The Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway propagates signals downstream of growth factor, hormone, and cytokine receptors. Thereby the JAK-STAT pathway regulates development, survival, proliferation, differentiation, and functionality of cells within the immune system [1,2,3,4,5].

The four JAKs, JAK1-3, and TYK2, are non-receptor tyrosine kinases which associate with their respective receptors. While some receptors associate with a pair of a specific JAK-family member, others are linked to more than one JAK. Ligand binding induces oligomerization of transmembrane receptors and activation of JAKs by auto- and/or trans-phosphorylation events. Subsequently, JAKs phosphorylate tyrosine residues on the receptors representing docking sites for STATs (STAT1-6) via their Src homology 2 (SH2) domain. Upon phosphorylation, STATs undergo a conformation change and switch from an anti-parallel to a parallel homo/heterodimer and translocate to the nucleus to regulate gene expression [2,3,5,6,7,8,9]. Additional posttranslational modification, formation of multimeric complexes with co-factors, and epigenetic remodeling processes add to the complexity of the transcriptional regulation initiated by the JAK-STAT signaling pathway and allow for its diverse functional consequences [5,6,10]. JAK1 is key for interferon (IFN) signaling, in combination with JAK2 in case of IFN-γ and with TYK2 in case of type I and type III IFN signaling. Together with JAK3, which is directly associated with the common γ chain (γc) of cytokine receptors, JAK1 signals downstream of the γc-dependent cytokines interleukin (IL)-2, IL-4, IL-7, IL-9, IL-15, and IL-21. JAK2 is responsible for signaling downstream of erythropoietin (EPO), thrombopoietin (TPO), growth hormone, prolactin, leptin, IL-3, IL-5, and granulocyte-macrophage colony-stimulating factor (GM-CSF). JAK1, JAK2, and, in some cases, TYK2 are activated by gp130 cytokine family members (IL-6, IL-11, IL-31, oncostatin M (OSM), ciliary neurotrophic factor (CNTF), leukemia inhibitor factor (LIF), cardiotropin-1 (CT-1), neurotrophin-1 (NNT-1)) and IL-13. TYK2 is involved in signaling of IL-10-like cytokines (IL-10, IL-19, IL-20, IL- 22, IL-24, and IL-26) in conjunction with JAK1 and mediates IL-12/IL-23 cytokine signaling together with JAK2 [8,11,12].

Deregulated JAK-STAT signaling in the hematopoietic system is associated with a wide range of alterations, including immunodeficiency, autoimmunity, and transformation [4,5,13,14]. The discovery of hyperactivation of JAK-STAT pathway members in inflammatory pathologies (such as rheumatoid arthritis (RA), psoriasis, and inflammatory bowel disease) and hematological cancers triggered the development of JAK inhibitors (JAKinibs) as potential treatment options [4,15,16]. In cancer, aberrant activation of the JAK-STAT pathway is achieved by different means, including deregulated upstream signals or feedback loops, gene amplifications, generation of fusion proteins as well as gain-of-function (GOF) mutations [4,17]. The latter will be discussed in more detail below focusing on hematological malignancies. Dysregulated JAK-STAT signaling also plays a role in solid cancers, as reviewed before [4,18,19].

