Treatment of canine fucosidosis by intracisternal enzyme infusion

Descripción

Experimental Neurology 230 (2011) 218–226

Contents lists available at ScienceDirect

Experimental Neurology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y e x n r

Treatment of canine fucosidosis by intracisternal enzyme infusion Gauthami S. Kondagari a, 1, Barbara M. King b, Peter C. Thomson a, Peter Williamson a, Peter R. Clements c, Maria Fuller b, Kim M. Hemsley b, John J. Hopwood b, Rosanne M. Taylor a,⁎ a b c

Faculty of Veterinary Science, University of Sydney, NSW 2006, Australia Lysosomal Diseases Research Unit, SA Pathology at the WCH, North Adelaide, South Australia 5006, Australia Genetics and Molecular Pathology, SA Pathology at the WCH, North Adelaide, South Australia 5006, Australia

a r t i c l e

i n f o

Article history: Received 15 December 2010 Revised 20 April 2011 Accepted 28 April 2011 Available online 6 May 2011 Keywords: Canine fucosidosis Lysosomal storage disease Central nervous system Intracisternal enzyme replacement therapy Cerebrospinal ﬂuid Tandem mass spectrometry

a b s t r a c t The blood brain barrier is the major obstacle to treating lysosomal storage disorders of the central nervous system such as canine fucosidosis. This barrier was overcome by three, monthly injections of recombinant canine α-L-fucosidase enzyme were given intracisternally. In dogs treated from 8 weeks of age enzyme reached all areas of central nervous system as well as the cervical lymph node, bone marrow and liver. Brainstem and spinal cord samples from regions adjacent to the injection site had highest enzyme levels (39– 73% of normal). Substantial enzyme activity (8.5–20% of normal controls) was found in the superﬁcial brain compared to deeper regions (2.6–5.5% of normal). Treatment signiﬁcantly reduced the fucosyl-linked oligosaccharide accumulation in most areas of CNS, liver and lymph node. In the surface and deep areas of lumbar spinal cord, oligosaccharide accumulation was corrected (79–80% reduction) to near normal levels (p b 0.05). In the spinal meninges (thoracic and lumbar) enzyme activity (35–39% of normal control) and substrate reduction (58–63% affected vehicle treated samples) reached levels similar to those seen in phenotypically normal carriers (p b 0.05).The procedure was safe and well-tolerated, treated (average 16%) dogs gained more weight (p b 0.05) and there was no antibody formation or inﬂammatory reaction in plasma and CSF following treatments. The capacity of early ERT to modify progression of biochemical storage in fucosidosis is promising as this disease is currently only amenable to treatment by bone marrow transplantation which entails unacceptably high risks for many patients. Crown Copyright © 2011 Published by Elsevier Inc. All rights reserved.

Introduction Fucosidosis is an inherited, progressive, neurodegenerative lysosomal disorder in human (OMIM 230000) and animals caused by deﬁciency of α-L-fucosidase (EC 3.2.1.51) (Occhiodoro et al., 1992). Intracellular accumulation of oligosaccharides, glycopeptides and glycolipids with terminal fucose residues occurs in brain, visceral organs and urine (Abraham et al., 1984; Ferrara et al., 1997). Delayed psychomotor development, progressive loss of motor skills, ataxia and mental deterioration occur and the disease is lethal in childhood. In dogs with this condition, the clinical signs are primarily neurological, commencing with subtle behavioral changes from weaning and

Abbreviations: CSF, Cerebrospinal ﬂuid; ERT, enzyme replacement therapy; ESS, English Springer Spaniels; CNS, central nervous system; MPS, mucopolysaccharidosis; AET, affected enzyme treated; AVT, affected vehicle treated; CVT, control vehicle treated; ELISA, enzyme linked immunosorbent assay; SEM, standard error of means; LC, liquid chromatography. ⁎ Corresponding author at: Faculty of Veterinary Science, JD Stewart Building (B01) The University of Sydney, NSW 2006, Australia. Fax: + 61 2 9660 1548. E-mail address: [email protected] (R.M. Taylor). 1 Current address: Faculty of Medicine University of New South Wales Level 2, Westﬁeld Research Laboratory Sydney's Children's Hospital High Street Randwick, NSW 2031, Australia.

proprioceptive deﬁcits from 8 to 12 months of age (Taylor et al., 1987). Affected dogs are usually euthanized by 30 months of age. In vitro (Bielicki et al., 2000) and in vivo (Pomroy, 2002) enzyme replacement in fucosidosis has been beneﬁcial in reversing fucoside storage, mirroring the ﬁndings of enzyme replacement therapy (ERT) in other lysosomal disorders. Following bone marrow transplantation (BMT) in fucosidosis dogs there was a rapid, sustained increase in α-Lfucosidase activity in plasma, leukocytes, visceral and a slower rise in neural tissues. This sustained enzyme delivery was stable for the life time of the transplant and ameliorated the central nervous system (CNS) pathology when undertaken early in disease. BMT delayed the onset of clinical signs (Taylor et al., 1992) and has been used to improve the quality of life of affected children (Vellodi et al., 1995). However, BMT is associated with signiﬁcant mortality, and not all patients have an appropriate donor available, therefore additional treatment options are required. The potential for effective ERT of CNS via the cerebrospinal ﬂuid (CSF) has been demonstrated in dogs, mice and patients with mucopolysaccharidosis (MPS) and other lysosomal disorders (Auclair et al., 2010; Chang et al., 2008; Dodge et al., 2009; Lee et al., 2007; Munoz-Rojas et al., 2010). Intrathecal ERT was used to circumvent the blood brain barrier in MPS I mutant dogs and produced high concentrations (23-fold normal) of enzyme in brain 48 hours (h)

0014-4886/$ – see front matter. Crown Copyright © 2011 Published by Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2011.04.019

