The cause: a missing part of the AP-4 delivery machine
AP-4 (adaptor protein complex 4) is made of four subunits, built by four genes: AP4E1 (SPG51), AP4B1 (SPG47), AP4M1 (SPG50) and AP4S1 (SPG52). All four are needed. In skin cells from a child with two faulty AP4S1 copies, the other three subunits were greatly reduced and no working complex formed; the authors concluded AP-4 function is "likely totally lost" [3].
AP-4's most important known job is to carry a protein called ATG9A from the cell's packaging centre (the Golgi) out to the far ends of nerve cells. ATG9A is a core part of the cell's recycling system (autophagy). Without AP-4, ATG9A piles up in the Golgi and is missing where it is needed [4,5]. Neurons suffer most because their fibres (axons) are so long: in an AP-4 knockout mouse the corpus callosum is thin, axons swell, and protein aggregates build up in them [7]. The working model is that without AP-4, recycling fails at the far end of the axon and the axon degenerates [6]. The longest axons run to the legs, which is why the legs are affected first and most.
Two useful consequences follow. The "ATG9A ratio" (how much ATG9A is stuck in the Golgi) is now a research test on patient skin cells that can confirm or reclassify an uncertain variant in about four to six weeks [8]. And the same readout is the screening test behind the drug-discovery work described on the research page [6].
How it is inherited
SPG52 is autosomal recessive: a child is affected only when both copies of AP4S1 are faulty, one from each parent. Each parent carries one faulty copy and is healthy. Each pregnancy of two carrier parents has a 25% chance of an affected child. Once a family's variants are known, prenatal and preimplantation testing are possible [2]. Compared with the other three AP-4 genes, AP4S1 patients were the least likely to come from related (consanguineous) parents, 47%, which fits with more children carrying two different variants [1].
Signs and timeline
Low muscle tone (floppiness) in infancy is the usual first sign: 92% (244 of 264) in the Boston registry [2]. The mean age at which symptoms are first noticed is 0.8 years [1]. Motor milestones are delayed in 99% [2]. Median ages in the registry: sitting 12 months, crawling 18, standing without support 24, walking with support 25. About 45% ever walk independently, at a median of 30 months [2].
Head growth slows after birth: 83% of children develop a small head size (microcephaly) after birth; only 8% are born with it [1]. For AP4S1 specifically the figure is 72%, the lowest of the four genes [1]. A normal head size at birth does not argue against the diagnosis.
Floppiness turns into stiffness (spasticity) in the legs during early childhood: 60% of children under 4 and 96% of those 4 and older have leg spasticity [2]. Brisk reflexes (92%) and an upgoing toe sign (88%) usually come before the stiffness [1]. The arms are involved later and more mildly, at a mean age of 16 [1].
Walking: at last follow-up in the largest published cohort, 11% walked unsupported, 36% walked with assistance (mean age 8.9) and 54% used a wheelchair. The mean age of becoming wheelchair-dependent was 9.0 years, with a very wide spread (standard deviation 5.3 years) [1]. The authors place the period of significant motor progression between ages 5 and 16 [1].
Speech is significantly impaired or absent in 97% [2]; about half are nonverbal [1]. Intellectual disability was recorded as severe in 58%, moderate in 33% and mild in 9% [1]; receptive understanding is not separately measured in the cohorts, and caregivers rate communication, comfort and health as what matters most [21]. GeneReviews recommends starting augmentative communication (AAC) early [2].
Seizures: about two thirds have had a seizure (67%; 61% for SPG52) and about 60% develop epilepsy (53% for SPG52) [1,2]. Fever is the trigger: about 70% of registry patients have had seizures with fever, often prolonged, in the first two years [2]. Status epilepticus (a seizure lasting more than five minutes, an emergency) occurred in 37%, "often at first seizure" [1]. The course is usually favourable: complete control on medication in 69%, seizures becoming less frequent with age, and many patients stopping medication [1,2]. In the only cohort that was mostly SPG52, no patient met criteria for drug resistance [17].
| Age | What the cohorts show |
|---|---|
| Birth to 12 months | Low tone (92%), slow milestones (99%), head growth slowing, brisk reflexes often already present. Fever-triggered seizures can start in the first year. |
| 1 to 3 years | Sitting (median 12 months), crawling (18), standing (24), supported walking (25); about 45% walk alone at a median of 30 months. Most seizure onsets happen before 3. Speech delay obvious. |
| 4 to 6 years | Leg spasticity in 96% from age 4. Foot deformities, drooling, toileting delays. |
| 5 to 16 years | The window of "significant progression": walking with assistance (mean 8.9 years), then wheelchair (mean 9.0, standard deviation 5.3). Seizures usually become less frequent. |
| Adolescence to adulthood | Arm involvement (mean onset 16.1 years), movement-disorder signs, late cognitive decline in advanced disease. Oldest reported person: 55. |
The MRI pattern
The best imaging study read 107 MRIs from 76 patients systematically [20]. Four core features: a thin corpus callosum, worst at the back (94%); an absent (82%) or thin (14%) anterior commissure, a small second bridge between the hemispheres that is relatively specific to AP-4; white matter changes around the ventricles (80%), including the "ears of the grizzly bear" sign in the forceps minor (48%); and enlarged ventricles at the back of the brain (65% in the 2020 cohort) [1,20]. At least two of the four core features were present in 99% of patients, all four in 45%, so an experienced neuroradiologist can suspect AP-4 from the scan alone [20].
