Miglustat Improves Purkinje Cell Survival and Alters Microglial Phenotype in Feline Niemann-Pick Disease Type C

Animals

Cats were raised in the animal colony of the School of Veterinary Medicine, University of Pennsylvania under the guidelines of the National Institutes of Health and USDA for care and use of animals in research. The study was approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Cats were housed in an environment with a temperature of 21°C, 12-hour light cycles, and 12 to 15 air changes per hour. Food and water were offered ad libitum. A PCR-based DNA test was used for diagnosis of the NPC1 missense mutation in peripheral blood leukocytes in cats on the first day after birth (20). Cats were classified as homozygous (NPC), heterozygous for the mutant allele (HET) or lacking the mutant allele (wild type [WT]). Previous studies showed no significant differences in neurological function or pathology between WT and HET cats; therefore, HET cats were included in clinical studies where indicated (21). Neurological examination, post-rotatory nystagmus (PRN), serum biochemical testing, and brainstem auditory evoked response testing were performed as previously described (21, 23).

Miglustat was administered at a dose of 25 mg/kg of powder every 12 hours dissolved in 0.5 ml of water and administered orally via syringe. Miglustat-treated cats (n = 6) received miglustat beginning at 3 weeks of age (the first age at which cats begin to take food other than the dam’s milk), and continued to receive miglustat until death. Both male (n = 2) and female (n = 4) cats were treated with miglustat.

Pharmacokinetics

Pharmacokinetic data on miglustat were obtained in 4 25-week-old WT cats and in 2 8-week-old WT cats following oral administration of 25 mg/kg miglustat. One ml of blood was collected in EDTA at 0 (pre-dose), 0.5, 1, 2, 4, 6, 8, 12, 16, 20, 24, 28, 32, and 36 hours following administration. Blood was centrifuged immediately at 4°C and plasma was collected and stored at −80°C. The concentration of miglustat in plasma was quantified using a validated LC-MS/MS assay with miglitol as internal standard. The limit of quantification was 10 ng/mL and the inter-day coefficient of variation was <12% and the inaccuracy <8%.

Postmortem Examination

Cats were killed using an overdose of barbiturates in accordance with the American Veterinary Medical Association guidelines. All untreated NPC disease cats (NPCuntreated) and 3 miglustat-treated cats (NPC-miglustat) were killed when they were no longer able to maintain sternal recumbency without assistance. Additionally, 3 NPC-miglustat cats were killed at 24 weeks of age (the mean age of death in untreated cats) (21) to compare tissue between age-matched untreated and treated cats. WT and HET cats were killed between 20 and 29 weeks of age for histological and biochemical comparisons. Immediately before death, each cat was given 0.5 mL of heparin (1000 units/mL) i.v. After death the cats were perfused with 500 mL of 0.9% cold saline and samples of brain, liver, spleen, and lung were acquired and frozen. Next, 750 mL of cold 4% paraformaldehyde was perfused into the left ventricle of the heart. After perfusion, further samples of brain, liver, spleen, and lung were collected and fixed in paraformaldehyde for 48 hours. Fixed samples were paraffin-embedded, sectioned, and stained with hematoxylin and eosin (H&E). Immunohistochemistry for calbindin (anti-rat CALB; Swant, Marly, Switzerland), CD18 (Leukocyte Antigen Biology Laboratory, University of California, Davis, CA), GM2 ganglioside (mab10-11; Progenics Pharmaceuticals, Tarrytown, NY) and GM3 ganglioside (DH2, GlycoTech, Gaithersburg, MD) gangliosides were performed. Filipin (Sigma-Aldrich, St. Louis, MO) histochemistry was utilized for identification of unesterified cholesterol sequestration in cells. Tissues were cut on a cryostat or sectioned by either vibratome or microtome for the immunohistochemical and histochemical procedures. Histological procedures were carried out according to published methods (3, 21, 24).

