Proteasome dysfunction in T cells causes immunodeficiency via cell cycle disruption and apoptosis

Abstract

Proteasomes are essential molecular complexes that regulate intracellular protein homeostasis by selectively degrading ubiquitinated proteins. Genetic mutations in proteasome subunits lead to proteasome-associated autoinflammatory syndromes (PRAAS) characterized by autoinflammation, partial progressive lipodystrophy, and, in certain cases, immunodeficiency. However, the molecular mechanisms by which proteasome dysfunction results in these phenotypes remain unclear. Here, we established a mouse model carrying a mutation in β5i (encoded by Psmb8) along with T-cell-specific β5 (encoded by Psmb5) deficiency (KIKO mice). The KIKO mice presented severe loss of mature T cells in the spleen but not in the thymus, with reduced proteasome activity leading to the accumulation of ubiquitinated proteins. The CD4⁺ T cells of KIKO mice presented impaired proliferative activity with cell cycle arrest in the G₀/G₁ phase following T cell receptor (TCR) engagement. T cells from KIKO mice underwent rapid cell death through apoptosis, as treatment of T cells with the caspase inhibitor Z-Val-Ala-Asp(Ome)-fluoromethylketone (Z-VAD-FMK) rescued cell viability. Moreover, proteasome dysfunction induced apoptosis in T cells without affecting either mitochondrial functions or endoplasmic reticulum (ER) stress responses. Thus, our data provide insight into the molecular mechanisms underlying not only immunodeficiency in PRAAS patients but also T-cell deficiency associated with other disorders.

T-cell deficiency in the β5i proteasome subunit causes immunodeficiency

Graphical Abstract

Graphical Abstract.

Introduction

Intracellular proteins are degraded by the ubiquitin‒proteasome system to maintain homeostasis in a variety of cell types (1–3). The 26S proteasome consists of a 19S regulator that regulates protein unfolding and the 20S proteolytic core complex, which is composed of 28 subunits in 4 rings with 7 different α subunits (α1–α7) and 7 different β subunits (β1–β7) (1, 2, 4). Among the 7 β-subunits, the β1, β2, and β5 subunits are responsible for three types of peptidase activities: caspase-like, trypsin-like, and chymotrypsin-like activities, respectively (5). In mammals, there are four different types of proteasomes: the standard proteasome, immunoproteasome, thymoproteasome, and spermatoproteasome (1, 6). Most cells express the standard proteasome with β1, β2, and β5, and immune cells, as well as interferon gamma (IFN-γ)-stimulated cells, express the immunoproteasome, in which β1, β2, and β5 are replaced with β1i, β2i, and β5i (7). The thymoproteasome, in which β5 is replaced with β5t, is specifically expressed by cortical thymic epithelial cells and plays an important role in the positive selection of CD8⁺ thymocytes (8). The spermatoproteasome is a proteasome subtype in which the standard α4 subunit (PSMA7) is replaced by α4s (PSMA8). An analysis of Psma8-deficient mice revealed that PSMA8 is essential for spermatogenesis (9).

We and other groups have reported that a missense mutation (G197V) in β5i (PSMB8) causes Japanese autoinflammatory syndrome with lipodystrophy (JASL), also known as Nakajo-Nishimura syndrome, which is characterized by recurrent fever, nodular erythema, hypergammaglobulinemia, and progressive partial lipodystrophy (10, 11). Several variants of not only β5i but also other proteasome subunits were subsequently detected in patients with autoinflammatory diseases with symptoms similar to those of JASL. These syndromes are collectively called proteasome-associated autoinflammatory syndromes (PRAAS) (12–17). Furthermore, recent studies have shown that variants of certain proteasome subunits cause immunodeficiency. The heterozygous de novo β1i (PSMB9) mutation (G156D) causes autoinflammation with immunodeficiency (PRAAS-ID) (18). Immunodeficiency caused by a mutation in β2i (Psmb10) was identified in a mouse model (19), and one paper showed that six unrelated infants with severe combined immunodeficiency (SCID)-Omenn syndrome had de novo heterozygous missense variants in PSMB10 (20). Although the variants in β1i and β2i disrupted the α-ring/β-ring interaction, leading to impaired 20S proteasome complex formation, 26S proteasome activity was largely preserved (18, 20).

Regarding the mechanisms by which proteasome dysfunction causes those phenotypes, we previously reported that activation of the p38 pathway partially contributes to the development of autoinflammation through IL-6 production (10). Another study showed that proteasome dysfunction induces the accumulation of cytoplasmic IL-24, resulting in the production of IFN-α through the activation of protein kinase R (PKR) (21). To further explore the mechanism underlying the various phenotypes of PRAAS, we established mice carrying Psmb8 genes containing a mutation (G197V) (Psmb8-KI) found in PRAAS patients. However, Psmb8-KI mice exhibited only mild lipodystrophy and showed no signs of autoinflammation (22). Detailed analysis of proteasome subunit expression in lymphocytes from Psmb8-KI mice revealed marked upregulation of the β5 subunit, a component of the standard proteasome. Thus, we hypothesized that this upregulation of β5 compensates for the dysfunction of mutated β5i. Consistent with this hypothesis, chymotrypsin-like proteasome activity in Psmb8-KI lymphocytes was significantly greater than that in wild-type (WT) lymphocytes (22).

In this study, we aimed to evaluate the effects of β5i dysfunction on murine T cells while avoiding compensatory β5 upregulation. To this end, we generated Psmb5^flox mice, which were subsequently crossed with either CD4-Cre (KO) or CD4-Cre and Psmb8-KI (KIKO). Our findings revealed that KIKO mice exhibited severe T-cell immunodeficiency. Notably, this immunodeficiency could not be solely attributed to the absence of β5 in T cells, as the KO mice did not display any signs of immunodeficiency. Upon TCR engagement, T cells from KIKO mice demonstrated impaired proliferative activity, characterized by cell cycle arrest in the G₀/G₁ phase. T cells from KIKO mice induced rapid cell death, through apoptosis, without affecting mitochondrial functions or endoplasmic reticulum (ER) stress responses. These findings shed light not only on the molecular mechanisms underlying T-cell deficiency in PRAAS patients but also on the regulatory roles of proteostasis in T-cell proliferation.

