Understanding brain calcification via N-terminal acetylation at the Golgi apparatus

Abstract

Primary familial brain calcification (PFBC) provides valuable insights into the mechanisms underlying brain calcification as it singles out the proteins that potentially are involved in the relevant cellular pathways. To date, seven genes have been linked to PFBC, and studying their encoded proteins marks an exciting new era in understanding the disease mechanisms, which may ultimately lead to therapeutic strategies to prevent brain calcification. With each new gene found to be associated with PFBC due to pathogenic variants, an additional level of understanding is achieved.

Here, we highlight the most recently discovered PFBC gene, encoding the Golgi-localized N-terminal acetyltransferase NAA60. We explore the novel perspectives gained from the understanding of this enzyme’s molecular, cellular and physiological properties. Interestingly, NAA60 shares a critical role with the most frequent PFBC gene, SLC20A2. Both these proteins seem to be involved in maintaining the structural integrity of the Golgi apparatus, as deficiency in either protein leads to Golgi fragmentation. Altered Golgi morphology is therefore emerging as a new significant topic in PFBC research, and we here discuss this topic in relation to existing knowledge regarding Golgi rearrangements and dysfunction as a factor in neurodegenerative diseases.

Primary familial brain calcification (PFBC) is a genetically heterogeneous disorder marked by bilateral calcium deposition in the brain parenchyma. Siggervåg et al. review the role of NAA60, the most recently identified PFBC gene, and show how studying its encoded protein can shed light on the pathways involved in brain calcification.

Brain calcification

Brain calcification involves deposition of calcium phosphate compounds, mainly hydroxyapatite, in the brain parenchyma.¹ Its occurrence may be secondary to various pathological conditions, such as metabolic disturbances, or it may correspond to a primary manifestation, in which the calcium deposits are directly caused by pathogenic gene variants.² Interestingly, brain calcifications are also relatively common in the general population where the disease frequency increases with age.³ This means that brain calcification frequently coincides with—and may synergistically impact—the prevalent age-related brain disorders with dementia symptoms. As such, brain calcification may offer an attractive perspective to better understand neurodegenerative diseases.

The genetically inherited form of brain calcification, known as primary familial brain calcification (PFBC), is a heterogenic condition associated with a growing number of genes.^2,4,5 Studying PFBC is extremely important for our understanding of the molecular and cellular pathways involved in brain calcification, particularly, in the context of therapeutic advancements. The current list of genes whose pathogenic mutation is associated with PFBC includes SLC20A2, PDGFRB, PDGFB, XPR1, MYORG, JAM2 and NAA60.^6,7 While pathogenic variants in any of these seven genes may cause the same disease, the pathological pathways and cellular mechanisms underlying brain calcification are only just beginning to emerge.

PFBC proteins offer insight into brain calcification mechanisms

The first gene shown to be associated with PFBC, SLC20A2, encodes a protein involved in transport of inorganic phosphate (P_i) at the cell plasma membrane.⁸ This introduced the idea that the occurrence of PFBC is rooted in an imbalance of P_i in the brain. And was later on supported by the linking of another gene encoding a P_i transporter, XPR1, to PFBC.⁹ This correlates with increased P_i levels in the CSF of PFBC-affected individuals.¹⁰ The involvement of the genes PDGFB¹¹ and PDGFRB,¹² which encode the platelet-derived growth factor and its receptor, initially raised the possibility of blood–brain barrier (BBB) dysfunction as a pathogenic mechanism of PFBC. This correlates with the observation of calcifications occurring along brain capillaries.¹³ Involvement of the junctional adhesion molecule JAM2¹⁴ is also compatible with the BBB model, as this protein may contribute to the physical integrity of the neurovascular unit (NVU).^15,16 Furthermore, revealing the astrocyte-enriched MYORG as a PFBC gene elicited a focus on astrocytes,^17,18 and pathological post-mortem analysis of a MYORG-PFBC individual revealed altered astrocytic end-feet morphology in calcified areas.¹⁹ Research along this track identified critical P_i-balancing roles of SLC20A2, XPR1 and MYORG in astrocytes.¹⁸ In addition, MYORG and SLC20A2 were found to be binding partners in the endoplasmic reticulum (ER).¹⁸ This work placed astrocytes as the leading candidate for a shared location where most, if not all, of the PFBC-associated proteins could be functionally connected in a common disease pathway.^7,18 Also emphasizing the role of astrocytes, a PFBC mouse model showed oxidative stress in astrocytes, as well as development of a neurotoxic astrocyte response.²⁰ Addition of MYORG, which is an α-galactosidase, to the list of PFBC proteins, additionally introduced the idea of the significance of post-translational modifications in the PFBC pathway.²¹ This was substantiated by the identification of the most recent PFBC gene encoding the N-terminal modifying enzyme NAA60.⁴ Figure 1 summarizes the PFBC proteins and their key contributions toward understanding the molecular and cellular mechanisms of PFBC.