Biochemical Analysis of Lipids

Lipid studies were conducted on frozen tissues essentially as in previous studies (25–27). Water homogenates (20%) were obtained from dissected cerebral gray matter and cerebellum or from liver and total lipids were extracted in chloroform:methanol 1:2 (v/v), as described (25). Protein content was measured on the total homogenates by a modified Lowry procedure (28). Part of the extract was desalted and separated into 2 fractions using 100 mg reverse-phase Bondelut C18 columns (Agilent Technologies, Englewood, CO) (29, 30). The acid fraction eluted with methanol-water 12:1 (v/v) contained all gangliosides and was used without further purification. Total sialic acid was measured by the Svennerholm resorcinol-HCl method and aliquots of this fraction spotted on high performance silica gel G thin layer chromatography (HPTLC) plates (Merck, Darmstadt, Germany) using a Linomat 5 device (Camag, Muttenz, Switzerland).

The plates developed in chloroform-methanol-0.2% CaCl2 55:45:10 (by volume) were sprayed with resorcinol-HCl to visualize the sialic acid moiety of individual gangliosides, and densitometric quantification at 580 nm was performed using a Camag TLCII scanner/Cats software system(25). After normalization to the total sialic acid content and eventual correction of water loss using the protein content, individual concentrations were calculated taking into account the number of sialic acids for each ganglioside. The study of neutral glycolipids (glucosylceramide, lactosylceramide, gangliotriaosylceramide) was done on the second column eluate, obtained with chloroform-methanol 1:2 (v/v). In the brain tissue, separation of glucosylceramide from galactosylceramide was achieved using borate-impregnated HPTLC plates. Densitometry of plates sprayed with the orcinol-sulfuric acid reagent was done at 650 nm (25, 30). Chromatographic systems have been described (25). Analysis of the free sphingoid bases sphingosine and sphinganine was carried out on total lipid extracts by high-performance liquid chromatography using eicosasphinganine (Matreya, Pleasant Gap, PA) as an internal standard. Separation of orthophtalaldehyde derivatives was achieved on a 200 × 5 mm Spherisorb ODS2 column (Waters) and monitored by fluorometry (31).

Biochemical Protein Studies

Western blot analysis of LC3 was carried out according to the following method: Frozen brain tissue was homogenized in ice cold lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Igepal CA-630, 1% deoxycholic acid, 0.1% SDS supplemented with protease inhibitor cocktail), centrifuged (15000 rpm) for 30 minutes at 4°C and the supernatants (soluble fraction) were collected. Protein concentrations were determined using a BCA protein assay kit. For immunoblotting, samples were analyzed by SDS-PAGE (16% gels) under reducing conditions and transferred to Immuno-blot PVDF membranes. Membranes were blocked in 1x TBS, 0.1% Tween-20, 5% non-fat dry milk, 1% BSA, followed by incubation with an antibody to microtubule-associated protein light chain 3 (LC3, 2 mg/ml) and subsequent incubation with Peroxidase Labeled anti-Rabbit IgG secondary antibody (1:5000). SuperSignal West Pico Chemiluminescent Substrate was used for protein detection on a KODAK 2000R imaging station. Protein quantification for each sample was performed by densitometric analysis using KODAK imaging software and normalized to actin in the same sample. This analysis was represented as LC3-II/Actin normalized to the WT control.

