Beta-Mannosidosis in a Domestic Cat Associated with a Missense Variant in MANBA

Abstract

A 6-month-old cat of unknown ancestry presented for a neurologic evaluation due to progressive motor impairment. Complete physical and neurologic examinations suggested the disorder was likely to be hereditary, although the signs were not consistent with any previously described inherited disorders in cats. Due to the progression of disease signs including severely impaired motor function and cognitive decline, the cat was euthanized at approximately 10.5 months of age. Whole genome sequence analysis identified a homozygous missense variant c.2506G>A in MANBA that predicts a p.Gly836Arg alteration in the encoded lysosomal enzyme β-mannosidase. This variant was not present in the whole genome or whole exome sequences of any of the 424 cats represented in the 99 Lives Cat Genome dataset. β-Mannosidase enzyme activity was undetectable in brain tissue homogenates from the affected cat, whereas α-mannosidase enzyme activities were elevated compared to an unaffected cat. Postmortem examination of brain and retinal tissues revealed massive accumulations of vacuolar inclusions in most cells, similar to those reported in animals of other species with hereditary β-mannosidosis. Based on these findings, the cat likely suffered from β-mannosidosis due to the abolition of β-mannosidase activity associated with the p.Gly836Arg amino acid substitution. p.Gly836 is located in the C-terminal region of the protein and was not previously known to be involved in modulating enzyme activity. In addition to the vacuolar inclusions, some cells in the brain of the affected cat contained inclusions that exhibited lipofuscin-like autofluorescence. Electron microscopic examinations suggested these inclusions formed via an autophagy-like process.

1. Introduction

Precision genomic analysis is becoming a standard diagnostic tool in veterinary medicine (Buckley and Lyons, 2020; Mauler et al., 2017), and has quickly contributed to the expanded identification of naturally occurring large animal models of human hereditary disorders (Awano et al., 2006; Gurda et al., 2017, 2016; Gurda and Vite, 2019; Katz et al., 2014; Whiting et al., 2016, 2014). Commercial laboratories now offer DNA testing for a high number of disease-causing or disease-associated variants in cats and dogs at very low cost. Many of the DNA sequence variants responsible for these disorders in dogs were discovered by performing whole genome sequence (WGS) analyses on affected dogs (Bullock et al., 2022; Gilliam et al., 2015; Guo et al., 2019, 2014b, 2014a; Kolicheski et al., 2017a, 2017b, 2016; Schmutz et al., 2019). Both whole genome and whole exome sequencing (WES) are now also available for domestic cats utilizing the 99 Lives Cat Genome sequencing dataset that incorporates a vast majority of the high quality, Illumina-based short-read genomic sequence data for domestic cats, including 362 WGS and 62 WES datasets (Buckley et al., 2020; Gandolfi et al., 2015; Katz et al., 2020; Rodney et al., 2021; Shelton et al., 2023). The genetic bases of a number of disorders in domestic cats have been identified using the available WGS and WES resources (Christen et al., 2023; Cogne et al., 2020; Hilton et al., 2023; Jenni et al., 2023; Katz et al., 2020; Kiener et al., 2023, 2022; Woelfel et al., 2022). The Online Mendelian Inheritance in Animals database has documented over 90 variants in cats since 2015, mainly discovered by WGS or WES, since one of the first cat WGS studies discovered a myasthenic syndrome caused by a variant in collagen like tail subunit of asymmetric acetylcholinesterase (COLQ) (Gandolfi et al., 2015).

WGS analysis tools and resources were utilized to investigate a potential genetic basis for a disorder in a domestic cat exhibiting progressive neurological signs. The cat was euthanized due to disease progression and tissues were obtained for morphological and biochemical analyses. Correlation of the WGS, phenotypic, and biochemical data was performed to identify the genetic basis of the cat’s disorder.

2. Materials and methods

2.1. Subject cat: clinical evaluations

A 6-month-old spayed female cat of unknown ancestry presented for a neurologic examination due to progressive motor impairment. Complete physical and neurologic examinations, including cerebrospinal fluid analysis, as well as the health history of the cat, suggested that the disorder was likely to be hereditary. However, the signs were not consistent with any previously described inherited disorders in cats. When the cat was 10 months of age, the owners completed a questionnaire in which they were asked to rate the cat on a list of 29 potential disease signs. Due to the progression of disease signs, the cat was humanely euthanized at approximately 10.5 months of age.

