ARCN1 Mutations Cause a Recognizable Craniofacial Syndrome Due to COPI-Mediated Transport Defects

Abstract

Cellular homeostasis is maintained by the highly organized cooperation of intracellular trafficking systems, including COPI, COPII, and clathrin complexes. COPI is a coatomer protein complex responsible for intracellular protein transport between the endoplasmic reticulum and the Golgi apparatus. The importance of such intracellular transport mechanisms is underscored by the various disorders, including skeletal disorders such as cranio-lenticulo-sutural dysplasia and osteogenesis imperfect, caused by mutations in the COPII coatomer complex. In this article, we report a clinically recognizable craniofacial disorder characterized by facial dysmorphisms, severe micrognathia, rhizomelic shortening, microcephalic dwarfism, and mild developmental delay due to loss-of-function heterozygous mutations in ARCN1, which encodes the coatomer subunit delta of COPI. ARCN1 mutant cell lines were revealed to have endoplasmic reticulum stress, suggesting the involvement of ER stress response in the pathogenesis of this disorder. Given that ARCN1 deficiency causes defective type I collagen transport, reduction of collagen secretion represents the likely mechanism underlying the skeletal phenotype that characterizes this condition. Our findings demonstrate the importance of COPI-mediated transport in human development, including skeletogenesis and brain growth.

Main Text

In eukaryotic cells, secretory and membrane proteins are generally synthesized on rough endoplasmic reticulum (ER). Vesicle trafficking is required for the correct intracellular transport of these proteins. There are three main intracellular protein trafficking systems: COPI, COPII, and clathrin systems.¹ These transport systems are composed of heteromeric proteins forming a lattice-like protein complex that coats vesicles for intracellular transport. COPI is a heptameric protein complex composed of alpha-COP, beta-COP, beta-prime-COP, gamma-COP, delta-COP, epsilon-COP, and zeta-COP subunits.^{1, 2} COPI plays an important role in retrograde transport from the Golgi apparatus to the ER; however, recent reports also implicate COPI in anterograde transport from the ER to the Golgi apparatus.^{1, 2, 3} In order for cells to perform highly complex cellular functions, intracellular transport achieved by COPI plays cardinal roles in cellular homeostasis.

Disruption of intracellular protein transport in human disease has been documented in various genetic disorders. COPII plays major roles in anterograde transport from the ER to the Golgi apparatus,¹ and mutations in COPII components are associated with several genetic disorders. These disorders include lipid absorption disorder (MIM: 246700),⁴ skeletal disorders (such as cranio-lenticulo-sutural dysplasia [MIM: 607812]⁵ and osteogenesis imperfecta [MIM: 616294]⁶), hematological disorders (such as congenital dyserythropoietic anemia type II [MIM: 224100)⁷ and combined deficiency of factor V and factor VIII [MIM: 227300, 613625]^{8, 9}), Cowden syndrome, and thyroid cancer (MIM: 616858).¹⁰ However, in contrast to the findings for COPII, disease-causing germline mutations in COPI components had not been described until lately. Recently, germline mutations of COPA encoding the alpha-COP subunit of COPI have been described in individuals with hereditary autoimmune-mediated lung disease and arthritis, demonstrating that the disruption of COPI transport can cause human disease (MIM: 616414).¹¹ Here, we report a clinically recognizable genetic disorder that we identified in four individuals; this disorder is characterized by facial dysmorphisms, severe micrognathia, rhizomelic shortening, microcephalic dwarfism, and mild developmental delay caused by germline loss-of-function mutations in archain 1 (ARCN1), which encodes the coatomer subunit delta of COPI.

