Identification of eight novel proteasome variants in five unrelated cases of proteasome-associated autoinflammatory syndromes (PRAAS)

Abstract

Mutations in genes coding for proteasome subunits and/or proteasome assembly helpers typically cause recurring autoinflammation referred to as chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperatures (CANDLE) or proteasome-associated autoinflammatory syndrome (PRAAS). Patients with CANDLE/PRAAS present with mostly chronically elevated type I interferon scores that emerge as a consequence of increased proteotoxic stress by mechanisms that are not fully understood. Here, we report on five unrelated patients with CANDLE/PRAAS carrying novel inherited proteasome missense and/or nonsense variants. Four patients were compound heterozygous for novel pathogenic variants in the known CANDLE/PRAAS associated genes, PSMB8 and PSMB10, whereas one patient showed additive loss-of-function mutations in PSMB8. Variants in two previously not associated proteasome genes, PSMA5 and PSMC5, were found in a patient who also carried the PSMB8 founder mutation, p.T75M. All newly identified mutations substantially impact the steady-state expression of the affected proteasome subunits and/or their incorporation into mature 26S proteasomes. Our observations expand the spectrum of PRAAS-associated genetic variants and improve a molecular diagnosis and genetic counseling of patients with sterile autoinflammation.

Keywords: PSMA5; PSMB10; PSMB8; PSMC5; interferonopathy; proteasome associated autoinflammatory syndrome; proteasomopathy; type I interferon.

Figure 1

Clinical features and pedigrees. (A) Pedigrees of five families (patients 1–5) showing digenic or trigenic inheritance for patient 1 who harbors a novel de novo variant in PSMA5 and a rare PSMC5 variant on the paternal and a known PSMB8 mutation on the maternal allele. Three unrelated Brazilian patients present with four novel PSMB10 variants. All share the p.Phe14del variant plus a different novel variant. Open squares and circles represent healthy males and females, respectively, whereas black squares and circles represent PRAAS affected males and females, respectively. ^ Patient 5 was previously reported (29). (B) The IFN score obtained from whole blood prior to treatment with JAKis was assessed by NanoString and presented as the sum of z-scores of 28 ISGs. Type I IFN scores for the five patients with PRAAS are depicted in comparison with healthy controls (HC), patients with the IL-1–mediated disease cryopyrin-associated periodic syndrome (CAPS), and patients with the type I interferonopathies chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperature (CANDLE) and stimulator of interferon genes (STING)–associated vasculopathy with onset in infancy (SAVI). Type I IFN score (mean ± SD): HC, −4.8 ± 9.53; NOMID, −9.37 ± 10.97; CANDLE, 738.2 ± 754.4; SAVI, 495.6 ± 225.9; patient 1, 763.27; patient 2, 283.57; patient 3, 144.76; patient 4, 106.75; patient 5, 209.47. Clinical skin and joint features in patients with CANDLE/PRAAS are shown in (C–G). (C) Facial features of nodular rash over cheeks and forehead and periorbital edema and swelling with mild lipoatrophy in temple and cheek area (Pt. 1). (D) Annular erythematous skin rash on left cheek at 9 months of age (Pt. 3). (E) Annular erythematous lesions involving the hand and arm at 12 months of age (Pt. 3). (F) Erythematous swollen fingers at the age of 3 years (Pt. 4). (G) Fixed flexion contractures of the interphalangeal joints at the age of 5 years (Pt. 4).

Figure 2

Proteasome variant positions within the 26S proteasome complex. (A) A lateral view of the 20S core particle (left) and 19S regulatory particle (right), indicating the localization of the PSMA5/α5, PSMB10/β2i, PSMC5/Rpt6 (yellow), and PSMB8/β5i (orange) affected subunits. (B) Left: bottom view of the upper 20S proteasome α-ring (brown) showing the localization of the missing C-terminal part of the p.(Arg168*) PSMA5/α5 variant (yellow). Right: Lateral view of the upper 20S proteasome β-ring (purple) together with the p.(Arg168*) PSMA5/α5 variant (yellow) from the upper α-ring. (C) Left: Lateral view of the 20S proteasome α-ring (brown) together with PSMC5/Rpt6 (yellow) and its exon 10 (red) from the 19S regulatory particle. Right: Lateral view of the of the ring base of 19S regulatory particle (blue) containing the PSMC5/Rpt6 with exon 10 marked in red. (D) Left: Top view of the lower 20S proteasome β-ring (purple) showing the localization of the G167 residue, exon 7 (red) within the PSMB10/β2i subunit (yellow). The theoretical frameshift PSMB10/β2i variant emerging from the c.247_248insT alteration is highlighted in pink. Right: View of the 20S immunoproteasome subunit β2i (dark gray) together with the β6, β3, and β7 of the proteasome β-ring showing the localization of β2i exon7 (red). (E) Top (left) and lateral (right) views of the lower 20S proteasome β-ring showing the localization of the T75 and S118 residues (marked in red) within the β5i (PSMB8) subunits (yellow).