2.2. Tissue morphological evaluation

Following euthanasia of the proband, the eyes were enucleated and the central areas of the corneas were removed. One eye was then placed in 2.5% glutaraldehyde fixative (2.5% glutaraldehyde, 0.1 M sodium cacodylate, pH 7.4) and the other eye was placed in “Immuno” fixative (3.5% formaldehyde, 0.05% glutaraldehyde, 0.12 M sodium cacodylate, 1 mM CaCl₂, pH 7.4). The brain was removed from the skull and slices of the parietal lobe cerebral cortex and cerebellar cortex were placed in Immuno fixative. A second slice of each tissue was placed in EM fixative (2.0 % glutaraldehyde, 1.12% formaldehyde, 0.13 M sodium cacodylate, 1 mM CaCl₂, pH 7.4). The remainder of the brain was divided in half along the rostral-caudal midline, and one half was placed in 10% buffered formalin (Fisher Scientific, Cat. No SF93-4). The other half was frozen and stored at −80°C. Slices of the heart ventricular wall were also obtained and placed in the Immuno and EM fixatives. All fixed tissues were incubated at room temperature until being further processed for microscopic evaluations.

The fixed tissue samples were prepared for light and electron microscopic evaluations and for immunohistochemistry as described previously (Bullock et al., 2022). Slices of the Immuno-fixed tissues were embedded in OCT medium (Tissue-Tek, Sakura Finetek, Torrence, CA) and cryo-sectioned at a thickness of 8 μm. The unstained sections were examined for autofluorescence as described previously (Bullock et al., 2022). Images were obtained using a Zeiss Axiophot microscope equipped with a Prior Lumen 200 light source, Zeiss filter set 487705, which consisted of a 400–440 nm bandpass excitation filter, an FT 460 dichromatic beam splitter, and an LP 470 barrier filter to block emission below 470 nm from reaching the measuring device. In addition, a 515-nm barrier filter was also placed in the emission light path. Fluorescence images were acquired using a Zeiss Neofluor 40x objective with a numerical aperture of 0.75 and an Olympus DP72 digital camera. Slices of both Immuno-fixed and EM-fixed tissues were paraffin embedded and sections were stained with hematoxylin and eosin. Slices of the EM-fixed cerebral cortex and cerebellar cortex were also processed for electron microscopy. This processing included secondary fixation in osmium tetroxide and embedding in epoxy resin (Katz et al., 1982). Sections of the resin-embedded tissues were cut at a thickness of 0.2 μm, mounted on glass slides, and stained with toluidine blue. The tissue blocks were then trimmed, and sections were cut at thicknesses of 70 to 80 nm, mounted on copper grids, and stained with lead citrate and uranyl acetate. The sections mounted on glass slides were examined with a Zeiss Axiophot microscope, and the sections mounted on grids were examined with a JEOL 1400 transmission electron microscope.

2.3. Molecular genetic analyses

After obtaining informed consent from the owner, approximately 3 ml of EDTA whole blood was collected by medial saphenous venipuncture. DNA was isolated using organic extraction with phenol:chloroform and evaluated for quality and quantity by agarose gel (1.5% in TAE) electrophoresis (80V for 90 min) and visualization with ultra-violet light after ethidium bromide staining (Buckley et al., 2020; Sambrook and Russell, 2001). Approximately 1 μg of high molecular weight DNA was submitted to the University of Missouri Genomics Technology Core to construct a 550 bp sequencing library using a Tru-Seq DNA PCR-free kit (Illumina, San Diego, CA). The barcoded case sample library was pooled with approximately 21 additional cat genomic sequencing libraries and loaded onto a single S4 flow cell on a NovaSeq 6000 (Illumina, San Diego, CA) per manufacturer’s recommendations to produce ~ 30x sequencing coverage of 150 bp paired-end reads. Data was processed using a custom Nextflow workflow following best-practices for the Genome Analysis Toolkit (GATK) version 4.2 (Di Tommaso et al., 2017; VanderAuwera G and O’Connor B, 2020). Reads were mapped to Felis_catus_9.0 (GCF_000181335.3) using Minimap version 2 (Li, 2018). Duplicate reads were marked using Picard version 2.27 Specific tools from GATK 4.2 for genotyping, variant database construction, and hard-filtering were completed as previously described (Broad Institute, 2022). The produced 22 cat dataset was combined with the previously produced 340 cat variant call file (Buckley et al., 2020; Kopke et al., 2022; Lyons et al., 2021;) and the 61 cat WES dataset for the 99 Lives Cat Genome Consortium(Katz et al., 2020; Rodney et al., 2021). The NCBI RefSeq Felis catus annotation 104 (10 December 2019) and GATK Variant Effect Predictor (McLaren et al., 2016) were used to characterize the variants. Exonic variants and 10 bp flanking each exon were filtered and visualized using VarSeq software (GoldenHelix, Bozeman, MT). The WES and WGS data are available in the NCBI short read archive under project accession numbers PRJNA308208, PRJNA627536, and PRJNA844099 and others (Supplementary File 1).