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

after treatment (Kakkis et al., 2004). The ﬁndings of reduced glycosaminoglycan (GAG) storage (57% of normal) in CNS were promising, however MPS I dogs had previously been sensitized to recombinant enzyme, which may have inﬂuenced GAG and histological storage in tissue (Kakkis et al., 2004). Frequent intrathecal and intravenous infusions in MPS I dogs normalized brain substrates (Dickson et al., 2007). In MPS IIIA mice intrathecal ERT replaced the missing enzyme in neural tissues but did not eliminate the early axonal pathology (Hemsley et al., 2008). High dose intrathecal ERT reduced substrate accumulation in a dose-dependent manner with a signiﬁcant reduction in neuropathology (Hemsley et al., 2007, 2008). In MPS IIIA dogs, similar treatment resulted in widespread enzyme distribution in superﬁcial and deep areas of brain with reduced substrates (Hemsley et al., 2009). The treatment has proved safe in one MPS VI patient (Munoz-Rojas et al., 2008) and clinical trials of intra-lumbar CSF ERT are presently underway for MPS I (NCT00852358), MPS II (NCT00920647) and MPS IIIA (NCT01155778). This is the ﬁrst study of intra-CSF enzyme replacement therapy in a glycoproteinosis disorder. Whilst the efﬁcacy of intra-CSF enzyme replacement is evident in several other neurodegenerative conditions, we cannot presume that this treatment will be efﬁcacious in all. Other therapeutic avenues, for example BMT are efﬁcacious in some LSD e.g. αmannosidosis and fucosidosis (Krivit et al., 1999; Sivakumur and Wraith, 1999) but not all. In the present study we report increased enzyme activity and reduced storage after repeated intrathecal recombinant canine α-L-fucosidase infusion in the developing canine brain and spinal cord. We also report that CSF enzyme and substrate levels were sensitive, accessible markers for monitoring treatment. Material and methods Animals Unaffected and fucosidosis English Springer Spaniels (ESS) were bred, maintained, utilized and prepared in accordance with approved Animal Ethics Committee protocols of Westmead Hospital (#1000, #4007). Adult carrier dogs were bred to produce pups which were genotyped by α-L-fucosidase assay as affected, normal and carrier (Healy et al., 1984). Experimental animals were grouped as affected enzyme treated (AET, n = 3), affected vehicle treated (AVT, n = 3) and control vehicle treated (normal dogs) (CVT, n = 3) (Table 1). Production of recombinant canine α-L-fucosidase Canine α-L-fucosidase was synthesized in Chinese hamster ovary and Madin–Darby canine kidney cells and puriﬁed using an afﬁnity chromatography method on fucosylamine-linked agarose (Bielicki et al., 2000). Two batches were produced, with an enzyme activity assay determining the activities to be 46 and 33.8 U/mL, with speciﬁc activities of 7.2 and 11.3 U/mg, respectively. The enzyme was stored at 4 °C until use. Table 1 Experimental animal groups. Adult dogs are carriers for fucosidosis and were used for substrate analysis. Group

Age at necropsy (months)

8–12 Adult carrier English Springer spaniel CVT 4 AET 4 AVT

Number Infusion Clinical observation (n) 3

No

Normal

3 3

PBS Enzyme

Normal No anxiety, quiet, accepted restraint Anxiety, apprehension, unwilling to accept restraint

3 3–4 (one anestheticrelated death at 3 months)

PBS

219

Intracisternal infusions Dogs received intracisternal infusion via the cisterna magna. Animals were fasted for 12 h and premedicated 30 min prior to anesthesia with subcutaneous atropine (Pharmacia Perth Private Ltd, Perth, Australia) 0.04 mg/kg and buprenorphine [(Temgesic) Rickitt Benckiser Health Care Ltd, NSW, Australia] 0.01 mg/kg. Each dog was masked down quietly using 1 L/min oxygen and 5% isoﬂuorane and connected via Tpiece to the vaporizer. Anesthesia was maintained on 1.5–2% of isoﬂuorane given to effect and 0.5–1.0 L/min oxygen. Infusion procedure The head was positioned with the dog in lateral recumbency and a 22 G pediatric CSF needle (B.D. Spinal needle, 0.7 × 38 mm, Becton Dickson SA, Madrid Spain) was inserted under sterile conditions into the cisterna magna. CSF was collected (1–3 mL) and dogs were then administered a total of 6.9 U of recombinant canine α-L-fucosidase or vehicle per injection (in a volume of 150 or 204 μL, depending on enzyme batch), over a period of 2 min. Blood samples were collected before the infusion and weekly through the experimental period and at necropsy. Clinical monitoring Daily clinical and monthly neurological examinations were undertaken. Body weights were measured weekly. Necropsy and sample collection Forty-eight hours after the third infusion blood was collected and dogs were euthanized with an overdose of pentobarbitone (150 mg/ kg). Cisternal CSF, brain, visceral organs, meninges and spinal cord were collected and frozen at −20 °C for biochemical studies. Tissues were also obtained at necropsy from three adult carrier dogs and one affected newborn pup. Routine laboratory studies Blood samples were collected weekly and at post-mortem. Complete blood counts with differential leukocytes and vacuolated lymphocytes were conducted on blood smears stained with Haematoxylin and Eosin at 1000× magniﬁcation. CSF samples were analyzed for protein, glucose, α-L-fucosidase activity and substrates. Cytospin CSF smears were stained with Haematoxylin and Eosin at 1000× magniﬁcation and evaluated for vacuolated lymphocytes. Blood contaminated CSF samples were not analyzed. Detection of anti-fucosidase antibodies in blood plasma and CSF Antibodies to α-L-fucosidase were measured in plasma and CSF samples from the experimental animals and from age-matched animals from the same litter which did not receive infusions (negative control). A positive-control sample was obtained from a dog which developed a low level antibody response following repeated IV injections of α-L-fucosidase for one year (Pomroy, 2002). Ninety-six well Maxisorp plates (Nunc technologies Inc., Chantilly, VA, USA) were coated with a 5 μg/mL solution of α-L-fucosidase diluted in 0.1 M NaHCO3 (pH 9.6) and incubated overnight at 4 °C. They were then blocked with blocking buffer (0.5% (w/v) gelatine, 0.2% (v/v) Tween 20 in 0.02 M TRIS/0.25 M NaCl), the plate was washed and plasma (diluted 1:5 in block buffer) was added and incubated at 37 °C for 60 min. The plate was washed and rabbit anti-dog IgG-horse radish peroxidase (Sigma Aldrich, NSW, Australia), diluted 1/500 in block buffer was applied for 60 min at 37 °C. Fifty microlitres of 2,2′ azinobis-3-ethylbenzthiazoline-6-sulfonic acid microwell peroxidase substrate (KPL, Doncaster, Victoria) was added and incubated at 37 °C for

220

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

Fig. 1. Body weight (1a), antibody titres in plasma samples (1b) (mean ± SEM). Where a signiﬁcant difference between the groups was identiﬁed at P b 0.05, they were allocated a different letter, so that groups with the same letter are not signiﬁcantly different. PM = post-mortem.

