PML bodies in reactive sensory ganglion neurons of the Guillain–Barré syndrome

www.elsevier.com/locate/ynbdi Neurobiology of Disease 16 (2004) 158 – 168

PML bodies in reactive sensory ganglion neurons of the Guillain–Barre´ syndrome Nuria T. Villagra´, a Jose´ Berciano, b Marcos Altable, b Joaquı´n Navascue´s, a In˜igo Casafont, a Miguel Lafarga, a,c and Marı´a T. Berciano a,c,* a

Department of Anatomy and Cell Biology, University Hospital Marque´s de Valdecilla, Santander, Spain Neurology Service, University Hospital Marque´s de Valdecilla, Santander, Spain c Unit of Biomedicine-CSIC, University of Cantabria, Santander, Spain b

Received 8 December 2003; revised 28 January 2004; accepted 18 February 2004 Available online 2 April 2004 Acute inflammatory demyelinating polyneuropathy (AIDP) is a type of Guillain – Barre´ syndrome (GBS) characterized by primary nerve demyelination sometimes with secondary axonal degeneration. Studies on the fine structure of dorsal root ganglia in AIDP are lacking. Our aim was to investigate the cytology and nuclear organization of primary sensory neurons in AIDP with axonal injury using ultrastructural and immunohistochemical analysis. The light cytology of the L5 dorsal ganglion showed the characteristic findings of neuronal axonal reaction. The organization of chromatin, nucleolus, Cajal bodies, and nuclear pores corresponded to transcriptionally active neurons. However, the hallmark of the nuclear response to axonal injury was the formation of numerous nuclear bodies (NBs; 6.37 F 0.6, in the AIDP, vs. 2.53 F 0.2, in the control, mean F SDM), identified as promyelocytic leukemia (PML) bodies by the presence of the protein PML. In addition to PML protein, nuclear bodies contained SUMO-1 and the transcriptional regulators CREB-binding protein (CBP) and glucocorticoid receptor (GR). The presence of proteasome 19S was also detected in some nuclear bodies. We suggest that neuronal PML bodies could regulate the nuclear concentration of active proteins, a process mediated by protein interactions with PML and SUMO-1 proteins. In the AIDP case, the proliferation of PML bodies may result from the overexpression of some nuclear proteins due to changes in gene expression associated with axonal injury. D 2004 Elsevier Inc. All rights reserved. Keywords: AIDP; GBS; Dorsal root ganglion neurons; Promyelocytic leukemia (PML) bodies; SUMO-1; Glucocorticoid receptor; Proteasome

Abbreviations: AIDP, acute inflammatory demyelinating polyneuropathy; GBS, Guillain – Barre´ syndrome; DRGNs, dorsal root ganglion neurons; NBs, nuclear bodies; PML, promyelocytic leukemia; SUMO-1, small ubiquitin-like modifier 1; CBP, CREB-binding protein; GR, glucocorticoid receptor. * Corresponding author. Departamento de Anatomı´a y Biologı´a Celular, Facultad de Medicina, Universidad de Cantabria, Avd. Cardenal Herrera Oria s/n, 39011 Santander, Spain. Fax: +34-942-201903. E-mail address: [email protected] (M.T. Berciano). Available online on ScienceDirect (www.sciencedirect.com.) 0969-9961/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.nbd.2004.02.005

Introduction There is increasing evidence that the cell nucleus is a highly dynamic organelle with components organized in structural and functional compartments. They include chromosome territories, nucleolus, splicing factor compartments, and nuclear bodies (NBs), particularly Cajal bodies and promyelocytic leukemia (PML) bodies (Lamond and Earnshaw, 1998; Matera, 1999). Regarding PML bodies, they appear as nuclear dots that concentrate the signature protein PML by immunocytochemical analysis (for review, see Maul et al., 2000). The PML gene is consistently disrupted by t(15;17) in patients with acute promyelocytic leukemia (APL) in which the PML gene fused to the retinoic acid receptor a (RARa) gene and PML bodies are disrupted (de The´ et al., 1990; Weis et al., 1994). Interestingly, PML bodies may be restored after treatment with all-trans-retinoic acid, a cellular event that is associated with clinical remission of the APL (for review, see Melnick and Licht, 1999). It has been proposed that the ultrastructural counterpart of PML bodies is a subpopulation of ‘‘nuclear bodies’’ (NBs) initially reported by de The´ et al. (1960) and subsequently defined and classified by Bouteille et al. (1974) as simple and complex NBs. This possibility is supported by the ultrastructural observation that PML immunostaining of NBs usually presents a doughnut-like pattern (Blondel et al., 2002; Koken et al., 1994; Weis et al., 1994), which seems to correspond to the peripheral capsule of the complex NBs of Bouteille. In addition to PML protein, immunocytochemical studies have demonstrated the presence of other proteins such as Sp100, CREB-binding protein (CBP), Daxx, pRB, p53, nuclear DNA helicase II (NDH II), and SUMO-1 (Borden, 2002; Fuchsova´ et al., 2002; Hodges et al., 1998; Ishov et al., 1999; Zhong et al., 2000a). A variety of functions have been proposed for PML bodies, including regulation of transcription, resistance to virus infection, protein storage, and nuclear defense mechanism (for review, see Borden, 2002; Maul et al., 2000; Muratani et al., 2002; Negorev and Maul, 2001; Zhong et al., 2000b). Moreover, the accumulation of the transcriptional co-activator CBP in PML bodies (LaMorte et al., 1998), as well as the presence of nascent RNA at the periphery