The MANBA variant was validated in the case cat by PCR and fluorescence-based Sanger sequencing as previously described (Shelton et al., 2023). Cat reference assembly F.catus_Fca126_mat1.0 (GCF_018350155.1; BioProject: PRJNA684600; BioSample: SAMN19729387) and RefSeq Annotation 105 was used to design PCR primers for the MANB variant ([XM_023252901.1; ENSFCAT00000007560:c.2506G>A; p.Gly836Arg] at position B1:121800342 Felis_Catus 9.0 ([ca126 position B1:119398498]) using Primer3Plus (Rosen et al., 2022). The PCR forward primer was 5’-TGAAGGACACAATGCTTTGC-3' and reverse primer 5’-GCGATGATTCAGATGTGTGG-3'. PCR optimization and amplicon production was performed using ChoiceTaq^™ DNA Polymerase (Denville Scientific, Inc., Metuchen, NJ) according to manufacturer protocol and a primer annealing temperature of 58°C. Amplified fragments from the PCR products (~20 ul) of genomic DNA template from the proband and two control cats were purified using QIAquick^® PCR Cleanup Kit (Qiagen, Redwood City, CA, USA) and ~ 45 ng of each amplicon was submitted to the University of Missouri Genomics Technology Core (dnacore.missouri.edu), which conducted Sanger sequencing using a BigDye^™ Terminator v3.1 Cycle sequencing kit (Applied Biosystems, Waltham, MA) and an ABI 3730XL DNA Analyzer (Applied Biosystems). The generated sequences were aligned to the cat reference genome assembly F.catus_Fca126_mat1.0 and visually inspected using Sequencher 5.1 (GeneCodes, Ann Arbor, MI) to confirm the amplification of MANBA and the presence of the variant in the case cat.

Genomic tools and resources at the National Center for Biological Information (NCBI) website, including ClinVar (Landrum et al., 2020), the Genome Data Viewer (Rangwala et al., 2021), and the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990; Zhang et al., 2000), and the Genome Aggregation Database (gnomAD) (Chen et al., 2022) were used to correlate the variant in the cat to human MANBA (XM_047415692.1; XP_047271648.1) variant. Multi-species alignment was performed with CLUSTAL, prediction of alteration on protein function with SIFT (Ng and Henikoff, 2001)(https://sift.bii.a-star.edu.sg/), Align-GVDV (http://agvgd.hci.utah.edu/about.php) and Polyphen-2 (http://genetics.bwh.harvard.edu/pph2/) (Adzhubei et al., 2010).

2.4. Biochemical analyses

Slices of brain cerebral cortex temporal lobe and cerebellar cortex were dissected from brain samples from the proband and from an unaffected 7 year, 8-month-old intact male cat with no known disease. The tissues were thawed and homogenized in 154 mM NaCl, 0.2% Triton X-100 at approximately 50 mg of tissue per mL of solution and the homogenates were frozen at −80°C overnight. The homogenates were then centrifuged at 4°C for 10 minutes at 10,000xg, and enzyme activities were determined in aliquots of the supernatants. Total protein contents of the samples were determined using the Bio-Rad Protein Assay Kit II with bovine serum albumin (BSA) and the absorbance measured at 595 nm (Bio-Rad Laboratories cat. no. 5000002). The activities of 3 α-D-mannosidase [EC 3.2.1.24] and β–D-mannosidase [EC 3.2.1.26] were measured using 4-methylumbelliferyl α-D-mannopyranoside (GlycoSynth cat. no. 44081) and 4-methylumbelliferyl β-D-mannopyranoside (Sigma-Aldrich, cat. no. M0905) substrates respectively. Enzyme activities were measured by monitoring fluorescence emission at 470 nm with excitation at 325 nm over a 30 min incubation with the reaction mixtures maintained at 37°C. Activities are reported as pmol substrate hydrolyzed per min per mg total protein.

2.5. Data availability

All sequence variants unique to the affected cat are included in Supplementary File 1. The WGS data are available in the NCBI short read archive under project accession numbers PRJNA308208, PRJNA627536, PRJNA844099, PRJEB7401, PRJNA478778, PRJEB34047, and PRJEB34077 with this case cat (Fcat_23309) as accession SAMN35992035.