30 min for color development. The reaction was stopped with peroxidase stop solution. A micro-plate spectrophotometer (Spectra max 250, Molecular Device, USA) read the color intensity at 405 nm and ﬁnal antibody levels were obtained from the standard curve. All the samples were batch-run in triplicate.

analysis or tris sodium chloride buffer for substrate analysis (0.02 M tris, 0.5 M sodium chloride). Samples were freeze-thawed six times and centrifuged at 13000 rpm for 5 min at 4 °C. The supernatants were collected for enzyme assay and substrate analysis and stored at −20 °C. Total protein was measured (Bradford, 1976).

Biochemical analysis of tissues and ﬂuids

Blood sample preparation Blood samples were centrifuged, plasma was harvested and stored at −20 °C. The packed cells were harvested from the centrifuged blood (Healy et al., 1984). Leukocyte preparations were stored at −20 °C and analyzed in batches.

Tissue preparation A tissue sample of 0.1–0.2 g was homogenized (Dispenser T10) in 1 mL triton saline (2% triton, 0.9% sodium chloride) for enzyme

Fig. 2. Enzyme activity and substrate accumulation. Enzyme activity in cerebral cortex (2a), mid- and hind-brain (2b), brain stem and cerebellum (2c), spinal cord (2d) and meninges (2e) is shown. Substrate accumulation in cerebral cortex (2f), mid- and hind-brain (2g), brain stem and cerebellum (2h), spinal cord (2i) and meninges (2j) is shown (SC-spinal cord). Letters indicate statistically signiﬁcant differences between groups (p b 0.05). Where a signiﬁcant difference between the groups was identiﬁed at P b 0.05 they were allocated different letter, so that groups with the same letter are not signiﬁcantly different.

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

221

222

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

Measurement of α-L-fucosidase activity and protein

Results

Enzyme assays (Healy et al., 1984) used 2.84 nM/L 4-methylumbelliferyl standard (Sigma Aldrich, NSW, Australia) and 4-methylumbelliferyl fucoside substrate (Sigma Aldrich, NSW, Australia) with glycine stop buffer. The samples were read in Cary Eclipse Fluorescence Spectrophotometer (Varian Australia Pty. Ltd., VIC, Australia) at 365 nm excitation and 455 nm emission and 10 nm slit. The results were calculated as the mean of multiple enzyme readings from the standard curve. The enzyme activity was expressed as units/mg of protein. In AET the levels of enzyme activity was expressed as % of control = AET × 100/CVT.

Clinical response to ERT procedure

Liquid chromatography electrospray ionization tandem mass spectrometry (LC ESI-MS/MS) of fucoside derived oligosaccharide To each brain supernatant (0.1 mg protein), 1 nmol of internal standard (ISTD, O-(2-acetamido-2-deoxy-D-glucopyranosyl)-L-serine) was added and the supernatants were de-proteinized by the addition of 0.6 mL ice cold methanol. The supernatant was removed and 0.1 mL was lyophilized and derivatized with 0.2 mL of derivatizing solution (1:1:1:0.15 of H2O:ethanol:pyridine:phenylisothiocyanate) for 2 h at 45 °C. For CSF samples, 0.5 nmole of ISTD was added to 5 μL of CSF followed by lyophilization and derivatization. Samples were cooled by the addition of 0.3 mL H2O and excess derivatizing reagent was removed with two 0.75 mL CHCCl3 washes. Following the second CHCCl3 wash, 0.3 mL of the aqueous layer was removed, lyophilized and resuspended in 1% (v/v) acetonitrile in H2O. LC separation of the fucose derived oligosaccharide (fucose-nacetyl glucosamine-aspartic acid) was performed on a 3 μm Alltima C18 column (50 × 2.1 mm) by injecting 20 μL in mobile phase A (1% (v/v) acetonitrile in H2O with 0.025% (v/v) formic acid). A linear elution gradient of from 5% to 30% mobile phase B (0.025% (v/v) formic acid in acetonitrile) was established between 0 and 6.5 min and then stripped with 100% B, 0.5 min, and reequilibrated with 5% mobile phase B. Mobile phases were delivered at 0.2 mL/min and a divert valve directed salts eluting from the column to waste between 0 and 3.5 min and a ﬂow of 0.2 mL/min was delivered to the mass spectrometer during this time. Mass spectrometric analysis of the fucose-derived oligosaccharide was performed on a PE SCIEX API 3000 triple quadrupole mass spectrometer with an ionspray source. Nitrogen was used as curtain gas, collision gas, nebuliser gas and auxiliary gas. Ion source temperature was set to 350 °C and ion spray voltage to +5500 V. Quantiﬁcation of the fucose-derived oligosaccharide was performed using multiple-reaction monitoring in positive ion mode for the ion pair 617/251 and 444/204 for the ISTD. Quantiﬁcation was achieved by relating the peak area of the fucose-derived oligosaccharide to the peak area of the ISTD. The substrates accumulated in AET were expressed in terms of % of affected = AET X 100/AVT. Statistics GenStat 11th edition software (GenStat 11 edn, VSN International Ltd, Hertfordshire, UK) was used for statistical analysis. The raw data of enzyme assay, substrates and ELISA were log transformed for statistical analysis (analysis of variance). Descriptive summaries are expressed as back-transformed means ± standard error of means (SEM). CSF glucose and microprotein, body weight were analyzed using analysis of variance and expressed as means ± SEM. A repeated measure of analysis of variance was used for CSF substrate analysis. In Figs. 1–2 where a signiﬁcant difference between groups was identiﬁed at p b 0.05 they were allocated a different letter. Groups with the same letter are not signiﬁcantly different. Where a difference of p b 0.01 was identiﬁed it is noted in the text.

