dendrites (D). Planar image encompasses