A time domain-selective kinase modulating pharmaceutical combination for treating systemic lupus erythematosus

Abstract

The present application provides a time-domain selective kinase modulating drug combination for treating systemic lupus erythematosus. The combination comprises a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor, and strategically leaves JAK1, JAK2 and ITK untouched; it deeply inhibits the type I interferon axis and the IL-21 / Tfh axis in the peak window, and preferentially rebounds the protective STAT3 / STAT5 over the pathogenic STAT1 in the recovery window, so as to decouple the IL-10 / STAT3 homeostasis and the pathogenic axis inhibition in the time dimension. The maintenance of NK cell survival in the recovery window can be evaluated through Bcl-2 / Mcl-1 anti-apoptotic proteins and NK cell activity readings. IL-10 / STAT3 is retained to measure the net function pSTAT3-Y705, and the criterion is that the baseline is greater than or equal to 70%.

Description

Technical Field

[0001] This invention relates to the fields of immune regulation, molecular pharmacology and drug preparation, and particularly to a combination of kinase-modulating drugs for systemic lupus erythematosus, its time-selective dosing regimen and uses. Background Technology

[0002] Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by simultaneous dysregulation of both innate and adaptive immunity. JAK inhibitors have shown significant efficacy in various immune-mediated inflammatory diseases, but their application in SLE is subject to a structural limitation: JAK1 has a dual role in SLE, participating in pathogenic signaling (IFN-γ / IL-6 / IL-21 axis) and also carrying protective immune homeostasis (IL-10 / STAT3 anti-inflammatory circuit and IL-15 / STAT5 NK cell survival signaling). Since IL-10 (via JAK1 / TYK2→STAT3) and IL-21 (via JAK1 / JAK3→STAT3) share the STAT3 resource pool, any approach that inhibits JAK1 will damage the host's endogenous anti-inflammatory circuit while blocking pathogenic signaling; this invention refers to this as the "IL-10 paradox."

[0003] Furthermore, T follicular helper cells (Tfh) and their secreted IL-21 in systemic lupus erythematosus (SLE) participate in germinal center B cell differentiation, pathogenic autoantibody production, and disease activity; the IL-21 receptor complex uses the common γ chain (γc) as an indispensable subunit and transmits signals via the JAK1 / JAK3-STAT pathway (Gensous et al., Rheumatology, 2017, PMID:27498357; Asao et al., JImmunol, 2001, PMID:11418623; Habib et al., Biochemistry, 2002, PMID:12093291). Therefore, simple TYK2-JH2 allosteric inhibition mainly covers the upstream axis of type I interferon and IL-12 / IL-23, but has insufficient coverage of the JAK3 / γc-dependent pathogenic bypass of IL-21 / Tfh.

[0004] The evolution of existing JAK inhibitor research and development—from tofacitinib (pan-inhibition of JAK1 / 2 / 3) to baricitinib (JAK1 / 2) to utpatinib (highly selective for JAK1) and then to tymetinib (dual inhibition of TYK2+JAK1)—all have focused on strengthening JAK1 inhibition.

[0005] A recent static sub-mechanism study further exposed two aspects of the aforementioned limitations. This study independently evaluated utpatinib (JAK1 inhibitor), deuterocelexitinib (TYK2 allosteric inhibitor), and litexitinib (JAK3 / TEC family co-inhibitor) in human spinal attachment tissue: JAK1 inhibition, through negative feedback disruption of the IL-10 receptor, resulted in an approximately 10-fold anomalous increase in myeloid IL-23 / TNFα; notably, TYK2 allosteric inhibition (deuterocelexitinib) also induced this IL-10 antagonistic effect, while the JAK3 / TEC family co-inhibitor litexitinib did not (Giryes et al., Arthritis & Rheumatology, 2025, PMID:40320905). This literature exposes two structural flaws in existing technologies: first, the protective IL-10 / STAT3 circuit is sensitive to inhibition of both JAK1 and TYK2 arms; second, litexitinib-type JAK3 / TEC co-inhibitors impair NK cell effector surveillance due to associated ITK inhibition. It should be noted that this study was a static (continuous) sub-evaluation, using a human spinal cord attachment tissue system rather than a systemic lupus erythematosus system, and did not evaluate the therapeutic combination of TYK2-JH2 allosteric inhibitors and selective JAK3 inhibitors, nor did it involve pulsed time-domain protocols. This invention merely identifies the vulnerability of IL-10 negative feedback to JAK1 / TYK2 interference within the existing continuous inhibition framework as a source of technical problems to be solved, rather than identifying the closest existing technology to the combination or time-domain protocol of this invention.

[0006] The "γc arm" problem for NK cell survival constitutes the mechanistic root of the second line of defense. IL-15 in NK cells activates STAT5 through the synergistic activation of both the IL-2 / 15Rβ (JAK1-coupled) and γc (JAK3-coupled) arms. Deficiency of γc or JAK3 in humans leads to severe combined immunodeficiency of T-B+NK-type, demonstrating that JAK3 / γc is essential for NK cell development and maintenance (Notarangelo et al., Immunological Reviews, 2000, PMID:11213805). Therefore, any approach that inhibits JAK3 to sever the IL-21 / Tfh axis, under continuous (static) inhibition, simultaneously suppresses the protective arm of IL-15→γc-JAK3→STAT5; simply preserving JAK1 only supports the β arm and cannot independently replenish the γc arm that has been severed by JAK3 inhibition.