An overview of GOF mutations in the JAK-STAT signaling pathway occuring in hematological malignancies is presented in Table 1. JAKs consist of multiple domains, including an N-terminal FERM domain and an SH2-like domain that promote receptor interaction, a pseudokinase domain (JH2) with regulatory capacity, and the C-terminal catalytically active kinase domain (JH1) [4,20,21,22,23,24]. GOF mutations in JAKs are frequently located in the JH2 domain, where they lead to an altered auto-inhibitory function and render JAKs constitutively active [22,25]. The most prominent example is the somatic JAK2^V617F mutation that is highly prevalent in myeloproliferative neoplasms (MPNs). MPNs, including polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF), are a group of diseases in which myeloid precursors produce increased numbers of blood cells with the potential to evolve into acute myeloid leukemia (AML) [26,27,28,29,30,31,32,33,34,35]. The equivalent germline JAK2^V617I mutation, besides other pseudokinase domain mutations, has been described in hereditary essential thrombocytosis [36,37,38]. The corresponding somatic JAK1 JH2 mutations (V658I or V658F) are present in patients with acute lymphoblastic leukemia (ALL), T-cell prolymphocytic leukemia (T-PLL), and myeloid malignancies [4,15,25,39,40,41,42]. Somatic JH2 mutations are also found in JAK3 (e.g., A572V and A573V) in T-cell malignancies (T-PLL, T-ALL, early T-cell precursor ALL (ETP-ALL)), AML, and malignancies of natural killer (NK) cell origin (e.g., extranodal NK/T-cell lymphoma (NKTCL)) [4,41,43,44,45,46,47,48,49,50,51,52,53]. Rare cases of somatic and germline JH2 mutations in TYK2 have been described that are associated with ALL [4,54,55,56]. Other activating JAK mutations associated with hematological malignancies are found in the FERM, linker, and JH1 domain [4,15,32,57].

GOF mutations in STATs have also been described in hematological malignancies [5,10,58] and most frequently occur in the SH2 domain, which is required for dimerization and transcriptional activity. SH2 domain mutations of STAT3 (e.g., Y640F, D661Y) and STAT5B (e.g. N642H, Y665F/H) are found in mature T-cell and NK-cell neoplasms (see Table 1 for details of cancer subtypes), T-ALL, diffuse large B-cell lymphoma (DLBCL) and myeloid diseases (eosinophila, AML, chronic neutrophilic leukemia) [59,60,61,62]. Activating STAT mutations do also exist outside of the SH2 domain in hematological cancers [60,63,64,65] and DNA binding domain mutations of STAT6 are associated with B-cell malignancies [66,67,68,69].

The identification of JAK2^V617F as a driver mutation in the majority of PV cases and in over 50% of ET and PMF cases was a major breakthrough and provided the rationale for the development of JAKinibs [26,27,28,70]. Ruxolitinib (JAK1/2 inhibitor) has been approved by the US Food and Drug Administration (FDA) and European Medicines Agency (EMA) for the treatment of intermediate and high risk PMF and secondary MF and second line for PV patients resistant or intolerant to hydroxyurea [71]. JAKinib treatment prolongs overall survival and reduces splenomegaly typically associated with MPNs, but does not induce complete remission. Importantly, Ruxolitinib triggers comparable therapeutic responses in MPN patients irrespective of the JAK2^V617F mutation. Patients harboring deletions in the calreticulin gene (CALR) or mutations in the MPL receptor also show therapeutic responses as the constitutive activation of the JAK-STAT pathway represents a common feature of MPNs. As a consequence, JAKinibs are extensively tested in (pre)clinics as therapies for a variety of malignancies with overactivation of the JAK-STAT pathway due to mutations (see Table 1) or due to constitutive activation of the pathway. As such, diseases driven by the STAT5B^N642H mutation are partially responsive to Ruxolitinib [72,73]. Table 2 summarizes ongoing clinical trials for JAKinib treatment of hematological malignancies.

Activation of the JAK-STAT pathway by mutations mostly increases sensitivity to JAKinibs (as exemplified in Table 1), but no JAKinib is available that specifically targets a mutated allele. Such inhibitors may circumvent potential side-effects of current JAKinibs that also hit non-transformed cells [4].