Nine pups received monthly infusions and all appeared to tolerate the procedure well. They showed no adverse reactions during or after the protein infusion, maintaining normal body temperature and heart rate. One dog was euthanized following an adverse anesthetic event on 2nd treatment (AVT 2) and CVT 1 showed mild neck pain for 24 h following the 1st infusion which required multiple cisterna magna punctures. The clinical signs were low at the end of the study for all groups (Table 1). AET dogs exhibited reduced anxiety in novel situations and increased willingness to accept restraint compared to AVT dogs (Table 1), however, these signs were not quantiﬁed further. A signiﬁcant difference developed between the rate of weight gain in AET compared to AVT by the 3rd treatment and at post-mortem (Fig. 1A) (P b 0.05). Antibody titers in CSF and plasma In the CSF samples no antibody titer was observed between the three groups (0.011 U in CVT, 0.011 U in AET and 0.010 U in AVT) and titers were comparable to the negative control (0.011 U). In this study of intracisternal ERT, the mean plasma titer increased up to 0.03 U during the 7th week following the second enzyme infusion (Fig. 1b). This was primarily due to the antibody levels seen in a single dog, who recorded OD readings of 0.071 U and decreased to 0.028 U during the 8th week of infusion. The antibody titer decreased and remained at negligible levels after this despite continued infusions (weeks 9 and post-mortem). CSF microprotein and glucose CSF microprotein levels were stable and within the reference laboratory range (0.15–0.35 g/L) (Kaneko, 1989) throughout treatment in all groups. CSF glucose levels were in the normal range (4.1– 4.16 mmol/L) (Kaneko, 1989) during treatment. The glucose levels in AET and AVT were not signiﬁcantly different (p b 0.05). In both groups of affected dogs the glucose levels were signiﬁcantly lower (3.9 mmol/L) (p b 0.05) than controls (5.78 mmol/L). Hematology Vacuolated lymphocytes were observed in affected animals from birth. Signiﬁcant differences in the percentage of lymphocytes which were vacuolated were observed among the three groups (p b 0.001) with 13–22% in CVT and 52–68% in the affected dogs, but were not altered by enzyme treatment. Effect of treatment on canine α-L-fucosidase levels in CSF, plasma and leukocytes Signiﬁcant enzyme activity was found in CSF 10 min after the third infusion (fresh sample from a fresh cisterna magna puncture) at 998 times higher than the CVT mean (Table 2). In the samples collected 48 h after the third infusion the enzyme activity was 2.5 times above CVT (p b 0.05). Infused enzyme was not detected in AET plasma and leukocytes (data not shown). Effect of treatment on canine α-L-fucosidase levels in CNS and visceral tissues Following intracisternal injection of canine α-L-fucosidase, highly signiﬁcant differences in enzyme activity were observed among all groups in CNS, liver, lymph node and bone marrow (p b 0.001) and in corpus callosum (p = 0.002), thalamus (p = 0.008) (Fig. 2b), surface brain stem (p = 0.004) and deep brain stem (p = 0.002). Carrier ESS

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226 Table 2 CSF enzyme activity (units of enzyme activity/mL of CSF) in CVT, AET and AVT dogs (mean ± SEM). Letters indicate statistically signiﬁcant differences between groups (p b 0.05). NS = sampled. Infusion Immediately prior to 1st Immediately prior to 2nd Immediately prior to 3rd 10 min after third infusion At post-mortem (48 h post-third injection)

CVT (n = 3)

AET (n = 3) a

0.203 ± 0.06 0.200 ± 0.06 a 0.190 ± 0.06 a NS 0.202 ± 0.08a

0.001 ± 0.00 0.000 ± 0.00 0.001 ± 0.00 223.00 0.543 ± 0.52

AVT (n = 3) b b b

b

Table 3 Enzyme activity in peripheral organs (mean ± SEM). Data was expressed as Units of enzyme activity/mg of protein. Letters indicate statistically signiﬁcant differences between groups (p b 0.05). Tissue

b

0.004 ± 0.00 0.002 ± 0.00 b 0.002 ± 0.00 b NS 0.003 ± 0.00c

group had a mean enzyme activity of approximately 50% of CVT in all tissue. In treated dogs, higher enzyme activity was present in surface brain tissue relative to deep brain (corpus callosum, thalamus and mid brain). Deep brain is deﬁned as being N15 mm from the surface exposed to CSF. Enzyme levels in treated tissues ranged from 2 to 72% of normal control activity, with higher enzyme activity observed in regions closer to the injection point, particularly in cerebellum, brainstem and in the distant regions of spinal cord and meninges. In contrast, in deep regions of brain such as occipital cortex, internal capsule, striatum, thalamus as well as in peripheral nerve tissue (dorsal root ganglion, vagus and sciatic nerves) the enzyme activity was b5% of control. Enzyme activity in the superﬁcial frontal cortex and medulla, was 5–10% of normal control. Superﬁcial regions of occipital cortex and corpus callosum, had more than 10% of control enzyme activity (Figs. 2a and b). The enzyme activity in superﬁcial frontal cortex and deep cortex had a mean which was of 8.5% and 5.5% of control respectively. In the superﬁcial occipital cortex higher enzyme activity was observed (20% of control) compared to 2.6% in deep tissue (Fig. 2a). Intracisternal enzyme dispersed through large distances reaching the internal capsule, striatum, thalamus, midbrain and medulla (mean enzyme activity varied from 2 to 24% of control). A signiﬁcant amount of enzyme (24.6% control) reached the midbrain (p b 0.05) and dorsal and lateral cerebellum (35.8% and 24.7% control) and pons (2b and c). Signiﬁcant differences in enzyme activity were observed in all regions of cerebral cortex and cerebellum in AET compared to AVT (p b 0.05). Enzyme uptake was 5.7% of control in deep medulla of AET (Fig. 2b) (p b 0.05). The superﬁcial and deep regions of brainstem showed enzyme activity of 39.4% and 12.3% of control (Fig. 2c). The meninges had more than 10% of mean enzyme activity (Fig. 2e) and spinal meninges (thoracic and lumbar) reached control levels. Cerebral meninges had lower (9.6% of control) enzyme activity compared to spinal meninges (p b 0.05). Higher enzyme activity was observed in the lumbar meninges (39.9%) and this coincided with high enzyme activity in the lumbar cord, which exceeded the levels in thoracic (35.6%) and cervical (32.9% of control). Spinal cord enzyme activity was the highest of all CNS tissues (11–73% of control) reﬂecting its close contact with the infused enzyme (Fig. 2d). The enzyme activity in superﬁcial areas of spinal cord cervical, thoracic and lumbar levels and deep lumbar spinal cord approached or exceeded 50% of CVT. In deep regions of cervical and thoracic cord the enzyme activity (mean of 11.3%, 22.4% of control respectively) in the AET group was signiﬁcantly different from AVT and CVT (Fig. 2d) at p b 0.05. Infused enzyme reached some peripheral tissues with a signiﬁcant increase (2.45% of control) in dorsal root ganglion compared to AVT (p b 0.05). In sciatic and vagus nerves a trend towards increased enzyme activity (1.4% and 1.1% of control) was observed (but was not signiﬁcant) (Table 3). Visceral organs with signiﬁcant increases were those with reticuloendothelial clearance, immune surveillance and phagocytic functions, particularly liver, cervical lymph node and bone marrow (Table 3). Effect of enzyme replacement level on substrate accumulation The substrate quantiﬁed in fucosidosis brain was fucosyl-N-acetyl glucosaminyl aspartate (a fucose-derived oligosaccharide). For quality

