Associations of the circulating levels of cytokines with the risk of myeloproliferative neoplasms: a bidirectional mendelian-randomization study

The chronic hematological malignancies known as myeloproliferative neoplasms (MPN), which include polycythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis (MF), advance at varying rates [1]. The incidence rates of PV, ET, and PMF are estimated to be 0.5 to 4.0, 1.1 to 2.0, and 0.3 to 2.0 per 100,000 people, respectively. It is reported that nearly 10–15% of patients with MPN progress to AML [2], more than 20% will develop thrombosis during the disease, and approximately 6.2% of newly diagnosed patients will suffer hemorrhage [3, 4]. The presence of these symptoms mentioned above raises the rate of disability and mortality in MPN patients [5] and imposes a huge economic burden on the family and society. The most common feature of MPN is hyperactivation of Janus kinase 2 (JAK2) signaling, which is caused by acquired mutations in JAK2, MPL, and CALR [6]. However, clinically used JAK2 inhibitors such as Ruxolitinib and Fedratinib have limited efficacy, high toxicity, and are prone to drug resistance [6‐8]. Therefore, increased awareness of the pathogenic components may offer clues for halting the disease's course and creating novel treatments.

The chronic inflammatory environment is one of the typical features of myeloproliferative neoplasms, where inflammation is tightly intertwined with tumor clones, providing a permissive micro-environment for disease progression [9‐11]. Inflammatory cytokines are essential immune mediators in the physiology and disease process of MPN and not only play a significant role in inflammatory pathology but are also inextricably linked to the development of the disease [9, 12]. GM-CSF, IL-1, IL-4, IL-5, IL-6, IL-10, IFN-2, MIP-1, IL-12, and TNF-α were shown to have higher cytokine levels in treatment-naive patients in all three MPN groups when compared to age-matched control participants, according to an observational study [9]. In addition, serum IL-2 and soluble IL-2 receptor alpha (sIL-2rα) increased as patients with MPNs progressed to advanced clinical stages [13], and serum IL-2, sIL-2rα, and IL-6 levels were positively correlated with bone marrow neovascularization, indicating that increased inflammatory responses may be connected to the course of MPN disease [14], suggesting that MPN patients may benefit from using cytokines as a tool for illness monitoring [15]. However, little is known about the mechanisms and duration of inflammation in MPNs [16]. The origins of the increased cytokine production in MPNs (alterations, others?) and whether inflammation may occur before the development of JAK2/CALR/MPL gene mutations are still up for dispute. Observational studies are prone to common biases such as reverse causality and residual confounding [17] and have limitations such as small sample sizes and short follow-up periods. These studies, however, only addressed a small subset of inflammatory cytokines and did not take into account how other physical factors can affect changes in inflammatory cytokine levels. Determining whether variations in inflammatory cytokines cause the development of MPN or whether MPN development influences the microenvironment and causes variations in inflammatory cytokines is crucial. Investigating the precise nature of the connection between inflammatory cytokines and MPN is crucial from a therapeutic standpoint given the lack of knowledge regarding the etiology of MPN.

To establish a link between inflammatory cytokines and MPN, we applied Mendelian randomization (MR).MR has the advantage of reducing confounding variables and measurement error, as well as addressing the limitations of traditional observational studies mentioned above. This approach can effectively avoid bias caused by reverse causality [18]. The greatest level of evidence hierarchy outside of randomized controlled trials is provided by MR, which uses genetic variation as an instrumental variable (IV), which has been a dependable tool for getting reliable estimates of the causal influence of numerous risk variables on health [19]. In the current investigation, we used a two-sample MR design to methodically evaluate the potential causal link between inflammatory cytokines and MPN risk. Additionally, reverse MR analysis was done to determine how MPN affected cytokines.

Using publicly available pooled data from GWAS, we conducted a bidirectional two-sample MR analysis of the potential causative relationship between inflammatory cytokines and MPNs, and our study supported a causal association between inflammatory cytokines and MPNs. We found suggestive evidence that levels of the genetically predicted circulating cytokines IL-2rα, and IP-10 have a risk effect on MPNs. Reverse MR analysis found suggestive evidence of a positive causal effect of MPN on levels of the circulating cytokines IL-10, MIG, and RANTES. These findings passed sensitivity analyses and were not affected by heterogeneity or horizontal pleiotropy. To our knowledge, this investigation is anticipated to be the broadest and most thorough MR evaluation of links between genetically inflammatory cytokines and MPN risk to date.

