Among children, ALL is the most common type of cancer [11]. Over the last decades, considerable improvement in the treatment of patients with ALL has been achieved [12]. However, outcomes of relapsed or refractory ALL remain poor [10], and in terms of cancer-related deaths, ALL still ranks as the leading cause [13]. Evasion from chemotherapy-induced apoptosis could be the potential mechanism causing ALL chemoresistance or relapse [14, 15]. Thus, exploiting treatments that induce cell death independent on classical apoptosis may improve ALL outcomes.
Necroptosis is a form of PCD characterized by a molecularly and genetically defined necrotic program that is mechanistically similar to apoptosis and morphologically similar to necrosis [7, 16]. Numerous studies have demonstrated that necroptosis plays a crucial role in the initiation, progression, and metastasis of cancer [17, 18]. Scott McComb et al. showed [19] that in samples from highly resistant ALL patients, necroptosis activation could eliminate refractory leukemia cells. However, the effects of necroptosis on chemoresistance and prognosis in ALL are still unclear.
In the present study, we identified genes that are affected by necroptosis and assessed the correlation with prognosis in ALL patients. A genetic risk-scoring model was developed to predict the survival of patients. A functional analysis of this prognostic model revealed differences between high-risk and low-risk subgroups in pathways and biological functions, including drug resistance and immune microenvironment.
Using publicly available ALL datasets, we showed that NRGs could stratify patients into high- and low-risk patient subgroups. A prognostic risk model based on 7 NRGs was constructed using LASSO regression analysis. Depending on each gene's risk value, BIRC2, PKP3, MERTK, KL, ESR2, and TLE6 were regarded as risk genes related to poor prognosis in patients with ALL, whereas TET2 were associated with favorable prognosis.
BIRC2, acts as an E3 ubiquitin-protein ligase, regulating nuclear factor kB (NF-kB) signaling, with the ability to regulate apoptosis, necroptosis, proliferating, migrating, autophagy, and immunity [20–22]. BIRC2 degradation in cells leads to RIPK1 de-ubiquitination and autocrine TNF-α signaling, which trigger either necroptotic or apoptotic cell death[23]. Studies have reported that BIRC2 is overexpressed in many cancers including cervical cancer, breast cancer, lung cancer, and bladder cancer, which can lead to resistance to treatment or significantly worse outcomes[21, 22, 24, 25].
Armadillo-related proteins are involved in signal transduction, cell adhesion, and tumorigenesis. [26]. In various kinds of cancer, PKP3 plays an important role as the most widely expressed member of the armadillo protein family [26–28]. Furukawa et al. reported that upregulation of PKP3 expression was observed in lung tumor cells and RNAi-mediated knockdown suppressed cell growth, while overexpression enhanced it in vitro[29]. Researchers found that PKP3 was highly expressed in ovarian cancer cells, affecting cell proliferation, formation, and invasion [28].
First cloned from a human B lymphoblastoid expression library by Graham et al. [30], MerTK is a major macrophage receptor involved in the clearance of apoptotic cells [31]. In malignant tumors such as melanoma, gastric cancer, leukemia, and lung cancer, aberrant MerTK expression plays a pivotal role in oncogenesis [32–36]. Hematopoietic malignant tumors derived from B or T cells may be promoted by MerTK. Ectopic expression of MerTK has been reported in B-ALL, T-ALL, and acute myeloid leukemia (AML)[34, 37, 38], whereas inhibition of MerTK reduces downstream signaling, inhibits proliferation and invasion, promotes apoptosis, and induces chemosensitivity in tumor cells [39–41].
KL has been considered as an anti-aging gene[42], and also has a close relationship with cancer, by regulating tumorigenesis, cell survival, and metastasis[43] [44]. Previous studies showed that overexpression of KL inhibited cancer cell proliferation by regulating the insulin-like growth factors-1 (IGF-1) pathway in breast and lung cancer cells indicating positive relation with cancer prognosis [45, 46], which is inconsistent with our research.
Estrogen receptors β (ERβ), one of mediator of estrogen signaling, is encoded by ESR2 gene [47]. ERβ is mostly expressed in bone marrow stem cells, prostate tissues, lung, and colon [48–51]. Conflicting reports exist regarding ERβ's prognostic value. On one hand, amount of studies reported that ERβ activation resulted in pro-death signaling in solid tumors, such as prostate, colon, and breast cancer tissues [52, 53]. As for hematopoietic malignancies, ERβ activation strongly inhibited tumor growth in lymphoma [54], and low levels of ERβ or knockdown of ERβ confirmed resistance to diosmetin in AML cells [55]. In chronic lymphocytic leukemia (CMML) cells, selective ERβ agonists may influence growth and induce apoptosis of these cells [56]. On the other hand, some studies reported the opposite prognostic value of ERβ, where increased expression of ERβ in breast cancer patients predicted poor prognosis [57]. In our study, ESR2 was considered as a negative prognostic gene for ALL patients.