223

Pituitary gland Dorsal root ganglion Sciatic nerve Vagus nerve Liver Spleen Lymph node Bone marrow Thyroid gland Salivary gland

CVT (n = 3)

AET (n = 3) (% CVT) a

39.18 ± 4.69 89.82 ± 10.9 a 49.22 ± 6.8 a 44.08 ± 4.0 a 157.00 ± 10.3 a 263.74 ± 10.2 a 94.04 ± 10.0 a 432.30 ± 6.8 a 518.01 ± 5.6 a 651.76 ± 8.3 a

14.27 ± 10.1 2.20 ± 1.3 b 0.69 ± 0.3 b 0.49 ± 0.6 b 1.98 ± 1.0 b 0.21 ± 0.3 b 2.67 ± 3.2 b 0.10 ± 0.0 b 0.71 ± 1.2 b 0.55 ± 0.8 b

b

(36.42) (2.45) (1.39) (1.12) (1.26) (0.07) (2.84) (0.02) (0.14) (0.08)

AVT (n = 3) 0.08 ± 0.0 0.05 ± 0.0 0.32 ± 0.4 0.06 ± 0.1 0.04 ± 0.0 0.00 ± 0.0 0.03 ± 0.0 0.00 ± 0.0 0.02 ± 0.0 0.02 ± 0.0

c c b b c b c c b b

control, one brain sample was included in each assay, giving an intraassay CV of 5% (n = 10) and an inter-assay CV of 10% (n = 6). Intracisternal enzyme infusion decreased the oligosaccharide accumulation in CSF in two of the three treated dogs. The CVT dogs had near zero levels of substrate (0.0–3.7 peak area ratio/mL of CSF). In carrier ESS group the substrate levels were not signiﬁcantly different from CVT in all tissues (data not shown). The greatest substrate accumulation was observed in AVT and was in the range of 39–63 peak area ratio/mL of CSF which was signiﬁcantly different to control (P b 0.05). This did not change in the 3 month course of treatment. No differences were observed between AET and AVT in the samples collected immediately before the 1st and 2nd infusions. At the 3rd infusion, a 42% reduction was observed in AET compared to AVT (P b 0.05) (Fig. 3). Substrate accumulation was present at birth in the cerebral cortex (frontal and occipital), corpus callosum, thalamus, cerebellum, medulla and spinal cord of a one day-old affected dog (0.03–1.34 peak area ratio/mg of protein; data not shown). The substrate levels in AVT were 2.9–5.2 peak area ratio/mg of protein in brain, 2.1–4.6 peak area ratio/mg of protein in spinal cord and 3.9–5.4 peak area ratio/mg of protein in meninges compared to CVT (0.00–0.05 peak area ratio/ mg of protein). Intracisternal delivery of enzyme led to statistically signiﬁcant reductions in the accumulation of substrate in affected enzymetreated animals (Figs. 2f–j). In AVT tissues, substrates were in the

Fig. 3. Substrate in CSF during enzyme replacement therapy. Data was expressed as peak area ratio/mL of CSF. PM = post-mortem, PBS = phosphate buffer saline, CVT = control vehicle treated, AET = affected enzyme treated and AVT = affected vehicle treated.

224

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

Table 4 Oligosaccharide accumulation in organs (mean ± SEM). Data was expressed as peak area ratio/mg of protein. Letters indicate statistically signiﬁcant differences between groups (p b 0.05). Tissue Dorsal root ganglion Sciatic nerve Lymph node Liver