Four individuals with ARCN1 mutations were independently enrolled in our study after approval by the institutional review boards of The University of Tokyo, Nagano Children’s Hospital, KK Women’s and Children’s Hospital, and the genetics departments of the Centre Hospitalier Universitaire and the University of Liège. Informed consent for genetic and molecular studies was obtained from the guardians of the subjects. Shared clinical features among these four individuals are facial dysmorphisms, including severe micrognathia, rhizomelic shortening, and microcephalic dwarfism (see case reports in the Supplemental Note, Figure 1, Table 1). Subject 1 was referred to the genetics clinic of Nagano Children’s Hospital at the age of 1 month with the chief complaints of facial dysmorphism and intrauterine growth retardation. On physical examination, several facial dysmorphisms, including a prominent forehead, downslanted palpebral fissures, and severe micrognathia, were noted. In addition, shortening of upper arms and legs and small joint laxity were observed (Figure 1A). A skeletal survey revealed widening of the metaphysis and a wide femoral neck (Figure 1B). Subject 2, a boy with developmental delay and mild autism, was first seen in the genetics clinic at the KK Women’s and Children’s Hospital at the age of 1 year and 8 months. Subject 2 had micrognathia, scaphocephaly, and hypotelorism, as well as small joint laxity (Figure 1C). The clinical history of subject 3 was partly reported previously by Verloes et al.¹² At the age of 25 years, his weight was 98 kg, his length was 152 cm (−3.5 SD), and his head circumference was 50 cm (−5 SD) (Figure 1D). Subject 4, the daughter of subject 3, shared common features with subject 3, which included intrauterine growth retardation followed by postnatal growth failure (length 79.5 cm, −4 SD), microcephaly (head circumference 45 cm, −5 SD), and dysmorphic features, including microretrognathia, hypotelorism, shortening of upper arms and legs, and muscular hypertrophy (Figure 1E).

Figure 1.

Clinical Features of the Individuals with ARCN1 Mutations

(A) Physical features of subject 1. Note downslanted palpebral fissures and micrognathia.

(B) Skeletal survey of subject 1. Note micrognathia, widening of the metaphysis, and a wide femoral neck.

(C) Facial features of subject 2.

(D) Physical features of subject 3. Note rhizomelic shortening of the limbs.

(E) Physical features of subject 4. Note micrognathia and rhizomelic shortening of the limbs.

The guardians of these individuals provided consent for the publication of their photographs.

Table 1.

Clinical Features of ARCN1-Related Syndrome

	Subject 1	Subject 2	Subject 3	Subject 4
IUGR	+	+	+	+
Micrognathia	+	+	+	+
Cleft palate	−	−	−	+
Tracheostomy	+	+	−	−
Congenital heart disease	+, VSD	−	−	−
Cryptorchidism	NA	+	−	NA
Short stature	+	+	+	+
Rhizomelic shortening	+	−	+	+
Joint laxity	+	+	−	+
Developmental delay	+	+	+	+
Autism	−	+	−	−
Seizure	−	+	−	−
Microcephaly	+	−	+	+
ARCN1 mutation	p.Ser87∗	p.Val212Trpfs∗15	p.Ser53Cysfs∗39	p.Ser53Cysfs∗39

+, present; −, absent; IUGR, intrauterine growth retardation; VSD, ventricular septal defect; NA, not applicable.

Exome sequencing was performed with genomic DNA extracted from peripheral blood cells. Exome sequencing of the probands revealed heterozygous ARCN1 loss-of-function mutations (Figure 2). Subject 1’s mutation in ARCN1 (GenBank: NM_001655) resulted in a premature stop codon in exon 2 (c. 260C>A [p.Ser87^∗]). Exome sequencing of subject 2 revealed two de novo mutations in ARCN1 and SYT1. The ARCN1 mutation was a frameshift variant (c.633del [p.Val212Trpfs^∗15]) predicted to create a frameshift in exon 4 starting at codon Val212 and ending in a stop codon 14 positions downstream (Figure 2A and Figure S2B). The SYT1 (c.697G>A [p.Asp233Asn] [GenBank: NM_001135805.1]) variant is a missense variant that substitutes aspartic acid at codon 233 with asparagine (Figure S3C). Subjects 3 and 4 were found to have a common ARCN1 mutation, c.157_158del, leading to p.Ser53Cysfs^∗39 (Figure S2C).

Figure 2.

