|Home | About | Journals | Submit | Contact Us | Français|
Paraneoplastic cerebellar degeneration accompanying gynecological and breast cancers is characteristically accompanied by a serum and cerebrospinal fluid (CSF) antibody response, termed “anti-Yo,” which reacts with cytoplasmic proteins of cerebellar Purkinje cells. Because these antibodies interact with cytoplasmic rather than cell surface membrane proteins, their role in causing Purkinje cell death has been questioned. To address this issue, we studied the interaction of anti-Yo antibodies with Purkinje cells in slice (organotypic) cultures of rat cerebellum. We incubated cultures with immunoglobulin G (IgG)–containing anti-Yo antibodies using titers of anti-Yo antibody equivalent to those found in CSF of affected patients. Cultures were then studied in real time and after fixation for potential uptake of antibody and induction of cell death. Anti-Yo antibodies delivered in serum, CSF, or purified IgG were taken up by viable Purkinje cells, accumulated intracellularly, and were associated with cell death. Normal IgG was also taken up by Purkinje cells but did not accumulate and did not affect cell viability. These findings indicate that autoantibodies directed against intracellular Purkinje cell proteins can be taken up to cause cell death and suggest that anti-Yo antibody may be directly involved in the pathogenesis of paraneoplastic cerebellar degeneration.
Sera and cerebrospinal fluid (CSF) from patients with cerebellar degeneration in the setting of gynecological or breast cancers have repeatedly been shown to contain immunoglobulin G (IgG) antibodies (Abs) reactive with cytoplasmic components of cerebellar Purkinje cells (1–5). This Ab response, termed “anti-Yo” or “anti–Purkinje cell Abs” (“APCA”), reacts with proteins of 34 and 62 kd molecular weight in Western blots of Purkinje cell lysates (4, 6). Studies using electron microscopy show that anti-Yo Abs bind predominantly to Purkinje cell ribosomes and rough endoplasmic reticulum without detectable labeling of Purkinje cell surface membranes (7, 8). Anti-Yo Ab also labels cells within the patients’ tumors, suggesting that this Ab response is elicited by the underlying neoplasm but that it is also cross-reactive with Purkinje cell antigens (9).
Despite their repeated detection in sera and CSF of affected patients, the role of anti-Yo Abs in the pathogenesis of paraneoplastic cerebellar degeneration is unknown. Attempts to duplicate paraneoplastic cerebellar degeneration in experimental animals using anti-Yo Abs have been uniformly unsuccessful. These attempts have included systemic and intraventricular administration of Ab derived from patient sera, passive transfer of Ab raised by immunization with expression products of complementary DNA clones encoding Yo antigens, and immunization with the complementary DNA clones themselves (10–16). The presence of a blood-brain barrier complicates exposure of Purkinje cells to IgG over time in intact animals, however, and none of these studies has duplicated the central nervous system (CNS) synthesis of anti-Yo IgG that occurs in affected humans (5, 17).
An alternate approach to studying the interaction of anti-Yo Abs with Purkinje cells is through the use of cerebellar slice (organotypic) cultures because this culture system preserves anatomical relationships present in vivo and allows exposure of Purkinje cells to Abs without interposition of the blood-brain barrier. We recently found that live Purkinje cells in cerebellar slice cultures can incorporate and later clear normal IgG and that the intracellular presence of IgG does not affect Purkinje cell viability (18). In these studies, we also confirmed IgG uptake by Purkinje cells by demonstrating that incubation of cerebellar slice cultures with an IgG-daunorubicin immunotoxin resulted in Purkinje cell uptake of the immunotoxin and targeted Purkinje cell death (18). The present study was conducted to determine whether IgG containing anti-Yo Ab is similarly taken up by Purkinje cells in cerebellar cultures and whether this uptake is associated with Purkinje cell death.
