Overexpression of Ribosomal Proteins Is Associated with Resistance to Immunochemotherapy in Diffuse Large B-Cell Lymphoma Patients

Introduction: Diffuse Large B-cell Lymphoma (DLBCL) patients with early relapse or refractory disease have a very poor outcome. The major prognostic tool used today is still the clinically based International Prognostic Index (IPI). Yet, IPI seems to be less useful in the identification of individual patients with high risk of failing immunochemotherapy. Despite the higher risk of treatment failure in patients with biological subtypes such as ABC-DLBCL and double-hit DLBCL, about 30-40 % are long term progression-free with current therapy. So, novel biomarkers that more accurately can identify the high-risk patients are warranted, and there is a need to further explore underlying tumor biology features causing treatment resistance.

Aims: We compared the proteome in tumor tissue from two groups of DLBCL patients treated with modern immunochemotherapy, i) early relapse/refractory patients and ii) long-term progression-free patients, with the aim to determine if differences in clinical outcome could be correlated to diverse proteomic profiles.

Patients and Methods: Formalin-fixed paraffin-embedded tumor tissue, collected at diagnosis, from 93 DLBCL patients, divided into two groups: A) early relapse, i.e. within 12 months, or primary refractory patients, n=44 (REF/REL), and B) cured patients, i.e. with a follow-up of at least 5 years without relapse, n=49 (CURED). Micro-dissected tissue was analyzed using TMT-based mass spectrometry (MS): equal amounts of protein solutions from each patient sample and a reference pool were trypsin-digested into peptides using the FASP-method. NanoLC-MS/MS analysis was performed using a QExactive-Easy-nLC 1,000 instrument combination. Raw data were merged during the database search for protein identification and relative quantification using Proteome Discoverer.

Results: In total, we identified 1996 proteins of which 443 were found in all patient samples. Of the 443 proteins, 86 were differentially expressed between the two patient groups. Fifty-seven proteins were overexpressed in the REF/REL group, among which 36 were ribosomal proteins (RP), e.g. RPS2, S3, S5, S6, S7, S13, S15a, S23, L3, L5, L6, L7a, L10, L13a, L18, L31 and L32. In contrast, only 1 RP was overexpressed in the CURED group (RPS10) (p<0.0000001). Among individual proteins overexpressed in the REF/REL group we found several proteins that previously have been described as negative prognostic factors in other malignancies, e.g. IMPDH2, MARCKS, NPM1, Prohibitin, RACKS1 and SF3B3. Interestingly, we also found that 10 of 27 overexpressed proteins in the CURED group were associated with the actin-regulating network, while no such proteins were found in the REF/REL group (p<0.00001). Additionally, we performed a multivariate analysis (OPLS-DA), including the significant 87 peptides, to discriminate between the REF/REL and CURED groups: we observed a separation, and a variance analysis suggested a valid model (CV-ANOVA; p=3.7 x 10^-7).

Discussion: We found that a number of RP was overexpressed in DLBCL patients with early relapse or refractory disease. RP play a fundamental role not only in ribosome biogenesis and protein translation but there is also accumulating evidence that dysregulated RP could contribute to tumorigenesis, tumor progression and metastasis. Furthermore, multidrug resistance in some other cancer forms has been associated with overexpression of RP. In DLBCL, apart from overexpression of RPS6 (also found in this study), there is no previous available data on RP dysregulation.

Conclusion: Overexpression of multiple ribosomal proteins appears to be associated with failure to immunochemotherapy in DLBCL, and further understanding of their role could potentially lead to new therapeutic strategies in this disease.