[0007] The two aforementioned wounds are difficult to avoid simultaneously within a static (continuous) inhibition framework: adequate blockade of the pathogenic axis requires sufficient inhibitory depth, which, under continuous exposure, often results in partial attenuation of protective IL-10 / STAT3 and IL-15 / STAT5 signals—pathogenic axis inhibition and protective axis maintenance are forcibly coupled to the same exposure dimension. Existing static sub-studies (such as the aforementioned Giryes study) quantify the differences in negative feedback to IL-10 from the three inhibitors, but they neither disclose a combination of TYK2 allosteric inhibitors and selective JAK3 inhibitors, nor any techniques for decoupling pathogenic axis inhibition and protective axis maintenance over time. This lack of temporal decoupling is the underlying reason why existing technologies struggle to overcome the efficacy-steady-state bottleneck, and it is also the technical necessity of this invention. It should be pointed out that existing technologies not only converge to enhancing JAK1 inhibition in terms of target selection, but also consistently converge to a continuous (continuous exposure) dosing paradigm in terms of administration: all generations of JAK inhibition regimens since tofacitinib have aimed to maintain steady-state blood drug concentration and continuously occupy pathogenic kinases as the dosing target. There is no technical inspiration in the field to separate the pathogenic axis inhibition and the protective axis recovery in time in an intermittent or pulse manner. Existing dosing paradigms consistently point to continuous (continuous exposure) inhibition and do not provide technical inspiration for the pulsed time-domain decoupling scheme described in this invention. Summary of the Invention

[0008] The technical problem this invention aims to solve is that in systemic lupus erythematosus (SLE), JAK-related signaling nodes, which mediate pathogenic signals, also carry protective homeostasis functions. Existing continuous inhibition protocols, while inhibiting the pathogenic axis (type I interferon axis and IL-21 / Tfh axis), simultaneously weaken the protective axis (IL-10 / STAT3 immune homeostasis, IL-15 / STAT5-maintained NK cell survival, and ITK-dependent NK cell surveillance function), making it difficult to fully block the pathogenic axis while preserving the protective axis. To solve this technical problem, the core of this invention is a time-selective kinase regulation protocol: utilizing the differences in STAT dephosphorylation-recovery kinetics between pathogenic and protective signaling pathways, pathogenic axis inhibition and protective axis maintenance are separated over time through pulsed drug administration. The dephosphorylation recovery of pathogenic STAT1 (inducible by type I interferon) is relatively slow and locked by SOCS negative feedback; the recovery of protective STAT3 (inducible by IL-10) and STAT5 (inducible by IL-15) is relatively rapid. The pulsed approach deeply inhibits and blocks the slow-recovering pathogenic axis during the peak window, while prioritizing the rebound of the fast-recovering protective axis during the recovery window. Thus, it uses "which STAT line recovers faster" as the pathogenic-protective discrimination dimension, which is different from the spatial discrimination dimension of "which kinase to inhibit and to what depth" under the static framework.

[0009] As a basis for the scope of protection independent of the above reasonable time-domain expectations, the present invention simultaneously provides a purely spatial target selectivity base: the combination of a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor, together with the strategic reservation of JAK1, JAK2 and ITK activities, jointly constitute the target selectivity manipulation window of the present invention (the core dual windows are set forth in claim 1, and the component selectivity and reservation windows are set forth in claim 3 and claim 4 respectively). This spatial base can define the protection scope of the present invention even without implementing a pulse regimen; the time-domain regimen, as its progression, provides mechanistic self-consistent closure and enhanced steady-state protection. The present invention further provides an in vitro method for evaluating the applicability of a candidate drug combination to systemic lupus erythematosus, and a detection kit supporting the above regimen, both of which take systemic lupus erythematosus indicator-specific readouts such as type I interferon gene score and NK cell CD107a triggering function as essential components.

[0010] Through the triple strategic reservation of JAK1, JAK2 and ITK activities, the present invention systematically remedies the structural defects of existing JAK inhibition regimens: JAK1 reservation protects the IL-10 / STAT3 steady-state axis and maintains the β-arm baseline of IL-15→STAT5; JAK2 reservation (IC50>500 nM) avoids EPO / TPO-related hematopoietic toxicity; ITK reservation (the IC50 of the JAK3 inhibitor against ITK>1000 nM) preserves the ITK-dependent NK cell CD107a degranulation triggering function. ITK is a positive regulator of Fc receptor-initiated granule release, calcium mobilization and degranulation in human NK cells (Khurana et al., Journal of Immunology, 2007, PMID: 17339454); accordingly, the CD107a degranulation triggering function described in the present invention specifically refers to the ITK-mediated, IL-15-independent degranulation initiation layer, and is not equivalent to the granule loading (perforin / granzyme) and survival layer maintained by IL-15→γc-JAK3→STAT5.

[0011] This invention reattributes NK cell survival, a reattribution crucial to the self-consistency of its temporal protocol mechanism. NK cell survival is maintained above the apoptosis threshold by a pool of anti-apoptotic proteins (Bcl-2 / Mcl-1 / Bcl-XL), which are downstream products of the IL-15→STAT5 signaling pathway. STAT5 signaling maintains NK cell survival; withdrawal of IL-15 leads to decreased Bcl2 expression, increased caspase activity, and apoptosis in NK cells (Lin et al., Nature Communications, 2017, PMID:29105654); blocking IL-2 / 15Rβ decreases pSTAT5 and Bcl-XL in NK cells, accompanied by early apoptosis (Yamasaki et al., Leukemia Research, 2004, PMID:15289013). Therefore, NK cell survival does not depend on continuous STAT5 signaling, but rather on the anti-apoptotic protein pool acting as a time integrator of the STAT5 signal, maintaining it above the threshold: JAK1 refractory continuously supports the baseline signal in the β-arm; JAK3's periodic recovery during the recovery window replenishes STAT5 via pulsed γ-c-arm, "replenishing" the anti-apoptotic protein pool. Under the condition that the trough duration is shorter than the pool's decay time constant, the pool does not fall below the apoptosis threshold, and NK cell survival is maintained; the trough duration and recovery threshold can be determined through pharmacokinetic-pharmacodynamic evaluation. This time-integration model decouples JAK3 inhibition (peak period) from NK cell survival (trough recovery) in time, thus closing the contradiction between JAK3 inhibition and NK cell survival within the framework of continuous inhibition.

[0012] Regarding the net retention of the IL-10 / STAT3 homeostatic axis, this invention no longer relies on the assumption that any single component is absolutely neutral to the IL-10 receptor signal. Given that the aforementioned Giryes literature has empirically demonstrated that both the TYK2 allosteric inhibition arm and JAK1 inhibition partially weaken the IL-10 negative feedback, this invention instead uses the net functional effect of the drug combination—i.e., a measured retention rate of pSTAT3-Y705 relative to baseline ≥70% after IL-10 stimulation—as the sole criterion. Net retention is supported by two pathways: first, complete JAK1 retention (functional inhibition rate ≤30%) maintains the dominant contributing arm of IL-10Rα / JAK1 in the IL-10 receptor complex (its mechanistic breakdown is detailed in the specific implementation); second, during the recovery window of the pulsed protocol, the decreased occupancy of JAK3 and TYK2 allows the IL-10 / STAT3 pathway to gain periodic deinhibition space, thus increasing the time-averaged retention rate. In other words, even though the TYK2 arm itself weakens IL-10, the combination of the present invention still achieves the net IL-10 / STAT3 retention target, constituting a functional net retention effect that is not directly indicated by the prior art relative to static components.