Figure 1.

Seven proteins are associated with primary familial brain calcification (PFBC), each adding a novel layer of understanding of the underlying molecular mechanisms and cellular processes. The seven proteins are sodium-dependent phosphate transporter 2 (SLC20A2, also known as PiT2), xenotropic and polytropic retrovirus receptor 1 [XPR1; also known as solute carrier family 53 member 1 (SLC53A1)], myogenesis-regulating glycosidase (MYORG), N-alpha-acetyltransferase 60 (NAA60), platelet-derived growth factor receptor beta (PDGFRβ), platelet-derived growth factor subunit B (PDGFB), and junctional adhesion molecule B (JAM2). These have pointed out cellular P_i transport/homeostasis, astrocyte function and blood–brain barrier (BBB) integrity as key factors in the PFBC disease mechanism. SLC20A2 and XPR1 may exert their critical P_i-balancing function in astrocytes. NAA60’s disease mechanism is unknown but is likely to entail loss of protein N-terminal acetylation, and possibly Golgi structural and functional defects, potentially in astrocytes. Ac = acetyl group (-COCH₃); ECM = extracellular matrix; ER = endoplasmic reticulum; P_i = inorganic phosphate. Created in BioRender. Aksnes, H. (2025) https://BioRender.com/47r39gf.

NAA60: N-terminal acetylation of membrane proteins enters the brain calcification stage

The latest addition to the list of PFBC proteins is the N-α-acetyltransferase 60 (NAA60 also known as NatF).⁴ Thus far, 13 individuals with PFBC caused by a total of seven biallelic variants of NAA60 have been described.^4,22,23 NAA60 is a protein-modifying enzyme within the family of N-terminal acetyltransferases (NATs).²⁴ These enzymes catalyse the transfer of an acetyl group from Acetyl-CoA to the N-terminus of proteins.²⁵ Most NATs have multiple targets, as about 80% of human proteins undergo this modification.^26,27 Acetylation of the N-terminus reduces its charge and may alter the properties of the proteins greatly impacting their function.^28,29 Although many NATs are multi-unit complexes with auxiliary subunits, so far only the catalytic unit NAA60 has been identified for NatF.⁴ NAA60 is the only organelle-based human NAT described thus far. Namely, it localizes to the cytosolic side of the Golgi apparatus through two peripherally binding amphipathic α-helices located at the C-terminal tail (Fig. 2A).^30,31 Pathogenic variants of NAA60 causing truncation of the protein C-terminus, are unable to bind to Golgi membranes.^4,22 NAA60’s residency at the Golgi also suggests post-translational protein-modifying activity, in contrast to the majority of the NAT family, which attach to ribosomes. The crystal structure of NAA60 highlights the positioning of the critical Ac-CoA binding motif and the NAT-conserved β6-β7 loop, which is important for anchoring of the substrate protein N-terminus (Fig. 2B and C).^32,33 The structure also suggests homodimerization of NAA60 mediated by the same loop in the absence of a substrate.³² However, it is unknown whether dimerization of the enzyme also occurs in membrane-bound physiological conditions. All the missense variants of NAA60 described so far (L17R, R44C, H131Y and N143T)⁴ have single amino acid substitutions within a conserved catalytic domain shared by all NATs, referred to as the GNAT fold.³⁴ The frameshift/early stop-introducing variants have large truncations in this fold (Fig. 2A, D and E).^4,22

Figure 2.