Microglia Isolation and Characterization

Feline microglial cells were isolated via density gradient centrifugation using previously described methods with slight modifications (32). Briefly, samples of brain were collected immediately following saline perfusion and transported in Hankś solution (Gibco Invitrogen, Carlsbad, CA) with 3% fetal calf serum adjusted to pH 7.36, on ice. The meninges were carefully removed and 12 g of brain tissue were mechanically dissociated by mincing through a stainless-steel sieve and enzymatically digested with type II collagenase (Roche Diagnostics, Mannheim, Germany) and DNAse I (Sigma-Aldrich) for 60 minutes at 37°C. The digested CNS preparation was washed twice, resuspended in 45 ml of isotonic Percoll (Amersham Biosciences, Piscataway, NJ) and diluted in Hankś buffer resulting in a density of 1.030 g/ml. For the initial gradient this suspension was underlayered with 5 ml of isotonic Percoll of the density 1.124 g/ml. After centrifugation (1,250 × g for 25 minutes at 20°C) cells from the 1.124 g/ml-density surface were collected, washed, and for further purification transferred to a major gradient consisting of 5 consecutive Percoll-densities (5 ml of the density 1.124 g/ml, 12 ml of 1.072 and 1.060 g/ml each, 8 ml of 1.050, and 1.030 g/ml Percoll) in a 50-ml tube. After centrifugation, feline microglia accumulated specifically on the surfaces of the densities 1.072 and 1.060 g/ml Percoll (33). The cells were collected, washed and subsequently used for the ex vivo examinations. Cell viability was assessed by Trypan blue exclusion.

Phenotypic-expression of feline microglial cells was evaluated by flow cytometry with several monoclonal antibodies either specific for, or cross-reactive with, feline cells. Staining of microglial cells was performed as previously described for the canine species (32). Briefly, after Fc-receptor blockage with normal goat serum IgG (dilution 1:11; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), primary monoclonal antibodies against the surface molecules CD18, CD11b, CD11c, CD45, CD1c, intercellular adhesion molecule-1 (CD54), B7-1 (CD80), major histocompatibility complex (MHC) class I, MHC class II, CD44, CD14 (directly labeled), CD4, CD8α, CD21, and CD3 (Table 1) were incubated for 30 minutes at 4°C and washed twice with Hankś buffered saline. Secondary antibodies (Table 2) were diluted 1:100 and incubation continued for 30 minutes. Staining against CD3 required fixation and permeabilization of the cells. After washing, cells were resuspended in FACSFlow (BD Biosciences, San Jose, CA) and analyzed immediately ex vivo after staining without fixation with a FACSCalibur (BD Biosciences) using the Cell-Quest software. Negative control samples were stained with the secondary antibodies only. Analysis gates were adjusted to not exceed 2% positive staining with negative controls. For each sample 10,000 live-gated events were analyzed.

Microglial generation of reactive oxygen species (ROS) was analyzed as described with slight modifications (34). Briefly, the cell suspension was adjusted to a concentration of 2 × 10⁵ cells in 100 μl and pre-incubated for 15 minutes at 37°C and 5 % CO₂. This incubation step was repeated after adding either phorbol myristate acetate (PMA, Sigma-Aldrich) to result in a final concentration of 100 nM or phosphate buffered saline (PBS), which served as negative control. An additional tube with microglial cells only was used as negative reference to evaluate morphology and background fluorescence of the cells. Dihydrorhodamine 123 (Marker Gene Technologies Inc., Eugene, OR) was diluted in dimethyl sulfoxide to a concentration of 15 μg/ml. After adding 20 μl of the diluted dihydrorhodamine 123, incubation was again repeated. Cells were then put on ice in dark surroundings to stop ROS generation, minimize the amount of tube surface adhering cells, and to protect the fluorochromic particles. After addition of FACSFlow samples were measured immediately by flow cytometry in duplicates for each approach and mean values were used for evaluation. The results were evaluated on the basis of percentage of positive (ROS-producing) cells and mean fluorescence intensity (measured by means of fluorescent channel numbers, log values) as a relative means of the intensity of ROS-generation seen as a shift in FL1. Results were evaluated using the multiple histogram analysis of the FACS Cell-Quest Software provided.

