|
Fig. 1 Structure and function of the COP1 complex. a COPI consists of a scaffold “B-subcomplex” (blue) and an adaptor “F-subcomplex” (green). When GTP-bound, two ARF1 small GTPase molecules associate with the membrane and bind COPI via the β-COP and γ-COP subunits. A number of subunits of this complex have been implicated in human disease as shown. b COPI complexes and their ARF1 molecules associate into triads. Cargo, such as ER-resident proteins which need to be returned from the Golgi, are selected by binding directly with COPI subunits or indirectly through transmembrane receptors which in turn bind with COPI. COPI polymerises on the membrane enabling its deformation/curvature, and eventually budding and scission of the transport vesicle. When released, the vesicle’s coat is shed and ARF1 and COPI dissociate. Adapted from Nickel et al. [12]
|
|
Fig. 2 Pedigrees and clinical presentation of individuals with COPB1 mutations. a Family 1 includes two affected individuals IV2 and IV3. Individuals IV5, IV6 and IV7 are also suspected to have been affected with microcephaly and developmental delay. They died in early childhood and no further details were available. IV2 and IV3 have minor dysmorphism with up-slanting palpebral fissures. b Family 2 includes four affected individuals from two nuclear families all from the same Saudi tribe. c Gel electrophoresis of RT PCR amplicons in Family 1 demonstrates 2 bands in the homozygous state (IV2, an affected individual) and in the heterozygous state (III1, an unaffected parent). d Electropherograms of band A and band B from the heterozygous parent and homozygous proband. In the proband’s trace, a G has effectively been deleted due to the creation of a new donor site by the G>T mutation. In band B, Exon 8 has been skipped (proband and parent)
|
|
Fig. 3 Structure and conservation of COPB1 protein, and structural effect of missense variants. a Simple schematic diagram of COPB1 (β-COP) structure showing two main structural domains; trunk domain and appendage domain, and relative location of the c.957+1G>T and p.Phe551Val variants located towards the N terminal of the trunk domain. b COPB1 (β-COP) structure with the location of the 12 amino acids (in red) which are deleted due to exon 8 skipping caused by c957+1G>T. c COPB1 (β-COP) structure in context of COPI complex. Note exon 8 (red) makes up an important link between the scaffold and adaptor subcomplexes. d Alignment of COPB1 amino acid sequence from H. sapiens to yeast showing very high conservation of Phe551 across all species tested. e 3D structural modelling of COPB1 in complex with COPB2 and COPG1 showing the location of Phe551 near the site of interaction with COPG1. f Higher resolution image showing the location of Phe551Val mutation in a turn in the trunk domain
|
|
Fig. 4 Targeted CRISPR/Cas9 disruption of copb1 exon 8 generates a range of insertion-deletion changes in vivo. Xenopus tropicalis copb1 has 21 exons, including an untranslated first exon (a). CRISPR/Cas9 directed indel formation disrupting X. tropicalis exon 8 using 2 sgRNAs (red: sgRNA3 and sgRNA4 (b)) results in homozygous frameshift and splice changes akin to those identified within the patient subpopulation. Genotyping analysis of X. tropicalis crispants details the range of indels across three groups of 10 pooled tadpoles (NF41) injected with sgRNA3 following PCR amplification and Sanger sequencing of subclones (c). The 41-bp deletion observed in gDNA subclones extends into the intronic region and is proposed to induce exon 8 skipping in crispant X. tropicalis tadpoles. Total RNA was obtained from 4 groups (un-injected control (1), copb1 sgRNA 3 crispant set 1 (2) and copb1 sgRNA 3 crispant set 2 (3), injected control (4)) of 10 pooled individuals as outlined in Guille 1999 [51] and cDNA synthesised using Primer Design’s Reverse Transcription Premix 2. Amplification across the copb1 region of interest revealed a band at 382 bp (exons: 7, 8 and 9) and an additional band in crispants at 262 bp (exons: 7 and 9, Additional file 1: Fig. S1D)
|
|
Fig. 5 F0 free-feeding wild-type and transgenic [Xtr.Tg (tubb2b:GFP) Amaya, RRID: EXRC_3001] Xenopus tropicalis crispants phenocopy key clinical hallmarks. Images show tadpole head structural morphology under normal conditions (a–c) and across the range of phenotypes observed: mild microcephaly (d–f), microcephaly with cataract(s) (g–i) and microcephaly with absent or missing eye structure(s) (j–l). White arrows on GFP fluorescence images show normal forebrain structures in un-injected control animals (b) and altered forebrain structures in mutants, with an increasing severity of microcephaly (e, h, k). Red arrows demonstrate the same forebrain structural trends in higher resolution MicroCT imaging (1% phosphotungstic acid contrast stain: c, f, i, l). Cataract formation is indicated in bright-field and MicroCT images by yellow arrows (g, i) and can be seen as a loss of GFP expression in fluorescence imaging (h), whilst green arrows highlight absent or missing structures of the eye (j, l). Re-occurring eye abnormalities were documented in 30 un-injected and 30 crispant tadpoles (exon 8, sgRNA3) to show the prevalence of cataract formation (mean, 14 tadpoles) and missing eye structures (mean, 5 tadpoles (m)). Further, 8 transgenic X. tropicalis crispant tadpole brains were imaged and measured (mm) 3 days post-fertilisation (control mean 1 mm, SD 0.02; Crispant mean 0.845 mm, SD 0.12 (t = 3.701; p = 0.007)) and 5 days post-fertilisation (control mean 6.37 mm, SD 0.29; Crispant mean 5.34 mm, SD 0.20 (t = 8.317; p = 0.000)) revealing sustained, significant reduction in brain length. Brain length measured as the distance from the forebrain to the hindbrain (Additional file 1: Fig. S5a) was expressed as a percentage of the mean of the control (N). Kaplan-Meier survival analysis of 50 un-injected control (median survival time 4.4 days) and crispant X. tropicalis tadpoles, injected at the one-cell stage with either sgRNA1 (exon 3 (Additional file 1: Fig. S4, median survival time 2 days) or sgRNA3 (exon 8, median survival time 3 days) show a significant difference in survival by log-rank comparison p = 0.000 (o)
|
|
Fig. 6 Effect of expression of COPB1 mutation on in vitro localisation and stability of β-COP protein. a Western blot image of protein extracts from HEK293 cells transfected with wild-type, and c.1651T>G (COPB1 tagged with c-myc, immunoblotted with c-myc antibody to detect COPB1, and with beta-actin antibody as loading control). c.1651T>G shows a slightly reduced normalised level. b Immunofluorescence images of hTERT-RPE1 cells nucleofected with wild-type COPB1 (top row), and COPB1 c.1651T>G p. p.Phe551Val (lower row) tagged with c-myc. Cells were stained with Golgi stain Bodipy-TR-ceramide (red), nuclear stain DAPI (blue) and immunostained with anti-myc (green). Mutant expression shows more diffuse β-COP staining throughout the Golgi and outside the Golgi in the wider cytoplasm. Scale bar = 5 μm. c Immunofluorescence images of hTERT-RPE1 cells nucleofected with wild-type COPB1 (top row), and COPB1 c.1651T>G p. p.Phe551Val (lower row) tagged with c-myc. Cells were stained with nuclear stain DAPI (blue) and immunostained with an antibody to resident Golgi protein giantin (green) and anti-β-COP (red). This clarified that wild-type β-COP tended to localise to the edges of the Golgi stacks, whereas mutant β-COP was localised throughout the Golgi. Scale bar = 10 μm
|