3. Results

3.1. Clinical presentation and neurological findings

The subject cat was adopted from a rescue organization and was of unknown ancestry but had the appearance of having mixed domestic shorthair and Siamese ancestry (Figure 1). At approximately 6 months of age, the proband presented for neurological evaluation due to ataxia in all four limbs, with more pronounced hind limb deficits. Onset of these motor deficits was reported to be at approximately 3 months of age, and they progressed over time. By 6 months of age, the cat was unable to stand while defecating, was unstable on her feet at all times, and often fell while walking. She seldom played and slept 90% of the time. The owners also reported the cat appeared to be cognitively dull and clumsy. The signs did not improve with administration of antibiotics.

Figure 1.

Images of the female shorthaired cat of unknown ancestry. The subject cat had a body shape and physical appearance suggestive of a domestic shorthair but the “pointed” coloration suggested Siamese ancestry. (Photographs courtesy of the owner.)

Upon examination at 6 months of age, the cat exhibited apparent retinal degeneration based on a fundus appearance suggestive of chorioretinitis. Muscle tone was normal and symmetric. Upon neurological examination the cat was found to have dull mentation, a tetraparetic gait, and delayed conscious proprioception in all four limbs. Spinal reflexes and muscle tone were normal in all four limbs. The cat was visual in both eyes and exhibited bilateral positive menace responses and symmetric pupillary light and palpebral reflexes. The cat did not exhibit abnormal nystagmus or strabismus and had normal face sensation, lip curl, gag and tongue movements, and oculocephalic reflexes. Cerebrospinal fluid analysis did not reveal any evidence of active central nervous system infection or inflammation. Due to the progression of motor and cognitive impairment, the owners elected to have the cat euthanized at approximately 10 months of age.

3.2. Tissue Pathology

The clinical signs exhibited by the proband were consistent with those of animals with a group of lysosomal storage diseases classified as neuronal ceroid lipofuscinoses (NCLs). The NCLs are characterized by accumulation of autofluorescent storage material in neurons and other cell types, including cardiac muscle in some forms of NCL (Katz et al., 2020, 2017; Mole et al., 2011). Therefore, unstained cryostat sections of retina, cerebellar cortex, cerebral cortex, and cardiac muscle were examined for the presence of intracellular inclusions with the fluorescence properties characteristic of the NCLs. No detectable autofluorescence was present in the retina or cardiac muscle. In the cerebellar cortex, large autofluorescent inclusions were present in the Purkinje cell bodies, but not in other cell types (Figure 2A). In the cerebral cortex, almost all neurons contained autofluorescent inclusions that varied in intensity between cells (Figure 2B).

Figure 2.

Fluorescence micrographs of unstained cryosections of cerebellar cortex (A) and cerebral cortex (B) from the proband cat. Autofluorescent inclusions in the cerebellar cortex were present primarily in the Purkinje cells. Minimal autofluorescence was present in the molecular (m) or granule cell (g) layers of the cerebellar cortex or in non-Purkinje cells in the Purkinje cell layer (p). In the cerebral cortex (B) autofluorescent inclusions were present in most of the neurons.

Examination of stained paraffin sections of the cerebellar cortex and cerebral cortex revealed that the somas of most of the cells in these tissues contained abundant vesicular inclusions that in many cases filled almost the entire cell bodies (Figure 3). In the cerebellar cortex, cells in every layer of the tissue were affected. In the cerebral cortex, the pericuclear area of almost every cell was filled with these vesicular inclusions. Similar clear vesicles were scattered in the neuropil of the cerebral cortex. Clear vesicular inclusions were also abundant in all layers of the retina, particularly at the level of the photoreceptor inner segments (Figure 4).

Figure 3.

Light micrographs of paraffin sections of cerebellar cortex (A) and the cerebral cortex (B) of the proband cat. In both tissues, the cell bodies of almost all cells contained abundant clear vesicular inclusions that, in many cases, filled almost the entire cell body. In the cerebellum the affected cells were in the molecular (m), Purkinje cell (p) and granule cell (g) layers.

Figure 4.

Light micrograph of paraffin section of retina from the proband cat. Clear vesicular inclusions were present throughout the retina, with particularly large vesicles in the ganglion cells, inner nuclear layer, and photoreceptor inner segments (arrows).

Electron microscopic examinations revealed that the contents of the vesicular inclusions varied in ultrastructural appearance (Figures 5 and 6). The contents of many of these inclusions had a uniform fine granular appearance. In other inclusions, materials embedded in this uniform matrix were quite variable in appearance, and included vesicular, flocculant, and bead-like components. The vesicular components appear to have arisen, at least in part, from invaginations of the membranes surrounding the vesicles (Figure 6).