Methods

Mice

Mice harboring the Psmb5 allele with exon 2 flanked with two loxp sites introduced in intron1 and in intron2 (Psmb5^flox) were generated via the CRISPR/Cas9 system (C57BL6/J background). The following sequences were selected to disrupt Psmb5: 5′-gtgaccctagagcactaggg-3′ for guide RNA (gRNA) and 5′-cgtgtgtagtgctcgcaggc-3′ for gRNA (Supplementary Figure 1A). Psmb8-KI mice were previously established and reported (22). Cd4-cre transgenic mice were provided by Dr M. Kubo (Riken, Yokohama, Japan). Modified Psmb5 loci were detected by PCR with the following primers. Primer1: 5′-GTGAGGGAAGGAGACGGACAGC-3′; Primer2: 5′-CTAAAGGCGTGAGCCACCACCACC-3′. All the mice were used for the experiments at 6–8 weeks of age. All the mice were maintained under pathogen-free conditions and handled in accordance with the Guidelines for Animal Experimentation of Tokushima University.

Cell preparation, T-cell isolation, activation, proliferation, and cell cycle analysis

Thymic and splenic single-cell suspensions were first depleted of CD8⁺ T cells and B cells using BioMag goat anti-rat immunoglobulin G (IgG) separation beads (QIAGEN, 310107). CD4 single-positive (SP) thymocytes and naïve CD4⁺ T cells from splenocytes were sorted into CD45⁺CD8⁻CD4⁺ lymphocytes and T cell receptor (TCR)β⁺CD8⁻CD4⁺CD44^lowCD62L^high lymphocytes, respectively, by a fluorescence-activated cell sorting (FACS) Aria III (BD Biosciences). For T-cell activation, the sorted T cells were stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 (Gibco, 11452D). To analyze T-cell proliferation, the sorted T cells were incubated with 5 µM CellTrace Violet (Thermo Fisher Scientific, C34557) in phosphate-buffered saline (PBS) for 20 min at 37°C. The cells were subsequently washed with Roswell Park Memorial Institute (RPMI) 1640 medium (Nacalai Tesque) supplemented with 10% heat-inactivated fetal calf serum (FCS) (Invitrogen), 50 U/ml penicillin (Nacalai Tesque), 50 µg/m streptomycin (Nacalai Tesque), and 50 µM 2-ME (Sigma), hereafter referred to as R10, before 72 h of stimulation. For cell cycle analysis, the sorted T cells were stimulated with 1 µg/ml plate-bound anti-CD3ε in R10 supplemented with 1 µg/ml anti-CD28 mouse antibody (mAb) for 72 h. The stimulated cells were subsequently stained with 5 μM Vybrant DyeCycle Violet Stain (Invitrogen, V35003) for 30 min at 37°C before analysis.

Cell lysates, glycerol gradient analysis, and western blotting

The cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (Nacalai Tesque) containing protease inhibitors (Roche) and clarified by centrifugation at 20 000 × g for 15 min at 4°C. For glycerol gradient centrifugation analysis, clarified cell lysates were fractionated by linear density gradient centrifugation (22 h, 26 000 rpm, 4°C) as described previously (10). The following antibodies were used: anti-β1i (Abcam, ab3328), anti-β5i (Proteintech, PTG14859), anti-β5 (Enzo Life Sciences, PW8895), anti-Ump-1 (Proteintech), anti-ubiquitin (Santa Cruz Biotechnology, P4D1), anti-β-actin (Proteintech, 81115-1-RR or Sigma-Aldrich, A2066) mAbs, and horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG antibodies (Bio-Rad). The bands were detected with ECL Prime Chemiluminescent Substrate (GE Healthcare) and an ImageQuant LAS-4000 mini system (GE Healthcare).

Proteasome activity assay

The proteasome activity of splenic naïve CD4⁺ T cells was analyzed via a Proteasome-Glo Cell-Based Assay (Promega) according to the manufacturer’s protocol.

Flow cytometry

Single-cell suspensions of splenocytes and thymocytes were incubated with 2.4G2 supernatant to block Fc receptors. Then, the cells were stained with fluorochrome-conjugated mAbs specific for mouse TCRβ-APC (clone H57-597), CD4-APC-Cy7 (clone GK1.5), CD4-PE-Cy7 (clone GK1.5), CD4-PB (clone GK1.5), CD8-FITC (clone 5H10-1), CD8-APC (clone 53-6.7), CD8-APC-Cy7 (clone SK1), CD44-PE-Cy7 (clone IM7), CD62L-PB (clone MEL-14), CD3ε-APC (clone 145-2C11), CD3ε-PE (clone 17A2), and CD45-PB (clone S18009F) from BioLegend. The samples were analyzed with a CytoFLEX (Beckman Coulter). Dead cells were excluded by staining with either propidium iodide (PI) or 7-aminoactinomycin D (7-AAD). Flow cytometry data were analyzed with FlowJo software (BD Bioscience).