Effects of ARCN1 Mutation

(A) Structure of ARCN1 (GenBank: NM_001655) and the location of ARCN1 mutations identified. Subject 1 has a stop mutation in exon 2, c. 260C>A (p.Ser87^∗). Subject 2 has a frameshift mutation in exon 4, c.633del (p.Val212Trpfs^∗15). Subjects 3 and 4 have a common ARCN1 mutation in exon 2, c.157_158del (p.Ser53Cysfs^∗39).

(B) mRNA expression of ARCN1 in HCT116 cell lines with and without ARCN1 mutation. RNA extraction was performed with TRIzol reagent (15596-018, Ambion) and Nucleospin RNA kits (Macherey-Nagel). The RNA obtained was reverse transcribed with a SuperScript VILO cDNA Synthesis Kit (11754-050, Invitrogen). The PCR primer sets used for qRT-PCR are listed in Table S2. ARCN1 expression was normalized against GAPDH. ARCN1 expression relative to that of the control sample (WT #1) is demonstrated. Data represent mean ± 2 SD. ^∗∗∗p < 0.001, two-tailed t test. n = 3 technical replicates per group. Three biological replicates for each sample demonstrated consistent results. WT, wild-type.

(C) Amounts of ARCN1 in HCT116 cell lines with and without ARCN1 mutation. The ARCN1 antibody (ab96725) and histone H3 antibody (ab1791) were from Abcam. Three biological replicates for each sample demonstrated consistent results.

In order to evaluate the effect of loss-of-function mutations of ARCN1, genome editing was performed with the CRISPR/Cas9 system on the colon-cancer-derived HCT116 cell line. After the introduction of the ARCN1 mutation into these cells, single-cell cloning was performed with the limiting dilution technique. The gRNA empty vector (catalog no. 41824) and hCas9 vector (41815) were obtained from Addgene. The gRNA target sequence was cloned into the gRNA empty vector with Phusion polymerase (M0530S, New England Biolabs) and the Gibson assembly system (E5510S, New England Biolabs). The gRNA target sequence was CTAAGGCTCTTCTCAAGAG; it is located at exon 2 of ARCN1. Cas9 and gRNA vectors were transfected into cell lines by electroporation via a Neon electroporation system (ThermoFisher Scientific). Genomic DNA was extracted with NucleoSpin tissue kits (Macherey-Nagel). The presence of the ARCN1 mutation was confirmed by Sanger sequencing. A total of 68 clones were screened, and 5 clones with ARCN1 mutations were identified. None of the clones possessed biallelic loss-of-function mutations of ARCN1, suggesting the importance of ARCN1 in maintaining cell viability. Two independent clones possessing 1-bp duplication leading to a frameshift that resembled the mutations identified in subject 1 (c.262dupA, in ARCN1 [GenBank: NM_001655]) were used for subsequent experiments (Figure S2D). Analysis with the ExPASy Translate tool indicated that this 1-bp duplication leads to p.Arg88Lysfs^∗5. Similar to the mutation found in four individuals with ARCN1 mutations, this frameshift mutation in the HCT116 cell line likely causes a loss-of-function effect, given that the mRNA in the case of this 1-bp insertion is predicted to undergo nonsense-mediated decay.

To test whether such a truncating mutation caused actual dosage reduction of ARCN1, we quantified ARCN1 mRNA by using qRT-PCR. Introduction of frameshift mutations in exon 2 resulted in reduction in ARCN1 mRNA expression, confirming that truncating mutations cause a reduction in the expression of the ARCN1 transcript (Figure 2B). Reduction of ARCN1 mRNA expression was also observed in the skin fibroblast samples obtained from subjects 3 and 4 (Figure S4A).

In order to test the amount of ARCN1, HCT116 cells with and without ARCN1 mutations were lysed with SDS sample buffer for immunoblotting. Western blotting of HCT116 mutant clones showed a reduction of ARCN1, in comparison to that in the wild-type clones (Figure 2C). Mild reduction of ARCN1 was also demonstrated in the skin fibroblast sample obtained from subject 4 (Figure S4B). These observations confirmed that truncating mutations of ARCN1 cause reduction of ARCN1 mRNA and protein amounts.

