Siddiqi syndrome (SIDDIS, OMIM #618635) is a rare autosomal recessive neurodevelopmental disorder caused by biallelic loss-of-function variants in the FITM2 gene (Fat Storage Inducing Transmembrane Protein 2). This gene is located on chromosome 20q13.12, contains two exons, and encodes a 4,628 bp transcript that translates into a 262-amino-acid protein, named FIT2. The FIT protein family (FIT1 and FIT2) comprises evolutionarily conserved endoplasmic reticulum (ER) - resident transmembrane proteins involved in lipid droplet biogenesis and energy homeostasis (Gross et al. 2011).

SIDDIS presents as a severe, childhood-onset multisystem disorder with marked neurological involvement. Affected individuals typically exhibit failure to thrive, together with neurological features of dystonia, motor delay and regression, sensorineural hearing loss and limited speech development. This syndrome was first reported in 2017 in five siblings from a consanguineous Pakistani family harboring a homozygous truncating variant in the FITM2 gene (c.4G > T, p.(Glu2Ter). These individuals presented with a deafness-dystonia syndrome characterized by progressive sensorineural hearing loss, global developmental delay with motor regression, ichthyosis-like features and signs of sensory neuropathy (Zazo Seco et al. 2017). Since its initial description, only a limited number of SIDDIS cases have been reported in the literature (Table 1; Fig. 1). Among them, three involve pediatric patients (age range 1- 8years old), highlighting the rarity of this syndrome and underscoring the need for further genetic studies.

Although SIDDIS was initially described as a deafness-dystonia disorder, a recently published preprint has suggested a link between FITM2 and a hereditary spastic paraplegia (HSP)-like phenotype (Salazar-Villacorta et al. 2025) (Table 1; Fig. 1). In this report, we describe the second case of SIDDIS due to a compound heterozygous mutation in the FITM2 gene, identified in a 29-year old woman who exhibits the characteristic features of the disorder (deafness and intellectual disability) and also supports the recently proposed association between FIMT2 and HSP. In addition, this case provides noticeable, previously unreportedbrain MRI signal abnormalities, thereby contributing to a more comprehensive definition of the clinical spectrum associated to FITM2 alteration.

a Graphical representation of FITM2 gene (NM_001080472.4) and its encoded protein FIT2. b Location of the pathogenic/likely pathogenic variants identified in FIMT2 patients relative to the transmembrane domains. In bold are highlighted the variants identified in the patient herein reported

The patient is a 29-year-old woman, born to non-consanguineous healthy Spanish parents with no significant family history. She was referred to the Neurology Service at the Hospital Clinic of Barcelona (Barcelona, Spain) in 2017, at age 21, due to progressive gait and balance difficulties. She was delivered via cesarean section at 36 weeks’ gestation, with a birth weight of 1,200 g, consistent with intrauterine growth restriction. Early developmental milestones were delayed, particularly in motor skills (e.g., gait) and speech. She also presented with short stature, mild intellectual disability, attention deficit/hyperactivity disorder and sensorineural hearing loss diagnosed at age 5 that required hearing aids. Although her clinical picture initially appeared nonprogressive, around age 17 she experienced a marked and progressive deterioration in gait and balance, evolving from unassisted walking to using a cane and eventually a wheelchair for long distances. Over time, she also developed mild dysphagia, particularly with liquids, with occasional choking and coughing episodes, especially when distracted.

Neurological evaluation revealed a spastic-ataxic gait with slight inward deviation of both feet, predominantly the left one, requiring assistance. The inward deviation of both feet appears to be more related to the severity of spasticity than to dystonic posture because there were no dystonic spasms and an electromyographic recording did not reveal a co-contraction pattern in the calf muscles. She was able to stand without support in a natural stance but not with her feet together. There was hyperreflexia, bilateral Babinski sign, and inextinguishable clonus at the ankles. Increased tone was noted in the lower limbs, with Ashworth scores of 2–3. Weakness was mild to moderate, predominantly proximal, while upper limb strength remained largely preserved. Speech was characterized by moderate, yet intelligible, scanning dysarthria. Coordination testing revealed bilateral, but left predominant, dysmetria and dysdiadochokinesia, along with end-point tremor during finger-nose test. Heel-shin slide was clearly abnormal bilaterally. Fine motor tasks such as buttoning or tying shoelaces were impaired due to cerebellar incoordination. Ocular pursuit and saccadic movements were normal, and .there was no nystagmus. Muscle bulk was preserved, but she had pes cavus with equinovarus deformity of both feet, and recurvatum of the knees. Skin was normal in appearance, with no signs of ichthyosis.