Analysis of mitochondrial function

For determination of mitochondrial size, T cells were stained with MitoTracker Green (Thermo Fisher Scientific, M7514) according to the manufacturer’s protocol. To determine the mitochondrial membrane potential, T cells were incubated in R10 supplemented with 100 nM tetramethylrhodamine methyl ester perchlorate (Thermo Fisher Scientific, T668) for 30 min at 37°C and then washed and analyzed for the mitochondrial membrane potential. T-cell mitochondrial superoxide production was measured with MitoSOX Green (Invitrogen, M36005) according to the manufacturer’s protocol. For analysis of adenosine triphosphate (ATP) production, T cells were activated with Dynabeads™ Mouse T-Activator CD3/CD28 for 16 h. For b-AP15 treatment, T cells were first treated with 1 μM b-AP15 for 2 h before activation. ATP production was measured with the CellTiter-Glo Luminescent Cell Viability Assay (Promega).

Real-time PCR

Total RNA was extracted from T cells via an RNeasy Mini Kit (QIAGEN, 74104), and cDNA was generated via ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, FSQ-301) in accordance with the manufacturer’s instructions. Real-time PCR was performed using Fast SYBR Green Master Mix (Thermo Fisher Scientific, 4385612) on a QuantStudio 5 (Thermo Fisher Scientific) with the following primers:

• CHOP-forward: 5′-GCGACAGAGCCAGAATAACA-3′,

• CHOP-reverse: 5′-GATGCACTTCCTTCTGGAACA-3′,

• Bip-forward: 5′-TCATGACATTCAGTCCAGCAA-3′,

• Bip-reverse: 5′-CTGAGGCGTATTTGGGAAAG-3′,

• Hprt-forward: 5′-CAGTCCCAGCGTCGTGATTA-3′,

• Hprt-reverse: 5′-TCGAGCAAGTCTTTCAGTCC-3′.

The Hprt mRNA level was used as an internal control.

Apoptotic cell analysis and Z-VAD treatment

First, thymic single-cell suspensions were depleted of CD8⁺ T cells using BioMag goat anti-rat IgG separation beads (QIAGEN, 310107). The remaining T cells were then stimulated for 24 h with Dynabeads™ Mouse T-Activator CD3/CD28. T cells from KIKO mice were stimulated in the presence or absence of the caspase inhibitor Z-Val-Ala-Asp(Ome)-fluoromethylketone (Z-VAD-FMK) (Funakoshi, FMK001) at a concentration of 150 μM. After stimulation, the T cells were stained with the surface markers CD45 APC (clone 30F-11), CD4 PECy7 (clone GK1.5), and CD8 violetFluor (clone 2.43), washed twice with PBS, and subsequently stained with Annexin V-FITC (BioLegend, 640906) or Annexin V-PE (BioLegend, 640908) and 7-AAD (BioLegend, 420404) in Annexin V binding buffer (BioLegend, 422201) for 15 min at room temperature (RT) in the dark. Positive control cells for single color compensation of fluorescein isothiocyanate (FITC), phycoerythrin (PE), and 7-AAD were generated by heating whole thymocyte cell suspensions for 3 min at 70°C. Apoptotic populations were evaluated by gating on CD45⁺CD4⁺CD8⁻ cells. The cells were analyzed using CytoFlex (Beckman Coulter). Flow cytometry data were analyzed with FlowJo software (BD Bioscience).

Electron microscopy

CD4 SP thymocytes were sorted and stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 for 16 h. The cells were then fixed with 2% fresh formaldehyde and 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 2 h at room temperature. After the cells were washed with 0.1 M cacodylate buffer (pH 7.4) three times, they were collected on nanopercolator filters (JEOL) and embedded in low-melting-point agarose (Sigma-Aldrich). The samples were further fixed with 1.5% potassium ferrocyanide and 1% osmium tetroxide in the same buffer at 4°C for 1 h and then washed five times with Milli-Q water. The samples were incubated in 1% thiocarbohydrazide solution at 60°C for 1 h and washed five times with Milli-Q water. The samples were further reacted with 1% osmium tetroxide in Milli-Q water at 4°C for 1 h and washed five times with Milli-Q water. The cell samples were incubated in 4% uranyl acetate in Milli-Q water at 4°C for 2 h, washed five times with Milli-Q water, and then immersed in Walton’s lead aspartate solution at 60°C for 1 h. After being washed five times with Milli-Q water, they were dehydrated with an ethanol series (30%, 70%, 85%, 95%, and 99.5% for 15 min each and three times in 100% for 20 min) and propylene oxide for 20 min twice, followed by infiltration with Epon 812 resin (TAAB). Polymerization was performed at 65°C for 72 h. Semithin sections (500 nm) were cut with a diamond knife on an ultramicrotome (Leica EM UC6, Leica Microsystems, Wetzlar, Germany) and placed on a piece of silicon wafer. The sections were observed with a field emission scanning electron microscope (JSM-IT800SHL; JEOL, Tokyo, Japan) used at 3 kV with a back-scattered electron detector.

RNA sequencing and downstream analysis

The splenic naïve CD4 T cells (TCRβ⁺CD4⁺CD44^lowCD62L^high) were activated for 16 h with Dynabeads™ Mouse T-Activator CD3/CD28. Total RNA isolation, library preparation, RNA-seq, and downstream analysis were conducted through Genewiz. Total RNAs were prepared by RNeasy Mini kit (QIAGEN). Isolation of poly(A)⁺ RNAs from total RNAs was performed by NEBNext Poly(A) mRNA Magnetic Isolation Module (NEW ENGLAND Biolabs) according to the manufacturer’s instructions. Library preparation was performed using a TaKaRa SmartSeq Stranded kit (TaKaRa) according to the manufacturer’s instructions. Libraries were sequenced using a 150-base pair pair-end run on the DNBSEQ G-400 platform. The quality of the raw sequence fastq-formatted files was assessed using fastQC (v0.11.9). The sequences were trimmed by Cutadapt (v2.9) to obtain high-quality clean data. For mapping, trimmed sequences were aligned with HISAT2 (v2.2.0) via the mouse mm10 (UCSC) genome. For expression analysis, HTseq (v0.6.1) was used to estimate gene expression levels from clean paired-end data.