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

of these bodies, suggests that PML bodies create a favorable environment for transcription in their immediate vicinity (Boisvert et al., 2001). Regarding the presence of PML bodies in neurons, previous studies indicate that they are absent in large neurons and other cells of neuronal lineage such as neuroblastoma cells in which the PML protein Sp100 is absent (Lam et al., 1995; Negorev and Maul, 2001). However, the presence of NBs of Bouteille has been reported in both neurons and glial cells in several pathological conditions by conventional electron microscopy (Bouteille et al., 1974; Grunnet, 1975: Lafarga et al., 1991). In the Guillain – Barre´ syndrome (GBS), we have previously reported the presence of NBs in reactive Schwann cells associated with de-remyelinated nerve fibers by conventional electron microscopy (Berciano et al., 1996), but primary sensory neurons were not examined in this study. Recent immunocytochemical studies have reported the presence of PML protein in one type of neuronal intranuclear inclusions (NII) in polyglutamine diseases (Takahashi et al., 2002, 2003; Yamada et al., 2001). These inclusions also contain ataxin-7 and CBP, and seem to be PML bodies (Takahashi et al., 2002). In transfected HeLa cells that express mutant forms of ataxin-1 or ataxin-3, and also in PC12 cells that express a mutant SCA3 protein, aberrant proteins form nuclear inclusions that co-localize with the PML protein to PML bodies (Chai et al., 1999; Yasuda et al., 1999). All these immunocytochemical studies suggest that PML bodies may be involved in the formation of NIIs. The GBS is an acute or subacute evolving paralytic disease of unestablished aetiology with characteristic pathological features of macrophage and lymphocytic infiltration of peripheral nerve with either myelin or axonal destruction (Asbury et al., 1969). GBS includes at least three disease patterns (Griffin et al., 1996): acute inflammatory demyelinating polyneuropathy (AIDP), acute motor or motor – sensory axonal neuropathy (AMAN and AMSAN), and Fisher’s syndrome. AMAN and AMSAN are characterized by primary axonal degeneration probably due to immune attack on the same or related axonal epitopes. Demyelination and inflammatory infiltrates in the spinal roots and nerves are the hallmark of AIDP (Asbury et al., 1969; Honavar et al., 1991). In a variable proportion of cases, however, demyelination is accompanied by axonal degeneration (Albers et al., 1985; Hadden et al., 1998). Such axonal pathology was initially correlated with a bystander effect in inflammatory foci (Asbury et al., 1969) and more recently with an increase of endoneurial fluid pressure in nerve trunks possessing epi-perineurium (Berciano et al., 1997, 2000). Pathological descriptions in GBS have mainly been focused on lesions in spinal roots and peripheral nerve trunks. Posterior root ganglia have scarcely been studied (Asbury et al., 1969; Haymaker and Kernohan, 1949; Honavar et al., 1991), and the ultrastructural analysis of the posterior root ganglia in GBS is lacking. This is an extremely important issue both in AMSAN and AIDP with secondary axonal pathology to understanding the process of axonal reaction in primary sensory neurons to degeneration of their axons. In the present work, we demonstrated by ultrastructural and immunocytochemical analysis that PML bodies are present in human dorsal root ganglion neurons (DRGNs) from control and AIDP samples. We investigate the structural and molecular organization of these PML bodies, and the compartmentalization of PML and SUMO-1 proteins within these nuclear organelles. Furthermore, DRGNs of the AIDP provide an excellent neuronal

159

system to investigate the reorganization of PML bodies in reactive neurons to the distal axonal injury. Our results indicate that neuronal PML bodies are dynamic structures that increase in number in response to axonal injury. The high concentration of PML and SUMO-1 proteins in PML bodies of DRGNs also suggests that these organelles can contribute to modulate the nucleoplasmic concentration of active nuclear proteins, including transcriptional regulators, by molecular interactions with PML or SUMO-1.

Patients and methods Case history and neuropathological findings Detailed clinical, electrophysiological, and autopsy findings have been reported elsewhere (Berciano et al., 2000). Briefly, a 79-year-old man had a 2-day history of acroparesthesias and ascending paralysis culminating in quadriplegia, facial palsy, and mechanical ventilation. Five intravenous immunoglobulin cycles were administered with no response. Steroids were not given. He died on day 60. Serial electrophysiological studies (days 4, 17, and 50) initially showed normal nerve conduction velocities with further slowing down, progressive attenuation of compound muscle action potentials, and profuse denervation. The density of myelinated nerves was preserved in L5 ventral and dorsal roots and reduced in the sural nerve. Dorsal columns in the spinal cord were preserved. Mild de-remyelination was observed in lumbar roots. In both lumbar nerves and their branches, there were extensive de-remyelination and centrofascicular or wedge-shaped areas with marked loss of large myelinated fibers. Axonal degeneration was the predominant lesion in the sural nerve. Inflammatory infiltrates, mainly composed of T lymphocytes and macrophages, were mainly observed in the lumbar plexus. Inflammation was scanty in the spinal roots and absent in L5 posterior root ganglion; Nageotte nodules were not noted. In short, these findings are characteristic of AIDP with distally accentuated axonal damage in post-foraminal nerve trunks. Dorsal root ganglia material At autopsy, the posterior root ganglia L3 – L5 were dissected, but just the L5 posterior root ganglion was used here given that the morphometry of L5 ventral and dorsal roots and complete transverse section of the fifth lumbar nerve were available in this case (Berciano et al., 2000). Control material came from a man aged 77 with no evidence of neurological disorder. In both cases, autopsy was performed 6 h after death. Conventional light and electron microscopy For conventional light microscopy and immunofluorescence, the ganglia were fixed with 3% glutaraldehyde and 1% paraformaldehyde in 0.12 M phosphate buffer (pH 7.2) for 6 h. The ganglia were then postfixed with 2% osmium tetroxide in 0.12 M phosphate buffer, stained in block in uranyl acetate, dehydrated in gradient concentrations of acetone, and embedded in Araldite (Fluka). Semithin sections, 1 Am thick, stained with toluidine blue were used for light cytology study. Ultrathin sections stained with uranyl acetate and lead citrate were used for the ultrastructural analysis.