Microglial phagocytosis was assessed with heat-killed and lyophilized FITC-labeled Staphylococcus aureus (Invitrogen, Molecular Probes, Eugene, OR) that were resuspended in PBS, sonicated, and adjusted to a concentration of 8 × 10⁸ bacteria/ml. Bacteria were incubated with 20 μl pooled cat serum diluted with PBS (1:5) for 60 minutes at 37°C and 5% CO₂ in a humid atmosphere for opsonization. Microglia were mixed with either opsonized, non-opsonized bacteria suspension resulting in a ratio of 1:100, or with PBS (negative control). Incubation was performed as described before with gentle resuspension after 30 minutes to ensure optimal cell-to-cell-contacts. To stop the phagocytosis reaction and minimize adhesion of cells on the tube surface after incubation, cells were put on ice for 15 minutes. After addition of 100 μl FACSFlow into each tube, the percentage of cells performing phagocytosis and phagocytosis intensity (measured as relative means by mean fluorescent channel numbers, values shown in log) was determined by immediate flow cytometric measurement with the FACSCalibur and Cell-Quest software. Phagocytosis measurements were performed in duplicates and mean values were used for evaluation.

Statistical Analysis

Mean values and standard deviations were calculated to describe the findings. The unpaired 2-tailed t test was used to compare data of untreated cats with NP-C with miglustat-treated and normal cats. Values were considered statistically significant at the p < 0.05 (* or †) or p < 0.001 (**) level.

Pharmacokinetics of Miglustat in WT Cats

The plasma half-life of miglustat was 6.6 ± 1.1 hours in 3 WT 25 wk old cats in which it was orally administered. t_max, C_max, and area under the plasma concentration-time curve in these cats were 1.7 ± 0.6 h, 20.3 ± 4.6 μg/mL, and 104.1 ± 16.6 μg*h/mL respectively. Mean half-life, T_max, C_max, and area under the plasma concentration-time curve (AUC) in 2 WT 8-week-old cats were 7.5 hours, 1.29 hours, and 24.7 μg/mL, and 104.1 μg*h/mL, respectively.

Clinical Outcome Data in NPC Disease Cats Treated with Miglustat

The mean time to euthanasia due to disease progression was significantly longer (35.7 ± 0.6 weeks) for NPC-miglustat cats than for the NPC-untreated cats (20.5 ± 4.8 weeks). Increased survival time was due to a delay in onset of the more severe signs of neurological dysfunction typically observed late in disease (Fig. 1A). Specifically, a significant difference in onset of clinical signs of disease was identified in treated cats when compared to untreated 17- and 19-week-old cats, while no significant difference was found in clinical signs between untreated and treated cats at 6 and 12 weeks of age. Remarkably, treated cats maintained the ability to walk for 11 weeks longer than untreated cats but all treated cats eventually developed the full spectrum of clinical signs that were of the same severity as those in untreated cats.

Untreated NPC-affected cats showed a significantly greater number of beats of PRN vs. WT cats beginning at 12 weeks of age and continued to show greater beats of PRN until death at 24 weeks of age (Fig. 1B). In contrast, treated cats were not significantly different from WT cats until 20 weeks of age. When treated animals were compared to untreated animals, significantly fewer PRN beats were found at 12 and 20 weeks of age (Fig. 1B). Quantitative testing of the auditory system showed no significant differences between NPC-miglustat cats and untreated cats in measures of central conduction time, wave V/I amplitude, or in hearing threshold (data not shown).

NPC-untreated cats showed significant elevations in serum ALT, AST, cholesterol, and chitotriosidase compared to WT cats whereas no differences were identified between NPC-miglustat and NPC-untreated cats in these measures. Finally, no significant differences in body weight were identified between NPC-untreated cats and NPC-miglustat cats (data not shown), which were each significantly lower than control cats beginning at 5 weeks of age and continuing until the time of death.