Serum and CSF samples from 11 patients with paraneoplastic cerebellar degeneration and anti-Yo Abs were studied. These included 3 patients with adenocarcinoma of the ovary, 3 with endometrial malignancies, 2 with adenocarcinoma of the breast, 1 with transitional cell carcinoma of the bladder, and 2 in whom neoplasms had not been identified before the patients were lost to follow-up. Presence of anti-Yo Abs and absence of other paraneoplastic autoantibodies was confirmed in all patients by demonstrating immunohistological staining of Purkinje cells typical for anti-Yo Ab in frozen and fixed sections of human and rat cerebellum and by Ab labeling restricted to the 34- and 62-kd proteins characteristic of Yo antigens in Western blots of Purkinje cell lysates (6). Specificity of the autoantibodies was also confirmed by labeling of the cloned PCD17 protein, homologous with the 62-kd Yo antigen, in Western blots ( and data not shown). Serum Ab titers of anti-Yo Ab as defined by limiting dilution ranged from 1:320 to 1:20,480; CSF Ab titers ranged from 1:40 to 1:640. Control sera were obtained from individuals without neurological disease. Control CSF samples were from individuals without abnormalities of CNS IgG synthesis and also from patients with nonparaneoplastic neurological disorders whose CSF contained both elevated IgG levels and oligoclonal bands. All materials were obtained and studied under institutional review board guidelines (University of Utah and Veterans Affairs Salt Lake City Health Care System). Purified IgG from anti-Yo+ patients was prepared from patient sera by protein G column chromatography, as previously described (20).
All aspects of animal handling and care were conducted with local Institutional Animal Care and Use Committee approval in an Association for Assessment and Accreditation of Laboratory Animal Care–approved facility (Veterans Affairs Salt Lake City Health Care System Veterinary Medical Unit). Cerebellar slice cultures were prepared at 200 μm thickness from 10- to 12- or 23- to 27-day-old Sprague-Dawley rats (Charles River, Germantown, MD) after killing with CO2, using the method of Rothstein et al (21). Horse sera used in tissue culture medium were heated to inactivate complement. Cultures were incubated at 37°C in a 5% CO2/95% humidified air environment with twice-weekly changes of medium. Resulting cultures exhibited typical cerebellar morphology, with Purkinje cells clearly identifiable by morphology and by positive immunostaining of fixed cultures with both anti-Yo Abs and antisera to the Purkinje cell marker calbindin-28 k (Millipore, Temecula, CA) (22, 23).
A variety of compounds including propidium iodide (PI), ethidium homodimer (EH), and dyes such as SYTOX green or SYTOX orange (Invitrogen, Springfield, OR) are normally excluded from living cells. After cell membrane injury, these dyes can enter cells and subsequently bind to cellular nucleic acids, allowing their use as markers of cell death (24). Viable Purkinje cells in rats, however, take up a variety of molecules from the CSF, including PI (25–31). Because of this property, development of a reliable method of determining Purkinje cell death was essential to this study. We compared uptake of PI with that of EH and SYTOX dyes to determine their usefulness as markers of Purkinje cell death. Cerebellar cultures were first maintained in medium for 24 hours and then incubated with medium containing decreasing concentrations of PI (from 7.5 mmol/L to 75 nmol/L), EH (from 2 to 0.02 μmol/L), or SYTOX green (50, 25, or 15 nmol/L). Cultures were harvested at time intervals between 1 and 24 hours, fixed in 2% paraformaldehyde, and examined by confocal microscopy. To evaluate the utility of each compound as a marker for dead or dying cells, cultures were treated with N-methyl-d-aspartic acid (NMDA), kainic acid, or the detergent, Triton X-100 (Sigma-Aldrich, St Louis, MO) or were exposed to reduced temperature (4°C). Cultures were then incubated with PI, SYTOX green, or SYTOX orange for varying periods.
Cultures were incubated with sera from 8 different patients with an anti-Yo Ab response, IgG purified from plasma from 1 of these patients, and anti-Yo+ CSF from 3 patients (20). Controls consisted of purified normal IgG, normal sera, or normal CSF, and CSF not containing anti-Yo Abs but containing increased amounts of IgG and oligoclonal bands. All sera were used at dilutions of 1:800 to approximate concentrations of anti-Yo Ab seen in CSF of most affected patients. Purified IgG was adjusted to provide an end point titer of anti-Yo staining similar to that found in patient CSF. Cerebrospinal fluid samples were used at a 1:10 to 1:30 dilution. Cultures were harvested at 24-hour intervals from 24 to 144 hours. To identify dead or dying cells, SYTOX green, a reliable marker of Purkinje cell death (18), was added to cultures 2 hours before harvesting. All studies were performed in triplicate unless otherwise indicated. To visualize patient or control IgG incorporated within Purkinje cells, cultures were fixed in 2% paraformaldehyde, permeabilized with 0.2% Triton X-100, and incubated overnight at 4°C with 1:800 dilutions of Cy5-conjugated donkey anti-human IgG (Jackson ImmunoResearch, West Grove, PA), as previously described (18).