[0013] The inventive core of this invention lies in the overall combination and time-domain selection architecture described above, rather than any single reason for leaving it blank. Within the existing continuous inhibition framework, static component studies suggest that TYK2 allosteric inhibition may weaken the IL-10 negative feedback. Therefore, existing static component results do not provide direct technical inspiration for continuing to include TYK2 allosteric inhibition while retaining IL-10 / STAT3. This invention still includes TYK2-JH2 allosteric inhibition in the combination because the core pathogenic driver of SLE is the type I interferon axis, and non-TYK2 allosteric inhibition cannot block this axis without deeply inhibiting JAK1. Existing technologies do not disclose therapeutic combinations of TYK2 allosteric inhibitors and selective JAK3 inhibitors, nor do they disclose ITK isolation conditions, nor do they disclose time-domain selection schemes utilizing STAT differential recovery kinetics. The JAK2 whitespace to avoid hematopoietic toxicity and the NK retention to avoid monitoring incompetence are both known security logics and cannot support inventiveness on their own. The technical contribution of this invention is carried by the overall architecture of combination, ITK isolation, integrated quantization window and time domain selection principle (as shown in Figure 6).

[0014] The beneficial effects of the present invention include: (1) By combining a TYK2-JH2 allosteric inhibitor with a selective JAK3 inhibitor, the pathogenic signaling pathway inhibition rate is ≥70%, while the net functional effect is measured to maintain the IL-10 / STAT3 homeostatic axis retention rate at ≥70% baseline, thus achieving functional decoupling between pathogenic axis inhibition and protective axis maintenance; (2) By strategically leaving blanks for JAK1, JAK2 and ITK, the IL-10 / STAT3 homeostatic signal is preserved, the risk of hematopoietic toxicity is expected to be reduced, and the ITK-dependent NK cell CD107a degranulation surveillance function is preserved; (3) By using pulsed administration to restore kinetics by utilizing the difference between pathogenic STAT1 and protective STAT3 / STAT5, the pathogenic axis inhibition and protective axis maintenance are further separated in the time dimension, thereby improving the time-averaged retention rate of the protective axis; (4) The scope of protection is defined by a measurable functional operating window and net functional reading, which facilitates in vitro evaluation of the applicability of candidate drug combinations to systemic lupus erythematosus. Attached Figure Description

[0015] Figure 1 is an immunological logic comparison diagram of the static (continuous) inhibition architecture and the temporal (pulse) selective architecture of the present invention. The left side illustrates that under continuous inhibition, the depth of pathogenic axis blockade and the weakening of the protective axis (IL-10 / STAT3, IL-15 / STAT5 NK survival) are coupled to the same exposure dimension; the right side illustrates that the present invention deeply inhibits the pathogenic axis in the peak window and allows the protective axis to rebound preferentially in the recovery window, thereby achieving temporal decoupling between pathogenic axis inhibition and protective axis maintenance.

[0016] Figure 2 is a schematic diagram of the time integral of the anti-apoptotic protein pool for NK cell survival. In the figure, the Bcl-2 / Mcl-1 anti-apoptotic protein pool is represented by a reservoir, which is continuously replenished by the baseline of the β arm maintained by the JAK1 blank and periodically replenished by the γc arm pulsed by the recovery from the trough of JAK3; the dashed line represents the apoptosis threshold; when the trough duration is shorter than the pool decay time constant, the pool level is maintained above the threshold.

[0017] Figure 3 shows the kinetic curves of the difference between pulse peaks and troughs and STAT. The horizontal axis represents time after drug administration, and the vertical axis represents the signal level relative to the baseline. In the figure, pathogenic STAT1 (type I interferon) is represented by a slow recovery curve, while protective STAT3 (IL-10) and STAT5 (IL-15) are represented by fast recovery curves. The shaded area represents the recovery window.

[0018] Figure 4 is a schematic diagram of the time-domain selective operation window of the present invention. The functional operation window of the present invention is represented by the area jointly defined by two threshold lines: the peak window pathogenic axis inhibition rate ≥70% and the recovery window protective axis recovery rate ≥70%.

[0019] Figure 5 shows the triple strategic white space (JAK1 / JAK2 / ITK) architecture, illustrating that JAK1 white space protects the β arm of IL-10 / STAT3 and IL-15→STAT5, JAK2 white space protects hematopoiesis, and ITK white space preserves the CD107a degranulation triggering function of NK cells.

[0020] Figure 6 illustrates the decoupling of the net functional window and time domain under the risk of IL-10 negative feedback. The upper section lists the technical constraints suggested by existing technologies for static sub-items: both JAK1 inhibition and TYK2 allosteric inhibition may weaken IL-10 negative feedback, while JAK3 / TEC co-inhibitors, although not showing disruption of this IL-10 negative feedback, have ITK / NK functional limitations; the middle section shows that existing technologies do not disclose the combined use of TYK2-JH2 allosteric inhibition, ITK-isolated selective JAK3 inhibition, IL-10 / STAT3 net functional criteria, and peak-valley recovery window; the lower section illustrates that the present invention achieves decoupling of IFN / Tfh pathogenic axis inhibition and IL-10 / NK protective axis maintenance through JAK1 blanking, ITK isolation, measured criterion of net pSTAT3-Y705 ≥ 70%, and recovery window protection axis rebound. Detailed Implementation

[0021] The present invention is described below with reference to specific embodiments, but the scope of the invention is not limited thereto. The JAK family includes four members: JAK1, JAK2, JAK3, and TYK2. In SLE, the pathogenic hierarchy is driven by different JAK pairings: type I interferon via JAK1+TYK2→STAT1 / STAT2; IL-21→Tfh→autoantibodies via JAK1+JAK3→STAT3; IFN-γ / IL-6 via JAK1+JAK2→STAT1 / STAT3. The present invention precisely blocks the type I interferon axis through allosteric inhibition of TYK2, blocks the IL-21 / Tfh axis through selective inhibition of JAK3, and fully preserves the homeostatic protection function of JAK1.