Overview of functional domains and positions of the seven pathogenic variants of NAA60 that have currently been described. (A) Primary structure of NAA60 with key domains and positions of PFBC variants indicated. (B) Crystal structure of NAA60 amino acids 5–184 (PDB: 5ICV) with bound bisubstrate analogue consisting of an NAA60-type peptide N-terminus (MKAV) bound to acetyl coenzyme A (Ac-CoA) in the NAA60 catalytic cleft. (C) Structure as in B, but colour-coded as in A, highlighting the binding motif for the acetyl donor Ac-CoA (residues 108–113), which is essential for all enzymes in this family. Also note the β6-β7 loop (residues 165–173) important for anchoring of the substrate N-terminus. (D) Structure of wild-type NAA60 with the site of pathogenic variants indicated. Missense variants are labelled in magenta (Leu17, Arg44, His131, Asn143, Asp154), whereas for variants involving frameshift and truncation/early stop codon, the residue where the frameshift occurs is labelled in cyan (Arg108, Gly113). (E) Truncated parts are shown in semi-transparent overlay for two of the variants involving frameshift and truncation. PFBC = primary familial brain calcification. Figure made using PyMOL and Adobe Illustrator.

NAA60 likely has many substrates and it shares a theoretical substrate subclass with NatC and NatE. These enzymes are able to acetylate methionine-starting N-termini, typically followed by hydrophobic or amphipathic-type amino acids (like ML-, MI-, MF-, MY-, MK- and MA-).³⁴ However, NAA60 demonstrates an additional specificity, as it selectively targets membrane proteins with their N-termini facing the cytosol (N-in topology). An N-terminomic analysis of NAA60 knockdown A-431 cells identified 23 NAA60 substrates which localize to membranes of various organelles.³¹ Thus, Golgi residency does not seem to be an essential requirement for a protein undergoing NAA60-mediated Nt-acetylation. With Golgi being a major site for protein modification, NAA60 could be responsible for N-terminally modifying many membrane proteins that pass through this organelle. Initially, NAA60 was found to colocalize extensively with the cis-Golgi marker GM130, leading to the hypothesis that it resides mainly at the cis-face of the Golgi.³¹ Later, the utilization of more advanced microscopic technologies, however, has revealed that NAA60 and GM130 do not perfectly overlap.⁴ A recent Bio-ID analysis searching for proteins with proximity to NAA60 pointed to a possible location in later stage Golgi compartments.³⁵ Subcellular localization of NAA60 is dependent on lipid composition of membranes³⁰ and seems to vary among eukaryotes. In the plant Arabidopsis thaliana, NAA60 resides at the plasma membrane.³⁶

Several pathogenic NAA60 variants entail loss of enzymatic activity.⁴ Therefore, it is reasonable to hypothesize that NAA60-PFBC may be elicited by loss of Nt-acetylation of one or more membrane proteins critical for brain calcium phosphate homeostasis or BBB integrity (Fig. 3, top). The NAA60 substrates identified so far,³¹ may be evaluated as potential novel PFBC genes. Additional putative NAA60 substrates may be predicted based on the following characteristics. The candidate should (i) be a membrane protein; (ii) show N-in topology; and (iii) contain an N-terminal sequence starting with methionine, typically followed by a hydrophobic or amphipathic amino acid.³¹ A functional connection between NAA60 and SLC20A2 has been demonstrated based on the finding that SLC20A2 levels at the plasma membrane are reduced in cells derived from humans with NAA60-PFBC. Moreover, SLC20A2 is Nt-acetylated by NAA60 in vitro.⁴ However, SLC20A2 is not a good theoretical cellular substrate of NAA60 due to its N-terminal topology,³⁷ and a possible functional connection of these proteins needs further investigation. Identification of novel NAA60 substrates, for example using N-terminomics,^38-40 or Nt-acetylation-specific antibodies,⁴¹ may therefore be a valuable approach to further decipher the molecular underpinnings of brain calcification.

Figure 3.

Two alternative or overlapping hypothetical models of molecular and cellular processes underlying NAA60-PFBC.  Top: Loss of NAA60-mediated Nt-acetylation of key PFBC proteins (or their close associates) may abolish or impair their cellular function in maintaining the structural integrity of the neurovascular unit/blood–brain barrier or in balancing brain P_i homeostasis. A potential molecular disease mechanism might involve the galactosidase MYORG, since several of the PFBC proteins are glycan modified. Bottom: Loss of NAA60-mediated Nt-acetylation of protein(s) important for the structural integrity of the Golgi apparatus cause Golgi fragmentation (perhaps due to loss of protein-protein interactions), which affects its functions such as trafficking and glycan processing, resulting in mislocalization or defective function of key PFBC proteins and cellular P_i imbalance. Notably, P_i imbalance may reciprocally affect Golgi structure. Ac = N-terminal acetyl group; BBB = blood–brain barrier; Met = methionine; PFBC = primary familial brain calcification; ɸ = hydrophobic amino acid; ψ = amphipathic amino acid; hexagon indicates sugar group of oligosaccharides; red zigzag icon and red inhibition icon indicate disrupted molecular interactions and processes. Created in BioRender. Aksnes, H. (2025) https://BioRender.com/uy2xvkq.