Figure 3

The proteasome subunit variants PSMB8/β5i, p.Thr75Met, PSMA5/α5, p.Arg168*, and PSMC5/Rpt6 p.Ala324_Lys360del (patient 1) are not fully incorporated into 20S/26S proteasome complexes. (A) Schematic representation of the PSMB8, PSMA5, and PSMC5 constructs encoding wild-type (WT) and β5i/LMP7, α5, or Rpt6 variants, respectively, used in this study. Unlike the PSMB8 and PSMA5 constructs that are C-terminally tagged with a combined V5/HIS epitope, the PSMC5 constructs are N-terminally fused with a HA-tandem repeat. Scissors represent propeptide processing, which does not occur for α5 or Rpt6 variants. (B) HeLa cells expressing WT or β5i/LMP7, α5, or Rpt6 variants were subjected to denaturing protein extraction prior to SDS-PAGE/Western blotting analysis using antibodies specific for β5i/LMP7, α5, Rpt6, V5, myc, and β-actin (loading control), as indicated. (C) HeLa cells transiently transfected with WT or variants of the α5 proteasome subunit were assessed for their content in α5-V5/HIS transcripts by RT-PCR using specific primers. Equal sample loading was ensured by amplifying the RPLP0 housekeeping gene, as indicated. The quantification of the PSMA5 bands from three replicates was normalized using the obtained RPLP0 housekeeping gene signals and subsequently analyzed and visualized using GraphPad software (GraphPad Software, Inc., California, USA). (D) HeLa cells engineered to express WT and variants of the β5i/LMP7, α5, or Rpt6 subunits were subjected to non-denaturing protein extraction and subsequently analyzed for their amounts of V5- and HA-containing proteasome complexes by native-PAGE/Western blotting, as indicated. Densitometric evaluation (n = 3) of impaired incorporation of β5i/LMP7 T75M-V5 (p = 0.006 for incorporation into 20S complexes; p = 0.04 for incorporation into 26S complexes) or HA-Rpt6del324-360 variants (p = 0.005) in relation to the WT as fold change. WT incorporation was set to 1. ns, not significant; * p-value <0.05; ** p < 0.01.

Figure 4

Both PSMB10/β2i variants, p.Phe14del and p.Gly167Asp (patient 2), cannot mature and incorporated. (A) Cartoon depicting the C-terminally V5/HIS-tagged PSMB10 constructs encoding wild-type (WT) and β2i/MECL1 variants, as indicated. Scissors represent propeptide processing. (B) HeLa cells engineered to express WT and β2i/MECL1 variants were subjected to protein extraction prior to Western-blot analysis using antibodies specific for V5 and β-actin (loading control). (C) HeLa cells ectopically expressing WT and PSMB10 variants were subjected to RNA extraction and subsequent RT-PCR using primers specific for the overexpressed PSMB10 species, as indicated. Amplification of RPLP0 housekeeping was used a control to ensure equal loading of samples. The quantification of the PSMB10 bands from three replicates was normalized using the obtained RPLP0 housekeeping gene signals and subsequently analyzed and visualized using GraphPad software (GraphPad Software, Inc., California, USA). (D) HeLa cells transfected with WT and β2i/MECL1 variants were assessed for their content of V5-containing native complexes by native-PAGE/Western blotting using an antibody directed against V5, as indicated. Densitometric evaluation (n = 3) of failed incorporation of β2i/Mecl-1 variants into 20S or 26S complexes in relation to the WT variants as fold change (p-values < 0.0001). WT incorporation was set to 1. ns, not significant; **** p-value <0.0001.

Figure 5

Both PSMB8 variants, p.Gln55* and p.Ser118Pro (patient 5), fail to assemble into mature 20S and 26S proteasomes. (A). Depiction of the PSMB8 wild-type and mutant constructs shows the amino acids (AA) with in the prodomain and the functional domain of PSMB8/β5i/LMP7. All constructs were fused at the C-terminus in frame with a combination of V5 and HIS tags, as indicated. Scissors represent propeptide processing. (B) HeLa cells transiently transfected with PSMB8 constructs were assessed for their V5 content by Western blotting, as indicated. Equal protein loading was ensured by probing the membrane with a monoclonal antibody specific for β-actin. (C) HeLa cells engineered to express wild-type and PSMB8 variants were subjected to RNA extraction and subsequent RT-PCR using primers specific for the overexpressed PSMB8 species, as indicated. Amplification of RPLP0 housekeeping was used a control to ensure equal loading of samples. The quantification of the PSMB8 bands from three replicates was normalized using the obtained RPLP0 housekeeping gene signals and subsequently analyzed and visualized using GraphPad software (GraphPad Software, Inc., California, USA). (D) HeLa cells expressing wild-type or mutant PSMB8 were subjected to non-denaturing protein extraction prior to native-PAGE analysis by Western blotting using an antibody specific for V5, as indicated. Densitometric evaluation (n = 3) of failed incorporation of β5i/LMP7 S118P-V5 or β5i/LMP7 Q55*-V5 variants into 20S complexes or 26S complexes in relation to the WT as fold change (β5i S118P-V5: 20S incorporation, p = 0.01; 26S incorporation, p = 0.0004; β5i Q55*-V5: p-values < 0.0001 for 20S and 26S incorporation). WT incorporation was set to 1. ns, not significant; ** p < 0.01; *** p < 0.001; **** p < 0.0001.