The JAK-STAT pathway: from structural biology to cytokine engineering

Abstract

The Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway serves as a paradigm for signal transduction from the extracellular environment to the nucleus. It plays a pivotal role in physiological functions, such as hematopoiesis, immune balance, tissue homeostasis, and surveillance against tumors. Dysregulation of this pathway may lead to various disease conditions such as immune deficiencies, autoimmune diseases, hematologic disorders, and cancer. Due to its critical role in maintaining human health and involvement in disease, extensive studies have been conducted on this pathway, ranging from basic research to medical applications. Advances in the structural biology of this pathway have enabled us to gain insights into how the signaling cascade operates at the molecular level, laying the groundwork for therapeutic development targeting this pathway. Various strategies have been developed to restore its normal function, with promising therapeutic potential. Enhanced comprehension of these molecular mechanisms, combined with advances in protein engineering methodologies, has allowed us to engineer cytokines with tailored properties for targeted therapeutic applications, thereby enhancing their efficiency and safety. In this review, we outline the structural basis that governs key nodes in this pathway, offering a comprehensive overview of the signal transduction process. Furthermore, we explore recent advances in cytokine engineering for therapeutic development in this pathway.

Fig. 1

Timeline for milestones in the JAK-STAT signaling pathway: from fundamental research to therapeutic applications. Abbreviations: IL interleukin; ISG interferon-stimulated gene; TYK tyrosine kinase; ISGF interferon-stimulated gene factor; IRF interferon regulatory factor; SCID severe combined immunodeficiency; KO knockout; EPO erythropoietin; EPOR erythropoietin receptor; LOF loss of function; MPN myeloproliferative neoplasm; cryo-EM cryogenic electron microscopy; HBV hepatitis B virus; HCV hepatitis C virus; G-CSF granulocyte colony-stimulating factor; GM-CSF granulocyte-macrophage colony-stimulating factor; γc, common gamma chain

Fig. 2

The JAK-STAT pathway cytokines and their receptors. a Exemplary structures of typical cytokines, colored as a spectrum from the N- (blue) to the C-terminus (red). b The structure of the cytokine-binding homology region (CHR). c Domain composition of common cytokine receptors

Fig. 3

Crystal Structure of the EPO:EPOR complex (PDB ID: 1EER). a Overall Structure of the EPO:EPOR complex from both “front” (left) and “back” views (right). The interfaces between the cytokine and receptor are boxed. H1, H2, H3, and H4 correspond to helices 1, 2, 3, and 4 of EPO, while L1 and L2 designate loop 1 and loop 2 of EPO, respectively. The N-terminal helix (Nter-H) of EPOR plays a crucial role in maintaining receptor stability, forming numerous contacts with both the upper and lower domains of the receptor. The lower domains of EPORs are in close proximity, and a hydrogen bond forms between S135 of the first EPOR and D133 of the second EPOR. b Cytokine-receptor interaction details for site 1. c Cytokine-receptor interaction details for site 2

Fig. 4

Structure of the cytokine-receptor complex of βc cytokines exemplified by the GM-CSF-receptor complex (PDB ID: 4NKQ). Domain 1 (D1, orange) of GM-CSFRα is modeled using D1 of GM-CSFRα from the GM-CSF:GM-CSFRα complex (PDB ID: 4RS1). a Hexamer of the GM-CSF-receptor complex from a side view. The CHR modules of GM-CSFRα (D2 and D3, orange) interact with GM-CSF (colored in magenta) at site 1, while the CHR modules of βc (D4 from itself and D1 from the partner βc) interact with GM-CSF at site 2. Each βc dimer is bound by two molecules of JAK2, but two JAK2 molecules are distant from each other, as βc D4 domains are approximately 120 Å apart. b Cytokine-receptor hexamer of the GM-CSF-receptor complex from a top view. The CHR modules of βc are bridged by the D2 and D3 domains of βc. c Signaling complex for βc cytokines exemplified by the GM-CSF-receptor complex. The intracellular domains of both GM-CSFRα and βc associate with JAKs, which is critical for signaling

Fig. 5

Cytokine-receptor complex structurers of γc cytokines and dimeric cytokines. Upper domains (domain 1) of receptors, are labeled D1, while lower domains (domain 2) are labeled D2. Helices 1, 2, 3, and 4 are labeled as H1, H2, H3, and H4, respectively. a The structure of IL-2 in complex with IL-2Rα, IL-2Rβ, and γc (high-affinity form, PDB ID: 2B5I). b The structure of IL-2 mutant D10 in complex with IL-2Rβ, and γc (PDB ID: 3QAZ). c Structure of the IL-15 cytokine-receptor complex (PDB ID: 4GS7), which signals in a trans manner. d The structure of the signaling complex of dimeric cytokines, exemplified by IFNγ (PDB ID: 6E3K)