The CXC chemokine family member interferon gamma-induced protein 10 (IP-10) is crucial for cell growth and proliferation [29].IP-10 combines with the CXCR3 receptor, being a key driver in cancer and autoimmune regulation [30]. Several observational studies have demonstrated the presence of aberrant IP-10 expression in MPN patients, especially in PMF and PV, where IP-10 expression is significantly elevated [31, 32]. Meanwhile, the serum level of IP-10 was also correlated with the disease progression of MPN [32]. Our MR analysis suggests that elevated IP-10 levels may contribute to MPN disease progression, which is consistent with results derived from observational studies. Previous basic research can explain our findings and the phenomena of observational studies in terms of pathogenesis.IP-10 expression is reported to be required for the activation of the JAK signaling pathway [33]and its level correlates with JAK2V617F status [9, 34]. Therefore, JAK inhibition can reduce downstream chemokine IP-10 production by disrupting T cell-induced macrophage activation [35]. However, stromal cells in the microenvironment can protect MPN clonal cells from JAK2 inhibitors by secreting IP-10, which can promote disease progression. These discoveries underscore the importance of researching IP-10 as a potential therapeutic target in the MPN tumor microenvironment and highlight the necessity of further studies on the exact mechanism of its role in MPN oncogenesis.

Our study also reveals a potential association between IL2rα and increased risk of MPN disease.IL2rα is an important component of IL-2R, a high-affinity receptor molecule highly expressed by activated T lymphocytes [36], and plays an important role in the regulation of T cell differentiation. Increasing IL-2rα expression on antigen-presenting cells (APCs) enhances the formation of memory T cells [37], and mutations in IL-2rα lead to decreased T cell function [38]. Accordingly, IL-2rα levels are associated with T-cell, B-cell, and immune system activation [36]. It has been demonstrated that conditions linked to cellular immune activation correlate with increased IL-2rα [39, 40]. Additionally, a few observational studies have shown a connection between IL2r and MPN. Katerina et al. found that serum levels of IL-2rα were significantly elevated in patients with MPN compared to normal individuals [14], which was confirmed by further studies, where IL2rα levels were correlated with overall survival in patients with MF in MPN [41], and levels of IL-2ra in patients with MPN were positively correlated with disease progression and bone marrow angiogenesis [42]. According to the findings of our MR investigation, elevated levels of IL-2rα in the circulatory system may accelerate the development of MPN disease. This finding is not only consistent with the results of observational studies but also compensates for the shortcomings of small sample sizes and potential confounders in the observational studies mentioned above and provides more reliable evidence for the association between IL2rα and MPN at the level of genetics, emphasizing the importance and necessity of further investigating the role ofIL2rα in the development of MPN disease.

Both our analyses and previous studies suggest that IP-10 and IL2rα may play an important role in MPN disease development. Considering the heterogeneity of the three subtypes of MPN, we sought to explore the effects of IP-10 and IL2rα on the disease risk of each subtype of MPN. Unfortunately, due to the limitation of the dataset, we had no way to further explore the relationship between IP-10 and IL2rα and different subtypes of MPN. Therefore, we initially analyzed the expression and diagnostic value of IP-10 and IL2rα in each subtype of MPN using the GEO database. We were surprised to find that IP-10 and IL2rα not only had elevated expression in the three subtypes of MPN compared to healthy individuals, but also had the potential to serve as independent biomarkers. This is consistent with our MR analysis that high expression of IP-10, IL2rα increases the risk of MPN disease. This greatly encourages our confidence in further exploring the role of IL2rα, IP-10 in MPN at a later stage.