Figure 5.

Electron micrographs showing vesicular inclusions (V) in cells of the cerebellar cortex from the proband cat. The contents of many of the vesicles had only a fine granular appearance (A). However, numerous vesicles also contained heterogenous mixtures of material that included smaller vesicles, flocculent material, and bead-like structures (B, C, and D).

Figure 6.

Electron micrograph showing vesicular inclusions in a cell of the cerebral cortex from the proband cat. The appearances of these inclusions were similar to those seen in cells of the cerebellar cortex. In this example, an invagination of the vesicular membrane was present (arrow). The appearance of this structure suggests that the smaller vesicles within some of the larger vesicles may form by in-budding of the vesicular surrounding membrane.

In addition to the abundant vacuole-like inclusions, cells of the cerebellar and cerebral cortexes contained smaller numbers of inclusions; the contents of which consisted either exclusively of whorls of membranes (Figure 7A), or of mixtures of components with heterogeneous morphologies (Figure 7B and Figures 8 and 9). In the cerebellar cortex, the latter types of inclusions were present primarily in Purkinje neurons. In the cerebral cortex, it was not apparent whether the abundance of these inclusions was most pronounced in a specific cell type.

Figure 7.

Electron micrographs of storage bodies (SB) in cells of the cerebellar cortex from the proband cat. Some storage bodies consisted primarily of whorl-like arrays of membranous structures (A). In addition to these membrane-like arrays, other storage bodes also contained a heterogeneous mixture of inclusions (B).

Figure 8.

Electron micrographs of storage bodies (SB) in Purkinje cells of the cerebellar cortex from the proband cat. The contents of these storage bodies consisted of aggregates of a heterogenous mixture of inclusions.

Figure 9.

Electron micrographs of storage bodies (SB) in cells of the cerebral cortex from the proband cat. These storage bodies were similar in appearance to those seen in the cerebellar cortex. In both (A) and (B), in some areas the storage bodies could be seen to be bounded by double membranes characteristic of autophagosomes.

3.3. Molecular genetic findings

DNA sequence variant filtering was performed based on the hypothesis that any variant underlying the disorder is rare among cats. Rare variants in the proband were identified by comparing WGS data from this cat to sequence data in the cat 99 Lives dataset. Relative to the reference cat genome, the 99 Lives dataset has 2,289,566 variants from 413 additional unique cats (10 cats had both WES and WGS data). In analyzing the data from the proband, both dominant and recessive modes of inheritance were considered with no a priori consideration of a candidate gene. The unique variants for the proband relative to the reference 99 Lives dataset are presented in Supplementary File 1. Considering the hypothesis of recessive inheritance and unique homozygous variants with a passing Variant Quality Score Recalibration (VQSR) Tranche score (Buckley et al., 2020) and predicted to alter the gene coding sequences, 10 variants were identified, including four within splice regions, one inframe insertion and five missense variants, including a missense in MANBA (XM_023252901.1; ENSFCAT00000007560:c.2506G>A; p.Gly836Arg) at position B1:121800342 Felis_Catus 9.0 (Fca126 position B1:119398498) (Supplementary File 1). The variant had 37X sequencing reads for the case, suggesting good sequencing coverage of the variant region. Considering the hypothesis of dominant inheritance, 626 heterozygous variants were identified in the affected cat, including several loss of function variants, splice variants and missense variants (Supplementary File 1). None of these heterozygous variants had an obvious causal relationship to the disease phenotype. All variants detected in 99 Lives dataset for MANBA are presented in Supplementary File 1.

Based on the disease phenotype, including the tissue pathology, homozygosity for the MANBA variant appeared most likely to be the cause of the cat’s disorder. For variant validation, a single PCR product of the expected size, 648 bp, was produced from DNA of the affected cat and two unrelated unaffected cats (Figure 10). Both the male control cat, Fcat_4406, and the female control cat, Fcat_12682, were homozygous for the wildtype allele and the affected female cat, Fcat_23309, was homozygous for the MANBA variant (Figure 10). This variant was unique relative to 413 unique cats within the 99 Lives dataset. MANBA (ENSFCAT00000058358; XM_023252901) is at position 121685862-121800542 on the B1 chromosome in Felis Catus 9.0 reference assembly using Ensembl annotation 104 and at position chromosome B1 (NC_058371.1): 119275197 - 119411044 in the Felis_Catus Fca126 assembly using NCBI annotation 105.

Figure 10. Left panel: MANBA PCR product for validation by Sanger sequencing.