Recent advances have provided a peek into the molecular mechanisms of PFBC and suggest possible treatments for targeting PFBC via increasing SLC20A2 protein levels with viral-based gene therapy or oligonucleotide approaches.^7,18,42 Given that NAA60-PFBC involves a reduced cell surface level of SLC20A2,⁴ potential future SLC20A2-boosting treatments might be considered also for individuals with NAA60-PFBC. Antisense oligonucleotide treatment to stabilize gene expression could potentially be relevant for all intronic variants involving aberrant mRNA processing.⁴² Thus far, one intronic variant has been described for NAA60.⁴

The functions of NAA60 in cells and model organisms

Only a few cellular functions of NAA60 have been described. In Drosophila melanogaster cells, knockdown of NAA60 led to lagging chromosomes and chromosome bridges during anaphase.²⁴ In cells derived from humans with NAA60-PFBC, a defect in the balance of intra/extracellular P_i was identified,⁴ and impaired phosphate homeostasis was further demonstrated in HEK293FT NAA60 knockout (KO) cells.²² Additionally, loss of NAA60 in A. thaliana caused reduced tolerance of the plant cells to high salt conditions.³⁶ Whether this plant phenotype shares some common underlying biology with defective brain homeostasis of calcium and phosphate remains unknown. However, a possible common factor may lie in NAA60 having a direct or indirect role in determining the proper function of ion transporters at the plasma membrane. Further connection between NAA60 and functionality of plasma membrane proteins was suggested by the finding that NAA60 is necessary for productive influenza A virus infection. Here, NAA60 depletion reduced virus growth via IFNα signalling, whereas the opposite was observed for NAA60 overexpression.⁴³

An Naa60 KO and Naa60 deleterious knock-in mouse model exhibited reduced body weight compared to wild-type mice at 3–5 weeks postpartum. Surprisingly, however, no brain calcifications were identified during the 19-month period of examination.²²  Pdgrfb KO and Pdgfb KO mice both die perinatally, but a homozygous Pdgfb mutant mouse model with only partially lost protein function show brain calcifications after 4 months.⁴⁴ Heterozygous Xpr1 KO mice show no calcification after 12 months.¹¹ Calcifications are detected in Slc20a2 KO after 5 months,⁴⁵ and in Myorg KO mice after 9 months.¹⁷ Thus, 19 months should cover the typical brain calcification window in PFBC mice models.

In humans, although PFBC symptoms vary greatly between individuals,^2,5 it may seem like NAA60-PFBC can have a relatively early onset. Among the few NAA60-PFBC individuals described so far, the median age at examination (22 years, n = 11) is much lower compared to other PFBC groups: SLC20A2 (51 years, n = 240), PDGFB (35 years, n = 54), PDGFRβ (46 years, n = 23), XPR1 (45,5 years, n = 24), MYORG (47 years, n = 66), JAM2 (37 years, n = 11).^2,4,22

NAA60 has been linked to Golgi maintenance based on the finding that NAA60 knockdown,^4,31 and to some extent knockout,³⁵ causes Golgi fragmentation. Thus, rather than pathogenic NAA60 variants causing PFBC via loss of Nt-acetylation of a PFBC pathway protein (Fig. 3, top), an alternative possibility is that NAA60-PFBC manifests through a structural collapse of the Golgi, which may disturb its functions in a manner that affects PFBC proteins (Fig. 3, bottom). Interestingly, Golgi fragmentation has recently been connected to another of the seven brain calcification proteins, SLC20A2.⁴⁶ This strengthens the idea that Golgi structural damage or impaired functionality may be involved in the pathological accumulation of hydroxyapatite deposits in the brain. As discussed below, this offers a new cellular perspective to understand brain calcification, which seems especially relevant considering that Golgi fragmentation and dysfunction has been suggested to play an important role in neurodegenerative diseases.^47-49