To study the interaction of Purkinje cells with anti-Yo Abs in real time, IgG containing high titers of anti-Yo Ab was purified as described previously and conjugated to Cy5 or to fluorescein isothiocyanate using DyLight Ab labeling kits (Pierce Biotechnology, Rockford, IL). Cultures incubated with IgGs conjugated to each fluorochrome were observed at 2-hour intervals through 8 hours and then at 24 hours using a microscope stage incubating chamber (SmartSlide micro-incubation chamber; WaferGen Biosystems, Fremont, CA). To confirm that Purkinje cells in these cultures were viable and that uptake of Ab seen in fixed cultures corresponded to stages of Ab uptake observed in living cultures, SYTOX green was added to each culture. The cultures were then fixed with 2% paraformaldehyde after 4, 8, and 24 hours of incubation and subjected to confocal analysis. To minimize artifactual labeling of cells resulting from diffusion of extra-cellular IgG, fixed cultures incubated with fluorochrome-conjugated IgG in these real-time studies were not subjected to permeabilization with Triton X-100. Distribution of IgG within Purkinje cells after fixation was compared with that seen in the same cultures during the observation period.
Cultures were incubated with control or anti-Yo sera, CSF, or purified IgG. SYTOX green (25 nmol/L) was added 2 hours before fixation and immunofluorescence confocal microscopy. The amount of cell death was quantified by an observer unaware of treatment of the cultures and consisted of counting of the number of cells labeled by Cy5-conjugated antihuman IgG that contained or lacked SYTOX green. Live cells were recorded as containing Cy5 IgG and lacking SYTOX green. Dead cells were scored as cells colabeled for IgG and SYTOX green. Approximately 40 to 90 cells were counted for each field, and the mean percentage of cell death was obtained from a minimum of 8 fields captured at 40× magnification for each time point (18). Statistical significance between groups was determined by nonparametric Mann-Whitney U statistical analysis using GraphPad Instat statistical software (GraphPad Software, Inc, La Jolla, CA).
Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay for apoptosis was carried out in replicate cultures using an in situ cell death detection kit, TMR red (Roche Applied Science, Indianapolis, IN), and a 1:3 dilution of terminal deoxynucleotidyl transferase enzyme. Cultures were incubated with 1:800 dilutions of sera from patients with anti-Yo Abs or with control sera for 72, 96, and 120 hours. SYTOX green was added to cultures 2 hours before harvesting. Cultures were fixed in 2% paraformaldehyde, permeabilized, and incubated with TUNEL assay mix at 37°C for 2 hours. Ab uptake was confirmed by immunofluorescence staining using Cy5-conjugated donkey antihuman IgG; cell death was confirmed by SYTOX green staining. Positive controls for apoptotic cell death included permeabilized cultures treated with DNase I (Sigma) to induce nicks in the DNA to allow TUNEL staining. Negative controls included cultures maintained without IgG, with normal IgG, or with omission of conjugated secondary Ab during postfixation staining. Negative TUNEL controls were cerebellar cultures incubated with the TUNEL assay mix without the addition of DNase. As an additional assay for apoptosis, parallel cultures were incubated with anti-Yo or control antisera as described previously and then incubated with the pan-caspase substrate carboxyfluorescein-labeled fluoromethyl ketone peptide inhibitor of caspases (FLICA; Immunochemistry Technologies, LLC, Bloomingdale, MN).
To acquire confocal images, we used a Nikon Eclipse E800 upright microscope (Nikon Biosciences, Melville, NY) and the Personal Confocal Microscopy PCM-2000 using argon ion and green and red HeNe lasers. Simple personal confocal image software program (Compix, Cranberry Township, PA) was used to acquire digital images and image analysis. A green HeNe laser with a 543-nm excitation filter and 605-nm long-pass (LP) filter was used to visualize PI, and with a 565-nm LP filter to visualize SYTOX orange. A red HeNe laser with a 633-nm excitation filter and 675-nm LP filter was used to visualize Cy5. The argon ion laser with a 514-nm excitation filter was used with a 605-nm LP filter to visualize EH and with a 510-nm LP filter to image SYTOX green and calcein green. All filters were matched to the peak emission spectra of the fluorochromes used. General procedures used individual fluorochromes with x, y, and z scans of 14 to 20 focal planes. Identical focal plane settings for each fluorochrome were used for single visual field analysis to ensure that each corresponding fluorochrome was imaged in the same focal plane. Stringent uniform experimental parameters and computer software setting were maintained for the respective image analyses in all studies. Because the vibratome preparation techniques used to prepare cerebellar slice cultures invariably resulted in death of neurons on the cut surfaces of culture slices, image analyses were confined to the interior portions of the cultures.