Given the phenotypic overlap of the individuals with ARCN1 mutations and individuals with collagenopathies such as Stickler syndrome (MIM: 108300), we hypothesized that reduction of ARCN1 might have caused intracellular collagen transport defects. To evaluate the effect of loss-of-function ARCN1 mutations, siRNA-mediated gene knockdown (KD) of ARCN1 was performed on a control skin fibroblast cell line, GM02036, given that skin fibroblast cell lines synthesize type I collagen.^{13, 14} The greatest reduction in ARCN1 mRNA and ARCN1 protein amounts was achieved 4 days after ARCN1 KD (Figure 3A and Figure 4A).

Figure 3.

ER Stress Response with the Reduction of ARCN1

Skin fibroblast ARCN1-KD experiments were performed three times, and consistent results were obtained. Two different siRNAs (ARCN1HSS100628 [KD1] and ARCN1HSS100629 [KD2], ThermoFisher Scientific) were used in the study. Stealth siRNA was transfected with Lipofectamine RNAiMAX reagent (ThermoFisher Scientific), according to the manufacturer’s instructions. For ARCN- KD in skin fibroblasts, two siRNA transfections were performed 2 days apart; 4 days after the initial siRNA transfection, RNA and cellular lysates were harvested.

(A) ARCN1 mRNA expression with ARCN1 KD. qPCR was run in triplicates, and mean and SD were calculated. ARCN1 expression was normalized against GAPDH expression. Gene expression relative to that of the control sample is shown. Data represent mean ± 2 SD. ^∗∗∗p < 0.001, two-tailed t test. n = 3 technical replicates per group.

(B) Elevation of ER stress response genes’ expression upon ARCN1 KD. Gene expression was normalized against GAPDH expression. Gene expression relative to that of the control sample is shown. Data represent mean ± 2 SD. ^∗∗∗p < 0.001, two-tailed t test. n = 3 technical replicates per group.

Figure 4.

Collagen Transport Defects with ARCN1 KD

(A) Reduction of ARCN1 caused accumulation of type I collagen in total cellular lysates and reduction of type I collagen secretion. The type I collagen antibody (AB758) was from Millipore.¹³ For the measurement of culture supernatant type I collagen, skin fibroblast cell numbers were counted, and equal numbers were plated in culture flasks. The cells were changed to serum-free culture media for the final 24 hr, the time point at which the medium was harvested. One-tenth the volume of TCA was added and placed on ice for 30 min. The samples were centrifuged and precipitates were washed with ethanol. After the ethanol wash, the supernatants were dried and the remaining cell pellets were dissolved with SDS sample buffer for immunoblotting. TCA precipitation was performed four times, and consistent results were obtained. Two bands of type I collagen represent the bands for alpha 1 and alpha 2 type I collagen. Top larger band represents alpha 1 type I collagen, and lower smaller band represents alpha 2 type I collagen.

(B) Brefeldin A (BFA; B5936, Sigma-Aldrich) treatment caused the intracellular accumulation of type I collagen. 10 uM of BFA was used. DMSO was added for control. The BFA treatment experiment was performed three times, and consistent results were obtained.

(C) COL1A1 and COL1A2 expression with ARCN1 KD. Slight reduction of COL1A1 expression was elicited with ARCN1 KD; however, the expression of COL1A2 remained unchanged. Gene expression was normalized against GAPDH expression. Gene expression relative to that of the control sample is shown. Data represent mean ± 2 SD. ^∗p < 0.05, ^∗∗∗p < 0.001, two-tailed t test. n = 3 technical replicates per group.