CVT (n = 3) 0.00 ± 0.6 0.02 ± 0.0 0.02 ± 0.0 0.01 ± 0.0

a a a a

AET (n = 3) 4.9 ± 0.0 2.6 ± 0.7 4.0 ± 1.8 2.4 ± 0.3

a b ab b

AVT (n = 3) 4.0 ± 2.5 3.3 ± 1.4 5.3 ± 2.6 4.0 ± 0.5

a b b c

range of 2.1–5.6 peak area ratio/mg of protein and the AET tissues had lower substrates in all the regions of CNS except corpus callosum and thalamus. Intracisternal infusion of enzyme led to a statistically signiﬁcant 32.2% reduction in substrate in the superﬁcial frontal cortex, compared to AVT (p b 0.05). Decreased substrate levels following ERT were observed in other deep forebrain regions, including the striatum (Fig. 2g). Mid-brain enzyme activity was substantial (24.6% control) and was linked with a signiﬁcant substrate reduction of 22.5% (p b 0.05). Areas closer to the injection point, including brainstem (surface and deep) and cerebellum (dorsal and lateral), had substantially less (23–45%) oligosaccharide accumulation following treatment (Fig. 2h). Signiﬁcant reductions in substrates were observed in dorsal and lateral cerebellum (44.7%, 45.1%) (p b 0.05) (Fig. 2h) and reﬂected the increased enzyme activity (35.8 and 24.7% control). More substantial enzyme correction and substrate reduction were observed in spinal cord and meninges. Lumbar spinal cord had the greatest enzyme replacement and a ﬁve-fold decrease in superﬁcial and deep substrates with the levels which were not signiﬁcantly different from normal tissue (Figs. 2d and i). There were substantial reductions in substrate in the superﬁcial thoracic and cervical (67.2%, 64.5%) and deep cervical and thoracic spinal cord (21.6%, 43.9%) (Fig. 2i). These changes were signiﬁcantly different from AVT (P b 0.05) and associated with higher levels of enzyme activity (11–72% of control). Cervical spinal meninges had the greatest substrate reduction (77.3%) followed by thoracic and lumbar spinal meninges (63% and 58%). In cerebral meninges 35.6% substrate reduction, associated with 9.6% of control enzyme activity. In sciatic nerve a trend towards reduction of substrate accumulation (18.6%) was associated with low levels of enzyme activity (1.4% of control) but was not signiﬁcant. In dorsal root ganglion despite replacing 2.4% of control enzyme levels in AET, there was no reduction in oligosaccharide accumulation (Table 4). A signiﬁcant reduction in substrate was observed in AET liver (40.5%; Table 4; p b 0.05), and was associated with low but signiﬁcant levels of enzyme replacement. A 24.2% reduction in substrate accumulation was noted in lymph node (Table 4). Discussion This study provides ﬁrst evidence of the effects of very early intracisternal enzyme administration in fucosidosis dogs. This is the only established animal model of this disease, and the ﬁrst case of glycoproteinosis disorder to be treated with intrathecal ERT. The intracisternal approach was effective in delivery of therapeutic levels of enzyme to CNS and in reducing lysosomal storage. The infused enzyme penetrated a substantial distance from the site of injection, reaching areas such as the corpus callosum and thalamus, however it appears that the levels were not therapeutic as there was no reduction in substrate. An increase above 2% of normal enzyme activity in tissue resulted in diminished storage in most regions of the fucosidosisaffected dog brain, however the biological signiﬁcance of this level of substrate reduction, i.e. its effects on neuropathology and neural dysfunction, have yet to be described. We report limited enzyme replacement in peripheral nervous system (dorsal root ganglion, sciatic and vagus nerve), at levels approaching those previously

reported to ameliorate storage after BMT (Taylor et al., 1992), and in visceral organs including lymph node and liver following intracisternal delivery. A mean enzyme activity of 20, 21 and 58 (% of control) resulted in gradual amelioration of central nervous system pathology in a 6–18 month period following BMT in fucosidosis dogs (Taylor et al., 1992). Substrate analysis conﬁrmed that a biochemical response to enzyme replacement occurred across all regions of CNS (except thalamus and corpus callosum) but was insufﬁcient to completely normalize storage in the disease in regions other than the spinal cord and meninges. Treatment with recombinant canine α-L-fucosidase or vehicle was safe, did not produce signs of inﬂammation (determined by low cell and protein levels following repeated treatment) or did not induce sustained antibody production in CSF or in circulation. This study differs to others in dogs in its use of same species recombinant enzyme, a strategy intended to reduce immune responses. Antibodies which develop following repeated recombinant enzyme treatment may impede the effects of ERT in the periphery and neural tissue. Antibody formation following the infusion of recombinant human arylsulfatase in knock out metachromatic leukodystrophy mice altered the neuronal uptake of enzyme by blocking the mannose-6phosphate receptors and reducing the stability of enzymes (Matzner et al., 2008). In MPS I and MPS IIIIA dogs, intracisternal infusion of human recombinant enzymes to adult dogs led to the development of circulating antibodies and a cellular immune response (Kakkis et al., 2004; Hemsley et al., 2009). Seizures or nervous system irritability occurred in MPS I dogs receiving undiluted enzyme while those dogs receiving enzyme in Elliott's B solution (artiﬁcial CSF) had fewer reactions (Dickson et al., 2007). While PBS was a safe vehicle in fucosidosis dogs, with no adverse reactions on repeated solution, Elliotts B solution may be preferable (Dickson et al., 2007). The dose of 6.9 U fucosidase produced high enzyme levels in CSF immediately after injection, which diminished by 48 h later and penetrated all areas of brain, demonstrating that a high CSF to tissue concentration gradient was achieved for a short period after delivery. The enzyme activity varied across the CNS in a pattern which reﬂected proximity to the site of infusion and patterns of CSF ﬂow for dogs in lateral recumbency, with considerable enzyme activity reaching caudal spinal cord. Positional change after infusion was not attempted, however in MPS IIIA dogs (prone, supine and lateral after enzyme infusion for 20 min period) it contributed to a 30-fold difference with increased enzyme distribution and uptake in cerebral hemispheres (Marshall et al., 2010). The extent of retrograde ﬂow of enzyme into CSF in the lateral and third ventricles is unknown and may contribute to the enzyme penetration observed here into some regions that were greater than 2 cm from brain surface such as deep cortex. In corpus callosum (close to lateral ventricles) and thalamus (near to third ventricle) tissue enzyme activity may have been produced by attachment and absorption of enzyme from the ventricles, rather than deep tissue penetration as it was not accompanied by any change in the substrate accumulation. The lumbar spinal cord (surface) showed the highest enzyme replacement in CNS (72.9% control) and greatest substrate reduction (80.7%) with more substantial correction achieved than in other intracisternal ERT dog studies. This reﬂects the ﬂow of CSF caudally towards the cauda equina. Similar ﬁndings of substrate reduction were observed in lumbar spinal cord in MPS IIIA mice with intracisternal ERT (Hemsley et al., 2008). The small enzyme activity in dorsal root ganglion and sciatic nerve (2.4% and 1.4%) was not associated with reduction in substrates but is an intriguing ﬁnding which may reﬂect transport through neuronal processes. Intracisternal enzyme reached liver, cervical lymph node at low levels most likely by diffusion into lymphatic drainage which enters systemic circulation. In cervical lymph node, low levels of enzyme activity (mean 2.8%) were linked to decreased oligosaccharide accumulation, which might reﬂect the impact of sustained enzyme leakage from CNS.