160

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

Confocal microscopy and quantification For immunofluorescence, semithin sections from tissue samples fixed with the glutaraldehyde and paraformaldehyde solution were mounted on poly-L-lysine-coated slides. After araldite extraction with sodium etoxide and treatment with sodium borohydride and 3% H2O2, semithin sections were pre-incubated for 1 h in a blocking solution of phosphate-buffered saline (PBS) containing 0.2% Triton X-100, 4% normal goat serum, and 0.1 M glycine. Sections were then incubated overnight at 4jC with the primary antibody washed with 0.01% Tween 20 in PBS, incubated for 45 min in the specific secondary antibody conjugated with FITC or TexasRed (Jackson, USA), washed in PBS, and mounted with the anti-fading medium Vectashield (Vector, USA). As controls, semithin sections were treated as described above but omitting primary antibodies. The following primary antibodies were used in this study: mouse monoclonal antibodies directed against the SMN (Clone 8, Transduction Labs.), neurofilament (2F11, Dako), and DNA (Clone AC-30-10, Boheringer); rabbit polyclonal serum 204.3 anti-coilin, rabbit polyclonal sera anti-PML (H-238), antiSUMO-1 (FL-101), anti-CBP (A-22), anti-glucocorticoid receptor (GR; E-20), anti-SP3 (D-20) (all from Santa Cruz Labs), and antiproteasome 19S (Tbp1, Affiniti Research Products). Samples were examined with a laser confocal microscope (BioRad MRC-1024) using argon ion (488 nm) and HeNe (543 nm) lasers. The quantitative analysis was performed on semithin sections immunolabeled with the polyclonal anti-PML antibody. The proportion of neuronal nuclear profiles containing PML bodies and the number of these bodies per nuclear section were estimated by direct examination using a 40 objective. Samples of two ganglia of both the control and AIDP were used and at least 50 nuclear sections containing nucleolus per ganglion were counted. Data

were analyzed by the StatView 4.5 software and using the statistical tests ANOVA and chi square. Significance was established at P < 0.05. Immunogold electron microscopy For immunoelectron microscopy, the ganglia were fixed with 4% paraformaldehyde in 0.1 M cacodylate buffer for 4 h at room temperature. The tissue samples were then washed with 0.1 M cacodylate buffer, dehydrated in increasing concentrations of methanol at 20jC, embedded in Lowicryl K4M at 20jC, and polymerized with ultraviolet irradiation. Ultrathin sections were mounted on nickel grids and sequentially incubated with 0.1 M glycine in PBS for 15 min, 5% BSA in PBS for 30 min, and the primary antibody (diluted in 50 mM Tris HCl, pH 7.6, containing 1% BSA and 0.1 M glycine) for 1 h at 37jC. After washing, the sections were incubated with goat anti-rabbit or anti-mouse antibodies coupled to 10 or 15 nm gold particles (BioCell, UK; diluted 1:50 in PBS containing 1% BSA). After immunogold labeling, the grids were stained with lead citrate and uranyl acetate and examined with a Philips EM208 electron microscope operated at 60 kV. As controls, ultrathin sections were treated as described above but omitting primary antibodies. Primary antibodies used were the same as indicated for immunofluorescence.

Results Light cytology of dorsal root ganglia Conventional, 1 Am thick, semithin sections of the AIDP dorsal ganglion stained with toluidine blue revealed that the cytology and

Fig. 1. Semithin sections of the GBS (AIDP) ganglion stained with toluidine blue. (A) Representative example of a DRGN showing a large nucleus containing dispersed chromatin, a prominent nucleolus, and several NBs (arrowheads). The cytoplasm displays some microvesicules and the moderately basophilic Nissl bodies appear extensively distributed excluding the marginal zone. Note the well-preserved organization of satellite glial cells and the absence of perineuronal inflammatory infiltrates. (B) Organization of myelinated nerve fibers passing through the ganglion. Note the great abundance of remyelinated fibers and the presence of some images of demyelination (asterisks). Scale bars = 5 Am.

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

organization of DRGNs and their associated satellite glial cells were well preserved (Fig. 1A). Neurons showed large and centrally located nuclei with smooth outlines and a predominant disperse chromatin configuration. Nucleoli were very prominent and some NBs appeared scattered throughout the nucleoplasm. The cytoplasm frequently showed microvesiculation, moderate chromatolysis, and occasionally focal accumulations of lipofuscin granules (Fig. 1A). Most fibers passing through the ganglion exhibited deremyelination (Fig. 1B), but images of axonal degeneration were exceptional. In conclusion, this neuropathological scenario reflects the reactive neuronal response to predominantly post-foraminal axonal injury. Immunofluorescence and quantitative analysis of PML bodies Semithin sections were also used for the immunofluorescence and confocal laser microscopy study of the organization of PML bodies in DRGNs. We performed double-labeling experiments to detect PML protein, a specific marker of PML bodies, and neurofilament subunits, as a marker of the general cytoskeletal organization of the neuronal cytoplasm. Neurofilament staining displayed the typical dense network of these intermediate filaments throughout the cytoplasm in both control and AIDP samples (Figs. 2A and B). Intense PML immunofluorescence was detected in nuclear foci of variable size identified as PML bodies. Whereas PML bodies tended to appear isolated in DRGNs of control ganglia, these

161

bodies were larger and preferentially clustered in nuclear domains in the AIDP (Figs. 2A and B). Neuronal PML bodies were clearly distinguishable from Cajal bodies (Gall, 2000). Double-immunolabeling experiments using antibodies that recognize coilin and the survival motor neuron (SMN) protein, two molecular markers of neuronal Cajal bodies (Carvalho et al., 1999; Pena et al., 2001), revealed that prominent Cajal bodies containing these two molecular constituents are commonly found in AIDP neurons (Figs. 2C – E). Because previous studies have demonstrated that Cajal bodies are transcription-dependent nuclear organelles (Gall, 2000; Lafarga et al., 1998), their common presence in AIDP neurons suggests that transcription is well preserved in these reactive neurons. To determine whether the reactive neuronal response to terminal axonopathy induces changes in the frequency of PML bodies, the percentage of DRGNs nuclear profiles containing PML positive NBs and the mean number of these bodies per nuclear section were estimated on nuclear sections (Figs. 3A and B). Thus, the percentage of nuclear sections containing at least a PML body was 81.54 F 1.52% (mean F SDM) in the control and increased to 100% in the AIDP neurons. Similarly, the mean number of PML bodies per nuclear section was significantly higher in the AIDP neurons than in the control ones (6.37 F 0.6 vs. 2.53 F 0.2, mean F SDM, P < 0.001). This indicates that the reactive response of the DRGNs is accompanied by a proliferation of PML bodies. Because PML protein and other constituents of PML bodies may be conjugated with SUMO-1, a posttranslational modification

Fig. 2. Confocal laser microscopy images of a control (A) and AIDP (B) neurons double immunolabeled for detection of PML (red staining) and neurofilaments (green staining). The extensive cytoplasmic network of neurofilaments is well preserved in both cases. The control neuron exhibits one small-size PML body (A), whereas clusters of several PML bodies of larger size are visible within the cell nucleus of the AIDP neuron (B). (C and D) Double immunostaining with antibodies against coilin and SMN in a neuron of the AIDP ganglion. Both proteins co-localize in a prominent Cajal body (E). Scale bars = 5 Am.