Golgi alterations in PFBC and neurodegenerative disease

The Golgi apparatus is characterized by its enigmatic stacked cisternal organization and positioning in the perinuclear region close to the centrosome. Operating at the intersection of the secretory and endocytic pathways, the Golgi plays a key role in post-translational processing of proteins and lipids, as well as sorting of the modified molecules into transport carriers that mediate their delivery to the cell surface or various intracellular destinations.^50,51 The best characterized biochemical activity of the Golgi deals with glycosylation, specifically the terminal processing of protein- and lipid-bound oligosaccharide side chains. Additional functions of the Golgi include protein phosphorylation and sulfation, as well as the more recently discovered N-terminal acetylation of membrane proteins.³¹

Mammalian cells contain multiple Golgi stacks that are laterally joined into a continuous Golgi ribbon (Fig. 3, top). Consequently, this extensive convoluted structure consists of the alternating cisternal Golgi stacks (compact zones) and the intervening linker regions (non-compact zones). A commonly held view is that the stacks are laterally connected via narrow membrane tubules.⁵² Alternatively, it has been proposed that the central domains of biosynthetic (pre-Golgi) and endocytic (post-Golgi) membrane networks meet at the non-compact zones of the Golgi ribbon and establish ‘linker compartments’ that communicate with each other and also dynamically connect the Golgi stacks (Fig. 3, top).⁵³

Golgi fragmentation is a process which involves reversible or irreversible unlinking of the interconnected stacks in the Golgi ribbon and their scattering throughout the cytoplasm. In some instances, it is accompanied by further morphological alteration or disassembly of the separated Golgi stacks. As part of normal cell physiology, Golgi fragmentation takes place during cell division (onset of mitosis), migration, polarization and differentiation, as well as autophagy.^52,53 It is also connected to cellular stress and apoptosis⁵⁴ and, accordingly, has also been observed in various pathological conditions.⁴⁹ For instance, it has been associated with neurodegenerative disorders like Alzheimer’s and Parkinson’s diseases, and amyotrophic lateral sclerosis.^49,55,56 The primary cause in these cases appears to involve alterations in intracellular transport, particularly affecting the early secretory pathway.⁵⁷ Vice versa, the fragmented Golgi may be functionally affected in several ways; for instance, in terms of the disruption of its normal protein processing activities and alterations in speed and accuracy of protein trafficking and secretion.^58-60 Of note, recent studies suggest that Golgi fragmentation may be an early event during the development of Alzheimer’s disease.⁶¹

Golgi fragmentation was recently connected to brain calcification, based on studies of two PFBC proteins.^{4  ,46} Knockdown of the most recently identified PFBC protein, NAA60, in HeLa and human primary dermal fibroblasts caused Golgi fragmentation.^4,31 Thus, a hypothesis for the aetiology of NAA60-PFBC might be that its occurrence is due to disrupted Golgi structure and function (Fig. 3, bottom). The study using HeLa cells did not reveal any changes in the relative localizations of the cis- and medial-Golgi markers GM130 and ManII, indicating that the cisternal stacks remain intact after NAA60 knockdown.³¹ Therefore, the Golgi scattering effect observed in NAA60-depleted cells is most likely due to unlinking of the Golgi ribbon, potentially reflecting dysregulation of the molecular machineries operating in ribbon integrity at the non-compact zones (Fig. 3, bottom). Since NAA60 mainly acetylates transmembrane proteins, and Nt-acetylation may impact protein-protein interactions,^28,29,62,63 it can be hypothesized that ribbon unlinking might be caused by the absence of Nt-acetylation of an NAA60 substrate that is specifically located at the non-compact zones. This defect may prevent it from participating in protein-protein interactions that are crucial for ribbon formation (Fig. 3, bottom).^31,64

For SLC20A2, disruption of Golgi structure was observed in induced pluripotent stem cell (iPSC)-derived astrocytes and neurons generated from cells from two PFBC-diagnosed individuals with different SLC20A2 variants.⁴⁶ The study suggests that Golgi alteration is a direct consequence of dysfunction of the SLC20A2 protein. Moreover, the study reported impaired autophagy as indicated by accumulation of the autophagy receptor p62 and decreased levels of GM130 and GRASP65 (also called GORASP1) in the SLC20A2-PFBC-derived cells. GM130 and GRASP65 are Golgi-associated proteins with important roles in Golgi ribbon formation.^65,66 Thus, SLC20A2 might participate in the regulation of Golgi proteins with crucial roles in maintaining Golgi ribbon integrity, such as these. The GRASP proteins GRASP65 and GRASP55 (also called GORASP1 and GORASP2, respectively) are involved in the linking of cisterna stacks, and their depletion has been shown to cause Golgi fragmentation.^67-71 This demonstrates the sensitivity of Golgi structure to the depletion of proteins involved in maintaining its integrity. Consequently, any impact on Golgi-associated proteins, whether through direct modification or indirect influence, could explain the Golgi phenotype observed in the SLC20A2-PFBC-derived cells.