Disruption of intracellular trafficking can cause the accumulation of protein, leading to an ER stress response. When the rate of protein synthesis exceeds the capacity of protein folding and protein degradation machinery, the ER stress response is induced, which can result in cell death if prolonged.¹⁵ Therefore, the degree of the ER stress response was evaluated in the skin fibroblast cell line after ARCN1 KD. Overexpression of stress response genes such as ATF4, DDIT3, and HSPA5 serves as a marker of ER stress. qRT-PCR analysis demonstrated the upregulation of all these ER stress response genes after ARCN1 KD (Figure 3B). Therefore, reduction in ARCN1 amounts triggered the ER stress response. In order to evaluate the consequence of ER stress, cells were treated with thapsigargin (2 μM and 5 μM)¹⁶ and tunicamycin (2 μg/mL and 5 μg/mL)^{17, 18} for 17 hr; then, the total cellular lysates and RNA were obtained. Thapsigargin is an inhibitor of the ER Ca²⁺ATPase and induces ER stress, and tunicamycin inhibits the initial step of glycoprotein synthesis in the ER and induces ER stress. Artificial induction of ER stress by the addition of thapsigargin and tunicamycin triggered ARCN1 overexpression, although the amount of ARCN1 remained unchanged, suggesting increased ARCN1 turnover during the ER stress response (Figures 5A and 5B). Collectively, these results suggest that ARCN1 plays a major role in ameliorating the induction of the ER stress response. Evaluation of the skin fibroblast sample from subject 4 further confirmed the presence of ER stress, indicated by the cytosolic accumulation of BiP, which is an ER stress marker (Figures S4C and S5). Therefore, reduction in the amount of ARCN1 triggered the ER stress response.

Figure 5.

Artificial Induction of ER Stress Caused ARCN1 mRNA Overexpression

(A) Treatment with tunicamycin (T7765, Sigma-Aldrich) and thapsigargin (T9033, Sigma-Aldrich) triggered overexpression of ARCN1. ARCN1 expression was normalized against GAPDH expression. Gene expression relative to that of the control sample is shown. Data represent mean ± 2 SD. ^∗∗p < 0.01, ^∗∗∗p < 0.001, two-tailed t test. n = 3 technical replicates per group.

(B) The amount of ARCN1 remained unchanged with tunicamycin and thapsigargin treatments. The alpha tubulin antibody (T6074) was from Sigma-Aldrich. Two different concentrations of tunicamycin treatment (2 μg/mL [Tuni2] and 5 μg/mL [Tuni5]) and thapsigargin treatment (2 μM [Tg2] and 5 μM [Tg5]) were used. DMSO was added for control. Two biological replicates demonstrated the consistent results.

Next, the amount of type I collagen was evaluated in the control skin fibroblasts with ARCN1 KD. Interestingly, ARCN1 KD caused the accumulation of type I collagen in the total cellular lysates (Figure 4A). Type I collagen secretion from skin fibroblasts was then quantitated by performing trichloroacetic acid (TCA) precipitation of culture supernatants. Immunoblotting performed on culture supernatants demonstrated the reduction of secreted type I collagen in the culture media, suggesting that ARCN1 reduction caused intracellular accumulation of type I collagen as a result of defective intracellular protein transport (Figure 4A). In order to confirm that collagen accumulation was a result of defective intracellular transport, 10 μM brefeldin A (BFA), an ER-Golgi transport inhibitor, was added to the culture media, and total cellular lysates were obtained 23 hr after treatment.¹⁸ Addition of BFA caused the accumulation of intracellular type I collagen, similar to what was observed with ARCN1 KD (Figure 4B). Interestingly, BFA treatment reduced the amount of ARCN1 in the cells (Figure 4B). Type I collagen is encoded by two genes, COL1A1 and COL1A2. The alpha 1 chain of type I collagen is encoded by COL1A1, and the alpha 2 chain of type I collagen is encoded by COL1A2. qRT-PCR demonstrated a slight reduction in COL1A1 mRNA with ARCN1 KD, whereas the expression of COL1A2 mRNA remained unchanged with ARCN1 KD (Figure 4C). The addition of thapsigargin and tunicamycin did not cause accumulation of type I collagen, suggesting that the ER stress response does not cause the collagen transport defect (Figure 5B and Figure S6). These experiments indicate that ARCN1 is directly responsible for the intracellular transport of type I collagen and that the collagen transport defect is not secondary to the ER stress response.