The DEseq2 Bioconductor package, a model based on a negative binomial distribution, was used for differential expression analysis. The estimates of dispersion and logarithmic fold changes incorporate data-driven prior distributions, and the P_adj values of genes were set to <.05 to detect differentially expressed genes. GO-TermFinder was used to identify Gene Ontology (GO) terms to annotate a list of enriched genes with a significant P_adj less than .05.

Proteomic analysis

CD4 SP thymocytes were sorted and stimulated or not stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 for 16 h. The cells were washed with PBS and lysed in 150 μl of a buffer containing 6 M guanidine-HCl, 100 mM HEPES-NaOH (pH 7.5), 10 mM Tris(2-carboxyethyl)phosphine (TCEP), and 40 mM chloroacetamide (CAA). The lysates were dissolved by heating and sonication, followed by centrifugation at 20 000 × g for 15 min at 4°C. The supernatants were recovered, and the proteins were purified by methanol–chloroform precipitation and solubilized in 20 µl of 8 M urea and 50 mM Tris-HCl (pH 8.0). After sonication, the protein solutions were diluted 8-fold with 50 mM Tris-HCl (pH 8.0) and digested with 1 µg of trypsin/Lys-C mixture (Promega) at 37°C overnight. The resulting peptide solutions were desalted using GL-Tip SDB (GL Sciences), evaporated in a SpeedVac concentrator, and redissolved in 0.1% trifluoroacetic acid (TFA) and 3% acetonitrile (ACN). Liquid chromatography-tandem mass spectrometry (LC‒MS/MS) analysis of the resulting peptides was performed on a nanoElute 2 coupled with a timsTOF HT mass spectrometer (Bruker). The peptides were separated on a 75-μm inner diameter × 150-mm C18 reversed-phase column (Nikkyo Technos) with a linear 5%–35% ACN gradient for 0–60 min, followed by an increase to 95% ACN for 1 min and a final hold at 80% ACN for 4 min. Data acquisition was conducted in dia-PASEF mode with an MS1 m/z range of 100–1700 and an ion mobility (1/K₀) range of 0.6–1.6. The ramp time was set to 100 ms with a 100% duty cycle. The MS2 polygon was manually defined on the basis of the region where peptides were predominantly detected, specifically using the following four vertices: Point 1 (m/z = 300, 1/K₀ = 0.64), Point 2 (m/z = 300, 1/K₀ = 0.8), Point 3 (m/z = 770, 1/K₀ = 0.97), and Point 4 (m/z = 770, 1/K₀ = 1.15). Within the defined polygon, an 8 Da isolation window was employed with an overlap of 0.5 Da for each window, resulting in 62 windows per cycle and a cycle time of 3.36 s. Protein identification was carried out in DIA-NN (version 1.9) against a mouse in silico spectral library. The parameters for constructing the library included trypsin as the digestion enzyme, one allowed missed cleavage, peptide lengths ranging from 7 to 45 amino acids, precursor charges of 2 to 4, and a fragment ion m/z range of 200–1800. Features such as library-free search, deep learning-based retention time and ion mobility predictions, N-terminal methionine cleavage, and cysteine carbamidomethylation were enabled. In the DIA-NN search settings, MS1 and MS2 mass accuracies were set to auto, neural network classifiers were configured for single-pass mode, and the quantification strategy was set to QuantUMS (high precision).

Statistical analysis

GraphPad Prism was used for statistical analysis. Statistical significance for the comparison of two groups was analyzed with an unpaired, two-tailed Student’s t test. One-way analysis of variance (ANOVA) coupled with Tukey’s multiple comparison test was employed for comparisons of more than three groups. Differences were considered significant at P < .05.

Data availability

Raw data from RNA-seq analysis were deposited in the NCBI Gene Expression Omnibus (GEO) under accession number GSE 291117. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the jPOST partner repository with the data set identifier PXD060794.

Results

Dysfunction of the proteasome β5i subunit in T cells leads to T-cell deficiency

To address the effect of mutated β5i (Psmb8) on T cells and avoid compensatory upregulation of β5 (Psmb5), we generated a conditional knockout allele of Psmb5 in which exon 2 of Psmb5 was surrounded by loxp sites (Psmb5^flox) (Supplementary Figure 1A). We crossed Psmb5^flox with Cd4-cre transgenic mice (CD4-cre^+/−Psmb5^flox/flox; hereafter referred to as KO) (Supplementary Figure 1B), and the deletion of the sequence surrounded by loxp was confirmed by PCR (Supplementary Figure 1C). KO mice were further crossed with Psmb8 (G197V)-KI mice (hereafter referred to as KI) (22) to generate mice in which T cells carried a mutation in Psmb8 (hereafter referred to as KIKO).

We first analyzed the expression of the β5i, β5, and β1i subunits in CD4 SP thymocytes from four different types of mice (WT, KO, KI, and KIKO) by western blotting (Fig. 1A). In WT T cells, mature β5i and β1i were detectable, whereas β5 expression was hardly detectable, which is consistent with a previous report (23). In KI T cells, both mature and insufficiently cleaved β5i and β1i were observed with elevated β5 expression, as we previously reported (22). In KO T cells, the expression of β5i was upregulated, probably due to compensatory expression in response to β5 deficiency. In KIKO T cells, β5 expression was barely detected, although both mature and insufficiently cleaved β5i and β1i remained detectable (Fig. 1A).

Figure 1.