Routine metabolic and biochemical tests (including lactate levels, ammonia, vitamin E, alpha-fetoprotein, immunoglobulins, ferritin, vitamin B12, and thyroid profile) were normal. Magnetic resonance imaging (MRI) studies of the brain at 21 and 27 years old showed bilateral signal alteration involving thalami, red nuclei and midbrain tegmentum that were hyperintense on T2 and FLAIR, and hypointense in T1-weighted sequences, as well as cerebellar vermian atrophy (Fig. 2). Spinal cord imaging was normal. Electromyography did not evidenced signs of peripheral neuropathy but suggested mild myopathic features in selected muscle groups (e.g., right deltoid), with no widespread findings. However, muscle biopsy ruled out pathological changes.

Brain MRI showing hyperintensities on T2-weighted sequences (a, b) and hypointensities on T1-weighted sequences (c, d) involving both thalami (red arrows), red nuclei (yellow arrows) and, forming a butterfly-shaped image in the midbrain

Firstly, a mitochondrial disorder was discarded by analyzing mitochondrial DNA (mtDNA) variants, mtDNA depletion and mitochondrial complex enzymatic activities. Afterwards, next generation sequencing (NGS) testing was performed focusing on genes associated with HSP and hearing loss, but no pathogenic variants were identified. Finally, trio-based exome sequencing was carried out as previously described (Alvarez-Mora et al. 2023), revealing two heterozygous variants in the FITM2 gene (NM_001080472.4). Segregation analysis by Sanger sequencing confirmed a paternally inherited delins variant in exon 1 (c.158_161delinsTCAT, p.(Arg53_Asn54delinsLeuIle) and a maternally inherited truncating variant in exon 2 (c.567del, p.(Thr190ProfsTer9). Both variants are novel, not reported in public databases and absent from the general population (ClinVar, HGMD and gnomAD v4.1.0, date of consultation 08/04/2025). The frameshift c.567del variant is predicted to truncate approximately 30% of the coding sequence. Although located in the last exon, this variant truncates approximately 30% of the protein removing the last 2 transmembrane domains, regions in which causative variants have previously been reported (Table 1). The delins variant affects two highly conserved residues (based on PhastCons100way and PhyloP100way scores). Existing prediction tools primarily focus on missense variants and this approach is not directly applicable to insertions or deletions (Choi et al., 2012). Following general recommendations for inframe altering variants the c.158_161delinsTCAT variant has been evaluated using PROVEAN and MutationTaster2025. Both suggest a high likelihood of deleterious effects (PROVEAN score: -12.147; MutationTaster2025 deleterious class) 3D structural modeling suggests that this variant may influence local helix packing and/or structural stability of FIT2 (Fig. 3). According to ACMG guidelines (Richards et al. 2015), both variants are classified as likely pathogenic (c.158_161delinsTCAT: PM2, PM3, PP3 and PP4; c.567del: PVS1_Strong, PM2 and PP4) and might explain the clinical manifestations observed in our patient.

Protein prediction of FIT2. The amino acids affected by the R53_N54delinsLI variant are highlighted in orange and green, respectively. a Wild-type structure prediction using TmAlphaFold with six transmembrane (yellow) and topological (grey) domains. The boundaries of the lipid bilayer are indicated by gray planes. b Visual representation of wild-type (blue) and mutant (gray) FIT2 predicted proteins using Alphafold2 and ChimeraX. The predicted local distance difference test (pLDDT) scores for each substitution are 78 and 81.06, respectively, indicating a high confidence for the predicted structure