Lane 1: 100 bp PLUS^™ DNA Ladder; lane 2: control cat 1; lane 3: control cat 2; lane 4: case cat. DNA from the two control cats and the case cat produce a 648 bp fragment for Sanger sequencing. The brightest fragment in the DNA ladder is 500 bp. Right panel: Electropherograms of MANBA Sanger sequencing. A: Control cat 1 coding strand; B: Control cat 1 complementary strand; D Control cat 2 coding strand; E: Control cat 2 complementary strand; E Case cat strand; F: Case cat complementary strand. Box where both unaffected cats were homozygous for the c.2506G variant and the case cat was homozygous for the c.2506A variant.

The NCBI cat reference assembly FCA_126 is annotated with three variant transcripts in the domestic cat, which code for 17 - 18 exons. The cat variant is positioned in the last exon of the transcripts, which has over 95% homology (91.63% identity) with the corresponding human MANBA exon. The nucleotide sequence and predicted amino acid sequence encoded by the MANBA exon 18 are highly conserved across diverse species (Figure 11).

Figure 11.

Multispecies alignment of Felis Catus MANBA exon 18. The arrows indicate the position of the MANBA missense variant (ENSFCAT00000007560:c.2506G>A, XM_023252901.1:c.2506G>A; p.Gly836Arg) at chromosome position B1:121800342 in the Felis_Catus 9.0 assembly. (A) Nucleotide alignment for diverse species (MANBA affected feline case, normal feline, human, canine, bovine, porcine, and murine) indicating the 100% conservation of the nucleotide and the high conservation of the remainder of exon 18 (lower histogram). Nucleotides are represented by different colors. (B) Amino acid sequence alignments for the same species depicting the amino acid change in exon 18 for the affected cat. The mammals have nearly 100% amino acid conservation within this region.

Effects of the p.Gly836Arg amino acid substitution on the protein function were predicted by SIFT, Align-GVGD, and Polyphen-2. The replacement of the small nonpolar amino acid glycine with the large basic arginine and was predicted as not tolerated by SIFT with a score of 0.00, implying deleteriousness. Polyphen-2 predicted alteration to be probably damaging with a score of 1.000 (sensitivity: 0.00; specificity: 1.00) and Align-GVGD predicted the variant to be class C65 and most likely to interfere with function (GV = 0.00; GD = 125.13; Prediction = Class C65).

3.4. Enzyme activity assays

The MANBA variant in the proband was associated with a complete absence of detectable β–mannosidase enzyme activity in the cerebellar cortex and cerebral cortex (Table 1). The limit of detection was less than 0.4% of the activities detected in the tissues from a control unaffected cat. In contrast, the activities of α–mannosidase, another lysosomal enzyme, were elevated in the affected cat relative to a healthy control animal (Table 1). In the affected cat, α–mannosidase activities in the cerebral and cerebellar cortex of the affected cat were 6.4- and 11.5-fold those in these tissues of the affected cat (Table 1).

Table 1.

Mannosidase Enzyme Activities in Cat Tissues

Enzyme	Dog	Tissue	Activity*
β–mannosidase	Proband	Cerebral Cortex	0
β–mannosidase	Normal cat	Cerebral Cortex	124
β–mannosidase	Proband	Cerebellum	0
β–mannosidase	Normal cat	Cerebellum	137
α-mannosidase	Proband	Cerebral Cortex	306
α-mannosidase	Normal cat	Cerebral Cortex	48
α-mannosidase	Proband	Cerebellum	356
α-mannosidase	Normal cat	Cerebellum	31

^*pmol substrate hydrolyzed per min per mg total protein; limit of quantitation: 0.5 pmol substrate hydrolyzed/min/mg protein

4. Discussion

Several lysosomal storage diseases have been genetically characterized in domestic cats, including α-mannosidosis (OMIA: 000625-9685; c.1749_1752, p.Q584Afs) (Berg et al., 1997), however β-mannosidosis has not been previously documented in the cat. Forms of the disease have been identified in dogs (Bolfa et al., 2019; Jolly et al., 2019), as well as in goats, cattle, and mice (Abbitt et al., 1991; Chen et al., 1995; Jones and Abbitt, 1993; Jones and Dawson, 1981; Leipprandt et al., 1999, 1996; Zhu et al., 2006). β-Mannosidosis has also been identified in human subjects, although it is extremely rare in the human population, with fewer than 60 cases associated with MANBA variants having been described in published reports (Blomqvist et al., 2019; Gort et al., 2006; Labauge et al., 2009; Levade et al., 1994; Lund et al., 2019; Renaud, 2023; Riise Stensland et al., 2008; Sabourdy et al., 2009; Safka Brozkova et al., 2020; Uchino et al., 2003). The ClinVar database (accessed 19 July 2023) predicts 41 of the known human variants to be pathogenic in MANBA and 19 to be likely pathogenic. The disease phenotype in human subjects is variable, even among patients with null mutations (Bedilu et al., 2002), but almost always include neurological signs.