We first examined the uptake of cell viability dyes by Purkinje cells in cerebellar slice cultures to determine their utility as indicators of cell death. Propidium iodide and EH, which are excluded from most living cells, could be detected within Purkinje cell dendrites and cell bodies within 7 hours of incubation (data not shown). Purkinje cell labeling by PI and EH was predominantly cytoplasmic (as with living animals injected intraventricularly with PI) and the time course of PI uptake in organotypic cultures paralleled that in live animals (25, 26). We then compared PI uptake by Purkinje cells during shorter periods in control cultures and in cultures treated with NMDA (18 μmol/L) to induce excitotoxic cell death. Incubation of cultures with PI for 2 hours resulted in nuclear and cytoplasmic staining of Purkinje cells and other neurons in NMDA-treated cultures but not in control cultures (data not shown). In contrast to PI and EH, both SYTOX green and SYTOX orange were excluded from Purkinje cells in control cultures at 7 hours (data not shown) and continued to be excluded for up to 24 hours (Fig. 1A). In contrast, labeling of Purkinje cells and other neurons was seen within 2 hours after the addition of SYTOX dyes to cultures treated with NMDA (18 μmol/L), kainic acid (20 μmol/L; not shown), Triton X-100 (0.1%), or reduced temperature (4°C; Figs. 1, B–D).
Because SYTOX dye staining was more intense than that produced by PI and because they were more reliably excluded from viable Purkinje cells, we conducted additional experiments to confirm that entry of SYTOX dyes into Purkinje cells accurately identified cells with compromised membranes. Cerebellar slice cultures were incubated for 90 minutes with the live-cell viability dye calcein AM (calcein green; Invitrogen; 2 μmol/L) and then incubated overnight with NMDA (18 μmol/L). SYTOX orange, used in place of SYTOX green to allow simultaneous visualization of calcein green, was added to cultures during the last 2 hours before harvesting. Control cultures were incubated with the same regimen of calcein AM and SYTOX orange but were maintained in medium lacking NMDA. Calcein AM readily crosses the plasma membrane of living cells and is then hydrolyzed by cell esterases to generate calcein green, which is retained within the cell (24). The dye is lost from cells after cell membrane disruption, so that intracellular presence of calcein green serves as an indicator of cell viability; loss of calcein green from cells provides a marker of cell death (24). Before NMDA treatment, intra-cellular retention of calcein green could be demonstrated in Purkinje cells throughout the cultures. After incubating with NMDA, virtually all Purkinje cells and other neurons were found to contain SYTOX orange and to have lost intracellular calcein green (data not shown). These observations confirmed the utility of SYTOX dyes as markers for Purkinje cell membrane injury and death. Here, SYTOX green (25 nmol/L) or SYTOX orange (30 nmol/L) was added to cultures 2 hours before harvesting and fixation.
Uptake of IgG by Purkinje cells was detected in cultures incubated with normal IgG controls as previously shown (18), as well as with anti-Yo antisera, or purified anti-Yo IgG. In these studies, however, the intensity of IgG staining in cultures incubated with anti-Yo IgG (Figs. 2B, D) was always much greater than that seen in control cultures (Figs. 2A, C), and visualization of Cy5-labeled normal IgG within Purkinje cells required increased imaging gain settings compared with those optimal for visualizing labeled anti-Yo Abs (Figs. 2A, C). Both normal and anti-Yo IgG were detected in Purkinje cell processes and occasional cell bodies within 4 to 8 hours, within the cytoplasm of most Purkinje cells by 24 hours, and in Purkinje cell nuclei within 48 hours (Fig. 3).
We previously demonstrated that Purkinje cells are able to incorporate IgG and that they are able to clear IgG once transferred to the medium lacking the Ab (18). It was thus of interest to determine whether serum or IgG containing anti-Yo Abs were similarly cleared or whether anti-Yo IgG persisted intracellularly after transfer to medium lacking the Ab. Purkinje cells in cultures incubated with IgG containing anti-Yo Abs continued to exhibit strong immunostaining for IgG through 120 hours after transfer to medium lacking the Ab (data not shown). In contrast, consistent with our previous studies, IgG staining became undetectable within 24 to 48 hours in cultures incubated with normal human IgG and then transferred to medium lacking the immunoglobulin (18).