Histological Studies

Three NPC-miglustat cats were euthanized at 24–25 weeks of age to compare brain histology and biochemistry to age-matched NPC-untreated cats and to 3 other NPC-miglustat cats at terminal stage (35–36 weeks of age). There was widespread accumulation of GM2 and GM3 gangliosides and cholesterol in the cerebral cortex was in the treated and non-treated 24- to 25-week-old NPC-affected cats (Fig. 2). Blinded analysis comparing the two groups for storage of GM2 and GM3 gangliosides revealed qualitatively less accumulation of both gangliosides in NPC-miglustat cats (Fig. 2A, B vs. D, E). As anticipated, WT brain exhibited no detectable GM2 or GM3 ganglioside accumulation (Fig. 2G, H). Unexpectedly, in addition to the reduced ganglioside storage, cholesterol storage, as revealed by levels of filipin staining also appeared to be qualitatively reduced in the treated vs. NPC-untreated cat (Fig. 2C, F). As anticipated, WT cerebral cortex showed no significant filipin labeling (Fig. 2I). NPC disease cats that survived to 35 weeks following treatment also showed GM2 and GM3 gangliosides and cholesterol at detectably lower levels than those observed in the 24-week-old untreated NPC disease cat at end-stage disease (not shown).

In the cerebellum, NPC-miglustat cats exhibited a remarkable preservation of Purkinje cells as compared to age-matched NPC-untreated cats (Fig. 3). Nonetheless, storage of GM2 ganglioside and cholesterol was prominent in all of the surviving Purkinje cells in both treated and untreated animals (Fig. 3B, C, E, F). Again, however, in terms of the relative levels of GM2 staining observed in the granule cell layer of the cerebellum, this appeared qualitatively less in the treated cat (Fig. 3E vs. 3B). The cerebellum of NPC-miglustat cats that survived to 35 weeks showed loss of Purkinje cells, whereas surviving cells containing GM2 and cholesterol storage similar to the untreated cat at 24 weeks of age (not shown). Finally, no differences were identified between NPC-miglustat and NPC-untreated cats on light microscopic evaluation of the liver.

Biochemical Lipid Studies in Brain

Biochemical studies of human patients have shown that sphingolipid abnormalities essentially affect the cerebral gray matter and much less the cerebral white matter (35). This is also true for the cat model (Vanier and Vite, unpublished data) and is well in line with the neuropathological studies. Analyses in cerebrum were, therefore, conducted in carefully dissected gray matter. Because of their limited sizes, cerebellar samples included some white matter. Cholesterol concentrations were not specifically studied because it has previously been shown that the distribution changes observed with filipin staining are quantitatively too small to affect the global levels of cholesterol in brain, even in gray matter (35).

In cerebral gray matter of NPC-untreated cats (aged 21–29 weeks; n = 8), GM2 and GM3 gangliosides, which are normally minor components, showed a 10- to 12-fold increase compared to WT cats (Fig. 4A, B). Glucosylceramide and lactosylceramide, only present in minute amounts in gray matter of unaffected humans (35) or cats (data not shown), were also markedly increased in NPC-untreated cats (Fig. 4C, D), as previously reported for affected patients and mice (30, 35). The rate of migration of the single band observed for each of these lipids is in line with a C18:0 fatty acid moiety, as documented in human NPC disease (35). In cerebral gray matter of miglustat-treated cats (n = 6) there was a minor reduction in the concentrations of GM2 (414 ± 62 vs. 483 ± 46 nmol/g; p = 0.034), and a mean 27% reduction in the concentrations of GM3 gangliosides, with large inter-individual variations (406 ± 129 vs. 563 ± 86 nmol/g; p = 0.017). By biochemical measurement, comparison between 24-week-old and 35- to 36- week-old treated cats did not disclose clear differences. In normal human subjects, cerebellar tissue is known to show a ganglioside pattern strikingly different from that observed in cerebrum, with predominantly polysialogangliosides of the b-series. This was also the case in cats (Fig. 4A). The WT concentrations of GM3 and GM2 gangliosides in cerebellum appeared slightly lower than those in cerebrum (Fig. 4B). In NPC-untreated cats (n = 4), the increase of GM2 was 10- to 12-fold that of WT (i.e. its concentration was about half of that in cerebral cortex), but GM3 was only 3- to 4-fold that of WT (Fig. 4B). In the cerebellum of NPC-miglustat cats (n = 5) there was a mean 35% decrease in storage of GM2 (185 ± 24 vs. 284 ± 63 nmol/g; p = 0.014), and there was no significant change in the concentration of GM3 (Fig. 4B). Again, no clear trend was found comparing 24-week-old and 35- to 36-week-old treated animals. Concentrations of minor neutral glycosphingolipids were only studied in cerebral cortex. There were differences in treated vs. untreated NPC disease cats for glucosylceramide and lactosylceramide (Fig. 4C, D). Miglustat-treated cats showed an approximately 3- fold increase in their concentrations of glucosylceramide and an approximately 2-fold decrease in their concentrations of lactosylceramide. No noticeable change was observed regarding the concentrations of gangliotriaosylceramide (asialo-GM2).