[0022] The spatial pedestal layer of this invention provides a combination of a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor. TYK2 possesses a pseudokinase domain (JH2), which can be specifically targeted by the allosteric inhibitor; the allosteric inhibitor binds to the JH2 regulatory domain rather than the ATP-binding site (JH1), thus lacking substantial affinity for the JAK1, JAK2, and JAK3 catalytic domains. The representative compound, deuterocelexitinib, exhibits a three-state inhibition mechanism for TYK2-JH2, encompassing both self-inhibited and two activated states (Wang et al., Journal of Investigative Dermatology, 2025, PMID:40378946). The binding affinity Ki of the TYK2-JH2 allosteric inhibitor described in this invention to the TYK2 pseudokinase domain is ≤0.1 nM, while the Ki of the representative compound, deuterocelexitinib, is 0.02 nM.

[0023] Deuterocelexitinib demonstrated potent inhibition of the TYK2-dependent type I interferon pathway in functional assays of human whole blood and PBMCs, but exhibited limited efficacy against JAK1 / JAK3-γc-dependent NK cells, cytotoxic T cells, helper T cells, and regulatory T cells. The functional inhibitory IC50 difference between the TYK2 and JAK1 / JAK3 pathways exceeded 100-fold (Johnson et al., Frontiers in Immunology, 2024, PMID:39403378). Under clinically relevant exposure, the steady-state peak concentration of deuterocelexitinib was 8 to 17 times lower than the IC50 of JAK1 / 3 and more than 48 times lower than the IC50 of JAK2, with daily TYK2 inhibition lasting approximately 9–18 hours (Chimalakonda et al., Dermatology and Therapy, 2021, PMID:34471993). The above data support the conclusion that, under standard clinical exposure, the TYK2-JH2 allosteric inhibition arm can block the type I interferon axis while maintaining functional inhibition of JAK1 within the range of ≤30% as claimed in the claims.

[0024] The JAK1 functional inhibition rate was measured by downstream STAT phosphorylation following stimulation with a JAK1-dependent cytokine that does not directly overlap with the JAK3 inhibitory arm, to avoid attribution confusion caused by cytokine readings such as IL-15 that are simultaneously dependent on JAK1 and JAK3. Preferably, pSTAT3-Y705 after IL-10 stimulation (via IL-10Rα / JAK1 + IL-10Rβ / TYK2→STAT3) is used as a specific reading for JAK1 functional integrity, as this reading is independent of JAK3.

[0025] The safety significance of the triple strategic whitespace: JAK2 whitespace (IC50 > 500 nM) avoids the risk of hematopoiesis associated with EPO / TPO signaling inhibition. In SLE, hematopoiesis is difficult to distinguish from disease activity itself, thus JAK2 whitespace constitutes a safety redundancy; ITK whitespace (IC50 > 1000 nM) preserves the CD107a degranulation triggering function of NK cells. A comprehensive safety analysis of JAK3 / TEC dual inhibitors showed that the incidence of major adverse cardiovascular events and venous thromboembolism was low in the absence of JAK1 / JAK2 co-inhibition (King et al., American Journal of Clinical Dermatology, 2024, PMID:38263353). This invention satisfies the ITK isolation requirement for selective JAK3, and is expected to have a lower risk of impaired NK degranulation triggering function compared to the JAK3 / TEC dual inhibition regimen. The above safety logics are known individually and are used as components of the integration window in this invention, rather than as independent sources of inventiveness.

[0026] On a spatial base, the present invention provides a pulsatile dosing regimen: the drug combination is prepared into a formulation or dosing regimen that generates a pulsatile blood drug concentration-time curve, achieving a JAK3 functional inhibition rate ≥70% in the peak window (0-6 hours after administration), and deeply inhibiting the type I interferon axis and the IL-21 / Tfh axis (deep clearance of the pathogenic axis); in the recovery window (8-12 hours after administration, with a cumulative recovery time of ≥4 hours every 24 hours, preferably ≥6 hours), the JAK3-dependent protective function is restored to a functional inhibition rate ≤30%, with γc cytokine-induced pSTAT5 recovery to ≥70% of baseline as the functional recovery reading.

[0027] Differential STAT recovery selectivity: Within the recovery window, the recovery rates of protective STAT3 (IL-10 induced) and STAT5 (IL-15 induced) were higher than that of pathogenic STAT1 (type I interferon induced). Existing studies suggest that TYK2-mediated type I interferon-induced gene expression exhibits different kinetic characteristics compared to other cytokine pathways (Wang et al., Journal of Investigative Dermatology, 2025, PMID:40378946). Combined with the SOCS-locked negative feedback and relatively slow dephosphorylation characteristics of STAT1, the preferential rebound of the trough-phase protective axis (STAT3 / STAT5) protects IL-10 homeostasis and NK cell survival signals over time, while pathogenic STAT1 remains suppressed during the trough phase due to its slow recovery. This differential recovery selectivity is a reasonable expectation of this invention and can be verified by the following in vitro assays; however, no assertion of proven extent is made.

[0028] NK cell survival pulse maintenance: According to the time integral model of

[0011] , the NK cell anti-apoptotic protein (Bcl-2 / Mcl-1) pool is supplemented by the β-arm baseline maintained by JAK1 blanking and the γc-arm pulse provided by the recovery of JAK3 trough. When the trough duration is shorter than the decay time constant of the anti-apoptotic protein pool, the pool is maintained above the apoptosis threshold, and NK cell survival is maintained; a cumulative recovery time of trough ≥ 4 to 6 hours is the preferred setting based on this model, and the specific value can be determined by pharmacokinetic-pharmacodynamic evaluation.

[0029] Turnover constraints of covalent JAK3: In a preferred embodiment, the selective JAK3 inhibitor is a covalent inhibitor targeting JAK3-Cys909 (the active site of the JAK3 kinase domain, Cys909, is the only cysteine ​​residue in the JAK family, enabling the covalent inhibitor to achieve high JAK3 / JAK1 selectivity). Under covalent inhibition, functional recovery within the recovery window is determined by the resynthetic turnover of JAK3 protein rather than drug clearance; therefore, the trough duration must be set to accommodate the resynthetic turnover time of JAK3 protein in order to achieve functional recovery of the γc arm during the trough.