Dysfunction of the proteasome β5i subunit in T cells leads to T-cell deficiency. (A) Immunoblot analysis of CD4 SP thymocyte lysates using Abs against the indicated proteins. A representative result from more than three experiments is shown. (B) Flow cytometry analysis of splenic T, B, and CD4⁺ T cells and CD8⁺ T cells from WT, KO, KI, and KIKO mice. The numbers in the dot plot represent the percentages of total splenocytes. The top and bottom bar graphs show the frequency and total number of the indicated cells, respectively. A representative result from more than three independent experiments is shown (n = 3 for each group). (C) Flow cytometry analysis of thymic T cells from the indicated mice. The numbers in the dot plot represent the percentages of total thymocytes. The top and bottom bar graphs show the frequency and total number of CD4 CD8 double-negative thymocytes (DN), CD4 CD8 double-positive (DP), CD4 single-positive (CD4 SP), and CD8 single-positive (CD8 SP), respectively. A representative result from more than three independent experiments is shown (n = 3 for each group). The bars indicate the means ± SDs. Statistical significance was determined using one-way ANOVA coupled with Tukey’s multiple comparison test. P values: n.s. not significant, *P < .05, **P < .01, ***P < .001, ****P < .0001.

To elucidate the impact of a dysfunctional β5i subunit in T cells, we first examined the frequencies and numbers of peripheral CD4⁺ and CD8⁺ T cells in KIKO mice. KIKO mice presented significantly reduced frequencies and numbers of splenic CD4⁺ and CD8⁺ T cells, whereas the numbers of splenic B cells remained unaffected (Fig. 1B). The numbers of other immune cells, such as neutrophils, monocytes, dentritic cells (DCs), and natural killer (NK) cells, in the spleen of KIKO mice were also unaffected (Supplementary Figure 2). In contrast, KI and KO mice presented unimpaired frequencies and numbers of splenic CD4⁺ and CD8⁺ T cells (Fig. 1B), indicating that neither β5 deficiency nor the presence of mutated β5i with compensatory upregulation of β5 affected T-cell homeostasis. In contrast to those in the spleen, the deficiency of CD4 SP and CD8 SP T cells in the thymus of KIKO mice was minimal (Fig. 1C). Collectively, these data demonstrate that combined dysfunction of the β5i and β5 subunits in T cells induces severe immunodeficiency.

Impaired proteasomal activity and accumulation of ubiquitinated protein in KIKO T cells

The G197V variant in β5i affects the assembly of immunoproteasomes in the cells of PRAAS patients (10). Impaired assembly of immunoproteasomes was also detected in KI cells (22). Therefore, we next examined whether KIKO T cells presented similar assembly defects in immunoproteasomes. The cell lysates from CD4 SP thymocytes from KIKO mice were separated by glycerol gradient centrifugation to determine the mode of immunoproteasome assembly. Assembly intermediates containing insufficiently cleaved β1i were detectable in cell lysates from KIKO mice, as well as in cell lysates from KI mice (Fig. 2A), indicating that in KIKO T cells, immunoproteasome assembly is disturbed. The impaired assembly of the immunoproteasome found in KIKO T cells was consistent with the finding that KIKO T cells presented reduced chymotrypsin-like, trypsin-like, and caspase-like activities (Fig. 2B). Although the accumulation of ubiquitinated proteins caused by proteasome dysfunction is common in PRAAS (10, 11), ubiquitin accumulation was not marked in Psmb8-KI cells, probably because of the upregulation of β5 expression (22). Therefore, we next analyzed whether the degradation of ubiquitinated proteins is impaired in KIKO T cells (Fig. 2C). Compared with KI T cells, KIKO T cells accumulated high-molecular-weight ubiquitinated proteins. Importantly, TCR stimulation increased the accumulation of ubiquitinated proteins in T cells (Fig. 2C). These data suggest that impaired protein homeostasis in KIKO T cells caused by proteasome dysfunction is responsible for severe immunodeficiency.

Figure 2.

Impaired proteasomal activity and accumulation of ubiquitinated proteins in KIKO T cells. (A) Proteasome subunit expression in CD4 SP thymocytes from the indicated mice. The cell lysates were fractionated by glycerol gradient centrifugation, and each fraction was analyzed by immunoblotting via Abs against the indicated proteins. A representative result from three independent experiments is shown. (B) Proteasome enzymatic activity was measured by a Proteasome-Glo Cell-Based Assay. A representative result from three independent experiments is shown (n = 3 for each group). The bars indicate the means ± SDs. Statistical significance was determined using one-way ANOVA coupled with Tukey’s multiple comparison test. P values: n.s. not significant, **P < .01, ***P < .001, ****P < .0001. (C) Accumulation of ubiquitinated proteins in CD4 SP thymocytes. The indicated samples were stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 for 16 h prior to analysis. A representative result from three independent experiments is shown.

Dysfunction of the proteasome β5i subunit in T cells induces cell cycle arrest

To investigate the molecular mechanism of T-cell deficiency in KIKO mice, we first analyzed their proliferative activity following TCR stimulation. Compared with WT T cells, both splenic CD4⁺ T cells and CD4 SP thymocytes from KIKO mice presented severe proliferative defects (Fig. 3A and B). The impaired proliferative activity observed in KIKO T cells was not simply due to defects in TCR signaling, as KIKO T cells upregulated the expression of early T-cell activation markers such as CD69 and CD25 following TCR engagement (Fig. 3C).

Figure 3.

