Diagnostic Utility of Proteogenomic Analysis for Inherited Bone Marrow Failure Syndrome

Background: Inherited bone marrow failure syndrome (IBMFS)-a heterogeneous group of disorders characterized by marked cytopenia in one hematopoietic cell lineage-may progress to pancytopenia and make the patient susceptible to hematological malignancies or solid tumors. No clinically useful screening tests have been established for IBMFS, with a few exceptions, including the chromosome fragility test in Fanconi anemia (FA) and telomere length measurement in dyskeratosis congenita (DC). Therefore, high-throughput sequencing is widely used for not only clinical diagnosis of IBMFS but also screening applications; however, a significant number of patients remain genetically undiagnosed. Recent technological advances in in-depth data-independent acquisition proteomic analysis have enabled comprehensive quantitative analysis of >9000 proteins, including very low-abundant proteins. Integration of proteomic analysis with genomic information has been performed in brain tumors and acute leukemia, showing the potential to bridge the gap between genetic information and clinical validation; however, to our knowledge, proteogenomics in IBMFS has not been fully evaluated thus far. Herein, an omics analysis of patients with various IBMFS was performed to assess the diagnostic utility of in-depth proteomic techniques.

Methods: We performed an omics analysis of the discovery cohort of 60 patients with IBMFS for in-depth proteomic analysis, targeted capture DNA sequencing, and RNA sequencing. For non-targeted proteomic analysis, digested peptides from the peripheral blood mononuclear cell samples were processed using a Q Exactive HF-X (Thermo Fisher Scientific, Waltham, MA, USA) data-independent acquisition mass spectrometer. Furthermore, diagnostic proteomic analyses targeting several IBMFS-related proteins were performed on a developmental cohort (417 samples; IBMFS [n = 229], aplastic anemia [n = 75], myelodysplastic syndrome/acute myeloid leukemia [n = 113], and healthy controls [n = 27]) using an Orbitrap Exploris 480 mass spectrometer (ThermoFisher Scientific) equipped with an EVOSEP ONE system (EVOSEP, Odense, Denmark). This study was approved by the ethics committee of the Nagoya University Graduate School of Medicine.

Results: Non-targeted proteomic analysis was performed on 74 samples obtained from 60 patients with IBMFS and 14 healthy controls. Unsupervised clustering identified eight independent proteomic clusters (C1-C8), with ribosomal pathway-associated protein expression specifically downregulated in C1 and C2. Patients with Diamond-Blackfan anemia, Shwachman-Diamond syndrome (SDS), FA, ADH5/ALDH2 deficiency, and DC were enriched in C1, C2/C8, C3/C4, C6, and C7 clusters, respectively. In particular, among the 74 samples, four patients with ADH5/ALDH2 deficiency showed significantly reduced ADH5 protein expression, whereas the remaining patients showed normal expression, with 100% diagnostic concordance. In addition, 6 of 74 samples had significantly decreased SBDS protein expression, including two patients with monoallelic pathogenic variants in the SBDS gene. This suggested that proteomic analysis may play a complementary role to genetic analysis. To provide a large-scale rapid screening system for IBMFS in a practical clinical setting, targeted proteomic analysis was performed using a small panel of ADH5, SBDS, and GAPDH proteins in a developmental cohort of 417 patients. We demonstrated the significantly decreased SBDS protein expression levels in patients with SDS ( P = 4.98 × 10 ^-9). ADH5 protein expression levels were also significantly reduced in ADH5/ALDH2 deficiency ( P = 1.66 × 10 ^-6). Regarding the diagnostic tests, the sensitivity and specificity values were 85.7% and 93.4% for SDS, and 100.0% and 97.5% for ADH5/ALDH2 deficiency, respectively, indicating the sufficient performance to identify patients who required early diagnosis and therapeutic intervention.

Conclusion: We performed the first proteogenomic analysis of patients with IBMFS and identified eight independent proteomic clusters associated with IBMFS subtypes and characteristic pathways. Furthermore, the clinical application of targeted proteomic assays constructed from these results may help diagnose and screen IBMFS, including SDS and ADH5/ALDH2 deficiency, for which appropriate clinical screening tests are lacking.