PRKRA associated with NF-κB activation in pulmonary involvement of idiopathic multicentric castleman disease

Abstract Idiopathic multicentric Castleman disease (iMCD) is a heterogeneous group of HIV- and HHV-8-negative lymphoproliferative disorders characterized by systemic hyperinflammation and multi-regional lymphadenopathy, often involving organs like the kidney, skin, hepatobiliary tract, and lung. Pulmonary involvement (PI), caused by ectopic lymphoid infiltration and tissue damage, is a significant clinical manifestation. However, the pathogenesis of pulmonary tropism in iMCD remains undefined, and biomarkers to monitor lung disease activity are lacking. Consequently, identifying accessible molecular and cellular targets in iMCD-PI represents an urgent unmet need. To investigate potential targets, we first performed a discovery-phase serum proteome screen of 11,000 analytes (Somalogic SomaScan) in 23 iMCD-PI and 16 iMCD-non-PI patients sampled during disease flare. Interferon-induced double-stranded RNA-dependent protein kinase activator (PRKRA) emerged as the most differentially elevated protein, showing a 2-fold higher relative fluorescence unit (RFU) value in iMCD-PI versus iMCD-non-PI. This finding was validated in an expanded flare cohort (iMCD-PI n = 37; iMCD-non-PI n = 27) by orthogonal ELISA. Serum PRKRA concentrations were significantly higher in iMCD-PI than in iMCD-non-PI (P=0.027). A linear correlation was also observed between absolute PRKRA concentrations measured by ELISA and log2-transformed SomaScan RFU values (Pearson r = 0.402, P=0.015). Notably, while PRKRA is a cytoplasmic protein physiologically highly expressed in the brain and moderately expressed in the digestive tract, bladder, and gonads, it is not expressed in normal lung tissue nor typically detectable at significant levels in serum. We therefore hypothesized that the elevated circulating PRKRA originates from damaged involved tissue, specifically the lungs of iMCD-PI patients. To test this, we then performed immunohistochemical staining for PRKRA on lung tissue sections from iMCD-PI patients (n=9) and control patients with primary spontaneous pneumothorax (n=4; histologically normal lung tissue >50% per section). Given that PRKRA functions upstream of NF-κB, we assessed NF-κB pathway activation within lung tissue from all 13 cases, using phosphorylated NF-κB p65 (ser536, phospho-p65) as a read-out for NF-κB activity, and simultaneously evaluated the intermediate signalling molecule protein kinase R (PKR) and total P65 levels. Whole-slide scanning of lung tissue sections was performed using the Pannoramic 250 digital scanner. Digital images were analysed with AIPathWell software (v3.2.1) for automated quantification across entire tissue sections. Immunohistochemical staining intensity was quantified using the H-score system applied to all tissue regions: H-score = (% weakly positive cells × 1) + (% moderately positive cells × 2) + (% strongly positive cells × 3). This scoring method evaluates the percentage of positively stained cells at each intensity level across the complete tissue section. PRKRA expression was significantly elevated in iMCD-PI versus controls (median H-score 169.83 vs. 21.15, P=0.006), localizing to bronchial/alveolar epithelium and tertiary lymphoid structure plasma cells. iMCD-PI lungs showed increased PKR (median H-score 238.26 vs. 151.23, P=0.034) and phospho-p65 (median H-score 232.64 vs. 184.11, P=0.020) but comparable total P65 (P=0.199). Spearman correlations revealed: PRKRA-PKR (r=0.73, P=0.007), PRKRA-phospho-p65 (r=0.62, P=0.029), and PKR-phospho-p65 (r=0.79, P=0.002). These findings demonstrate PRKRA-PKR-phospho-p65 axis coordination, indicating PRKRA might drive NF-κB-mediated pulmonary inflammation via PKR activation in iMCD-PI. Signalling pathways driving iMCD-PI pathogenesis remain poorly understood. Here we provide the largest quantitative assessment of aberrant NF-κB activity in iMCD-PI to date and, to our knowledge, the first systematic demonstration that NF-κB is selectively activated in affected lungs. These findings advance our understanding of organ-tropic inflammatory mechanisms in iMCD and suggest that circulating PRKRA concentrations may serve as an accessible biomarker for monitoring pulmonary disease activity and treatment response. More importantly, the PRKRA–PKR–NF-κB axis represents a previously unrecognised, druggable pathway for targeted therapy in iMCD-PI.