[0030] This invention defines IL-10 / STAT3 retention based on net functional effect: whether the target is met is solely determined by the baseline of ≥70% pSTAT3-Y705 retention after IL-10 stimulation in the target detection system, and does not depend on whether any single component of the TYK2-JH2 allosteric inhibitor or selective JAK3 inhibitor is absolutely neutral to the IL-10 receptor signal. This criterion directly addresses the fact that the TYK2 arm also partially attenuates IL-10 (see

[0012] ): net retention is achieved by the white space of the JAK1-dominant contributing arm and the deinhibition rebound during the pulsatile trough. Specifically, the IL-10 receptor complex consists of two arms: IL-10Rα (IL-10R1, coupled with JAK1) and IL-10Rβ (IL-10R2, coupled with TYK2), and the IL-10→STAT3 signal is the result of the synergistic effect of the two arms. The TYK2-JH2 allosteric inhibition of this invention partially weakens the IL-10Rβ / TYK2 arm, while JAK1 remains intact, preserving the IL-10Rα / JAK1 arm. Therefore, the IL-10 / STAT3 retention rate ≥70% as claimed is the net retention rate after superimposing the IL-10Rα (JAK1 preserved) arm and the IL-10Rβ (TYK2 partially weakened) arm. Given that the IL-10Rα / JAK1 arm makes a dominant contribution to the IL-10→STAT3 signal, and the recovery window of the pulsed scheme further provides periodic de-inhibition space, the superposition of the two is sufficient to maintain the net pSTAT3-Y705 retention rate at a baseline of ≥70%. Based on this, this invention uses a measurable net functional window to transform the phenomenon of the TYK2 arm partially weakening IL-10 into a quantifiable and achievable operating condition, rather than presupposing that either single arm is absolutely neutral to the IL-10 receptor signal.

[0031] The structural basis for allosteric selectivity: Deuterocelexitinib specifically binds to the TYK2 pseudokinase domain (JH2), whose molecular binding pocket differs significantly in atomic spatial arrangement from the corresponding region of JAK1, exhibiting a structural lack of affinity for JAK1. Its binding to TYK2-JH2 encompasses tri-state inhibition, including an autoinhibited state and two activated states (Wang et al., Journal of Investigative Dermatology, 2025, PMID:40378946). This structural difference supports maintaining ≤30% functional inhibition of JAK1 under clinically relevant exposure.

[0032] [Implementation Method - Spatial Base] Select a pair of compounds that satisfy the component selectivity described in claim 2 or claim 3: a TYK2-JH2 allosteric inhibitor (such as deuterated lexicitinib, Ki=0.02nM, IC50 > 1000nM for the JAK1 / 2 / 3 catalytic domain) and a covalently selective inhibitor targeting JAK3-Cys909 (IC50 < 100nM for JAK3, IC50 > 500nM for JAK1 / JAK2, and confirmed IC50 > 1000nM for ITK). Confirm that the following operating windows are simultaneously met in a human whole blood or PBMC functional evaluation system: (1) pathogenic axis inhibition ≥ 70%, (2) JAK1 functional inhibition ≤ 30% (measured by IL-10→pSTAT3), (3) JAK2 functional inhibition ≤ 30%, and (4) NK cell CD107a retention ≥ 70% baseline. When selecting or confirming the selective JAK3 inhibitor, it is preferable to use a counterscreen profile including JAK1, JAK2, JAK3, TYK2, and TEC family kinases ITK, BTK, TEC, RLK, and BMX for selectivity confirmation. The IC50 values ​​of each kinase are preferably measured under conditions where the ATP concentration is equal to the apparent Km of the corresponding kinase to avoid bias in selectivity evaluation due to differences in ATP concentration. A selective JAK3 inhibitor that meets the requirements of this invention should simultaneously satisfy the following: JAK3 IC50 < 100 nM, JAK1 and JAK2 IC50 > 500 nM, ITK IC50 > 1000 nM, and meet the JAK3 / γc pathway peak-trough regulatory window and the NK cell CD107a degranulation triggering function retention window in PBMC or whole blood functional systems.

[0033] [Implementation Method - Pulsating Regimen] The above combination is configured as a dosing regimen that generates a pulsating plasma concentration-time curve (e.g., once or twice daily, ensuring a peak JAK3 functional inhibition rate ≥70% within a peak window of 0–6 hours, a recovery window of 8–12 hours, and a cumulative recovery of ≥4–6 hours per 24 hours to ensure a JAK3 functional inhibition rate ≤30%). When using a JAK3-Cys909 covalent inhibitor, the trough duration is set to accommodate the JAK3 protein resynthesis turnaround time, so that γc arm function recovery is achieved through protein turnaround.

[0034] [Implementation Method: Differential Recovery and Pool Maintenance] During the recovery window, pSTAT5 was induced with IL-2 or IL-15, pSTAT3 with IL-10, and pSTAT1 with type I interferon. The recovery rates of the three relative to the baseline were measured to confirm that the recovery rate of protective STAT3 / STAT5 was higher than that of pathogenic STAT1. In vitro NK cell survival experiments confirmed that NK cell survival was maintained at the set trough duration. The above effects were defined by reasonable expectations. The above thresholds served as the criteria for determining whether a candidate regimen fell within the time-domain selective operating window of this invention. Specific drug concentrations, component ratios, exposure time points, and dosing cycle parameters could be determined through in vitro or pharmacokinetic-pharmacodynamic verification.

[0035] [Implementation Methods and In Vitro Validation Procedures] Cells and drugs are pre-incubated at 37°C and 5% CO2 for 60 minutes before the addition of stimulating factors; the final concentration of DMSO in each well should not exceed 0.05% (v / v). STAT3 retention validation: Stimulate with IL-10 (10 ng / mL) for 30 minutes, and detect pSTAT3-Y705 (intracellular flow cytometry). Retention rate = (MFI of drug-treated group / MFI of solvent control) × 100, ≥70% is considered satisfactory. IL-6 is strictly prohibited from replacing IL-10. IFN inhibition validation: Stimulate with recombinant type I interferon, and detect the qPCR scores of IFIT1, MX1, and IFI44L. Inhibition rate ≥80% is considered satisfactory; pSTAT1-Y701 intracellular flow cytometry can be used for rapid screening. JAK1 negative lock validation: Measure the JAK1 functional inhibition rate ≤30% using JAK1-dependent stimulation (IL-10→pSTAT3) that does not overlap with the JAK3 arm. NK CD107a validation: Human NK cells were co-cultured with K562 target cells, and the CD107a positivity rate was maintained at ≥70% baseline by flow cytometry.

