Infantile Cerebellar-Retinal Degeneration Associated With Novel ACO2 Variants: Clinical Features and Insights From a Drosophila Model

Abstract

Infantile Cerebellar-Retinal Degeneration (ICRD) is an autosomal recessive neuro-disability associated with hypotonia, seizures, optic atrophy, and retinal degeneration. Recessive variants of the mitochondrial aconitase gene (ACO2) are a known cause of ICRD. Here, we present a paediatric male patient with ICRD, where whole genome sequencing of the family trio revealed segregating heterozygous variants of unknown significance in ACO2. At 4 months, he displayed generalised hypotonia, and by 6 years, visual electrophysiology indicated bilateral optic atrophy. Magnetic Resonance Imaging (MRI) at age seven confirmed optic nerve and cerebellar atrophy, and together with symptoms of developmental delay, align with ICRD. We established a Drosophila animal model to explore the impact of ACO2 loss- and gain-of-function. Manipulating the fly ortholog, mAcon1, through pan-neuronal knock-down or over-expression negatively affected longevity, locomotion, activity, whilst disrupting sleep and circadian rhythms. Mis-expression of mAcon1 in the eye led to impaired visual synaptic transmission and neurodegeneration. These experiments mirrored certain aspects of the human disease, providing a foundation for understanding its biological processes and pathogenic mechanisms, and offering insights into potential targets to screen for future treatments or preventive measures for ACO2-related ICRD.

Keywords: Drosophila melanogaster; ACO2; mAcon1; ERG; MRI; infantile cerebellar‐retinal degeneration; locomotion; optic atrophy; sleep.

FIGURE 1

Clinical features and neurological manifestations in MRI imaging. (A) Clinical presentation of bilateral large angle esotropia (squint), hypotelorism (decreased distance between the orbits), plagiocephaly (asymmetrical flattening of the head) and facial telangiectasias (dilation of capillaries) observed in the patient. (B) Photo of Right retina showed optic atrophy in the form of severe optic disc pallor. (C–E) Sequential MRI images illustrating the progression of neurological changes: (C) a normal volume of cerebellar hemispheres in an axial T2 sequence at age 1, (D) development of cerebellar vermian hypoplasia (arrow) at age 7 in a sagittal T1 sequence, and (E) symmetrical volume loss of the cerebellar hemispheres accompanied by high intensity in the dentate nuclei in an axial T2 sequence. A subtle finding of minimal prominence of the central tegmental tract (arrow) was also noted in the MRI analysis. (F) Coronal STIR sequence, revealing slender intra‐orbital optic nerves with more severe atrophy on the left side (arrow).

FIGURE 2

Right thigh (quadriceps) muscle biopsy obtained at age 3. (A) HH&E stain demonstrating mild variation in myofibre diameters, with the overall morphology revealing slight hypotrophy in some fibres. There was also a subtle but notable increase in endomysial connective tissue, suggestive of early, mild myopathic changes. However, no inflammatory infiltrates, necrotic fibres, or regenerating fibres were visualised, indicating an absence of active myonecrosis. (B) Spectrin immunostain highlighting the variation in myofiber diameters (*), aiding in the differentiation of smaller and larger fibres. Spectrin, a cytoskeletal protein, outlines the sarcolemmal membranes in intact fibres. As this was well preserved in all myofibres, it confirms an absence of muscle fibre necrosis. This stain also helps highlight an absence of fibre splitting and a lack of central nucleation in this case, further supporting the presence of only mild non‐specific myopathic features. (C) MATPase stain distinguishes fibre types, with type I (slow oxidative) fibres appearing dark blue and type II (fast glycolytic) fibres light blue. In this case, the MATPase stain revealed type I fibre hypotrophy, contributing to the mild non‐specific myopathic pattern. (D) Gomori Trichrome stain displaying a normal staining pattern with no evidence of ragged red fibres, nemaline rods, or other inclusions, further excluding specific congenital myopathies or mitochondrial disorders. (E) Combined COX‐SDH reaction revealing a normal staining pattern of myofibres and normal oxidative phosphorylation enzyme activity. The absence of COX‐deficient, SDH‐positive fibres suggests that there are no detectable mitochondrial abnormalities within the muscle, consistent with the lack of metabolic dysfunction observed in laboratory investigations. (F) NADH staining showing normal myofibrillar architecture within muscle fibres. There were no visible structural abnormalities such as cores, targets or disruptions in myofibrils, supporting the non‐specific nature of the myopathic features observed. Scale bars, 50 μm.

FIGURE 3

Molecular characterisation and in silico modelling of patient ACO2 variants. (A) ACO2 family pedigree and sequencing chromatograms with Sanger confirmation of the maternally transmitted c.542A>C p.(Tyr181Ser) ACO2 variant (left panel) and the paternally transmitted c.1640 T>C p.(Pro547Leu) ACO2 variant (right panel). Variant nomenclature refers to ACO2 RefSeq (NM_00198.3). (B–E) In silico modelling of ACO2 variants with Missense3D predicted a deleterious effect of the p.Tyr181Ser substitution, with the wildtype buried tyrosine residue (blue, B) being replaced by an exposed serine residue (red, C). Missense3D did not provide evidence of a deleterious effect due to the p.Pro547Leu substitution, with the proline residue (blue, D) being replaced by a leucine residue (red, E). Nearby residues are labelled. (F) AlphaMissense predictions for the patient's ACO2 variants showed that the p.Tyr181Ser (Y181S, left panel) substitution is scored at 0.975, supportive of pathogenicity, whilst the p.Pro547Leu (P547L, right panel) variant is predicted to have a milder effect, scoring 0.553, falling within the range of uncertain pathogenicity. (G, H) SDS‐PAGE and western blotting demonstrated a marked reduction in ACO2 steady‐state levels in the patient's muscle homogenate lysates compared to two age‐matched controls (G), whilst the patient's steady‐state levels of various different OXPHOS proteins, including NDUFB8 (Complex I), SDHB (Complex II), UQCRC2 (Complex III), COXII (Complex IV) and ATP5A (Complex V), were comparable to those of the controls (H). This aligns with previous results indicating unremarkable respiratory chain enzymology. Images from consecutive immunodetection with ACO2, the OXPHOS cocktail (Abcam), and SDHA, used as a mitochondrial marker, are provided. For clarity, ACO2 is shown separately in (G), while the OXPHOS panel in (H) uses the same SDHA loading control. SDS‐PAGE and immunoblotting were performed in triplicate, with representative data shown. P, patient; F, father; M, mother; C1 and C2, controls.