We then examined the time course of anti-Yo Ab uptake and its ability to induce cell death compared with cultures treated with control IgG. Each anti-Yo serum, CSF, or purified anti-Yo IgG was studied using both cultures derived from 12-day-old animals and cultures from 24-day-old animals (whose brains more closely resemble those of adult animals). All anti-Yo Abs tested produced similar effects on cultures from pups of both ages, although the extent and rapidity of killing seemed to vary more with Ab titer in cultures from 24-day-old than from 12-day-old animals. There was only minimal Purkinje cell death, as evidenced by exclusion of SYTOX green in more than 90% of cells through 96 hours in control cultures (Fig. 3H). In contrast, incubation with anti-Yo serum or purified IgG resulted in progressive Purkinje cell death. By 24 to 48 hours, occasional Purkinje cells in these cultures began to exhibit abnormal morphology (Figs. 3C, D). Purkinje cell death (indicated by SYTOX green) was apparent by 72 hours (Fig. 3E) and was extensive by 96 and 120 hours (Figs. 3F, G). Loss of other cerebellar neuronal populations compared with controls was not observed.
The Purkinje cells in cultures that had been incubated with anti-Yo Ab for 48 hours and were then maintained in medium lacking anti-Yo Ab continued to exhibit labeling for human IgG for up to 120 hours. Purkinje cell death began at 72 hours after exposure to anti-Yo Ab, similar to findings in cultures continuously exposed to anti-Yo Ab over time (data not shown).
IgG uptake was observed in cerebellar cultures incubated with 1:10 or 1:30 dilutions of anti-Yo CSF from 3 different patients for intervals up to 192 hours. As with cultures incubated with serum or purified anti-Yo IgG, cell death began by 72 hours of incubation and was widespread by 96 to 120 hours (Fig. 4). Cultures incubated with equal dilutions of control CSF from healthy individuals or those who contained elevated IgG and oligoclonal bands but lacked anti-Yo specificity exhibited less than 10% cell loss (data not shown).
To confirm morphological observations, Purkinje cell death was quantified in cultures treated for 72 hours with anti-Yo or control sera or CSF. In these experiments, CSF from 1 anti-Yo+ patient and sera from 4 anti-Yo+ patients were studied in cultures derived from day 12 pups. To evaluate the effect of cerebellar development on the anti-Yo Ab–mediated cytotoxicity, we also compared the effect of one of these sera on cultures from both day 12 and day 24 animals. The extent of cell death in Purkinje cells containing IgG (indicated by entry of SYTOX green) was compared with that seen in serum or CSF controls; gain settings were adjusted for the control samples to confirm uptake of normal IgG (Fig. 2). Although the extent of Purkinje cell death varied somewhat from patient to patient, all cultures incubated with anti-Yo sera or CSF exhibited more than 70% Purkinje cell death, and there was no significant difference in Purkinje cell death between cultures from 12- and 24-day-old pups incubated with the same anti-Yo serum (Fig. 5).
A concern in our studies was that residual extracellular anti-Yo Ab present in cultures after incubation and extensive washing could enter Purkinje cells after fixation and permeabilization and could bind to intracellular Yo antigens at detectable levels (32). To exclude this possible artifact, cultures in a microscope stage incubation chamber were incubated with anti-Yo IgG conjugated to either Cy5 or fluorescein isothiocyanate, allowing these cultures to be studied by confocal microscopy in real time. In cultures incubated with each IgG fluorochrome, bright fluorescence for each fluorochrome was detected in most Purkinje cells beginning at 6 hours and was significantly increased by 24 hours (Figs. 6A, B). Analysis of these sections after addition of SYTOX green and fixation without permeabilization demonstrated identical Purkinje cell uptake of IgG and exclusion of SYTOX green at 6 and 24 hours (Figs. 6D, E), indicating that cells taking up IgG were viable. Distribution of IgG in Purkinje cell processes and subsequently in Purkinje cell cytoplasm and nuclei in fixed cultures studied at each time point was indistinguishable from that seen in the same cultures studied in real time (Figs. 6D, E). Thus, the uptake of IgG by Purkinje cells in cultures had occurred before fixation, indicating that the cells were viable and that observations in fixed cultures mirrored those in living cells before harvesting.