Delayed myelination was unchanged or caught up in 16 of 17 children with repeat scans; the authors conclude this is mainly a development problem, not ongoing destruction [20]. What does progress is white matter volume loss, which correlates with motor severity [20,23]. Polymicrogyria (too many small folds in the cortex) was found in 26% when looked for systematically, and those children had more and longer seizures [20]. Changes are visible before birth: 23% of prenatal ultrasounds in affected pregnancies showed abnormalities, usually enlarged ventricles [23]. A normal MRI does not rule out SPG52: siblings in two families had normal scans [11,15].
The four AP-4 types behave the same
The authors of the largest study concluded that "disease severity and major phenotypes were equally distributed among the four subtypes" and proposed one "AP-4 deficiency syndrome" [1]. No genotype-phenotype correlation has been reported for any of the four genes [2]. Small hints that SPG52 may sit at the milder end exist (lower microcephaly rate, fewer drug-resistant seizures) but none reach significance, and the SPG52 numbers are small: 23 of 156 patients in the 2020 cohort [1,17].
How many children are known
About 43 molecularly confirmed SPG52 cases were in the medical literature as of 2025 [11]. The Barcelona research team cites about 50 to 55 known children worldwide [25,26]. SPG52 is the rarest of the four AP-4 types: 15% of the 2020 cohort, against 34% for SPG47 and 38% for SPG50 [1]. The Boston registry for all early-onset hereditary spastic paraplegias held 303 people from 289 families in the 2025 GeneReviews update [2].
Diagnosis is late
Mean age at genetic diagnosis was 10.2 years in the 2020 cohort, about ten years after the first signs; 10.4 years for the 23 AP4S1 patients [1]. The Boston team says the typical journey is still "three to four years to a diagnosis" unless an older sibling is already known [9]. Children are very often labelled "cerebral palsy" first [2,10,11,12], and in one SPG52 family two siblings had spinal surgery for a presumed tethered cord before the gene was found [11]. Genetic testing (an exome, a genome, or an HSP gene panel) is what settles it, and children diagnosed today are increasingly identified at one to two years of age.
Everything else, with numbers where they exist
- Drooling: 70% [1].
- Foot deformities: 50% (most often clubfoot) [2]; 69% in the 2020 cohort [1]. Contractures 50%, mainly ankles and knees [1].
- Swallowing difficulty 24%, feeding tube 5% [2]; swallowing dysfunction or aspiration 29% [1].
- Short stature: height averaged 1.69 standard deviations below the mean, 35% below minus 2 [1].
- Urinary incontinence 81%, stool incontinence 72% [2].
- Reduced pain sensation: 38% [2], so injuries and infections can show as irritability rather than complaint.
- Stereotypic laughter (bursts of laughter without an obvious trigger): 46% in the registry, 62% of AP4S1 patients [1,2].
- Movement-disorder signs (dystonia, slowness, tremor): 28% of the registry [2].
- Scoliosis, hip displacement, strabismus, sleep, dental, breathing and life expectancy: no AP-4 rates are published. Hip and spine X-rays are recommended yearly from diagnosis [2].
- The oldest reported person with AP-4-HSP is 55 [2].
Sources
Numbers in brackets refer to this list. Last reviewed: 2026-10-21. Sources checked on 2026-10-20.