FIGURE 4

mAcon1, the fly homologue of human ACO2, shortens lifespan and reduces climbing ability and exploratory behaviour of Drosophila. (A) Pan‐neuronal mAcon1 knock‐down (elav > mAcon1 ^RNAi , solid red bar) led to a 52% decrease, while mAcon1 over‐expression (elav > mAcon1 ^OX , solid blue bar) led to a 48% increase in mAcon1 mRNA levels compared to both Gal4 (black bar) and UAS (open bars) controls. n, technical replicates indicated; N = three biological replicates with ~40 heads each. (B) mAcon1 knock‐down (red solid line) dramatically and over‐expression (blue solid line) significantly reduce the lifespan of flies, represented by the proportion of surviving flies over time, compared to controls (Gal4, black line; UAS, dotted lines). n, indicated. (C) In the negative geotaxis assay to assess the climbing proficiency of flies, groups of 10 flies were tapped gently towards the bottom and the number of flies that reached the line within a span of 10 s was counted. Knock‐down of mAcon1 (solid red bar) resulted in a reduced climbing performance compared to controls. Bars represent means; whiskers represent 95% CI; n represents numbers in bars (groups of 10 flies each). (D) Individual exploratory locomotor behaviour was studied using a solitary fly moving freely within a well‐lit arena (ø 55 mm) for a duration of 3 min. The fly's movements were recorded using a webcam connected to tracking software on a computer. The combined tracks of mutant and control flies (12 flies each) show their exploration patterns, with mAcon1 knock‐down flies (red traces) covering a smaller area within the arena, tending to avoid the centre. (E–G) The measured parameters identify a decrease in both walking distance (F) and speed (G) for mAcon1 knock‐down flies (solid red bars) while maintaining a consistent overall activity level (E). Bars, means; whiskers, 95% CI; n, numbers in bars.

FIGURE 5

Manipulation of mAcon1 increases locomotor activity and impairs circadian rhythmicity. (A) Diagram illustrating the setup for automated fly tracking. Each fly is placed individually in a tube containing food and loaded into a DAM monitor placed in an incubator. Inside the monitor, an infrared beam intersects the tube, enabling the recording of fly movements whenever the beam is interrupted. The DAM monitor is connected to a computer that records the total count of beam breaks, providing a measure of the fly's activity. The histograms show the five‐day average daily activity levels of all flies in LD for control and mutant genotypes (filled bars, lights off; open bars, lights on). (B–D) Total activity levels, measured as the average daily number of beam crosses, showed an increase in activity of mAcon1 over‐expressing flies (B). mAcon1 knock‐down flies were less active during the day (C) while both manipulations led to increased activity during the night (D). (E–G) Both mutant strains exhibited impaired circadian rhythmicity, leading to a decrease in morning anticipation (E) and weakened circadian rhythm strength (G). However, the evening anticipation (F) remained unaffected. Bars, means; whiskers, 95% CI; n, numbers in bars.

FIGURE 6

Manipulation of mAcon1 reduces and fragments sleep. (A) The sleep distribution profiles show the amount of time spent asleep in 30‐min intervals throughout the day, demonstrating a reduction in sleep during the lights off, night period (grey area) for both knock‐down (red trace) and over‐expression (blue trace) of mAcon1 compared to controls. (B–D) Both mutant strains exhibited decreased total sleep levels (B). Flies over‐expressing mAcon1, however, displayed increased daytime sleep (C), while both manipulations resulted in reduced night‐time sleep (D). (E, F) Both mutant strains displayed a modified sleep architecture characterised by an increase in the number of sleep episodes (E) and a decrease in the average length of sleep bouts (F), resulting in more fragmented sleep. Bars, means; whiskers, 95% CI; n, numbers in bars.

FIGURE 7

Manipulation of mAcon1 leads to eye neurodegeneration and reduced electroretinograms. (A) Photographs of the compound eyes of mutant flies were taken to compare them to control flies (GMR / +) exhibiting the regular alignment of ommatidia. mAcon1 knock‐down eyes showed a mild “rough eye” phenotype, which is typically associated with neurodegeneration. Scale bar, 100 μm. (B) Exemplary electroretinogram (ERG) recordings with increasing light intensity stimulation for control and mutant flies. Control flies (black traces) show the typical fly ERG composed of an ON‐transient, sustained photoreceptor response and OFF‐transient. While these responses were reduced for both mutants, mAcon1 knock‐down flies (red traces) nearly completely lacked both transients. (C) Both knock‐down (red traces) and over‐expression (blue traces) of mAcon1 resulted in reduced receptor potential (left) compared to controls (black traces) across all stimulus intensities as well as decreased ON‐transients (middle) and OFF‐transients (right). The effects were more pronounced in the knock‐down flies. Lines, means; shaded area, 95% CI; n, indicated.