The proband cat was homozygous for the c.2506G>A; p.Gly836Arg variant, which was not found in any other cat in the 413 cat 99Lives cohort. This variant is in the last of 17 exons of MANBA. The amino acid sequences encoded by exon 17 around the equivalents of feline p.836 have strong conservation across diverse species, and the effect on protein structure and function of substituting the small nonpolar glycine with the large basic arginine could be significant. SIFT, Align-GVGD, and Polyphen-2 all predicted the cat c.2506G>A p.Gly836Arg variant to be deleterious, likely affecting the function of MANBA. Even though missense variants near the 3’ terminus of genes should be cautiously considered pathogenic, no β-mannosidase enzymatic activity was detected in tissues from the affected cat, suggesting that the amino acid substitution had a profound effect on the structure of the protein. No identical variant, benign or pathogenic, is documented in humans. MANBA missense variants associated with clinical signs of disease in human predict the amino acid changes listed in Table 2. Determination of the structures of the β-mannosidase proteins from human, goat, cow, and mouse suggested the effects of most of these single amino acid changes on enzyme function were related to the proximity of the alterations to the active site of the enzyme (Gytz et al., 2019; Huynh et al., 2011). Some alterations not near the active site, such as p.Arg182Trp, were predicted to destabilize the active site (Huynh et al., 2011). However, based on the proposed protein structure, amino acid 836 is not predicted to be within or have any effect on the active site, so its role in determining enzyme activity is probably indirect. The finding that a single amino acid substitution at this site is associated with severe β-mannosidosis and abolition of enzyme activity is strong evidence that the carboxy terminal region of the protein in which this amino acid occurs plays a key role in its function. Truncation of the bovine protein at Trp858 also results in a severe lethal phenotype, consistent with the carboxy terminal region of the protein being critical to normal function (Leipprandt et al., 1999). Two documented human variants that flank the comparable p.Gly836Arg cat variant (p.Leu830Trp and p.Phe839Cys) are not associated with disease and both were predicted by SIFT, Align-GVGD, and Polyphen-2 to be of uncertain significance (VUS). Further research will be necessary to elucidate the mechanism by which the p.Gly836Arg amino acid substitution alters β-mannosidase enzyme activity.

Table 2.

Missense variants in MANBA associated with β-mannosidosis

Predicted Amino Acid Change	Species	Disease Phenotype	Reference
p.Gly836Arg	Cat	Severe neurological disease	This report
p.Ile187Asn	Dog (German Shepherd)	Severe neurological disease	(Jolly et al., 2019)
p.Arg641His	Human	Severe neurological disease	(Sabourdy et al., 2009)
p.Ser505Pro	Human	Severe neurological disease	(Riise Stensland et al., 2008)
p.Arg182Trp	Human	Mild neurological signs	(Gort et al., 2006)
p.Gly392Glu	Human	No neurological signs	(Riise Stensland et al., 2008; Rodriguez-Serna et al., 1996)

Homozygosity for deleterious variants, such as that observed in the proband, has been previously documented in random bred populations with historical influences of a specific breed. Cats independently ascertained from a shelter in Winnipeg, Canada, which was rescuing cats from a specific population, had phenotypes suggesting contributions from the Burmese breed (Gandolfi et al., 2014). These cats had a Burmese body type and suffered from myotonia congenita. Cats with the “pointed” coloration of the Siamese are often sold in pets stores and marketed as “Traditional” or “Applehead” Siamese and are usually cross-bred, non-pedigreed cats. Thus, this current case could have come from a “kitten mill” of non-pedigreed cats or a feral population, both of which must have had sufficient inbreeding for identity by descent of the disease variant. Indeed, at least 30% of 185 cat variants documented in the Online Inheritance in Animals (OMIA) are detected in random bred, domestic short/medium/long haired cats in which a majority are autosomal recessive, although identity by descent was not considered in most studies. The overall linkage disequilibrium of random bred cats is low (Gandolf et al., 2018), as it is in some populations showing signs of recent breed admixture (Irving-McGrath et al., 2021).