[0036] Example 1 (SLE Time-Selective Protocol, Expected Effects, Verifiable): SLE patients with a type I interferon gene signature score higher than 2 standard deviations above normal, elevated circulating Tfh (CXCR5+PD-1+CD4+) proportion, SLEDAI-2K score ≤ 6, and no active lupus nephritis were given a combination of deuterocelexitinib and a JAK3-Cys909 covalently selective inhibitor, administered via a pulse regimen. In one embodiment, the peak window can achieve ≥80% inhibition of the type I interferon gene signature score and deep inhibition of the IL-21 / Tfh axis; the recovery window can achieve a net IL-10 / STAT3 retention rate ≥70% of baseline, γc-pSTAT5 recovery ≥70% of baseline, and NK cell anti-apoptotic protein pool maintained above the threshold. Compared to continuous dosing regimens with equal exposure, pulsed dosing regimens achieve higher average retention of IL-10 / STAT3 and deeper Tfh clearance during the trough phase, and increase the rate of decrease in anti-dsDNA antibody titers. These effects can be verified by in vitro and clinical sample testing as described in this specification, and are not empirically asserted in this application with a fixed amplitude.

[0037] Example 2 (Pure spatial base scheme, as an implementation method independent of reasonable temporal expectations): Without implementing the pulse scheme, a combination of a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor, satisfying the core dual-window of claim 1 and the blank window of claim 4, is applied. Relying on spatial target selectivity and strategic blanking of JAK1, the type I interferon axis and the IL-21 / Tfh axis are suppressed, the IL-10 / STAT3 homeostatic axis is relatively preserved, and the NK cell CD107a degranulation triggering function is preserved. This implementation method defines the basis of the scope of protection of this invention.

[0038] The detection indicators described in this invention can be replaced by alternative detection technologies (including but not limited to single-cell sequencing and high-sensitivity proteomics) that have been confirmed to have equivalent or higher sensitivity, specificity, and reproducibility. Unless otherwise specified, the IC50 values ​​described in this invention refer to the half-maximal inhibitory concentration (IC50) measured under conditions where the ATP concentration is equal to the apparent Km of the corresponding kinase. For candidate compounds used to confirm JAK3 selectivity, it is preferable to simultaneously measure the IC50 of JAK1, JAK2, JAK3, TYK2, and TEC family kinases ITK, BTK, TEC, RLK, and BMX on the same experimental platform or in the same counterscreen spectrum to ensure consistency in selectivity evaluation.

[0039] [Implementation Methods, Product Forms, and Packaging] The pharmaceutical compositions of the present invention can be defined not only by component selectivity but also by their net functional effect in a human peripheral blood mononuclear cell function evaluation system, i.e., simultaneously achieving a pathogenic signaling pathway inhibition rate ≥70% and an IL-10 / STAT3 retention rate ≥70% baseline, thereby supporting compositions defined by functional characteristics. The compositions can be formulated as single-component combination preparations or as pharmaceutical combination packages containing separately packaged TYK2-JH2 allosteric inhibitor units and selective JAK3 inhibitor units, along with instructions for use in systemic lupus erythematosus, to accommodate differences in the pharmacokinetic properties of the two components. The TYK2-JH2 allosteric inhibitor, deuterocelexitinib, can be a pharmaceutically acceptable salt, solvate, prodrug, any crystal form, or amorphous form.

[0040] [Implementation Method - Test Kit] This invention further provides a test kit for evaluating the applicability of drug combinations to systemic lupus erythematosus (SLE), comprising: type I interferon, IL-10 and IL-15 stimulating reagents; anti-pSTAT1-Y701, anti-pSTAT3-Y705 and anti-pSTAT5-Y694 detection antibodies; nucleic acid detection reagents for measuring the expression scores of type I interferon-induced genes IFIT1, MX1 and IFI44L; K562 target cells or equivalent target cells or stimulation systems validated to induce CD107a degranulation in human NK cells; and anti-CD107a detection antibody. This kit uses SLE-specific readings, such as type I interferon gene scores and NK cell CD107a triggering function, as essential components, thus distinguishing it from general kinase timing detection platforms that do not specify indications. The kit's decision logic can output the ratio of the pathogenic axis recovery rate to the protective axis recovery rate from the data processing unit. Furthermore, its signal quantification is not limited to a specific detection platform and can be performed using flow cytometry, electrochemiluminescence, or mass cytometry. This kit is used for in vitro evaluation of the suitability of candidate drug combinations for systemic lupus erythematosus (SLE). The test results are not intended to make a diagnostic conclusion about the disease itself, nor are they intended to directly guide in vivo treatment.

[0041] [Implementation Methods, Technical Details, and Preferred Effects] In a preferred embodiment, for systemic lupus erythematosus patients with elevated type I interferon gene characteristic scores, the present invention's regimen can achieve a type I interferon gene characteristic score inhibition rate of ≥80% under clinically effective exposure. Regarding the maintenance of NK cell survival during the trough phase, the mechanism lies in the fact that the trough duration of the pulsed dosing regimen is set to be shorter than the decay time constant of the expression pool of NK cell anti-apoptotic proteins Bcl-2 and / or Mcl-1, so that the anti-apoptotic protein pool is maintained above the apoptosis threshold during the pulse cycle; this mechanism can be characterized by the experimental measurement of Bcl-2 / Mcl-1 expression levels and NK cell viability at the end of the trough phase (as defined in claim 10). In one embodiment, the selective JAK3 inhibitor is a covalent inhibitor targeting JAK3-Cys909, which forms a sustained high occupancy during the peak phase and achieves functional recovery through JAK3 protein resynthesis during the trough phase; however, the achievement of the recovery window is not limited to this specific recovery mechanism and can also be achieved through drug clearance.