Since disrupted cellular P_i homeostasis and Golgi fragmentation are shared phenotypes between the cells lacking either SLC20A2 or NAA60,^4,22,46,72 these processes could be intertwined in the cellular aetiology of PFBC. It has been suggested that dysregulation of cellular P_i concentration in PFBC may lead to Golgi fragmentation through impaired autophagy, mitochondrial dysfunction or apoptosis.⁴⁶ Interestingly, phosphate homeostasis was recently linked to the regulation of Golgi functions, based on studies using Erd1 deletion yeast. Erd1 is a yeast homologue of XPR1 where it functions as a Golgi-localized P_i exporter protein. P_i accumulation in the Golgi lumen of Erd1 deletion yeast under certain conditions impairs protein transport at the Golgi level.⁷³ Intriguingly, organellar functions have also been suggested for the human XPR1, which was shown to regulate cellular P_i fluxes by binding to SLC20A1 and limiting its presence on the PM via endocytosis (Fig. 3, bottom).⁷⁴ Additionally, the Golgi-associated protein UBIAD1 was found to decrease vascular cell differentiation and calcification, thereby linking the Golgi apparatus to calcification and perhaps to NVU function.⁷⁵ Another point worth noting is that the Golgi apparatus is known to participate in intracellular storage of Ca²⁺—an important signalling molecule—and the regulation of cytoplasmic Ca²⁺ levels.^{76  ,77} Furthermore, Ca²⁺ is crucial for normal Golgi function, and disrupted Ca²⁺ homeostasis has been linked to neurodegenerative disease and brain ageing.^78,79

Golgi fragmentation in PFBC could be a cell-specific effect. For NAA60, Golgi fragmentation has been investigated in HeLa and HAP1 cells, as well as in primary dermal fibroblasts.^4,31,35 The connection between Golgi fragmentation and SLC20A2-deficiency, on the other hand, was observed in iPSC-derived astrocytes and neurons.⁴⁶ Since the NVU model has been central in PFBC literature,^16,80-82 with a recent focus on astrocytes,^7,18 the Golgi scattering effect seen in cells from individuals with SLC20A2-PFBC might be specific to cells of the NVU. Neurons and astrocytes also display different distributions of the Golgi, where, in addition to the Golgi in the cell body, so-called Golgi outposts may be localized to dendrites and astrocyte end-feet.^83-85 Interestingly, astrocyte end-feet seem particularly relevant in the cellular disease mechanism of PFBC, since XPR1 was found to be non-uniformly distributed, preferentially localizing to the astrocyte end-feet associating with blood vessels.¹⁸ Furthermore, SLC20A2 and MYORG were found to be binding partners in the ER of astrocytes.¹⁸ MYORG loss of function shares the P_i imbalance phenotype, but Golgi morphology in cells carrying a pathogenic variant of MYORG has not yet been assessed. MYORG is an α-galactosidase, trimming the sugar groups on glycoproteins, perhaps processing some of the other PFBC proteins (Fig. 3).^17,21 This is interesting to consider in connection with the notion that Golgi fragmentation evidently disturbs many of its functions, including glycan processing.⁵⁸ Additionally, the XPR1 yeast homologue Erd1 was suggested to function in recycling P_i byproducts of glycosylation from the Golgi,⁸⁶ as well as to be involved in trafficking of glycosyltransferases.⁸⁷ Taken together, disrupted processing of glycans may therefore be a central feature of brain calcification, a topic that would deserve further experimental attention. Interestingly, a chaperone-activity drug already used to stabilize another α-galactosidase, whose deficiency causes Fabry disease, was shown to restore the in vitro activity of pathogenic variants of MYORG.^7,21 However, it remains to be investigated whether PFBC-relevant glycan processing, as well as P_i homeostasis, is restored by this drug in more physiological models.