Dysfunction of the proteasome β5i subunit in T cells induces cell cycle arrest. (A) Proliferation of CellTrace Violet-labeled splenic CD4⁺ T cells after 72 h of stimulation. A representative result from more than three independent experiments is shown. (B) Proliferation of CD4 SP thymocytes with CellTrace Violet labeling after 72 h of stimulation. A representative result from more than three independent experiments is shown. (C) Expression of CD69 and CD25 after 2 h, 4 h, and 6 h of stimulation. The number in the histograms indicates the frequency of the indicated marker among live, splenic CD4⁺ T cells. Filled and open histograms represent naïve and activated T cells, respectively. A representative result from two independent experiments is shown. (D) Volcano plot showing differentially expressed genes in naïve T cells (left panel) and stimulated T cells (middle panel) from KI and KIKO mice. Principal component analysis (PCA) of the indicated samples (right panel). (E) Gene set enrichment analysis of downregulated genes in stimulated T cells from KIKO mice compared with those from KI mice via the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway. (F) Heatmap representing genes downregulated in KIKO mice compared with KI mice, on the basis of RNA-seq data from stimulated samples shown in the middle panel of (D). (G) Cell cycle analysis of splenic CD4⁺ T cells from the indicated mice was performed via Vybrant DyeCycle Violet after 72 h of stimulation. The bar graph represents the frequency of the indicated cells among the live CD4⁺ T cells (n = 4 for each group). A representative result from more than three independent experiments is shown. (H) Cell cycle analysis of CD4 SP thymocytes from the indicated mice was performed via Vybrant DyeCycle Violet after 72 h of stimulation. The bar graph represents the frequency of the indicated cells among the CD4 SP thymocytes (n = 3 for each group). The bars indicate the means ± SDs. Statistical significance was determined using one-way ANOVA coupled with Tukey’s multiple comparison test for (G) and an unpaired, two-tailed Student’s t test for (H). P values: n.s. not significant, **P < .01, ***P < .001, ****P < .0001. A representative result from more than three independent experiments is shown.

To analyze the molecules associated with impaired proliferation in KIKO T cells, we performed transcriptomic analysis. We first compared the transcriptomes of naïve splenic CD4⁺ T cells between KI and KIKO mice. Among the 14 157 genes, 421 genes were significantly upregulated [false discovery rate (FDR) < 0.05, fold change ≥ 2.0] and 203 genes were downregulated in KIKO CD4⁺ T cells (Fig. 3D, left panel). Furthermore, more genes were differentially expressed when we compared the transcriptomes of activated splenic CD4⁺ T cells between KI and KIKO mice. We identified 1295 upregulated and 949 downregulated genes among 14569 genes in KIKO T cells (Fig. 3D, middle panel). Principal component analysis (PCA) also revealed clear segregation between activated KI and KIKO T cells (Fig. 3D, right panel). Gene set enrichment analysis of downregulated differentially expressed genes (DEGs) in activated KIKO T cells revealed enrichment of cell cycle-related genes along with genes involved in various metabolic pathways, including glycolysis, gluconeogenesis, pyruvate metabolism, and one-carbon metabolism (Fig. 3E).

As cell cycle-related genes were highly enriched among downregulated DEGs in activated KIKO T cells compared with KI T cells (Fig. 3F), we examined whether KIKO T cells presented defects in the cell cycle. At 72 h following TCR stimulation with anti-CD3/CD28 mAbs, the splenic CD4⁺ T cells from the WT, KO, and KI mice showed an unimpaired cell cycle (Fig. 3G). In contrast, splenic CD4⁺ T cells from KIKO mice exhibited cell cycle arrest at the G₀/G₁ phase, which was consistent with the observation that the proliferative activity of KIKO T cells was completely impaired (Fig. 3G). Cell cycle arrest in the G₀/G₁ phase was also evident in activated CD4 SP thymocytes from KIKO mice (Fig. 3H). Collectively, these findings suggest that severe T-cell deficiency in KIKO mice is attributable primarily to impaired proliferative activity with cell cycle arrest.

Proteasome dysfunction in T cells induces rapid cell death through apoptosis

Since the accumulation of ubiquitinated proteins was detectable in KIKO T cells, we next sought to evaluate which proteins accumulated in KIKO T cells because of proteasome dysfunction. To address this, quantitative proteomic analysis using data-independent acquisition mass spectrometry was performed. Among the 7664 proteins, 241 proteins were significantly upregulated and 242 proteins were downregulated in KIKO T cells after TCR stimulation (P < .05, fold change ≥ 1.5) (Fig. 4A). Only a few differentially expressed proteins were detected in naïve T cells between KIKO and KI T cells (Supplementary Figure 3A), which is consistent with the finding that the accumulation of ubiquitinated proteins was highly inducible in KIKO T cells following TCR engagement (Fig. 2C, Supplementary Figure 3B and 3C). Importantly, only eight proteins were overlapped between upregulated proteins in naïve KIKO T cells and activated KIKO T cells (Supplementary Figure 3D). Among the 241 upregulated proteins in stimulated KIKO T cells, the p53 protein, which plays a crucial role in regulating the cell cycle, DNA repair, and apoptosis (24, 25), was identified (fold change; 2.91) (Fig. 4A and B). We also confirmed the upregulated expression of p53 in activated KIKO T cells by FACS (Fig. 4C). Therefore, we next analyzed whether apoptosis is responsible for the rapid induction of cell death in KIKO T cells following TCR engagement. After 24 h of TCR stimulation, the frequency and absolute number of early apoptotic cells, defined as Annexin V⁺ cells, among KIKO T cells were greater than those among KI T cells (Fig. 4D). KIKO T cells also presented a greater frequency and number of late apoptotic cells, defined as Annexin V⁺ 7-AAD⁺ cells, than did KI T cells (Fig. 4D). Treatment of KIKO T cells with the caspase inhibitor Z-VAD completely suppressed cell death (Fig. 4D).

Figure 4.