The JAK-STAT pathway has a critical role in the development and function of immune cells. Loss-of-function (LOF) mutations in JAKs or STATs are associated with immune deficiencies or enhanced susceptibility to infections [5]. The most severe phenotype occurs in patients with LOF in signaling components downstream of γc-dependent cytokines, namely JAK3 and STAT5B, which manifest as severe combined immune deficiencies (SCIDs) [4,110]. The severe immune deficiency is predominantly linked to impaired IL-7 and IL-15 signaling [14]. Interestingly, STAT5B LOF mutations may also provoke autoimmune phenotypes due to non-functional T regulatory cells (Tregs) [111]. LOF mutations in STAT1 and TYK2 enhance the susceptibility to bacterial and viral infections [110,112], while LOF in STAT2 and STAT4 increase the incidence of viral and fungal infections, respectively [113,114,115,116]. Similarly, GOF mutations of STAT1 and STAT3 may suppress the immune system and manifest with impaired anti-viral, anti-bacterial, and for STAT1, also anti-fungal responses [110,117]. Knockout mice for JAKs and STATs generally recapitulate the phenotypes of patients with LOF mutations and provide a valuable tool to understand the molecular mechanisms of JAK-STAT signaling in diseases. The consequences of JAK and STAT knockouts in mice and LOF mutations in humans have been reviewed extensively and the most prominent immune phenotypes are summarized in Table 3. The availability of conditional knockouts provides the opportunity to overcome embryonic and perinatal lethality of JAK1, JAK2, STAT3, or STAT5A/B-deficient mice. Conditional knockouts for STAT3 and STAT5A/B allowed to uncover the effects of STAT3 and STAT5 in T-cell differentiation and memory, dendritic cell function, and NK-cell tumor surveillance (reviewed in [5]). Conditional JAK knockouts have not yet been studied in depth despite the severe immune defects of knockout mice. Inducible deletion of JAK1 markedly impairs stem cell homeostasis and reduces the frequencies of B cells [118]. Similarly, NK cell-specific deletion of JAK1 almost completely abrogates the presence of peripheral NK cells [119]. In contrast, inducible deletion of JAK2 fails to impact on lymphopoiesis [120] and NK cell-specific JAK2 deficiency does not interfere with NK-cell homeostasis [119]. JAK2-deficient T cells are skewed towards Th2 and Treg polarization resulting in reduced graft versus host disease (GvHD) [121]. In addition, NK cell-specific deletion of TYK2 decreases anti-bacterial responses while leaving NK cell-mediated tumor surveillance intact in contrast to global TYK2 knockout mice [122,123,124]. These studies highlight the importance of the JAK-STAT pathway for immune responses and point towards a potential risk for patients when drug-targeting the JAK-STAT pathway.

JAKinibs provide a powerful tool in targeted cancer therapy. In light of the above-described immune alterations, JAKinibs are expected to provoke immune suppression. We here provide an overview of currently available JAKinibs undergoing clinical trials for treatment of hematological diseases and focus on their immune effects.

The strong suppressive effects of Ruxolitnib on NK and T cells are now exploited for the treatment of hematological diseases originating from cytotoxic lymphocytes including NKTCL, aggressive NK-cell leukemia (ANKL), and T-cell lymphomas/leukemias. NKTCL and ANKL are aggressive diseases with poor prognosis. They harbor alterations of the JAK-STAT signaling pathway, including STAT3 mutations and mutations in epigenetic modifiers [86,91,100]. Drug screens revealed a synergistic activity of the JAK1/2 inhibitor Ruxolitinib in combination with the BCL2 inhibitor Venetoclax in NKTCL and ANKL cell lines [100]. A synergistic efficacy of Ruxolitinib and BCL2 inhibitors was also shown in T-cell lymphoproliferative diseases, including T-ALL and human T-cell leukemia virus type 1 (HTLV1)-associated adult T-cell leukemia (ATL) [233,234,235,236]. A synergistic growth inhibitory effect in NKTCL cell lines was also achieved by a combination of Ruxolitinib with the CDK4/6 inhibitor Ribociclib (LEE011) [237]. This indicates that Ruxolitinib might not only be efficient in myeloid malignancies, but could also be of therapeutic value for patients with lymphoid malignancies [100,238,239,240,241,242]. Clinical trials using Ruxolitinib are ongoing for treatment of different cancer entities, such as HTLV1-associated ATL (NCT01712659), relapsed/refractory ETP-ALL (NCT03613428), relapsed/refractory NK-cell or peripheral T-cell non-Hodgkin Lymphoma (NCT01431209, NCT02974647), and ALL (NCT03117751). Similarly, the inhibitory effects of Itacitinib on T-cell proliferation [226] may be repurposed to block transformed T-cell growth. Indeed, there is a clinical trial ongoing for the use of Itacitinib in combination with Alemtuzumab (anti-CD52) in T-PLL patients (NCT03989466) [243]. A therapeutic window remains to be determined at which Itacitinib exerts anti-tumor effects, while avoiding counterproductive immunosuppressive effects.