The emerging connections between Golgi function, P_i homeostasis and calcification support the theory that Golgi fragmentation and P_i imbalance may jointly participate in the pathogenic mechanisms of PFBC. Nevertheless, their exact roles remain unknown and further exploration of this link will be of great interest.

Additional NAA60-related neuropathology

The identification of NAA60 as a gene associated with PFBC provided the first definitive disease-link. However, a couple of previous studies had already hinted at NAA60’s involvement in various pathologies, including neurological disorders. A genetic screen in Drosophila indicated that NAA60 might play a role in cellular pathways relevant to neuronal ceroid lipofuscinosis (NCL), also known as Batten disease.⁸⁸ NCL is a hereditary neurodegenerative disorder that may manifest at around 6 months of age. The condition is characterized by a shortened lifespan, progressive vision loss, seizures, cognitive decline, motoric dysfunction and changes in personality and behaviour.⁸⁹ Interestingly, some of these symptoms overlap with those seen in PFBC and align with observations in young individuals with NAA60-related PFBC.⁴ NCL is caused by the accumulation of lipopigments (ceroid lipofuscin) due to lysosomal failure in removing this waste product.⁸⁹ This process may be sensitive to the correct modification of membrane proteins, as well as normal structure and function of the Golgi apparatus.

NAA60 is subject to genetic imprinting by means of a differentially methylated region (DMR) acting as a bidirectional paternal silencer for NAA60 and a neighbouring gene ZNF597.⁹⁰ The ZNF597/NAA60-DMR was identified as one of multiple loci linked to imprinting disturbances in a subset of individuals with a certain imprinting disease affecting growth and motoric development.⁹¹ The restricted maternal expression of NAA60, however, only implies an alternative isoform not yet characterized at protein level, whereas isoform 1, which is considered the canonical version, is biallelically expressed.⁹⁰  NAA60 was also flagged in a twin study examining the effect of body mass index (BMI)-related DNA methylation variation on blood pressure,⁹² and in a genome-wide study of fibromyalgia looking for differentially methylated sites.⁹³ Fibromyalgia has a prevalence of 2%–5% and is characterized by widespread muscle pain, fatigue and unrefreshing sleep.⁹⁴ Although not much is known about its cause, fibromyalgia is believed to involve an increased sensitivity to pain brought on by a disruption in CNS function and is shown to involve genes implicated in stress response epigenetically.^95,96 NAA60 may thus be subject to regulation that, if disturbed, may be pathogenic in various ways. Potential molecular mechanisms of regulation upstream of NAA60 have not been assessed. Nevertheless, it seems possible that such regulatory factors may exist, given that NAA60 has high gene expression variability during different phases of early human development.⁹⁷

Conclusive remarks

Many people diagnosed with PFBC still remain without a genetic explanation, suggesting that pathogenic variants in additional genes may exist.^98,99 The loss of NAA60 function in PFBC provides valuable insight in the search for new PFBC genes, as potential candidates may be found among validated or predicted substrates of NAA60. PFBC is also considered underestimated and underdiagnosed.⁹⁸ Thus, with evolving genetics,¹⁰⁰ new diagnostic approaches,¹⁰¹ and prospective treatment options,^7,18,42,102 the future of PFBC research is intriguing.

Identification of novel PFBC genes offers new layers of insight into molecular disease mechanisms. We here highlighted NAA60’s contribution to the understanding of the cellular and molecular pathways underlying brain calcification. Notably, NAA60 seems to be functionally interconnected with at least one other PFBC gene, which may have implications for personalized treatment strategies.

Golgi fragmentation has emerged as a new phenotype in the literature on brain calcification. Although research is still in its early stages, there is emerging evidence that P_i dysregulation and Golgi fragmentation may be interconnected. Changes in Golgi architecture represent a relevant and measurable phenotype that can be further explored in brain calcification research.

NAA60 may be subject to regulatory mechanisms and its dysregulation could be associated with other neurological diseases. Taken together, a deeper understanding of NAA60’s vital role in the brain has the potential to advance our knowledge significantly, shape research efforts and ultimately improve diagnosis and treatment for various neurodegenerative and other neuropathological conditions.

Funding

The MemBrain research group at the University of Bergen acknowledges funding from the Trond Mohn Foundation (TMF), grant no TMS2024STG01 to H.A.

Competing interests

The authors report no competing interests.