Proteasome dysfunction in T cells induces cell death through apoptosis. (A) Volcano plot showing differentially expressed proteins in stimulated T cells from KI and KIKO mice. (B) Heatmap representing the protein abundance levels in the indicated mice, on the basis of proteome data analysis of the stimulated samples shown in (A). (C) Representative histogram (left) and mean fluorescence intensity (MFI) (right) of p53 expression among splenic CD4⁺ T cells from the indicated mice after 16 h of stimulation. Color-matched isotype controls were used as negative controls (n = 3 pooled samples). The bars indicate the means ± SDs. Statistical significance was determined using an unpaired, two-tailed Student’s t test. P values: **P < .01. A representative result from three independent experiments is shown. (D) Flow cytometry analysis of Annexin V and 7-AAD staining of CD4 SP thymocytes from the indicated mice after 24 h of stimulation (left panel). The numbers in the dot plot represent the percentages of CD4 SP thymocytes. Annexin V⁺ cell frequency and number among CD4 SP thymocytes (middle panel). Apoptotic cell frequency and number among CD4 SP thymocytes (right panel) (n = 3 for each group). The bars indicate the means ± SDs. Statistical significance was determined using one-way ANOVA coupled with Tukey’s multiple comparison test. P values: n.s. not significant, *P < .05, **P < .01, ***P < .001, ****P < .0001. A representative result from three independent experiments is shown.

p53 is well known for inducing apoptosis by activating its downstream targets (26). p53 was one of the proteins that significantly accumulated in activated KIKO T cells but not in naïve KIKO T cells (Fig. 4A and Supplementary Figure 3A). Therefore, we next analyzed whether the treatment of KIKO T cells with a p53 inhibitor could inhibit cell death. We found that the treatment of KIKO T cells with a p53 inhibitor during TCR stimulation did not affect the frequency of apoptotic cells, suggesting that the apoptosis induced in T cells by proteasome dysfunction is p53-independent (Supplementary Figure 4A).

UCH-L1 (ubiquitin C-terminal hydrolase L-1) is another protein that accumulated at significant levels in activated KIKO T cells but not in naïve KIKO T cells (Fig. 4A and Supplementary Figure 3A). UCH-L1 is a deubiquitinating enzyme that is responsible for recycling free ubiquitin (27). UCH-L1 is expressed in neurons at high levels and is associated with α-synuclein, which is a protein whose accumulation is linked to Parkinson’s disease (28). Therefore, we speculate that the accumulation of UCH-L1 in activated KIKO T cells might be responsible for cell death. To address this, activated T cells from WT mice were transduced with retrovirus carrying UCH-L1. However, UCH-L1 overexpression in activated WT T cells did not induce apoptosis, suggesting that the accumulation of UCH-L1 is not required for the apoptosis of T cells induced by proteasome dysfunction (Supplementary Figure 4B). Collectively, these data indicate that p53-independent apoptosis is involved in immunodeficiency caused by proteasome dysfunction in T cells.

Proteasome dysfunction in T cells affects neither mitochondrial function nor the unfolded protein response

Naïve T-cell activation is accompanied by dynamic changes in mitochondrial biogenesis (29). It has been reported that proteasome inhibition in cells induces mitochondrial aberration, leading to cytosolic oxidation and cell death (30, 31). Therefore, we examined whether mitochondrial dysfunction is involved in apoptosis in KIKO T cells. The mitochondrial membrane potential and mitochondrial size were comparable between KI and KIKO T cells 16 h after TCR stimulation (Fig. 5A). Compared with KI T cells, KIKO T cells presented similar levels of mitochondrial superoxide production (Fig. 5B). Electron microscopic analysis also revealed that KIKO T cells presented no significant changes in mitochondrial size or the number of mitochondrial cristae (Fig. 5C). Consistent with unimpaired mitochondrial functions in KIKO T cells, ATP production was not affected (Fig. 5D). Importantly, ATP production in WT T cells treated with proteasome inhibitor b-AP15 was unimpaired (Fig. 5D).

Figure 5.

Proteasome dysfunction in T cells does not affect mitochondrial function or ER stress responses. (A) KI and KIKO CD4 SP thymocytes were analyzed for mitochondrial membrane potential (left panel) using tetramethylrhodamine methyl ester perchlorate (TMRM) and for mitochondrial size (right panel) using MitoTracker. A representative result from more than three independent experiments is shown. (B) Mitochondrial superoxide production was measured in the CD4 SP thymocytes of KI and KIKO mice via MitoSOX. A representative result from more than three independent experiments is shown. (C) Electron microscopy of KI and KIKO CD4 SP thymocytes stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 for 16 h. Scale bars represent 0.5 μm. (D) ATP production levels were measured via the CellTiter-Glo Luminescent Cell Viability Assay. b-AP15 was used as a proteasome inhibitor (n = 2 for each group). The bars indicate the means ± SDs. Statistical significance was determined using one-way ANOVA coupled with Tukey’s multiple comparison test. P values: n.s. not significant. A representative result from more than three independent experiments is shown. (E) The expression of CHOP and BiP by activated splenic CD4⁺ T cells was measured by real-time PCR (n = 3 for each group). The bars indicate the means ± SDs. Statistical significance was determined using an unpaired, two-tailed Student’s t test. P values: n.s. not significant.

Several studies have shown that proteasome dysfunction leads to the accumulation of unfolded proteins in the ER, leading to the initiation of the unfolded protein response (UPR) (32). Therefore, we analyzed whether a defect in β5i in T cells induces the UPR. We analyzed the expression levels of ER stress markers such as C/EBP Homologous protein (CHOP) and Binding immunoglobulin protein (BIP). Following TCR engagement, KI and KIKO T cells expressed similar levels of CHOP and BIP (Fig. 5E). Collectively, these findings indicate that proteasome dysfunction in T cells induces apoptosis via a caspase-dependent, p53-independent pathway, which is distinct from mitochondrial or ER stress-mediated cell death. This unique mechanism underscores the critical role of β5i-mediated proteasome function in T-cell survival and homeostasis.