162

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

PML bodies contained sumoylated proteins. Additionally, SUMO1 was diffusely distributed in the nucleoplasm and more concentrated at the nuclear envelope; however, the staining intensity in both localizations was higher in AIDP neurons compared to control ones (Figs. 3C and D). To determine whether the transcriptional regulator GR, which may be modified by SUMO-1 conjugation (Le Drean et al., 2002; Tian et al., 2002), is recruited to PML bodies in response to axonal reaction, we performed immunofluorescence analysis with the antiGR polyclonal antibody. A weak diffuse nuclear staining, excluding nucleolus, was detected in most control neurons, whereas neurons of the AIDP showed higher nucleoplasmic GR staining and intensely labeled NBs (Figs. 3E and F). This finding raises the possibility that a sumoylated form of the GR may be recruited in neuronal NBs of the AIDP. The transcriptional co-activator CBP was also detected in NBs from control and AIDP neurons by immunofluorescence (Fig. 3G). To determine whether neuronal PML bodies contain proteasomes, as has been reported in other cell types (Everett et al., 1997; Lafarga et al., 2002; Lallemand-Breitenbach et al., 2001), we studied the immunocytochemical distribution of the proteasome subunit 19S. Proteasome immunostaining was localized throughout the nucleoplasm in DRGNs of both control and AIDP ganglia. Interestingly, some nuclear sections showed strongly stained NBs (Fig. 3H). Electron microscopy observations

Fig. 3. Confocal laser microscopy images of control (A, C, and E) and AIDP (B, D, and F – H) neurons immunostained with antibodies against PML, SUMO-1, GR, CBP, and proteasome 19S. Note in B the proliferation of PML-immunolabeled NBs. (C and D) SUMO-1 is localized in NBs from both control (C) and AIDP (D) neurons. Note in D the moderate staining of the nucleoplasm and the strong labeling of the nuclear envelope in the AIDP neuron with the SUMO-1 antibody. (E and F) GR immunostaining reveals a diffuse nucleoplasmic staining in the control neuron (E), but nuclear immunoreactivity is more intense and shows GR-positive NBs in the AIDP neuron (F). (G) NBs are also immunoreactive for the transcriptional regulators CBP in an AIDP neuron. Labeling with the antibody against proteasome 19S stains the nucleoplasm and, more intensely, several NBs (H). Scale bars = 5 Am.

that is a signal for their cellular re-localization (Melchior, 2000), we next investigate the presence of SUMO-1 in neuronal PML bodies. Immunofluorescence analysis with the rabbit polyclonal antibody demonstrated that NBs of both control and AIDP neurons concentrated SUMO-1 (Figs. 3C and D), suggesting that neuronal

The fine structure of the cell nucleus of both control and AIDP neurons showed a similar organization of the nuclear organelles, with the exception of a higher number of PML bodies in the patient (Figs. 4A and B). Fig. 4B from an AIDP neuron shows a representative nucleus with a disperse chromatin configuration, small aggregates of chromatin throughout the nucleoplasm, and several NBs. The nucleoli of reactive neurons from the AIDP displayed numerous fibrillar centers, round fibrillar areas of low electron-density surrounded by dense fibrillar components (Fig. 4C), which are related to the rRNA transcription sites (Shaw and Jordan, 1995). Noteworthy is the great density of nuclear pores detected in both transversal and tangential sections of the nuclear envelope (Figs. 4D and E). The nuclear pore complex was particularly well preserved in AIDP neurons as demonstrated by the presence of the basket-like structures at the nuclear face (Fig. 4D). All these ultrastructural features are consistent with a well-preserved transcriptional activity in the AIDP neurons. By conventional electron microscopy, the majority of NBs in both control and AIDP neurons appears as round structures of variable size and morphologically similar to the complex NBs of Bouteille et al. (1974). They were predominantly composed of an outer shell that enclosed an inner core of variable size containing granules and small dense aggregates. The outer shell was formed by several concentric layers of fibrillar material (Figs. 5A and B). In AIDP neurons, NBS were more abundant and frequently exhibited very large size, up to 1.7 Am in diameter (Fig. 5B). NBs were enmeshed in euchromatin regions and no specific spatial relationship of these bodies with the nucleolus, interchromatin granule clusters, large heterochromatin masses, and nuclear envelope was found (Figs. 4B and 5A and B). Immunogold electron microscopy for the detection of PML protein revealed the presence of immunolabeling in all NBs of both

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

163

Fig. 4. Conventional electron microscopy images of control (A) and AIDP neurons (B – E). In both cases (A and B), the cell nucleus shows extensive areas of dispersed chromatin, a prominent nucleolus, and a cluster of interchromatin granules (arrows). The cell nucleus of the AIDP neuron (B) also exhibits two clusters of NBs (arrowheads). (C) Detail of a nucleolus illustrating the presence of several fibrillar centers, small round areas of low electron-density closely surrounded by a shell of very electron-dense fibrillar material. (D and E) Transversal (D) and tangential (E) sections of the nuclear envelope illustrating the great density of nuclear pores in AIDP neurons. Scale bars = 1 Am in A – C, and 400 nm in D and E.