Fig. 6

Structures of gp130 family cytokine-receptor complexes are illustrated, with domains labeled as letter ‘D’ plus a number, and the text highlighting correlating with the color of each domain. a The cryo-EM structure of the IL-6 signaling complex (side view, PDB ID: 8D82), comprising IL-6 (green), IL-6Rα (cyan), and two molecules of gp130 (magenta and orange). A bend between D4 and D5 of gp130 brings D6 close to each other at a distance of approximately 19 Å. b The top view of the IL-6 signaling complex (the same PDB ID as a). To maintain clarity, D3 of IL-6Rα and D3-6 of gp130 are not shown. c The cryo-EM structure of CNTF (magenta) in complex with CNTFα (blue), LIFR (cyan), and gp130 (green) (PDB ID: 8D74, EMDB ID: EMD-27229). Due to the cryo-EM density map quality, only part of the molecules has been modeled. d The cryo-EM structure of LIF (magenta) in complex with LIFR (cyan) and gp130 (green) (PDB ID: 8D6A, EMDB ID: EMD-27221), with only part of the molecules have been modeled for the same reasons as the CNTF signaling complex

Fig. 7

Receptor recognition and activation of JAKs. a Schematic diagram showing the domain organization of JAKs. b Interaction between JAKs and receptors, illustrated by the structure of the intracellular domain of IFNλR1(magenta) in complex with the FERM-SH2 domain of human JAK1 (PDB ID: 5L04). The two primary binding regions from the intracellular receptor domain correspond to the ‘box 1’ and ‘box 2’ motifs, which bind to the F2 subdomain and SH2-like domain, respectively. c Structure of the PK and TK domains from human TYK2 (PDB ID: 4OLI), representing an autoinhibitory state. d The structure of full-length mouse JAK1 (mJAK1), derived from the mJAK1 pair (shown in e). When the PK domains of the TYK2 PK-TK structure (shown in c) and mJAK1 are superimposed, the TYK2 TK domain is found closer to the membrane, representing an autoinhibited state. Dimerization induced by cytokine binding likely causes a conformational change in the TK domain, relieving it from the autoinhibited state. e Dimeric structure of the mJAK1 complex (without nanobody stabilization, PDB ID: 8EWY), where PK domains form a zipper-like structure held together by a hydrophobic cluster of phenylalanine residues and an antiparallel β-sheet. The V657F mutation (equivalent to the pathogenic V617F mutation in human JAK2 PK domain) enhances the PK-PK interaction. TK domains also form a protein-protein interface in this structure, with active sites facing each other. f Dimeric structure of the mJAK1 complex (with nanobody stabilization, PDB ID: 7T6F), where TK domains rotate and the active sites face outward, proposed as an active state of JAK

Fig. 8

Phosphorylated tyrosine recognition and DNA binding by STATs. a Schematic diagram illustrating the domain arrangement in STATs. b Interaction between STATs and receptors, occurring during the STATs recruitment step; illustrated by the structure of a phosphopeptide (⁴⁴⁰pYDKPH⁴⁴⁴, shown in magenta color and labelled in red color) derived from the α-chain of the human IFNγ receptor in complex with the SH2 domain (shown in green color and labelled in black color) of STAT1 (PDB ID: 1YVL). The phosphorylated tyrosine forms extensive hydrogen bond networks with the SH2 domain, contributing significantly to the binding affinity of the phosphopeptide. c Detailed structure of the SH2 region of phosphorylated STAT1 (PDB ID: 1BF5). The phosphorylated tyrosine and its surrounding residues from its partner (shown in magenta color and labelled in red color) bind to the SH2 domain (shown in green color and labelled in black color), similar to the interaction of receptor and unphosphorylated STATs (as shown in b). d The overall structure of the phosphorylated STAT dimer without DNA (left, PDB ID: 4Y5U) and with DNA (right, PDB ID: 4Y5W), exemplified by STAT6. e Molecular details of STAT-DNA interaction, illustrated by using the structure of the STAT6-DNA complex (PDB ID: 4Y5W)

Fig. 9

The signaling process of the JAK-STAT pathway is depicted using the IL-4:IL-4Rα:γc signaling complex as an example. The JAK1 and JAK3 models are homologous structures built based on the mJAK1 structure and the human TYK2 PK-TK structure