Borges et al (25) reported that uptake of PI in living rats could be prevented by pretreatment with colchicine, suggesting that PI uptake was dependent on microtubules. To determine whether uptake of IgG could be similarly prevented, cultures were incubated for 2 hours with serial dilutions (10, 5, 2.5, and 1.25 μg/mL) of colchicine and maintained in medium lacking colchicine and containing 7.5 μg/mL of Cy5-conjugated anti-Yo IgG. These cultures were then studied both in real time and after fixation. Colchicine inhibited IgG uptake in a dose-dependent manner. Whereas IgG was present at 6 hours in most Purkinje cells in control cultures not treated with colchicine, IgG uptake was completely inhibited through 6 hours of observation in cultures incubated with 10 μg/mL (data not shown) and 5 μg/mL of colchicine (Figs. 6C, F). Complete inhibition was observed through 4 hours in cultures incubated with 2.5 μg/mL of colchicine but was not evident in cultures exposed to 1.25 μg/mL (data not shown). Purkinje cells in colchicine-treated cultures remained viable as indicated by continued exclusion of SYTOX green.
Cultures incubated with anti-Yo Ab for 72, 96, and 120 hours were immunolabeled for IgG, evaluated for intracellular SYTOX green, and analyzed by immunofluorescent TUNEL methods (Fig. 7) and through use of the pan-caspase substrate, FLICA. Ab uptake was confirmed by immunofluorescent labeling using Cy5-conjugated anti-human IgG, and death of Purkinje cells throughout the cultures was visualized by intracellular entry of SYTOX green. Purkinje cells containing anti-Yo IgG and SYTOX green remained negative for TUNEL as well as FLICA staining (not shown), indicating that the Purkinje cell death in this model does not involve apoptosis.
Anti-Yo Ab recognizes 2 distinct proteins: a 62-kd protein containing leucine zipper and zinc finger motifs and a 34-kd protein that consists largely of a 6–amino acid consensus sequence (L/FLEDVE) not identified in other eukaryotic proteins (33, 34). Both antigens are encoded in tumors of affected patients (9, 35), and paraneoplastic cerebellar degeneration is thus thought to result from an immune response directed against tumor antigens immunologically related to normally sequestered Purkinje cell proteins. This immune response results in systemic and CNS Ab synthesis as well as appearance of sensitized T lymphocytes (36, 37). To date, however, the respective roles of Ab and T cells in Purkinje cell destruction have remained uncertain. Anti-Yo Abs react with internal cytoplasmic Purkinje cell proteins and have not been shown to bind to antigens present on Purkinje cell surface membranes (7, 8). Because neurons have been considered to exclude IgG, it has been believed that these Abs are unlikely to play a role in the pathogenesis of Purkinje cell injury. Although older studies suggested that Purkinje cells in live animals might incorporate IgG (14, 15), it has in general been thought that Purkinje cell death in paraneoplastic cerebellar degeneration is more probably mediated exclusively by T lymphocytes (38).
We have recently demonstrated that Purkinje cells in cerebellar slice cultures are able to incorporate and later clear normal IgG and that incubation of cultures with an IgG-daunorubicin immunotoxin results in IgG-immunotoxin uptake and targeted Purkinje cell destruction (18). The present study demonstrates that IgG containing anti-Yo Abs (at concentrations typically found in CSF of patients with paraneoplastic cerebellar degeneration) was readily taken up by viable Purkinje cells in organotypic cultures. Unlike control IgG, which is readily cleared from Purkinje cells (18), IgG containing anti-Yo Abs accumulated in this cell population and was associated with Purkinje cell death without appreciable involvement of other neurons. Equivalent Ab accumulation was not observed in cultures incubated with similar concentrations of control IgG, and increased rates of Purkinje cell death were not observed. Anti-Yo IgG uptake was also observed in live cultures studied in real time, thereby confirming that IgG uptake had occurred in viable cells and was not an artifact of fixation. Purkinje cell death was also seen in cultures incubated with CSF containing anti-Yo Abs but not with control CSF from normal individuals or from individuals with nonparaneoplastic disorders whose CSF contains elevated IgG concentrations and oligoclonal bands. Purkinje cell death was observed in cultures incubated with purified anti-Yo IgG, making it unlikely that a non-IgG component of human serum was involved in Purkinje cell death. Our data thus demonstrate that IgG containing anti-Yo Abs, in the absence of sensitized T lymphocytes, seem able to induce Purkinje cell death. To the best of our knowledge, the present study provides the first evidence that Abs directed against intracellular neuronal proteins may be taken up by neurons, gain access to their target antigens, and cause cell death.