As described above, Tofacitinib potently inhibits T and NK cells in a time and dose-dependent manner. Therefore, it might be an option to treat malignancies derived from innate and adaptive lymphocytes. As indicated in Table 1, BaF3 cells expressing JAK3 mutations as well as JAK3-mutant T-ALL and NKTCL in vivo models are sensitive to Tofacitinib treatment [44,53,82]. Similarly, NKTCL cell lines with constitutive activation of JAK-STAT signaling, partially associated with JAK3 GOF mutations, respond to Tofacitinib [45,50,53,100,244]. A phase 2 clinical trial currently includes patients with relapsed/refractory NKTCL that are treated with a combination of Tofacitinib and the histone deacetylase inhibitor Chidamide (Table 2). First case reports demonstrated a moderate activity of the combined use of Tofacitinib and Ruxolitinib in T-PLL [243,245,246]. Tofacitinib inhibits STAT5-regulated miR-21 expression in cutaneous T-cell lymphoma (CTCL) and thereby blocks anti-apoptotic effects of miR-21 in malignant T cells [238,247]. Combination of the BCL2 inhibitor Venetoclax with Tofacitinib induced therapeutic responses in some hematological patients with relapsed/refractory T-ALL with surface IL-7R expression or IL-7R-pathway mutations and BCL2 expression [248]. In contrast, Tofacitinib was not effective in a patient with DDX3X-MLLT10 T-ALL carrying an activating JAK3 mutation [249]. Therefore, the individual genetic make-up of a hematological malignancy might determine its responsiveness to Tofacitinib treatment.

Tofacitinib is a promising therapeutic strategy also for large granular lymphocytic leukemia (LGLL)—a rare subtype of mature T- and NK-cell neoplasms that is characterized by clonal expansion of cytotoxic NK or T lymphocytes. LGLL has a largely indolent course and is frequently associated with autoimmune disorders [250,251,252,253]. T-cell LGLL (T-LGLL) in particular is associated with rheumatoid arthritis in 10–30% of the patients [254,255,256]. Other autoimmune disorders associated with LGLL are autoimmune-related cytopenia including neutropenia [251,253,257]. Deregulation of pro-survival pathways, including a deregulated JAK-STAT signaling, have been implicated in disease pathogenesis and support the expansion of auto-reactive lymphocytes [59,250,251,252,253]. T-LGLL shows a high proportion of somatic STAT3 GOF mutations [83,84,85,106]. Based on its immunosuppressive activity, Tofacitinib was studied as a salvage therapy for highly refractory T-LGLL with RA, showing encouraging response rates with improvement of RA and cytopenia symptoms and limited side effects [87]. Tofacitinib selectively induces apoptosis in STAT3-mutant T-LGLL cells compared to healthy CD8+ T cells [87]. There is hope that the suppressive effects of Ruxolitinib, Tofacitinib, and potentially Itacitinib on cytotoxic lymphocytes will be repurposed as novel therapies in NK/T-cell malignancies with Tofacitinib holding promise for specific types of leukemias associated with autoimmune symptoms [258].