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

We focused on the potential of early treatments to inﬂuence storage during maturation of the brain and before the onset of clinical signs, as previous bone marrow transplantation studies demonstrated the effectiveness of very early enzyme replacement, with optimal results achieved before 6–8 months age. Other intracisternal enzyme replacement studies have been largely conducted in adult MPS I and MPS IIIA dogs with clinical signs (Kakkis et al., 2004; Dickson et al., 2007, 2010; Hemsley et al., 2009). MPS I dogs receiving monthly or quarterly intracisternal enzyme along with weekly IV therapy had reductions in substrates to normal levels in brain, irrespective of state of the disease (Dickson et al., 2007; Kakkis et al., 2004). In fucosidosis-affected dogs, monthly enzyme replacement reduced the oligosaccharide storage in brain, however the enzyme activity achieved was insufﬁcient to normalize substrate level in most regions and did not match the sustained enzyme replacement achieved following repopulation of the brain by donor marrow derived tissue macrophages (Taylor et al., 1992). Heparan sulfate-derived oligosaccharide levels were reduced in the spinal cord of MPS IIIA mice receiving monthly treatment (Hemsley et al., 2007) and a similar outcome was achieved here with the greatest impact observed in spinal cord and meninges. The corrective impact on substrates in the dorsal root ganglia and peripheral nerves are novel ﬁndings, not reported in the MPS canine models, and suggest leakage, or transneuronal transport of enzyme from the CNS to the peripheral nervous system. Canine fucosidosis is the only large animal model of intrathecal enzyme replacement in which the brain, spinal cord and peripheral nerves show extensive neuropathological ﬁndings of demyelination, in addition to storage, neuron loss and inﬂammatory change (Kondagari et al., 2011a, 2011b; Taylor et al., 1988), similar to human disease, enabling correlation of clinical, pathological and biochemical changes in response to therapy. This study demonstrates the safety of repeated intracisternal injections in developing brain for treatment of the neurodegenerative lysosomal diseases. This dose and delivery route was safe and improved the weight gain in treated fucosidosis animals, which were close in size (6–10 kg) and stage of maturity (infantile) to the point at which fucosidosis children are likely to receive treatment after postnatal diagnosis. This may have had a central effect on appetite and feeding levels however this was not determined in this study. Increased frequency of treatment, higher doses of enzyme, combined therapy (with IV) along with increased group size may improve effectiveness in canine fucosidosis. Higher enzyme doses are expected to produce larger beneﬁts which may approach the biochemical normalization reported in the MPS I dog and MPS IIIA mouse intracisternal studies. While ERT was not completely curative at this dose and treatment frequency, the potential for treating fucosidosis brain lesions with direct enzyme delivery has been established. Early treatment reduces substrate storage with the potential to minimize the impact of these biochemical derangements on development and brain maturation in fucosidosis. The effects could be further explored with behavioral, pathological, imaging and neurophysiological studies. These canine studies more closely replicate the methods likely to be used in infants with severe neurological storage and allow a clinically-relevant investigation of the safety issues. The clear, consistent impact of enzyme delivery on substrate accumulation in all regions of CNS in fucosidosis provides impetus for further investigation of the effects on the pathology. Acknowledgments We are grateful to Vivarium staff at Westmead Hospital for caring for the dogs. We thank the Faculty Bequest Research Fund (Margot Roslyn Flood, Stewart Legacy and Schnakenberg Bequest) and the Faculty of Veterinary Science, University of Sydney for facilities to undertake this study.

225

Appendix A. Supplementary data Supplementary data to this article can be found online at doi:10.1016/j.expneurol.2011.04.019.