control (Fig. 5C) and AIDP neurons (Fig. 5D), demonstrating that these structures are bona fide PML bodies. Interestingly, gold particles specifically decorated the fibrillar material of the periphery of PML bodies, particularly the concentric layers, whereas no labeling was found in the inner granular core (Figs. 5C and D). A few PML bodies appeared as small dense bodies labeled with the polyclonal anti-PML antibody (Fig. 5E). Clusters of PML bodies intensely decorated with PML immunogold particles were frequently observed in AIDP neurons (Fig. 5D). The fine structure of PML bodies was clearly distinguishable from the Cajal body. Thus, immunogold with the anti-PML antibody revealed that the prominent Cajal bodies of AIDP neurons, identified as dense aggregates of coiled threads, were free of immunolabeling (Fig. 5E), whereas

these bodies appeared intensely labeled with the anti-coilin antibody (Fig. 5F). Additionally, Cajal bodies were frequently found near PML bodies (Figs. 5E and F and 6A). To investigate the ultrastructural distribution of SUMO-1 in neuronal PML bodies, we performed immunogold experiments using a rabbit polyclonal anti-SUMO-1 antibody. Immunolabeling was detected in PML bodies of both control (Fig. 6A) and AIDP neurons (Fig. 6B). As illustrated in Figs. 6A and B, gold particles of SUMO-1 immunoreactivity preferentially decorated the outer concentric layers of PML bodies. This indicates that SUMO-1, like the PML protein, is highly concentrated in the periphery of neuronal PML bodies. Electron microscopy also showed the presence of SUMO-1 at the cytoplasmic side of the nuclear

164

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

Fig. 5. (A and B) By conventional electron microscopy, NBs are formed by several concentric layers that enclose an inner core of variable size. The nuclear body in B shows a large central core composed of granules and small very electron-dense aggregates. NBs are located in nuclear domains of dispersed chromatin. (C and D) Immunogold electron microscopy with the anti-PML antibody reveals the presence of gold particles in NBs of the control (C) and AIDP (D) neurons. The particles decorated the outer shell of these structures, particularly the concentric layers, whereas are conspicuously absent in the inner core of NBs (D). (E and F) Cajal bodies (CB) lack PML immunoreactivity (E) but appear intensely labeled with the anti-coilin antibody (F). Scale bar: A, B, E, and F = 300 nm, C = 250 nm, and D = 400 nm.

envelope in AIDP neurons (Fig. 6C). In addition to PML and SUMO-1 proteins, the immunoelectron microscopy study revealed the presence of the transcriptional co-activator CBP in both control (Fig. 6E) and AIDP (Fig. 6D) neuronal NBs, which exhibited typical ultrastructural features of PML bodies. We next investigated the ultrastructural localization of the transcription regulator GR in neuronal NBs. No immunolabeling

was detected in NBs of control neurons, whereas some NBs, with the typical morphology of PML bodies, exhibited a moderate GR immunostaining on the outer shell (Fig. 6F). Taken together, the immunocytochemical data demonstrate that the localization of PML, SUMO-1, CBP, and GR in neuronal PML bodies is compartmentalized in the outer shell of concentric fibrillar layers. Immunoelectron microscopy also confirmed the presence of pro-

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

165

Fig. 6. Immunocytochemical electron microscopy localization of SUMO-1 (A – C), CBP (D and E), GR (F), and proteasome 19S (G) from control (A and E) and AIDP (B – D, F, and G) neurons. (A – C) A high density of gold particles of SUMO-1 immunoreactivity decorate the concentric layers at the periphery of NBs (A and B). In C, gold particles decorate the cytoplasmic side of the nuclear envelope. (D – F) The outer shell of NB appears labeled with the anti-CBP (D and E) and anti-GR (F) antibodies. (G) Immunolabeling for the detection of proteasome 19S shows two intensely labeled NBs and scattered gold particles in the nucleoplasm from an AIDP neuron. Scale bar: A, B, C, D, and F = 1 Am; and E and G = 700 nm.

teasome 19S in some small dense NBs of both control and AIDP neurons (Fig. 6G).

Discussion Conventional histological study in this case revealed the classical AIDP features together with secondary and distally accentu-

ated axonal degeneration in post-foraminal nerve trunks (Berciano et al., 2000). The posterior L5 root showed pure signs of inflammatory de-remyelination while the histological hallmark in the fifth lumbar nerve was a combination of de-remyelination and axonal degeneration. The cytology and organization of the L5 posterior root ganglion were preserved with just subtle neuronal cytoplasmic changes. As no perineuronal inflammatory infiltrates occurred in the studied L5 posterior root ganglion, neuronal changes observed

166

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

here are accounted for by axonal reaction in the primary sensory neurons to degeneration of their peripheral axons (Thomas et al., 1993). The organization of nuclear compartments in reactive primary sensory neurons from the AIDP ganglion is consistent with a wellpreserved transcriptional activity, as indicated by the following features: (i) predominant dispersed chromatin configuration (euchromatin), associated with transcriptionally active chromatin (Cremer and Cremer, 2001); (ii) abundance of nucleolar fibrillar centers, related to pre-rRNA biosynthesis (Shaw and Jordan, 1995); (iii) presence of prominent Cajal bodies, which are transcription-dependent nuclear organelles (Gall, 2000; Lafarga et al., 1998); and (iv) the high density of nuclear pores, which is related to the neuronal transcriptional activity (Garcı´a-Segura et al., 1988) and to the neuronal response to axotomy (Lieberman, 1976). However, the characteristic nuclear hallmark of the DRGNs from the AIDP is proliferation of PML bodies, clearly identified by the presence of the signature protein PML. Taken together, all these nuclear features of AIDP neurons and the extensive remyelination observed within the ganglia suggest that these primary sensory neurons are involved in a reactive – regenerative response to distal axonopathy. Our results show that PML bodies may be normal constituents in DRGNs, but their number increase in reactive neurons of the AIDP. In human neuropathology, PML protein expression has been detected in a subtype of NII in polyglutamine diseases (Takahashi et al., 2002, 2003; Yamada et al., 2001), suggesting that PML bodies are involved in the formation of NIIs (Takahashi et al., 2003). However, our observation of PML bodies in DRGNs that lack NIIs seems to indicate that the proliferation of PML bodies in these neurons is part of a more general reactive response to axonal injury. The fine structure of most PML bodies found in DGRN from both control and AIDP ganglia corresponds to the complex NB of Bouteille et al. (1974), which is composed of an outer shell of fibrillar material that encloses a central granular core. Our electron microscopic findings demonstrate that PML, SUMO-1, CBP, and GR proteins are preferentially distributed at the periphery of PML bodies. These results confirm and extend previous data on the ultrastructural localization of the PML protein (Blondel et al., 2002; Koken et al., 1994; Weis et al., 1994) with the additional demonstration that PML and SUMO-1 are highly concentrated in the concentric layers of the PML bodies. It is well established that the modification of PML protein by SUMO-1 conjugation is essential for the recruitment of other NB-associated proteins in PML bodies (Ishov et al., 1999; Lallemand-Breitenbach et al., 2001; Seeler and Dejean, 2001; Zhong et al., 2000a). Our cytochemical data suggest that the concentric layers of neuronal PML bodies are molecular assemblages enriched in sumoylated PML protein. These concentric layers may provide the molecular scaffold for the recruitment of additional NB components, such as CBP and GR detected here by immunogold electron microscopy. Noteworthy is the high concentration of SUMO-1 localized at the nuclear envelope of the AIDP neurons by light and electron microscopy immunocytochemistry. This localization may correspond to SUMO-1 covalently attached to RanGAP1 (Ran GTPase activating protein 1), which is associated with the cytoplasmic side of the nuclear pore complex and participates in the control of the nucleocytoplasmic transport (Pichler and Melchior, 2002). Among the different theories currently proposed for functions of PML bodies (Borden, 2002; Negorev and Maul, 2001; Zhong et