Fig. 10

IL-2-based therapeutics development. a The immunostimulatory and immunosuppressive effects are balanced by IL-2. Decoupling the pleiotropic effects of IL-2 is essential for therapeutic purposes, utilizing its immunostimulatory effects for cancer therapy and its immunosuppressive effects for autoimmune disease treatment. b Strategies for decoupling the pleiotropy of IL-2 by disrupting IL-2Rα binding. c Strategies for enhancing the T_eff to T_reg cells ratio via ‘superkines’ (super IL-2) to modulate the immune response. d Wild-type IL-2 and IL-2 partial agonists, varying in their affinity to γc, can induce different cell fates. e Strategies for enhancing the T_reg-to-T_eff cell ratio using IL-2 partial agonists, which have greater dependency on IL-2Rα. f The principle of orthogonal (ortho) IL-2 cytokine-receptor pairs. Ortho-IL-2 is an IL-2 mutant designed to exclusively bind to the ortho-IL-2Rβ:γc pair, while showing no affinity for the wild-type IL-2Rβ:γc pair. Similarly, Ortho-IL-2Rβ refers to an IL-2Rβ mutant specifically engineered to interact solely with ortho-IL-2, with no binding to the wild-type IL-2. Ortho-IL-2Rβ can be employed to generate chimeric orthogonal receptors capable of transducing signals for cytokines like IL-2, IL-9, or IL-21, by simply replacing the intracellular domain of ortho-IL-2Rβ. g Structural basis for designing human ortho-IL-2 and ortho-IL-2 receptor pairs

Fig. 11

Engineering of IL-10 family cytokines. a Receptor binding of IL-10 family cytokines. b Conversion of dimeric IL-10 into a functional monomeric IL-10 variant. c Decoupling of pro- and anti-inflammatory effects of IL-10 through engineering it (with lower binding affinity to IL-10Rβ compared to wild-type IL-10) with cell type selectivity. This modification results in myeloid-biased activity by suppressing macrophage activation, while avoiding stimulation of inflammatory CD8⁺ T cells. d Engineered IL-22 partial agonists (with lower binding affinity to IL-10Rβ compared to wild-type IL-22) exhibit biased signaling, selectively activating STAT3 signals but not STAT1 signals in HEK-293 cells. e Mechanism for STAT3-biased functions of IL-22 partial agonists. IL-22 partial agonists induce suboptimal phosphorylation of the intracellular receptor domain, which affects STATs recruitment. STAT3 can bind to the non-phosphorylated receptor through an unconventional mechanism, thus is less affected. f IL-22 partial agonists elicit tissue-selective STAT responses in vivo that correlate to the IL-10Rβ expression level in different tissues

Fig. 12

Structure and tunability of type I IFNs. a Overall structure of IFNω in complex with IFNAR1 and IFNAR2 (PDB ID: 3SE4). b Superposition of IFNω-receptor complex structure (colored in magenta) and IFNα2(YNS) in complex with IFNAR1 and IFNAR2 (colored in cyan, PDB ID: 3SE3). The two structures exhibit remarkable similarity, underscoring a conserved cytokine-receptor recognition mechanism. However, the activities may diverge among different cytokines. c Tunability of antiproliferative and antiviral effects of type I IFNs. Both activities are enhanced as cytokine receptor affinity increases, but with a more significant impact on antiproliferative activity

Fig. 13

Antibody-based surrogate cytokines. a Diabody that binds to EPOR, inducing receptor dimerization and thereby activating the signaling pathway. b The principle for surrogate cytokines that induce signals from receptor heteromers. Covalently linking VHH or scFv that bind to distinct receptors can induce receptor dimerization and activate signaling. c Combining different VHH or scFv can generate a diverse pool of surrogate cytokines, potentially resulting in varying signal amplitudes within the pathway. d By fusing two different VHH or scFv antibodies in tandem, each in different orientations, the geometry of the receptor pair can be altered, leading to distinct signaling effects, exemplified by IL-2 receptors

Fig. 14

Design and engineering of artificial cytokines. a Design of Neo-2/15. The design begins with four helices from IL-2 as a starting point, with the loops removed to facilitate the reorganization of these helices. The helices are optimized, with a specific focus on redesigning helix 2 and idealizing the others. Short loops and helix segments enclose the helices to create early IL-2/15 mimics. The resulting cytokine mimics, composed of four helix bundles, are further optimized to enhance stability and affinity, ultimately leading to the development of Neo-2/15. b Development of a conditionally active IL-2/15 agonist. Neo-2/15 is divided into two segments, named Neo2A and Neo2B, both of which are inactive when separate. However, when brought together, they form a four-helix complex and become an active IL-2/15 agonist. When fused with target protein binders, such as nanobodies that target tumor antigens, they can translocate to the site where the antigen is presented, forming a functional IL-2/15 agonist complex that stimulates local immune activity. c Engineering an IL-4 agonist via Neo2/15 modification. Residues corresponding to those binding to IL-4Rα (shown in cyan) were grafted onto Neo-2/15, yielding an IL-4 agonist. Subsequent optimization efforts resulted in a high-affinity IL-4 agonist (new mutations are shown in yellow). d Design of DARPin-based surrogate cytokines that can control the angle and distance between EPOR pairs. e Activating non-natural signaling through non-natural receptor composition by fusion of two dominant negative (DN) cytokines