In this paper, we report a clinically recognizable genetic disorder characterized by facial dysmorphisms, micrognathia, rhizomelic shortening, microcephalic dwarfism, and mild intellectual disability due to heterozygous ARCN1 loss-of-function mutations. We propose the name “ARCN1-related syndrome” to denote this condition. The ARCN1 is approximately 30 kb in size and located on chromosomal region 11q23.3 with ten exons (Figure 2A). Germline mutations of ARCN1 have not been previously reported. The Exome Aggregation Consortium (ExAC) website gives a probability of loss-of-function intolerance score of 1 to ARCN1, the highest score, given the absence of loss-of-function ARCN1 mutations in their cohort.¹⁹ Therefore, ARCN1 likely represents a dosage-sensitive gene in humans. To the best of our knowledge, germline chromosomal microdeletions spanning the ARCN1 region have not been reported in the medical literature. Copy-number variations within the ARCN1 locus have also not been reported in the Database of Genomic Variants.

The clinical phenotype of ARCN1-related syndrome includes severe micrognathia, microcephalic dwarfism, joint laxity, and mild developmental delay (Table 1). There is some phenotypic overlap between ARCN1-related syndrome and Stickler syndrome, such as micrognathia, short stature, and joint laxity, although neurological, ocular, and audiology features differ. Stickler syndrome is caused by loss-of-function heterozygous mutations in COL2A1 and other collagen genes.²⁰ Given the type of the mutations identified in Stickler syndrome, it is presumed that generalized reduction in type II collagen production leads to a clinical phenotype of Stickler syndrome, including features such as micrognathia and short stature. This clinical overlap prompted us to evaluate the role of ARCN1 in intracellular collagen transport, where we demonstrated the importance of ARCN1. Phenotypic resemblance between ARCN1-related syndrome and Stickler syndrome could be explained by the reduction in collagen secretion from the cells. The importance of ARCN1-associated intracellular trafficking has been documented in many biological processes, such as immune function and influenza virus infection.²¹ The inability to create an ARCN1-null cell line with CRISPR/Cas9 underscores the importance of ARCN1 in cell viability. Our findings highlight the unexpected importance of COPI transport in skeletogenesis, particularly in mandibular bone formation. Previously, similar pathological disease mechanisms resulting in skeletal dysplasia were implicated in COPII transport defects, such as cranio-lenticulo-sutural dysplasia and osteogenesis imperfecta.^{6, 22}

COL2A1 mutations also cause other skeletal dysplasias, such as achondrogenesis type II or hypochondrogenesis (MIM: 200610) and spondyloepiphyseal dysplasia (MIM: 183900). These disorders are collectively termed type II collagenopathies.²³ Many of these conditions are caused by heterozygous mutations in COL2A1. However, the effects of COL2A1 mutations vary in different type II collagenopathies. As a result, there is a wide phenotypic spectrum of type II collagenopathies. Interestingly, some COL2A1 missense mutations found in type II collagenopathies cause abnormal protein folding of type II collagen, leading to the intracellular accumulation of mutant collagen, which elicits an ER stress response.²⁴ Furthermore, the artificial induction of ER stress in chondrocytes causes skeletal dysplasia in mice.²⁵ Therefore, it is proposed that ER stress itself is involved in the pathogenesis of skeletal dysplasia. Hence, in ARCN1-related syndrome, the ER stress response might also play a role in the resultant skeletal phenotype. However, our ARCN1-KD experiments indicate that defective COPI transport could play a direct role in intracellular collagen accumulation in ARCN1-related syndrome, rather than the skeletal phenotype, which is secondary to ER stress. This finding is consistent with a previous report that demonstrated the involvement of COPI in intracellular collagen transport.²⁶ In order to deepen the understanding of the role that ER stress and COPI transport play in skeletogenesis, further studies are warranted.