Discussion

Mutations in genes encoding proteasome subunits cause PRAAS, characterized by progressive partial lipodystrophy, panniculitis, and hepatosplenomegaly (10, 11). Recently, it has become evident that dysfunction of certain proteasome subunits can lead to severe immunodeficiency (18, 20). However, the precise mechanism by which defects in proteasome function induce immunodeficiency remains unknown. Furthermore, it remains unclear which proteasome subunits are responsible for immune cell development. In this study, we established a novel mouse model of immunodeficiency caused by a missense mutation (G197V) in the immunoproteasome subunit β5i (Psmb8) in conjunction with a deficiency in the standard proteasome subunit β5 (Psmb5) specifically in T cells. Using this model, we demonstrated that apoptosis induction is a key mechanism underlying immunodeficiency caused by β5i dysfunction in T cells.

It has been reported that mice with the β1i (G156D) mutation found in PRAAS-ID patients are immunodeficient (18). This β1i missense variant induces activity defects in the 20S complex but not in the 26S proteasome without increasing the accumulation of ubiquitinated proteins. Immunodeficiency has also been reported in mice harboring a missense mutation of β2i (G170W) introduced by N-ethyl-N-nitrosourea mutagenesis, which leads to proteasome activity defects similar to those associated with the β1i (G156D) mutation (19). In contrast, β5i G197 is located at the S1 substrate pocket, and the β5i G197V mutation results in impaired activity in both the 20S and 26S immunoproteasomes (10), which leads to the accumulation of ubiquitinated proteins, suggesting that insufficient degradation of ubiquitinated proteins in T cells triggers immunodeficiency in KIKO mice. Since the β1i and β2i subunits are also components of the thymoproteasome expressed in cortical thymic epithelial cells (cTECs), dysfunction of cTECs might be responsible for the T-cell immunodeficiency induced by those mutations.

The inhibition of proteasome-mediated protein degradation in CHO cells with the proteasome inhibitor MG132 results in mitochondrial aberration through intracellular oxidation, which eventually leads to cell death (31). However, we observed that T cells with proteasome dysfunction underwent apoptosis without exhibiting significant mitochondrial damage. These findings suggest that the cell death mechanism in these T cells differs from the mitochondrial-dependent pathways reported in other cell types. In fact, ATP production in T cells was not affected by proteasome inhibitor b-AP15 treatment. Furthermore, although prolonged and irreversible ER stress also induces apoptosis (33, 34), we did not detect any activation of ER stress upon evaluating various ER stress markers in KIKO T cells. Upon TCR stimulation, treatment with Z-VAD, a pancaspase inhibitor, rescued these cells from apoptosis, whereas treatment with a p53 inhibitor did not confer protection. These findings suggest that proteasome dysfunction in T cells leads to apoptosis via a caspase-dependent, p53-independent mechanism, which is distinct from the mitochondrial and ER stress-related pathways observed in other cell types.

To investigate the mechanisms underlying enhanced apoptosis in KIKO T cells, we conducted a proteomic analysis comparing protein expression profiles between naïve and TCR-stimulated control and KIKO T cells. Our analysis revealed 17 significantly upregulated proteins in naïve KIKO T cells and 241 in TCR-stimulated KIKO T cells, with only eight proteins overlapping between the two conditions. This limited overlap suggests that the majority of the upregulated proteins in stimulated KIKO T cells are dependent on TCR engagement, which aligns with our observation that apoptosis induction in these cells requires TCR stimulation. Despite these findings, we have been unable to identify specific proteins responsible for triggering apoptosis in KIKO T cells. Identifying such proteins would be instrumental in understanding the pathways leading to apoptosis due to proteasome dysfunction and could inform therapeutic strategies for related immunodeficiencies. In addition, we did not observe a significant reduction in T-cell numbers within the thymus of KIKO mice, although cell cycle arrest at the G₀/G₁ phase was evident in CD4 SP thymocytes following ex vivo TCR stimulation. This finding suggests that, unlike those in the periphery, T cells in the thymus do not actively divide in response to TCR engagement in vivo. Consequently, this distinction highlights the potential role of TCR-induced upregulated proteins as key mediators of apoptosis, specifically in peripheral T cells.

In this study, we demonstrated that mice with β5i dysfunction in T cells exhibit severe immunodeficiency, characterized by a significant reduction in T-cell numbers. Following TCR engagement, β5i dysfunction rapidly induces apoptosis in T cells, independent of mitochondrial damage or ER stress responses. These findings suggest that the underlying mechanism of cell death in these T cells is distinct from previously reported mitochondrial-dependent or ER stress-induced apoptotic pathways. Furthermore, our findings provide valuable insights into the role of immunoproteasomes in maintaining T-cell homeostasis and highlight β5i as a critical factor for T-cell survival and function. Given the central role of T cells in immune regulation, our study not only advances the understanding of immunodeficiency in PRAAS but also suggests potential links between proteasome dysfunction and T-cell deficiencies observed in other immune-related disorders. Future studies aimed at identifying specific apoptotic triggers in β5i-deficient T cells may identify novel therapeutic targets for treating proteasome-related immunodeficiencies.

Supplementary data

Supplementary data are available at International Immunology Online.

Funding

This work was supported by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS) KAKENHI (grant numbers JP22H05188 and JP22H05182), Moonshot R&D from the Japan Sciency and Technology Agency (JST) (grant number JPMJMS2025) and Support for Pioneering Research Initiated by the Next Generation from JST (grant number JPMJSP2113). This work was supported by the COI-NEXT Support Unit for Imaging Science at Kento and JSPS J-PEAKS.

Author contributions

E.S., J.M., Y.S. and K.Y. designed the experiments. E.S., J.M., R.K., and T.N.N. conducted the experiments. S.Y. conducted the electron microscopy analysis. H.K. conducted the proteomic analysis. E.S., J.M., and K.Y. wrote the paper. K.Y. supervised all the research.

Conflict of interest statement: The authors declare that they have no conflicts of interest.