[0042] [Implementation Methods: JAK3 Function Degraders and TYK2-JH2 Equivalent Allosteric Modulators] In some embodiments, the selective JAK3 inhibitor further includes JAK3 function degraders that selectively reduce JAK3 functional abundance or JAK3 / γc pathway functional output, including but not limited to reversible inhibitors, covalent inhibitors, allosteric inhibitors, targeted protein degraders, or molecular colloidal degraders; their common feature is that they bring the JAK3 / γc-dependent signal to the peak window inhibition threshold in the target detection system, while simultaneously satisfying the JAK1, JAK2, and ITK white windows. The TYK2-JH2 allosteric inhibitor is not limited to deuterocelexitinib, but also includes JH2 allosteric modulators, conformation-locking agents, or pharmaceutically acceptable forms thereof that can bind to the TYK2 pseudokinase domain and stabilize the TYK2 inactive conformation, reduce TYK2-dependent type I interferon signal output, and do not primarily act on the ATP site of the JAK1 / JAK2 / JAK3 catalytic domain.

[0043] [Implementation Method: Comparison of Peak and Trough Window with Iso-AUC Continuous Exposure] The peak and recovery windows described in this invention can be determined by blood drug concentration-time curves, intracellular target occupancy, pSTAT function readings, or combinations thereof. Without departing from the concept of this invention, the peak and recovery windows are not limited by a specific number of hours after administration. As long as the pathogenic axis is deeply inhibited during the peak period and the protective STAT3 / STAT5 function is restored during the trough period within one dosing cycle, it falls within the time-domain selective operating window of this invention. In one embodiment, this invention also includes a method for comparing candidate pulsed exposure protocols with iso-AUC continuous exposure protocols: under conditions of the same or similar total drug exposure, pathogenic STAT1 / ISG inhibition, IL-10-induced pSTAT3 retention, IL-15-induced pSTAT5 recovery, and NK CD107a degranulation function are measured respectively. When the pulsed exposure protocol improves the protective axis recovery or NK function retention without reducing pathogenic axis inhibition, the candidate protocol is determined to have a time-domain selective advantage.

[0044] [Implementation Methods: IL-21 / Tfh Axis Readings, NK Equivalent Detection System, and JAK3 / TEC Co-inhibitor Treatment] The IL-21 / Tfh axis inhibition can be characterized by a decrease in pSTAT3 levels in CD4+ T cells or circulating Tfh cells after IL-21 stimulation, a decrease in the proportion of CXCR5+PD-1+CD4+ circulating Tfh cells, a decrease in ICOS or Bcl-6-related Tfh phenotypes, a decrease in the proportion of plasmablasts, or a decrease in autoantibody-related B cell activation readings. The NK cell CD107a degranulation triggering function can be achieved using K562 target cells, or by using equivalent target cells that have been validated to induce human NK cell degranulation, antibody-dependent cytotoxic stimulation systems, CD16 cross-linking systems, or other standardized NK activation systems; the common criteria are CD107a positivity rate, granzyme / perforin release, or target cell killing function reaching a predetermined retention level relative to baseline. In a preferred embodiment, the selective JAK3 inhibitor is not a JAK3 / TEC family co-inhibitor that significantly inhibits ITK; if the candidate JAK3 inhibitor has TEC family cross-activity, then the retention rate of human NK cell CD107a degranulation trigger function must reach a predetermined threshold as a functional condition for falling into the selective JAK3 arm of the present invention.

[0045] [Implementation Methods: Patient Stratification and Responder Identification] In some implementation methods, applicable patients may also be stratified based on the ability of IL-10 to induce pSTAT3 response, the ability of IL-15 to induce pSTAT5 response, the number of NK cells, or the baseline function of CD107a degranulation in ex vivo samples; preferably, patients who still retain detectable IL-10 / STAT3 or IL-15 / STAT5 response before administration are selected to improve the feasibility of time-domain selective windows.

[0046] [Implementation Method: Digital Indicator for Combined Packaging] The indicator information commonly used for systemic lupus erythematosus may be one or more of the following: paper instructions, drug labels, electronic instructions, QR code links, drug information carriers, medication guidance cards, software interface prompts, or medication preparation instructions from medical institutions.

[0047] [Implementation Method - Physical Carrier of Data Processing Unit] The data processing unit can be a computer-readable medium storing a calculation program, a built-in analysis module of a detection instrument, a laboratory information management system module, a local server, or a cloud-based data processing module. It is configured to receive pSTAT1, pSTAT3, pSTAT5, ISG qPCR scores, and NKCD107a data, and output the pathogenic axis recovery rate, the protective axis recovery rate, the ratio between the two, or the time-domain selectivity determination result.

Claims

1. The use of a pharmaceutical combination in the preparation of a medicament for treating systemic lupus erythematosus, said pharmaceutical combination comprising a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor, characterized in that, The drug combination simultaneously meets the following core dual-window criteria in the in vitro functional evaluation system: (1) the pathogenic signaling pathway inhibition rate is ≥70%, and the pathogenic signaling pathway includes the TYK2-dependent type I interferon pathway and the JAK3 / γc-dependent IL-21 / T follicular helper cell pathway; (2) the IL-10 / STAT3 signaling pathway retention rate is ≥70% of the baseline.

2. A pharmaceutical composition for systemic lupus erythematosus, characterized in that, The formulation includes a TYK2-JH2 allosteric inhibitor and a selective JAK3 inhibitor. The TYK2-JH2 allosteric inhibitor achieves allosteric inhibition by binding to the TYK2 pseudokinase domain (JH2 regulatory domain), with a binding affinity Ki ≤ 0.1 nM for TYK2-JH2 and IC50 > 1000 nM for the catalytic domains of JAK1, JAK2, and JAK3. The selective JAK3 inhibitor has an IC50 < 100 nM for JAK3, an IC50 > 500 nM for both JAK1 and JAK2, and an IC50 > 1000 nM for ITK.

3. The application according to claim 1, characterized in that, The TYK2-JH2 allosteric inhibitor achieves allosteric inhibition by binding to the TYK2 pseudokinase domain (JH2 regulatory domain). Its binding affinity for TYK2-JH2 is Ki≤0.1 nM, and its IC50 for the catalytic domains of JAK1, JAK2, and JAK3 is >1000 nM. The selective JAK3 inhibitor has an IC50 <100 nM for JAK3, an IC50 >500 nM for both JAK1 and JAK2, and an IC50 >1000 nM for ITK.