Free sphingoid bases, sphingosine (Fig. 5), and sphinganine were also studied. In untreated affected cats, a 3-fold increase was observed both in cerebral gray matter and in cerebellum, compared to WT age-matched cats. A proportional or higher increase in sphinganine (not shown) was also observed. In cerebrum of miglustat-treated cats (n = 6) there was a small but significant reduction in sphingosine concentrations compared to those in untreated cats (n = 8) (113 ± 19 vs. 149 ± 14 pmol/mg protein; p = 0.001). A similar, although not significant trend (possibly due to the smaller number of samples studied) was observed in cerebellum.

Lipid Studies in Liver Tissue

As expected, miglustat has no effect on cholesterol, sphingomyelin or bis(monoacylglycero)phosphate (BMP) (Fig. 6A). This also applied to the levels of free sphingosine and sphinganine. Sphingosine showed an approximately 50-fold increase in affected untreated cats (1675 ± 155 pmol/mg protein (n = 4) vs. a normal value of 37 ± 15 (n = 7). Values of 1375, 1940 and 2250 pmol/mg protein were found in 3 miglustat-treated cats. With regard to glycosphingolipids, liver tissue from miglustat-treated cats showed no significant reduction vs. untreated cats in the very abnormal levels of glucosylceramide, lactosylceramide and globotriaosylceramide (Fig. 6B). GM3 ganglioside was increased approximately 5 times in affected untreated cats (1.1–1.5 μmol/g tissue) vs. normal cats (0.2–0.3 μmol/g) and remained unchanged in miglustat-treated cats (0.9–1.4 μmol/g). We were thus unable to demonstrate any effect of miglustat (histologically or biochemically) on liver lipid storage in cats.

Microglia Immunophenotype

Recent studies suggest that inflammation in CNS neuronal storage disorders might play an important role in the pathogenesis of neurodegeneration and its therapy (36, 37). Therefore, we characterized microglial immunophenotype and function ex vivo to evaluate a possible impact on the pathogenesis of NPC disease. Feline microglia cells were identified on the basis of size (FSC = forward scatter), complexity (SSC = side scatter), and their characteristic CD18-positive and CD45^low immunophenotype. For most parameters examined 10,000 live-gated events could be evaluated in flow cytometry. Microglial purities of 86.1 to 99.0% (× = 95.3) were used for the ex vivo examinations. The low percentage of CD4-positive (× = 2.9%), CD8α-positive (× = 2.6%), CD21_-positive (× = 2.4%), and CD3-positive (× = 1.9%) cells indicative of T and B lymphocytes revealed an insignificant contamination with these cells.

In healthy age-matched WT cats, the isolated microglial cells appeared as a homogenous population of relatively small cells (Fig. 7A). In contrast, the microglia cell population of NPC-untreated cats displayed a distinctly increased diversity in size and complexity (Fig. 7B, C); less diversity was apparent in miglustat-treated cats (Fig. 7C).