Deficiencies in single lysosomal enzymes are often accompanied by elevation in tissue levels of other lysosomal enzymes (Awano et al., 2006; Sleat et al., 2019, 2012, 1998). Indeed, a similar elevation in alpha-mannosidase activities to that seen in brain tissues from the affected cat was reported in goats with β-mannosidosis (Lovell et al., 1994). Affected goats also exhibited elevated levels of alpha-fucosidase and beta-hexosaminidase in brain tissues, suggesting, as for other lysosomal storage diseases, β-mannosidosis is likely to be characterized by a global increase in lysosomal enzyme levels.

The massive vacuolar accumulations in cells of the brain in the affected cat were similar to those reported in goats, cattle, dogs, and mice with β-mannosidosis (Bryan et al., 1993; Jolly et al., 2019; Jones et al., 1983; Lovell et al., 1994; Lovell and Jones, 1983; Patterson et al., 1991; Zhu et al., 2006). Cattle and goats with β-mannosidosis also exhibited massive accumulations in vacuolar inclusions in the retina, like those of the affected cat (Render et al., 1992, 1989). However, this is the first report that β-mannosidosis is also accompanied by the accumulation of cellular inclusions with lipofuscin-like autofluorescence properties. Accumulation of autofluorescent lysosomal storage bodies is a defining characteristic of a group of progressive neurodegenerative disorders known as the neuronal ceroid lipofuscinoses (NCLs) (Katz et al., 2017; Mole et al., 2011). β-Mannosidosis is distinguished from the NCLs by the fact that in the latter disorders most, if not all, of the storage bodies exhibit lipofuscin-like autofluorescence, whereas in the tissues from the affected cat, only a minority of the inclusions were autofluorescent. Electron microscopic examinations revealed that the storage bodies in the affected cat included not only vacuolar inclusions with uniform fine granular contents previously associated with β-mannosidosis, but also many inclusions that also contained mixtures of other electron-dense components. Some of the inclusions were filled with mixtures of electron-dense structures, many of which had membrane-like appearances. The latter inclusions were similar in appearance to those that accumulate in some forms of NCL. The spectrum of ultrastructural appearances of the disease-related inclusions suggests that at least some of the lipofuscin-like autofluorescent inclusions may evolve over time from the vacuolar inclusions. This possibility is supported by the fact that invaginations of the membranes surrounding the vacuolar inclusions could sometimes be seen. Engulfment of cellular components into the vacuolar inclusions by the process of invagination could lead to a progressive accumulation of the heterogeneous collections of materials seen in many of the storage bodies. In addition, the ultrastructural appearances of some of the non-vacuolar storage bodies suggest that they may be autophagolysosomes containing partially undegraded cellular components that accumulate secondarily due to impaired lysosomal function. Similar processes of secondary storage body formation may occur in other lysosomal storage disorders. For example, a dog with α-mannosidosis, in addition to the accumulation of vacuolar inclusions characteristic of enzyme substrate storage, exhibited accumulation of autofluorescent inclusions with ultrastructural features of autophagolysosomes (Bullock et al., 2023). The presence of lipofuscin-like autofluorescent inclusions in lysosomal storage diseases with progressive neurological signs has been used for classifying disorders as NCLs. The findings in this study and in canine α-mannosidosis indicate that these criteria are not sufficient for designation of diseases as NCLs. Rather, the classification of a lysosomal storage disease as a form of NCL should be restricted to those cases in which almost all of the storage bodies exhibit lipofuscin-like autofluorescence properties.

In summary, a homozygous sequence variant in MANBA that predicts a p.Gly836Arg single amino acid substitution has been identified in a cat that exhibited the hallmarks of severe β-mannosidosis disease. The variant was associated with an absence of detectable β-mannosidase enzyme activity in brain tissue. These findings suggest that the region of the protein containing the variant amino acid plays a critical role in maintaining enzyme function. The primary accumulation of vacuolar lysosomal storage bodies is accompanied by apparent secondary accumulation of lipofuscin-like storage bodies.

Highlights.

• A domestic cat with a progressive neurological disorder exhibited massive intracellular accumulations of vacuolar inclusions throughout the brain and retina, as well as smaller numbers of lipofuscin-like inclusions.

• The disease was associated with a novel missense variant in MANBA and the absence of detectable β-mannosidase enzyme activity, indicating that the cat suffered from β-mannosidosis.

• The predicted amino acid change in β-mannosidase indicates an important functional role of the C-terminal region of the protein.

Abbreviation List

WGS

whole genome sequencing

WES

whole exome sequencing

GATK

Genome Analysis Toolkit

NCBI

National Center for Biotechnology Information

NCL

neuronal ceroid lipofuscinosis

VQSR

Variant Quality Score Recalibration

PCR

polymerase chain reaction