4. The application according to claim 1, characterized in that, The drug combination also meets the following requirements in the in vitro functional evaluation system: functional inhibition rate of JAK1 catalytic activity ≤30%, wherein the JAK1 functional inhibition rate is determined as the ratio of downstream STAT phosphorylation level after stimulation by JAK1-dependent cytokines that do not directly overlap with the JAK3 inhibitory arm to the baseline without drug administration; functional inhibition rate of JAK2 catalytic activity ≤30%; and ITK-dependent NK cell CD107a degranulation triggering function retention rate ≥70% of baseline, determined by human NK cell CD107a degranulation assay.

5. The application according to claim 1, characterized in that, The drug is used for systemic lupus erythematosus patients with elevated type I interferon gene trait scores and / or elevated proportion of circulating T follicular helper cells (CXCR5+PD-1+CD4+).

6. The application according to claim 1, characterized in that, The retention rate of the IL-10 / STAT3 signaling pathway is based on a baseline of ≥70% pSTAT3-Y705 retention rate as measured by the net functional effect of the drug combination in the target detection system, and does not depend on whether any single component of the TYK2-JH2 allosteric inhibitor or the selective JAK3 inhibitor is absolutely neutral to the IL-10 receptor signal.

7. The application according to claim 1, characterized in that, The TYK2-JH2 allosteric inhibitor is deucravacitinib or its pharmaceutically acceptable salt, solvate, prodrug, any crystal form or amorphous form thereof.

8. The application according to claim 1, characterized in that, The drug is formulated as a formulation or dosing regimen suitable for generating a peak window and a recovery window within a dosing cycle, wherein: within the peak window, the drug combination inhibits the overall pathogenic signaling of the type I interferon axis and the IL-21 / Tfh axis by ≥70%, wherein the JAK3 inhibitory arm inhibits the JAK3 / γc-dependent IL-21 / Tfh signaling by ≥70%; within the recovery window, JAK3-dependent protective function is restored to a functional inhibition rate of ≤30%, with γc cytokine-induced pSTAT5 recovery to ≥70% of baseline level as the functional recovery reading; preferably, the peak window is 0–6 hours after administration, the recovery window is 8–12 hours after administration, and the cumulative recovery time is ≥4 hours every 24 hours; the achievement of the recovery window is not limited to a specific recovery mechanism and can be achieved through drug clearance and / or JAK3 protein resynthesis turnover.

9. The application according to claim 8, characterized in that, The pulsed dosing regimen is configured such that the recovery rate of protective STAT3 (IL-10 induced) and STAT5 (IL-15 induced) is higher than that of pathogenic STAT1 (type I interferon induced) within the recovery window, thereby causing the IL-10 / STAT3 homeostatic signal and NK cell survival signal to rebound relatively preferentially during the trough phase; the difference in recovery rate can be verified by in vitro detection.

10. The application according to claim 9, characterized in that, The pulsed dosing regimen is designed to maintain the expression levels of NK cell anti-apoptotic proteins Bcl-2 and / or Mcl-1 at the end of the trough period at no less than 70% of the baseline before dosing or the baseline of the solvent control, and to maintain NK cell survival by measuring the proportion of Annexin V negative and 7-AAD negative cells at the end of the trough period, based on the proportion of such cells at the end of the trough period.

11. A method for in vitro evaluation of the suitability of candidate drug combinations for systemic lupus erythematosus, characterized in that, All biological samples used were ex vivo samples. This method is not intended to obtain in vivo disease diagnosis or treatment results. It includes: (a) measuring the inhibition of type I interferon-induced pSTAT1-Y701 and IFN-induced genes (IFIT1, MX1, IFI44L) qPCR scores within the peak window of pulsed exposure; (b) measuring the recovery rate of IL-10-induced pSTAT3-Y705 and IL-15-induced pSTAT5-Y694 relative to baseline within the recovery window; (c) measuring the retention rate of CD107a degranulation triggering function in the co-culture system of NK cells and K562 target cells; when the recovery rate of pathogenic STAT1 in (a) is lower than the recovery rate of protective STAT3 / STAT5 in (b), and the CD107a retention rate in (c) is ≥70% of baseline, the candidate drug combination is determined to meet the time-domain selective operation window.

12. The pharmaceutical composition according to claim 2, characterized in that, The composition simultaneously achieves the following in the human peripheral blood mononuclear cell function evaluation system: ≥70% inhibition rate of pathogenic signaling pathways (including type I interferon pathway and IL-21 / T follicular helper cell pathway) and ≥70% baseline retention rate of IL-10 / STAT3 signaling pathway.

13. The pharmaceutical composition according to claim 2, characterized in that, The composition is formulated as a pharmaceutical combination package, which contains separately packaged TYK2-JH2 allosteric inhibitor units and selective JAK3 inhibitor units, as well as instructions in the form of paper instructions, electronic instructions, labels, QR code links, medication guidelines or drug information carriers indicating that the two are used together to treat systemic lupus erythematosus.

14. A diagnostic kit for evaluating the suitability of a drug combination for systemic lupus erythematosus, characterized in that, The kit contains: (a) type I interferon, IL-10 and IL-15 stimulation reagents; (b) anti-pSTAT1-Y701, anti-pSTAT3-Y705 and anti-pSTAT5-Y694 detection antibodies; (c) nucleic acid detection reagents for determining the expression scores of type I interferon-induced genes IFIT1, MX1 and IFI44L; (d) K562 target cells or equivalent target cells or stimulation systems that have been validated to induce CD107a degranulation in human NK cells, and anti-CD107a detection antibodies; the kit is used to measure pathogenic axis inhibition in the peak window, to measure the differential recovery of protective STAT3 / STAT5 relative to pathogenic STAT1 in the recovery window, and to measure the retention of NK cell CD107a degranulation triggering function.

15. The test kit according to claim 14, characterized in that, It also includes a data processing unit, which is a computer-readable medium storing a calculation program, a built-in analysis module of a detection instrument, a laboratory information processing module, or a cloud data processing module, configured to output the ratio of the pathogenic axis recovery rate to the protective axis recovery rate; and the signal quantification of the kit is not limited to a specific detection platform, but is applicable to any of flow cytometry, electrochemiluminescence, and mass spectrometry.

Images