References Abraham, D., Blackmore, W.F., Dell, A., Herrtage, M.E., Jones, J., Littlewood, J.D., Oates, J., Palmer, A.C., Sidebotham, R., Winchester, B., 1984. Biochemical evidence for a case of canine fucosidosis. Biochem. Soc. Trans. 12, 2031–2032. Auclair, D., Finnie, J., White, J., Nielsen, T., Fuller, M., Kakkis, E., Cheng, A., O'Neill, C.A., Hopwood, J.J., 2010. Repeated intrathecal injections of recombinant human 4sulphatase remove dural storage in mature mucopolysaccharidosis VI cats primed with a short-course tolerisation regimen. Mol. Genet. Metab. 99, 132–141. Bielicki, J., Muller, V., Fuller, M., Hopwood, J.J., Anson, D.S., 2000. Recombinant canine alpha-l-fucosidase: expression, puriﬁcation, and characterization. Mol. Genet. Metab. 69, 24–32. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Chang, M., Cooper, J.D., Sleat, D.E., Cheng, S.H., Dodge, J.C., Passini, M.A., Lobel, P., Davidson, B.L., 2008. Intraventricular enzyme replacement improves disease phenotypes in a mouse model of late infantile neuronal ceroid lipofuscinosis. Mol. Ther. 16, 649–656. Dickson, P., McEntee, M., Vogler, C., Le, S., Levy, B., Peinovich, M., Hanson, S., Passage, M., Kakkis, E., Dickson, P., McEntee, M., Vogler, C., Le, S., Levy, B., Peinovich, M., Hanson, S., Passage, M., Kakkis, E., 2007. Intrathecal enzyme replacement therapy: successful treatment of brain disease via the cerebrospinal ﬂuid. Mol. Genet. Metab. 91, 61–68. Dickson, P.I., Hanson, S., McEntee, M.F., Vite, C.H., Vogler, C.A., Mlikotic, A., Chen, A.H., Ponder, K.P., Haskins, M.E., Tippin, B.L., Le, S.Q., Passage, M.B., Guerra, C., Dierenfeld, A., Jens, J., Snella, E., Kan, S., Ellinwood, N.M., 2010. Early versus late treatment of spinal cord compression with long-term intrathecal enzyme replacement therapy in canine mucopolysaccharidosis type I. Mol. Genet. Metab 101, 115–122. Dodge, J.C., Clarke, J., Treleaven, C.M., Taksir, T.V., Grifﬁths, D.A., Yang, W.D., Fidler, J.A., Passini, M.A., Karey, K.P., Schuchman, E.H., Cheng, S.H., Shihabuddin, L.S., 2009. Intracerebroventricular infusion of acid sphingomyelinase corrects CNS manifestations in a mouse model of Niemann–Pick A disease. Exp. Neurol. 215, 349–357. Ferrara, M.L., Occhiodoro, T., Fuller, M., Hawthorne, W.J., Teutsch, S., Tucker, V.E., Hopwood, J.J., Stewart, G.J., Anson, D.S., 1997. Canine fucosidosis: a model for retroviral gene transfer into haematopoietic stem cells. Neuromuscul. Disord. 7, 361–366. Healy, P.J., Farrow, B.R., Nicholas, F.W., Hedberg, K., Ratcliffe, R., 1984. Canine fucosidosis: a biochemical and genetic investigation. Res. Vet. Sci. 36, 354–359. Hemsley, K.M., King, B., Hopwood, J.J., 2007. Injection of recombinant human sulfamidase into the CSF via the cerebellomedullary cistern in MPS IIIA mice. Mol. Genet. Metab. 90, 313–328. Hemsley, K.M., Beard, H., King, B.M., Hopwood, J.J., 2008. Effect of high dose, repeated intra-cerebrospinal ﬂuid injection of sulphamidase on neuropathology in mucopolysaccharidosis type IIIA mice. Genes Brain Behav. 7, 740–753. Hemsley, K.M., Norman, E.J., Crawley, A.C., Auclair, D., King, B., Fuller, M., Lang, D.L., Dean, C.J., Jolly, R.D., Hopwood, J.J., 2009. Effect of cisternal sulfamidase delivery in MPS IIIA Huntaway dogs-A proof of principle study. Mol. Genet. Metab. 98, 383–392. Kakkis, E., McEntee, M., Vogler, C., Le, S., Levy, B., Belichenko, P., Mobley, W., Dickson, P., Hanson, S., Passage, M., 2004. Intrathecal enzyme replacement therapy reduces lysosomal storage in the brain and meninges of the canine model of MPS I. Mol. Genet. Metab. 83, 163–174. Kaneko, J., 1989. Clinical Biochemistry of Domestic Animals, 4th ed. Department of Clinical Pathology, School Of Veterinary Medicine, University Of California, Davis, California. Kondagari, G.S., Yang, J., Taylor, R.M., 2011a. Investigation of cerebrocortical and cerebellar pathology in canine fucosidosis and comparison to aged brain. Neurobiol. Dis. 41, 15–23. Kondagari, G.S., Ramanathan, P., Taylor, R., 2011b. Canine fucosidosis: a neuroprogressive disorder. Neurodegener. Dis. 8, 240–251. Krivit, W., Peters, C., Shapiro, E.G., 1999. Bone marrow transplantation as effective treatment of central nervous system disease in globoid cell leukodystrophy, metachromatic leukodystrophy, adrenoleukodystrophy, mannosidosis, fucosidosis, aspartylglucosaminuria, Hurler, Maroteaux-Lamy, and Sly syndromes, and Gaucher disease type III. Curr. Opin. Neurol. 12, 167–176. Lee, W.C., Tsoi, Y.K., Troendle, F.J., DeLucia, M.W., Ahmed, Z., Dicky, C.A., Dickson, D.W., Eckman, C.B., 2007. Single-dose intracerebroventricular administration of galactocerebrosidase improves survival in a mouse model of globoid cell leukodystrophy. FASEB J. 21, 2520–2527. Marshall, N.R., Jolly, R.D., Hartman, A., Hemsley, K.M., Leanne, W.K., Hannoyche, H., 2010. Intracisternal enzyme replacement therapy: the effect of post therapy posture. Internation Symposium on Mucopolysaccharidosis and Related Diseases 2010: Proceedings of the 11th meeting of the mucopolysaccharidosis society. 2010101. Matzner, U., Matthes, F., Weigelt, C., Andersson, C., Eistrup, C., Fogh, J., Gieselmann, V., Matzner, U., Matthes, F., Weigelt, C., Andersson, C., Eistrup, C., Fogh, J., Gieselmann, V., 2008. Non-inhibitory antibodies impede lysosomal storage reduction during

226

G.S. Kondagari et al. / Experimental Neurology 230 (2011) 218–226

enzyme replacement therapy of a lysosomal storage disease. J. Mol. Med. 86, 433–442. Munoz-Rojas, M.V., Vieira, T., Costa, R., Fagondes, S., John, A., Jardim, L.B., Vedolin, L.M., Raymundo, M., Dickson, P.I., Kakkis, E., Giugliani, R., 2008. Intrathecal enzyme replacement therapy in a patient with mucopolysaccharidosis type I and symptomatic spinal cord compression. Am. J. Med. Genet. 146, 2538–2544. Munoz-Rojas, M.V., Horovitz, D.D., Jardim, L.B., Raymundo, M., Llerena Jr., J.C., de Magalhaes Tde, S., Vieira, T.A., Costa, R., Kakkis, E., Giugliani, R., 2010. Intrathecal administration of recombinant human N-acetylgalactosamine 4-sulfatase to a MPS VI patient with pachymeningitis cervicalis. Mol. Genet. Metab. 99, 346–350. Occhiodoro, T., Hopwood, J.J., Morris, C.P., Anson, D.S., 1992. Correction of alpha-Lfucosidase deﬁciency in fucosidosis ﬁbroblasts by retroviral vector-mediated gene transfer. Hum. Gene Ther. 3, 365–369. Pomroy, S., 2002. Canine fucosidosis experimental treatment by enzyme replacement therapy. Thesis submitted to University of sydney.

Sivakumur, P., Wraith, J.E., 1999. Bone marrow transplantation in mucopolysaccharidosis type IIIA: a comparison of an early treated patient with his untreated sibling. J. Inherit. Metab. Dis. 22, 849–850. Taylor, R.M., Farrow, B.R.H., Healy, P.J., 1987. Canine fucosidosis—clinical ﬁndings. J. Small Anim. Pract. 28, 291–300. Taylor, R.M., Farrow, B.R., Stewart, G.J., Healy, P.J., Tiver, K., 1988. The clinical effects of lysosomal enzyme replacement by bone marrow transplantation after total lymphoid irradiation on neurologic disease in fucosidase deﬁcient dogs. Transplant. Proc. 20, 89–93. Taylor, R.M., Farrow, B.R., Stewart, G.J., 1992. Amelioration of clinical disease following bone marrow transplantation in fucosidase-deﬁcient dogs. Am. J. Med. Genet. 42, 628–632. Vellodi, A., Cragg, H., Winchester, B., Young, E., Young, J., Downie, C.J., Hoare, R.D., Stocks, R., Banerjee, G.K., 1995. Allogeneic bone marrow transplantation for fucosidosis. Bone Marrow Transplant. 15, 153–158.

Lihat lebih banyak...

Treatment of canine fucosidosis by intracisternal enzyme infusion

Descripción

Comentarios