al., 2000b), neuronal PML bodies may serve as dynamic protein storage compartments. In this way, PML bodies in DRGNs may contribute to regulate the nucleoplasmic concentrations of active proteins (Negorev and Maul, 2001), including the transcriptional co-activator CBP. Recent experiments using fluorescence recovery after photobleaching (FRAP), to study the diffusion kinetics of the acetyltransferase CBP, revealed that this transcriptional co-activator moves rapidly into and out of PML bodies (Boisvert et al., 2001; Wiesmeijer et al., 2002) and illustrated the constant flux of proteins through NBs. The different nuclear immunostaining patterns for GR in DRGNs from control and AIDP ganglia are worth noting. Whereas control neurons exhibited a weak nuclear signal for GR without nuclear foci, notably higher nucleoplasmic labeling and the presence of GR positive PML bodies were commonly observed in AIDP neurons. This indicates that changes in gene expression associated with the reactive neuronal response to axonal injury promote the nuclear accumulation of GR, and also suggests the possible implication of this receptor in the transcriptional regulation of target genes involved in the neuronal response to distal axonopathy. The presence of GR in PML bodies has not been reported in other cell types. However, recent biochemical studies have shown that GR is conjugated with SUMO-1 (Le Drean et al., 2002; Tian et al., 2002), a modification that reduces GR transcriptional activity. In this context, the colocalization of SUMO-1 and GR in the outer shell of PML bodies reported here suggests that this PML domain may concentrate GR modified by SUMO-1 conjugation. We have found that the proteasome 19S is present in some dense NBs of both control and AIDP neurons. These proteasome positive bodies may be sites of intranuclear proteolysis closely related to clastosomes (Lafarga et al., 2002), a category of NBs that concentrates active proteasome and protein substrates for degradation, including the PML protein. In conclusion, the proliferation of PML bodies is a nuclear hallmark in the response of DRGNs to axonal injury in the AIDP. The combination of SUMO-1 modification, which transiently relocalizes sumoylated substrates in PML bodies, and proteolysis may together provide an important regulatory mechanism to control the nuclear concentration of active nuclear proteins. This mechanism might play an important role under pathological conditions, such as the AIDP, in which important changes in gene and protein expression must occur in response to distal axonal injury.

Acknowledgments The authors wish to thank Dr. A. I. Lamond for providing the anti-coilin (204.3) antibody and R. Garcı´a-Ceballos for technical assistance. This work was supported by the following grants: ‘‘Centro de Investigacio´n de Enfermedades Neurolo´gicas (CIEN)’’, ‘‘Instituto de Salud Carlos III (Madrid, Spain)’’; ‘‘Direccio´n General de Investigacion’’ from Spain (BFI2002-0454); and ‘‘Fundacio´n Marque´s de Valdecilla’’ (A04/03) from Santander, Spain.

References Albers, J.W., Donofrio, P.D., McGonagle, T.K., 1985. Sequential electrodiagnostic abnormalities in acute inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 8, 528 – 539.