Positive MR analysis has revealed the role of inflammatory cytokines, particularly IP-10 and IL-2rα, in MPN disease progression. Indeed, MPN cells can also release large amounts of pro-inflammatory products, which in turn cause genomic instability and drive clonal myeloproliferation [43, 44].To explore the effect of MPN disease on inflammatory cytokines, we performed a reverse MR analysis. The inverse MR analysis showed a potential positive correlation between genetically predicted MPN and the levels of cytokines IL-10, MIG, and RANTES, and that MPN could promote slightly increased levels of the above cytokines. Inverse MR analysis revealed a potential positive correlation between genetically predicted MPN and levels of the cytokines IL-10, MIG, and RANTES, with MPN promoting slightly elevated levels of the aforementioned cytokines, which is consistent with observational findings [9, 45‐47]. Our review of the literature revealed that aberrantly expressed IL-10, MIG, and RANTES are all associated with premature atherosclerosis, a devastating consequence of chronic inflammation in the MPN [48‐50].MIG binds to the receptor CXCR3 and not only participates in the recruitment of T cells to peripheral sites of inflammation [51]but also chemotactically recruits monocytes/macrophages to sites of inflammation. Activated inflammatory cells release pro-inflammatory factors to induce an inflammatory response [52], which promotes atherosclerosis. RANTES is one of the chemokines highly expressed upon platelet activation, and RANTES released by activated platelets facilitates the formation of atherosclerotic lesions by platelet-monocyte aggregation [53, 54], and RANTES also regulates local inflammatory processes and atherosclerosis progression by mediating CD4 + T-cell homing [55]. It is interesting to note that IL-10 appears to be a protective factor against atherosclerosis, a common clinical symptom of MPN, and that, as an anti-inflammatory cytokine, IL-10 can attenuate atherosclerotic lesions by preventing dilation of inflamed areas, decreasing the size of plaques, and other mechanisms [56]. Specifically, IL-10 attenuates atherosclerotic lesions by inhibiting macrophage activation, as well as inhibiting the expression of matrix metalloproteinases, proinflammatory cytokines, and cyclooxygenase-2 in lipid-loaded and activated macrophage foam cells [57, 58]. Therefore, it is necessary to investigate the correlation and mechanism between the elevated circulating levels of MPN-promoting inflammatory cytokines IL-10, MIG, and RANTES and the common clinical complications of MPN, and to provide the possibility of targeting the above cytokines to alleviate the clinical complications.

Our study has several advantages. (1) The link between inflammatory cytokines and MPN risk is explained for the first time in magnetic resonance research. (2) Unlike observational studies, our study minimized confounders and reverse causality, providing a reliable causal relationship between MPN and inflammatory cytokines. (3) Our research data were sourced from the openly available GWAS database, which houses a significant volume of original research data, and thus gives this study a solid guarantee.

There are also some limitations to our study. In the first place, all participants in the dataset we used were of European ethnicity, which limits our ability to generalise our findings to other ethnicities. It is well known that Mendelian randomisation investigates the effect of genotypic variation (exposure) on phenotype (outcome) from a genetic perspective. Bias caused by confounding variables or reverse causality is avoided [18].In reality, however, it is well known that the level of gene expression determines the unique characteristics of a cell, that differences in disease prevalence between populations are associated with the frequency of alleles that regulate polymorphisms, and that differences in allele frequencies between racial groups have highly significant phenotypic consequences [59].Significant differences in gene expression phenotypes have been reported for at least 25 percent of genes between Europeans and Asians, and specific genetic variants (allele frequencies) between populations are the main cause of these differences [59].Therefore, in Mendelian randomisation analyses, genetic differences in quantitative phenotypes between different ethnic groups may be functionally equally important. Environmental, genetic, dietary, and lifestyle factors in different racial groups may influence phenotypic results [60], so we think it is unavoidable that the results of MR may differ between races due to residual confounding and selection bias [61, 62].This is our limitation in this study. Therefore, whether elevated IP-10 and IL-2ra increase the risk of MPN prevalence in other populations requires specific analyses of gene expression variation for particular populations. Future studies will also need to enhance the analysis of gene expression variation between populations to improve understanding of the underlying genetics and population differences observed in complex genetic diseases. In the second, there are three subtypes of MPN, and due to the limitations of the GWAS dataset, we have not specifically stratified to explore the relationship between cytokines and the different subtypes of MPN. Finally, after Bonferroni correction, no cytokines showed statistically significant associations with MPN risk, and only two of them (IP-10, IL-2rα) showed suggestive associations.

In conclusion, our study suggests that elevated circulating levels of IP-10 and IL-2rα are associated with a high risk of MPN. Potential positive correlation between genetically predicted MPN and levels of the cytokines IL-10, MIG, and RANTES. Our results show that cytokines play a significant role in the pathophysiology of MPN. Further research is required on the potential use of these biomarkers for the prevention and treatment of MPN.