The ER stress response is elicited even during normal embryogenesis.^{27, 28} Given the role of ARCN1 in the ER stress response, reduction of ARCN1 likely triggers an exaggerated ER stress response, and such an enhanced ER stress response might lead to cell death. In addition to the direct influence of ER-Golgi transport defects secondary to ARCN1 mutations, such cell death most likely contributes to the pleiotropic phenotype of ARCN1-related syndrome.

Mice with a homozygous missense mutation in Arcn1, that is, nur17 mice, have been previously created by N-ethyl-N-nitrosourea (ENU) mutagenesis.²⁹ The nur17 mouse harbors a homozygous missense mutation in exon 10 of Arcn1 and is characterized by the presence of dilute coat color, neurological defects, and low body weight; however, the skeletal phenotype was not characterized in the original report. Mild loss-of-function effects were presumed to mediate disease mechanisms in the nur17 mutant mouse. There are several similarities between Arcn1 mutant mice and the individuals with ARCN1-related syndrome, such as low body weight and the neurological phenotype, including ataxia due to cerebellar degeneration. Interestingly, subject 4 was found to have ataxia and cerebellar atrophy. Hence, nur17 mice might serve as an animal model for ARCN1-related syndrome, although skin color and hair abnormalities were not detected in individuals with ARCN1 mutations.

In addition to the skeletal phenotype, another feature typical of ARCN1-related syndrome is mild intellectual disability. There are several lines of evidence supporting a role for COPI transport in brain function. First, COPI interacts with a subset of RNA molecules, and COPI has been implicated in intracellular RNA transport in neuronal cells.³⁰ Second, the COPI system is known to transport molecules that play major roles in neurotransmitter release, such as SNARE proteins.^{31, 32, 33} In fact, the importance of intracellular trafficking has been well known in neuronal function and homeostasis because the abnormalities of intracellular trafficking are linked to the pathogenesis of various neurodegenerative disorders, such as Alzheimer disease.³⁴ Abnormal ER-Golgi trafficking is demonstrated in Dyggve-Melchior-Clause syndrome (MIM: 223800), associated with developmental delay and microcephaly.³⁵ Furthermore, the presence of ER stress during embryogenesis is well documented in the CNS, supporting a role for ER stress in brain development.^{27, 28} Therefore, the findings strongly suggest that COPI is involved in neuronal function. The neurological phenotype of the nur17 mouse model further supports the notion that defective ARCN1 function impairs brain function.²⁹

Subject 2 experienced seizures in addition to intellectual disability. However, it remains to be determined whether the seizures were due to ARCN1 mutations given that subject 2 also has a de novo missense SYT1 variant, which was not previously reported. SYT1 encodes synaptotagmin-1, which plays important roles in synaptic vesicle recycling, and there have been two recent reports associating SYT1 mutations with dyskinetic movement disorder and developmental delay in one child and with developmental delay, facial dysmorphisms, and anomalous electroencephalogram (EEG) patterns in another child^{36, 37} (Table S3). Therefore, it is possible that the neurological phenotype of subject 2 is partially due to the SYT1 mutation. However, given that all four individuals with ARCN1 mutations demonstrated comparable degrees of developmental delay and intellectual disability, ARCN1 mutations most likely play a major role in the neurological phenotype of ARCN1-related syndrome, and ARCN1 is probably required for normal brain growth and cognitive development.

In conclusion, we report a recognizable craniofacial disorder, ARCN1-related syndrome, characterized by facial dysmorphisms, micrognathia, rhizomelic shortening, microcephalic dwarfism, and mild intellectual disability. Intracellular COPI transport has been implicated in various biological processes, and our results strongly indicate the importance of COPI transport in skeletogenesis, particularly in mandibular bone formation, during embryogenesis.

Published: July 28, 2016

Web Resources

• Database of Genomic Variants (DGV), http://dgv.tcag.ca/dgv/app/home

• ExAC Browser, http://exac.broadinstitute.org/

• ExPASy Translate tool, http://web.expasy.org/translate/

• GenBank, http://www.ncbi.nlm.nih.gov/genbank/

• GeneMatcher, https://genematcher.org/

• OMIM, http://www.omim.org/