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168 Asbury, A.K., Arnason, B.G., Adams, R.D., 1969. The inflammatory lesion in idiopathic polyneuritis. Its role in pathogenesis. Medicine 49, 173 – 215. Berciano, M.T., Calle, E., Andres, M.A., Berciano, J., Lafarga, M., 1996. Schwann cell nuclear remodelling and formation of nuclear and coiled bodies in Guillain – Barre´ syndrome. Acta Neuropathol. 92, 386 – 394. Berciano, J., Figols, J., Garcı´a, A., Calle, E., Illa, I., Lafarga, M., Berciano, M.T., 1997. Fulminant Guillain – Barre´ syndrome with universal inexcitability of peripheral nerves: a clinicopathological study. Muscle Nerve 20, 846 – 857. Berciano, J., Garcı´a, A., Figols, J., Mun˜oz, R., Berciano, M.T., Lafarga, M., 2000. Perineurium contributes to axonal damage in acute inflammatory demyelinating polyneuropathy. Neurology 55, 552 – 559. Blondel, D., Regad, T., Poisson, N., Pavie, B., Harper, F., Pandolfi, P.P., de The´, H., Chelbi-Alix, M.K., 2002. Rabies virus P and small P products interact directly with PML and reorganize PML nuclear bodies. Oncogene 21, 7957 – 7970. Boisvert, F.M., Kruhlak, M.J., Box, A.K, Hendzel, M.J., Bazet-Jones, D.P., 2001. The transcription coactivator CBP is a dynamic component of the promyelocytic leukemia nuclear body. J. Cell Biol. 152, 1099 – 1106. Borden, K.L., 2002. Pondering the promyelocytic leukemia protein (PML) puzzle: possible functions for PML nuclear bodies. Mol. Cell. Biol. 22, 5259 – 5269. Bouteille, M., Laval, M., Dupuy-Coin, A.M., 1974. Localization of nuclear functions as revealed by ultrastructural autoradiography and cytochemistry. In: Busch, H. (Ed.), The Cell Nucleus, vol. I. Academic Press, New York, pp. 3 – 71. Carvalho, T., Almeida, F., Calapez, A., Lafarga, M., Berciano, M.T., Carmo-Fonseca, M., 1999. The spinal muscular atrophy disease gene product, SMN. A link between snRNP biogenesis and the Cajal (coiled) body. J. Cell Biol. 147, 715 – 728. Chai, Y., Koppenhafer, S.L., Shosmith, S.J., Perez, M., Paulson, H.L., 1999. Evidence for proteasome involvement in polyglutamine disease: localization to nuclear inclusions in SCA3/MJD and suppression of polyglutamine aggregation in vitro. Hum. Mol. Genet. 8, 673 – 682. Cremer, T., Cremer, C., 2001. Chromosome territories, nuclear architecture and gene regulation in mammalian cells. Nat. Rev. 2, 292 – 301. de The´, G., Riviere, M., Bernhard, W., 1960. Examen au microscope e´lectronique de la tumeur VX2 du lapin domestique derive´e du papillome de Shope. Bull. Cancer 47, 570 – 584. de The´, H., Chomienne, C., Lanotte, M., Degos, L., Dejean, A., 1990. The t(15:17) translocation of acute promyelocytic leukaemia fuses the retinoic acid receptor alpha gene to a novel transcribed locus. Nature 347, 558 – 561. Everett, R.D., Meredith, M., Orr, A., Cross, A., Kathoria, M., Parkinson, J., 1997. A novel ubiquitin-specific protease is dynamically associated with the PML nuclear domain and binds to a herpesvirus regulatory protein. EMBO J. 16, 1519 – 1530. Fuchsova´, B., Nova´k, P., Kafkova, J., Hoza´k, P., 2002. Nuclear DNA helicase II is recruited to IFN-a-activated transcription sites at PML nuclear bodies. J. Cell Biol. 158, 463 – 473. Gall, J.G., 2000. Cajal bodies: the first 100 years. Annu. Rev. Cell Dev. Biol. 16, 273 – 300. Garcı´a-Segura, L.M., Lafarga, M., Berciano, M.T., Herna´ndez, P., Andres, M.A., 1988. Distribution of nuclear pores and chromatin organization in neurons and glial cells of the rat cerebellar cortex. J. Comp. Neurol. 290, 440 – 450. Griffin, J.W., Li, C.Y., Ho, T.W., Tian, M., Gao, C.Y., Xue, P., Mishu, B., Comblath, D.R., Macko, C., Mckhann, G.M., Asbury, A.K., 1996. Pathology of the motor-sensory axonal Guillain – Barre´ syndrome. Ann. Neurol. 39, 17 – 28. Grunnet, M.L., 1975. Nuclear bodies in Creutzfeldt – Jakob and Alzheimer’s diseases. Neurology 25, 1091 – 1093. Hadden, R.D., Cornblath, D.R., Hughes, R.A., Zielasek, J., Hartung, H.P., Toyka, K.V., Swan, A.V., 1998. Electrophysiological classification of Guillain – Barre´ syndrome: clinical associations and outcome. Ann. Neurol. 44, 780 – 788.

167

Haymaker, W., Kernohan, J.W., 1949. The Landry – Guillain – Barre´ syndrome. A clinicopathological study of fifty fatal cases and a critique of the literature. Medicine 28, 59 – 144. Hodges, M., Tissot, C., Howe, K., Grimwade, D., Freemont, P.S., 1998. Structure, organization, and dynamics of promeyelocytic leukaemia protein nuclear bodies. Am. J. Hum. Genet. 63, 297 – 304. Honavar, M., Tharakan, J.K.J., Hughes, R.A., Leibowitz, S., Winner, J.B., 1991. A clinicopathological study of the Guillain – Barre´ syndrome. Nine cases and literature review. Brain 114, 1245 – 1269. Ishov, A.M., Sotnikov, A.G., Negorev, D., Vladimirova, O.V., Neff, N., Kamitani, T.T.H., Yeh, E., Strauss III, J.F., Maul, G.G., 1999. PML is critical for ND10 formation and recruits the PML-interacting protein daxx to this nuclear structure when modified by SUMO-1. J. Cell Biol. 147, 221 – 234. Koken, M.H., Puvion-Dutilleul, F., Guillemin, M.C., Viron, A., LinaresCruz, G., Stuurman, N., et al., 1994. The t(15;17) translocation alters a nuclear body in a retinoic acid-reversible fashion. EMBO J. 13, 1073 – 1083. Lafarga, M., Berciano, M.T., Suarez, I., Viadero, C.F., Andres, M.A., Berciano, J., 1991. Cytology and organization of reactive astroglia in human cerebellar cortex with severe loss of granule cells: a study on the ataxic form of Creutzfeldt – Jakob disease. Neuroscience 40, 337 – 352. Lafarga, M., Berciano, M.T., Garcı´a-Segura, M., Andre´s, M.A., CarmoFonseca, M., 1998. Acute osmotic/stress stimuli a transient decrease of transcriptional activity in the neurosecretory neurons of supraoptic nuclei. J. Neurocytol. 27, 205 – 217. Lafarga, M., Berciano, M.T., Pena, E., Mayo, I., Castan˜o, J.G., Bhmann, D., Rodrigues, J.P., Tavanez, J.P., Carmo-Fonseca, M., 2002. Clastosoma: a subtipe of nuclear body enriched in 19S and 20S proteasomes, ubiquitin, and protein substrates of protasome. Mol. Biol. Cell 13, 2771 – 2782. Lallemand-Breitenbach, V., Zhu, J., Puvion, F., Koken, M., Honore, N., Doubeikovsky, A., Duprez, E., Pandolfi, P., Puvion, E., Freemont, P., de The´, H., 2001. Role of promyelocytic leukaemia (PML) sumoylation in nuclear body formation, 11S proteasome recruitment, and As(2)O(3)induced PML or PML/retinoic acid receptor alpha degradation. J. Exp. Med. 193, 1361 – 1371. Lam, Y.M., Ammerlaan, W., Kroese, F., Opstelten, D., 1995. Cell type- and differentiation stage-dependent expression of PML domains in rat, detected by monoclonal antibody HIS55. Exp. Cell Res. 221, 344 – 356. Lamond, A.I., Earnshaw, W.C., 1998. Structure and function in the nucleus. Science 280, 547 – 553. LaMorte, V.J., Dyck, J.A., Ochs, R.L., Evans, R.M., 1998. Localization of nascent RNA and CREB binding protein with the PML-containing nuclear body. Proc. Natl. Acad. Sci. U. S. A. 95, 4991 – 4996. Le Drean, Y., Mincheneau, N., Le Goff, P., Michel, D., 2002. Potentiation of glucocorticoid receptor transcriptional activity by sumoylation. Endocrinology 143, 3482 – 3489. Lieberman, A.R., 1976. Sensory ganglia. In: Landon, D.N. (Ed.), The Peripheral Nerve. Chapman & Hall, London, pp. 188 – 278. Matera, A.G., 1999. Nuclear bodies: multifaceted subdomains of the interchromatin space. Trends Cell Biol. 9, 302 – 309. Maul, G.G., Negorev, D., Bell, P., Ishov, M., 2000. Properties and assembly mechanisms of ND10, PML bodies, or PODs. J. Struct. Biol. 129, 278 – 287. Melchior, F., 2000. SUMO—Nonclassical ubiquitin. Annu. Rev. Cell Dev. Biol. 16, 591 – 626. Melnick, A., Licht, J.D., 1999. Deconstructing a disease: RARa, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukaemia. Blood 93, 3167 – 3215. Muratani, M., Gerlich, D., Janicki, S.M., Gebhard, M., Eils, R., Spector, D.L., 2002. Metabolic-energy-dependent movement of PML bodies within the mammalian cell nucleus. Nat. Cell Biol. 4, 106 – 110. Negorev, D., Maul, G.G., 2001. Cellular proteins localized at and interacting within ND10/PML nuclear bodies/PODs suggest functions of a nuclear depot. Oncogene 20, 7234 – 7242. Pena, E., Berciano, M.T., Fernandez, R., Ojeda, J.L., Lafarga, M., 2001.