Previous studies demonstrated uptake of IgG by Purkinje cells in vivo, and our finding that Purkinje cells take up IgG in cerebellar slice culture thus recapitulates events that have been shown to occur in living animals and possibly in humans (14, 15, 39). Our own early studies of rats showed that intraperitoneal injection of anti-Yo IgG after blood-brain barrier disruption was followed by the appearance of anti-Yo IgG, but not control IgG, in Purkinje cell processes and occasional Purkinje cell bodies (15). Graus et al (14) found that intraventricularly administered anti-Yo and control IgGs were taken up by Purkinje cells in guinea pigs, but Purkinje cell death was not observed either in our earlier study or that of Graus et al. However, our study was only carried out for 96 hours after intraperitoneal IgG injection (15), and the methods used to detect cell injury and death in the present study (incorporation of SYTOX dyes) may have provided earlier and more sensitive recognition of Purkinje cell injury than the detection of Purkinje cell dropout in histological sections as used by Graus et al. The observation by Graus et al (18) that Purkinje cells can also take up normal IgG is similar to our own findings in cerebellar slice cultures exposed to IgG over time.
The mechanisms by which anti-Yo Abs cause Purkinje cell death are not known. The ability of colchicine to block IgG uptake by Purkinje cells suggests that movement of IgG within the cells may involve microtubules, similar to their reported role in uptake of PI (25). Purkinje cells containing both anti-Yo IgG and SYTOX green did not exhibit TUNEL or FLICA staining, suggesting that cell death had not occurred by apoptosis. Basing their observations that Yo antigens are localized on membrane-bound and free ribosomes, Hida et al (8) suggested that Purkinje cell death might be caused by interference with protein synthesis. Attempts to produce anti-Yo–mediated Purkinje cell destruction in vitro or in vivo, however, are complicated by the fact that the anti-Yo Ab response (unlike that associated with most other paraneoplastic neurological syndromes) recognizes 2 unrelated antigens. Because the 62-kd component of the Yo antigens contains leucine zipper and zinc finger motifs commonly found in transcription factors, it has been suggested that this component of the anti-Yo Ab response might interfere with regulation of gene expression (40). However, neither have the functions of the 34- or 62-kd protein been elucidated by in vivo analysis of knockout animals or by other methods nor has the intracellular localization of the 2 antigens been individually determined by electron microscopy.
We demonstrate here that IgG containing anti-Yo Abs can cause Purkinje cell death, but we did not determine whether the Purkinje cell death induced by anti-Yo Ab represents the effect of Abs directed against the 34-kd antigen, the 62-kd antigen, both antigens together, or, less likely, Purkinje cell antigens consistently recognized by anti-Yo Abs but not labeled in Western blots. Experiments are currently in progress to determine the individual and combined effects of each known component of the anti-Yo Ab response on Purkinje cells in organotypic cerebellar cultures.
The rapidity with which anti-Yo IgG enters Purkinje cells and produces Purkinje cell death in vitro may have important implications for human disease. In our studies, entry of IgG into Purkinje cells occurred within 24 hours; cell injury and death were significant at 72 hours and were extensive by 96 to 144 hours. If entry of anti-Yo IgG into Purkinje cells and cell death follow a similar course in affected humans, there would seem to be very little time in which to institute therapy before Ab binding and cell death begins. Treatment of anti-Yo–related paraneoplastic cerebellar degeneration may thus require very early removal of Ab from CSF and/or as-yet undeveloped treatment strategies to block IgG uptake or intracellular IgG transport.
The authors thank Jeffrey Rothstein, MD, PhD, Dan Gincel, PhD, and Carol F. Coccia, the Johns Hopkins University School of Medicine, for their assistance in establishing the cerebellar slice culture system used in these experiments. The authors thank Dr John W. Rose and Dr James B. Burns for their helpful comments concerning the article.
This work was supported by a Merit Review Award from the United States Department of Veterans Affairs (Dr Greenlee) and an award from the Western Institute for Biomedical Research (Dr Greenlee). Dr Tsunoda is supported by an R21 award (R21NS059724) from the National Institutes of Health.