TLE proteins are transcriptional corepressors that maintain stem/progenitor cell state and inhibit differentiation in a variety of tissues [58]. It was reported that FOXG1 and TLE inhibition decreased brain tumor growth [59]. Also, in colon cancer cell lines, TLE6D overexpression increases cell proliferation, colony formation ability, cell migration, and tumorigenicity of xenografts [60].
The TET2 gene is involved in DNA demethylation, mainly catalyzing the conversion of 5-methylcytosine (5-mC) to 5-hydroxymethylcytosine (5-hmC) to facilitate DNA demethylation. [61]. It was reported one of the most commonly mutated genes in hematopoietic malignancies. TET2 mutations are seen in AML, T-ALL, myelodysplastic syndromes (MDS), chronic myelomonocytic leukemia (CMML), and peripheral T-cell lymphomas (PTCL) [62, 63]. It has been shown in vivo that knocking out TET2 causes myeloid or lymphoid cell spread, as well as the progression of fully infiltrating aggressive tumors, indicating that TET2 plays a crucial role in maintaining hematopoiesis [64].
Expectedly, the 7-gene signature can well predict the prognosis of ALL patients. According to risk score, ALL patients can be divided into high- and low-risk groups, and we found that the survival was notably different between the two risk groups. Importantly, the risk score was identified as the independent prognostic factor through univariate and multivariate analysis. The risk score was found to have the strongest impact on the risk category and survival of patients with ALL, which can be used to effectively determine prognosis. Moreover, our prognostic model has also been extended clinically with a prognostic nomogram combining clinical parameters with the risk model.
Currently, Tumor cells hijacking the immune system is believed to be an important cause of ALL progression or relapse [65, 66], and thus, The use of emerging immunotherapies to enhance the immune response against tumor cells is growing in the clinical treatment of ALL [14, 67, 68]. Tumor microenvironment (TME) mechanisms contribute to immune evasion. Previous studies showed that ALL blasts may increase immunosuppressive cells such as regulatory T Cells (Tregs), M2-type macrophages, and myeloid-derived suppressor cells (MDSCs) and decrease immunoreactive cells such as cytotoxic T cells (CTLs) and natural killer (NK) cells [69]. Moreover, monocyte abundance has connection with inferior relapse-free and overall survival in both pediatric and adult B-ALL patients [70]. Interestingly, our study also indicates that immune cell infiltration and risk score are correlated. The risk score was significantly positively related to plasma cells (PC). Previous studies showed that plenty of evidence supported a positive role for PC in antitumor immunity although the role of PC in tumors remains controversial [71], which is consistent with our study. Moreover, the 7 NRGs based risk score had significant positive correlation with the immune inhibitors CD244 and TIGIT, and showed significant negative correlation with the immune stimulator CXCR4 and TNFRSF9. The high-risk group showed higher levels of CCL2, CXCL10, CXCL3, XCL1, and XCL2, which may enhance ALL development and progression, leading to poor prognoses. It is likely that the NRGs risk model will be beneficial for precision immunotherapy in the future for ALL patients. However, to clarify this association between risk score and immune microenvironment in ALL, further research with clinical samples is needed.
Next, our study examined the association between NRGs-based risk scores and drug resistance in patients with ALL. It was predicted that high-risk patients would have a lower sensitivity to antitumor agents including Roscovitine (a CDKs inhibitor), A.770041 (a Src-family Lck inhibitor), AZ628(a pan-Raf kinase inhibitor), BMS.509744 (an Itk inhibitor), Dasatinib, and Docetaxel. Even though some of these drugs above are not being used for ALL treatments or are under investigation, using data on drug resistance can also help design better clinical trials for future anti-cancer treatments.
This study has some limitations as well. First, although we provide a nomogram for predicting survival in ALL, more experimental and clinical studies are needed to confirm its reliability. Second, for prognostic model construction, a single hallmark (necroptosis) has inherent weaknesses. Finally, to prove our conclusion and identify how these 7 genes contribute to ALL progression, immune therapy, and drug resistance, more experiments are needed.