Eleven antibodies were used to characterize microglial immunophenotype. Two parameters were assessed for each surface antigen examined: the percentage of positive microglia cells and the mean fluorescence intensity (measured by mean fluorescent channel numbers) as a relative mean for level of expression. The percentage of CD1c-positive cells and the expression intensity of CD1c were inversely affected by disease. NPC-untreated cats showed a decrease in CD1c-positive cells but an increase in expression compared to WT cats (p = 0.0164, Fig. 8A, B). In contrast, 35 wk old NPC-miglustat cats, with clinical signs as severe as those found in untreated 24 wk old cats, showed percentage and expression of CD1c that were more similar to those of WT than to those of NPC-untreated cats. Despite the variation within the group of 4 NPC miglustat-treated cats at the 2 ages examined, significantly lower expression intensity was found in the miglustat-treated cats compared to NPC-untreated cats (p = 0.04, Fig. 8B). In contrast, the expression intensity of CD45 was significantly upregulated in miglustat-treated cats (p = 0.02) compared to WT cats, although the percentage of CD45-positive microglia was unaltered (data not shown).

Upregulation of the expression intensity was also seen for the surface antigens B7-1 (CD80) (data not shown), CD14 (data not shown), and MHC I (Fig. 8C) in NPC– untreated cats vs. WT cats. The expression intensity of MHC I was also significantly less in NPC-miglustat cats at both 24 and 35 weeks of age vs. NPC-untreated cats (Fig. 8C; p = 0.02). Intercellular adhesion molecule-1 and CD44 expression intensity were significant upregulated in miglustat-treated cats vs. WT cats (p = 0.004 and 0.04, respectively). For CD44 there was a tendency for upregulation in NPC-untreated cats compared to the controls (p = 0.08). No difference in expression or expression intensity between NPC-untreated and WT cats could be demonstrated for the integrins CD18, CD11b, and CD11b, or for MHC II.

Generation of ROS

To evaluate microglial generation of oxygen radicals on the pathogenesis of NPC and their modulation by miglustat we performed an ROS generation test. Feline microglia could be triggered to produce ROS, which is reflected by the high percentage of positive cells (PBS and PMA) when compared to the negative control (Fig. 9). The percentage of ROS-generating microglia was significantly lower in NPC-untreated cats compared to WT (p = 0.02 for PBS both and PMA) and NPC-miglustat cats (p = 0.02 for both PBS and PMA; Fig. 9). There was no difference in the enhancement of ROS generation by stimulation with PMA. The intensity of ROS generation was significantly enhanced in NPC-untreated cats compared to WT cats (p = 0.04 for PBS and PMA; Fig. 9B). The difference between the groups of cats was already very distinct in comparison to the negative control preparation, which was 4.3-fold higher in untreated and 5.4-fold higher in NPC–miglustat cats (p = 0.0001 and p = 0.04, respectively), and compared to age-matched WT cats. The 2 35-week-old miglustat-treated cats demonstrated a distinctly higher ROS generation intensity than the 2 24-week-old cats (Fig. 9B), leading to a high standard deviation within this group of treated cats. Therefore, only a tendency for an upregulated ROS generation intensity could be recognized compared to healthy control cats (p = 0.1).

Phagocytosis

Feline microglia showed a distinct phagocytosis of FITC-labeled Staphylococci, reflected by the high percentage of positive cells (non-opsonized and opsonized bacteria) compared to the negative control (Fig. 9C). No differences were noted between age-matched WT and cats affected with NPC either treated or untreated (Fig. 9C).

A significantly enhanced phagocytosis of opsonized Staphylococci could be demonstrated in the NPC-untreated cats compared to the age-matched WT cats (p = 0.006, Fig. 9D). This significant difference could also be detected for the negative control preparation, which showed a 3.5-fold higher intensity in the NPC-untreated cats compared to the age-matched WT cats (p = 0.0001).