168

N.T. Villagra´ et al. / Neurobiology of Disease 16 (2004) 158–168

Neuronal body size correlates with the number of nucleoli and Cajal bodies, and with the organization of the splicing machinery in rat trigeminal ganglion neurons. J. Comp. Neurol. 430, 250 – 263. Pichler, A., Melchior, F., 2002. Ubiquitin-related modifier SUMO 1 and nucleocytoplasmic transport. Traffic 3, 381 – 387. Seeler, J.S., Dejean, A., 2001. SUMO: of branched proteins and nuclear bodies. Oncogene 20, 7243 – 7249. Shaw, P.J., Jordan, E.G., 1995. The nucleolus. Annu. Rev. Cell Dev. Biol. 11, 93 – 121. Takahashi, J., Fujigasaki, H., Zander, C., Hachimi, K.H., Stevanin, G., Du¨rr, A., Lebre, A.S., Yvert, G., Trottier, Y., de The´, H., Hauw, J.J., Duyckaerts, C., Brice, A., 2002. Two populations of neuronal intranuclear inclusions in SCA7 differ in size and promyelocytic leukaemia protein content. Brain 125, 1534 – 1543. Takahashi, J., Fujigasaki, H., Iwabuchi, K., Bruni, A.C., Uchihara, T., Hachimi, K.H., Stevanin, G., Du¨rr, A., Lebre, A.S., Trottier, Y., de The´, H., Tanaka, J., Hauw, J.J., Duyckaerts, C., Brice, A., 2003. PML nuclear bodies and neuronal intranuclear inclusion in polyglutamine diseases. Neurobiol. Dis. 13, 230 – 237. Tian, S., Poukka, H., Palvimo, J.J., Ja¨nne, O.A., 2002. Small ubiquitinrelated modifier-1 (SUMO-1) modification of the glucocorticoid receptor. Biochem. J. 367, 907 – 911. Thomas, P.K., Scaravilli, F., Belal, A., 1993. Pathological alterations in cells bodies of peripheral neurons in neuropathy. In: Dyck, P.J., Thom-

as, P.K., Griffin, J.W., Low, P.A., Podulso, J.F. (Eds.), Peripheral Neuropathy. Saunders, Philadelphia, pp. 476 – 513. Weis, K., Rambaud, S., Lavau, C., Jansen, J., Carvalho, T., Carmo-Fonseca, M., Lamond, A., Dejean, A., 1994. Retinoic acid regulates aberrant nuclear localization of PML-RARa in acute promyelocytic leukaemia cells. Cell 76, 345 – 356. Wiesmeijer, K., Molenaar, C., Bekeer, I.M.L.A., Tanke, H.J., Dirks, W., 2002. Mobile foci of Sp100 do not contain PML: PML bodies are immobile but PML and Sp 100 proteins are not. J. Struct. Biol. 140, 180 – 188. Yamada, M., Sato, T., Shimohata, T., Hayashi, Y., Igarashi, S., Tsuji, S., Takahashi, H., 2001. Interaction between neuronal intranuclear inclusions and promyelocytic leukaemia protein nuclear and coiled bodies in CAG repeat diseases. Am. J. Pathol. 159, 1785 – 1795. Yasuda, S., Inoue, K., Hirabayashi, H., Yamamoto, Y., Fuyuhiro, H., Komure, O., Tanaka, F., Sobue, G., Tsuchiya, K., Hamada, K., Sasaki, H., Takeda, K., Ichijo, H., Kakizuka, A., 1999. Triggering of neuronal cell death by accumulation of activated SEK1 on nuclear polyglutamine aggregations in PML bodies. Genes Cells 4, 743 – 756. Zhong, S., Mu¨ller, S., Ronchetti, S., Freemont, P.S., Dejean, A., Pandolfi, P.P., 2000a. Role of SUMO-1-modified PML in nuclear body formation. Blood 95, 2748 – 2753. Zhong, S., Salomoni, P., Pandolfi, P.P., 2000b. The transcriptional role of PML and the nuclear body. Nat. Cell Biol. 2, E85 – E90.