- [1] Ebrahimi-Fakhari D, Teinert J, Behne R, et al. Defining the clinical, molecular and imaging spectrum of adaptor protein complex 4-associated hereditary spastic paraplegia. Brain. 2020;143(10):2929-2944. PMID 32979048
- [2] Ebrahimi-Fakhari D, Behne R, Davies AK, Hirst J. AP-4-Associated Hereditary Spastic Paraplegia. GeneReviews, University of Washington. Last update February 6, 2025. www.ncbi.nlm.nih.gov/books/NBK535153
- [3] Hardies K, May P, Djémié T, et al. Recessive loss-of-function mutations in AP4S1 cause mild fever-sensitive seizures, developmental delay and spastic paraplegia through loss of AP-4 complex assembly. Hum Mol Genet. 2015;24(8):2218-2227. www.ncbi.nlm.nih.gov/pmc/articles/PMC4380070
- [4] Mattera R, Park SY, De Pace R, Guardia CM, Bonifacino JS. AP-4 mediates export of ATG9A from the trans-Golgi network to promote autophagosome formation. Proc Natl Acad Sci USA. 2017;114(50):E10697-E10706. PMID 29180427
- [5] Davies AK, Itzhak DN, Edgar JR, et al. AP-4 vesicles contribute to spatial control of autophagy via RUSC-dependent peripheral delivery of ATG9A. Nat Commun. 2018;9:3958. PMID 30262884
- [6] Saffari A, Brechmann B, Böger C, et al. High-content screening identifies a small molecule that restores AP-4-dependent protein trafficking in neuronal models of AP-4-associated hereditary spastic paraplegia. Nat Commun. 2024;15:584. PMID 38233389
- [7] De Pace R, Skirzewski M, Damme M, et al. Altered distribution of ATG9A and accumulation of axonal aggregates in neurons from a mouse model of AP-4 deficiency syndrome. PLoS Genet. 2018;14(4):e1007363. PMID 29698489
- [8] Agianda HAP, Kim HM, Battaglia N, et al. Diagnostic utility of the ATG9A ratio in AP-4-associated hereditary spastic paraplegia. Ann Clin Transl Neurol. 2026;13(4):834-839. PMID 41491634
- [9] Fliesler N. The journey to a treatment for hereditary spastic paraplegia. Boston Children's Answers, October 14, 2025. answers.childrenshospital.org/hereditary-spastic-paraplegia-treatment
- [10] Tessa A, Battini R, Rubegni A, et al. Identification of mutations in AP4S1/SPG52 through next generation sequencing in three families. Eur J Neurol. 2016;23(10):1580-1587. PMID 27444738
- [11] Günay Ç, Gazeteci Tekin H. Hereditary spastic paraplegy associated with the AP4S1 gene: a case series highlighting diagnostic pitfalls and phenotypic variability. Mol Syndromol. 2025;16(5):489-497. PMID 41059447
- [12] D'Amore A, Tessa A, Naef V, et al. Loss of ap4s1 in zebrafish leads to neurodevelopmental defects resembling spastic paraplegia 52. Ann Clin Transl Neurol. 2020;7(4):584-589. PMID 32216065
- [13] Teinert J, Behne R, Wimmer M, et al. The clinical, molecular and radiographic spectrum of AP-4-HSP: results from the AP-4-HSP International Registry. Mov Disord. 2019;34(suppl 2), abstract 1507. www.mdsabstracts.org/abstract/the-clinical-molecular-and-radiographic-spectrum-of-adaptor-protein-complex-4-associated-hereditary-spastic-paraplegia-ap-4-hsp-results-from-the-ap-4-hsp-international-registry
- [14] Abou Jamra R, Philippe O, Raas-Rothschild A, et al. Adaptor protein complex 4 deficiency causes severe autosomal-recessive intellectual disability, progressive spastic paraplegia, shy character, and short stature. Am J Hum Genet. 2011;88(6):788-795. PMID 21620353
- [15] Niyoyita S, Uwibambe E, Ndinkabandi J, et al. A rare homozygous AP4S1 variant in Rwandan siblings with autosomal recessive hereditary spastic paraplegia type 52 (SPG52). Genes (Basel). 2025;16(5):542. PMID 40428364
- [16] Abdollahpour H, Alawi M, Kortüm F, et al. An AP4B1 frameshift mutation in siblings with intellectual disability and spastic tetraplegia further delineates the AP-4 deficiency syndrome. Eur J Hum Genet. 2015;23(2):256-259. PMID 24781758
- [17] Bartolini E, Ferrari AR, Santorelli FM, et al. Combined generalized and focal epilepsy with reflex features in adaptor protein complex 4-associated hereditary spastic paraplegias: a cohort observational study. Seizure. 2024;121:186-193. PMID 39208719
- [20] Ebrahimi-Fakhari D, Alecu JE, Ziegler M, et al. Systematic analysis of brain MRI findings in adaptor protein complex 4-associated hereditary spastic paraplegia. Neurology. 2021;97(19):e1942-e1954. PMID 34544818
- [23] Brooks AK, Quiroz V, Schierbaum L, et al. Fetal and perinatal brain MRI findings in adaptor protein complex 4-associated hereditary spastic paraplegia. Ann Child Neurol Soc. 2025;3(3):226-231. PMID 41815511
- [25] Universitat Autònoma de Barcelona. First positive results of gene therapy for SPG52, suffered by three children in Spain. UAB Newsroom, December 9, 2024. www.uab.cat/web/newsroom/news-detail/first-positive-results-of-gene-therapy-for-spg52-suffered-by-three-children-in-spain-1345830290613.html?detid=1345938510885
- [26] Universitat Autònoma de Barcelona. Over 600,000 euros donated to the SPG52 gene therapy study thanks to crowdfunding campaign. UAB Newsroom, May 6, 2025. www.uab.cat/web/newsroom/news-detail/over-600-000-donated-to-the-spg52-gene-therapy-study-thanks-to-crowdfunding-campaign-1345830290613.html?detid=1345952770000
- [27] Alecu JE, Saffari A, Ziegler M, et al. Plasma neurofilament light chain is elevated in adaptor protein complex 4-related hereditary spastic paraplegia. Mov Disord. 2023;38(9):1742-1750. www.ncbi.nlm.nih.gov/pmc/articles/PMC10529494