ROCK2 inhibition suppresses cytotoxic T lymphocyte-mediated platelet destruction in primary immune thrombocytopenia

Juanjuan Song; Bingjie Ding; Qian Liu; Ao Xie; Mengjuan Li; Xuewen Song; Fengdi Wang; Junxia Hu; Liu Liu; Xinguang Liu; Hu Zhou

doi:10.1186/s13578-026-01623-x

. 2026 Jul 20;16:114. doi: 10.1186/s13578-026-01623-x

ROCK2 inhibition suppresses cytotoxic T lymphocyte-mediated platelet destruction in primary immune thrombocytopenia

Juanjuan Song ^1,^2,^#, Bingjie Ding ^1,^#, Qian Liu ², Ao Xie ¹, Mengjuan Li ¹, Xuewen Song ¹, Fengdi Wang ¹, Junxia Hu ¹, Liu Liu ³, Xinguang Liu ^2,^✉, Hu Zhou ^1,^✉

PMCID: PMC13670769 PMID: 42477845

Abstract

Cytotoxic T lymphocyte (CTL)-mediated platelet destruction represents an important pathogenic mechanism in primary immune thrombocytopenia (ITP) patients. Rho-associated coiled-coil kinase 2 (ROCK2) is an emerging regulator of immune balance, but its role in pathogenic CTL activation in ITP remains undefined. Here, we demonstrated that selective ROCK2 inhibition with KD025 potently suppressed CTL-mediated platelet destruction. In vitro, KD025 treatment of CTLs from ITP patients suppressed key effector functions, reducing degranulation as measured by CD107a expression, diminishing the secretion of cytotoxic molecules such as granzyme B and perforin, and decreasing CTL-platelet conjugate formation, resulting in reduced platelet apoptosis and activation. RNA-sequencing revealed downregulation of cytotoxic and glycolytic programs, with enrichment of JAK-STAT signaling. Mechanistically, KD025 reversed the pathogenic metabolic shift in ITP CTLs by lowering glycolytic flux and restoring mitochondrial respiration, accompanied by decreased STAT3 phosphorylation. IL-6-mediated STAT3 activation largely reversed these effects, indicating a ROCK2-STAT3-dependent mechanism. In vivo, both daily KD025 administration to an active ITP mouse model and transplantation of KD025-pretreated CD8⁺ T cells into irradiated Rag1^−/− mice alleviated CTL-mediated platelet apoptosis and increased platelet counts. Collectively, these findings supported a model in which ROCK2 contributed to STAT3-associated transcriptional and metabolic regulation of CTL pathogenicity, providing a rationale for further preclinical and translational evaluation of ROCK2 inhibition in ITP.

Graphical abstract

graphic file with name 13578_2026_1623_Figa_HTML.webp

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13578-026-01623-x.

Keywords: Primary immune thrombocytopenia, Cytotoxic T lymphocyte, ROCK2, KD025, STAT3

Introduction

Primary immune thrombocytopenia (ITP) is an acquired autoimmune disorder characterized by accelerated platelet destruction and impaired platelet production, leading to an increased risk of bleeding [1, 2]. Traditionally, autoantibody-mediated platelet clearance via Fcγ receptor (FcγR)-bearing macrophages has been considered the central pathogenic mechanism [3]. However, accumulating evidence highlights a complementary and clinically important antibody-independent pathway: cytotoxic T lymphocyte (CTL)-mediated platelet destruction [4, 5]. CTLs can kill platelets directly through granzyme/perforin-mediated lysis and death receptor signaling pathways [6, 7]. The formation of CTL-platelet conjugates and the ensuing polarized secretion of cytotoxic granules represent key mechanistic step in CTL-driven platelet lysis [8]. Despite the increasing availability of new therapeutic agents, a substantial proportion of patients remain refractory to treatments, underscoring the urgent need for novel strategies that target alternative pathogenic mechanisms, particularly CTL-mediated platelet destruction [9].

Rho-associated coiled-coil kinase 2 (ROCK2) is a serine/threonine kinase with dual functions in regulating cellular processes [10, 11]. While initially recognized for its canonical function in orchestrating actin cytoskeletal dynamics, which is crucial for cell adhesion and migration [12, 13], a growing body of work has revealed its non-canonical role as a direct modulator of key signaling pathway [14, 15]. This includes its ability to interact with and regulate transcription factors such as STAT, thereby controlling T cell differentiation and metabolism [16]. This dual functionality makes ROCK2 a particularly attractive therapeutic target in T-cell-driven autoimmune diseases. The selective ROCK2 inhibitor KD025, approved for the treatment of chronic graft-versus-host disease (cGVHD), exerts robust immunomodulatory effects by dampening inflammatory signaling pathways and facilitating the restoration of immune homeostasis [17, 18]. However, its direct effects on CTL activity—particularly cytotoxic molecule expression, degranulation capacity, and CTL-platelet conjugate formation—remain poorly characterized in ITP. Elucidating these mechanisms is crucial, as modulation of CTL-mediated cytotoxic pathways may provide therapeutic benefit for patients with refractory ITP who respond inadequately to conventional immunosuppressive therapy.

In this study, we examined how ROCK2 inhibition with KD025 affected the effector function and metabolic profile of CTLs from ITP patients, focusing on cytotoxicity, degranulation, and CTL-platelet conjugate formation. By integrating functional, transcriptional, and metabolic analyses, we delineated an immunometabolic program in which ROCK2 appeared to influence STAT3-associated regulation of pathogenic CTL activity in ITP, thereby supporting ROCK2 inhibition as a potential therapeutic strategy for CTL-mediated autoimmunity.

Materials and methods

Patients and healthy controls

Adult patients with ITP were prospectively enrolled at Qilu Hospital, Shandong University, and Henan Cancer Hospital, Zhengzhou University. Active ITP was defined according to the 2019 ASH guidelines [19]. Peripheral blood samples were obtained before initiation of ITP-specific therapy. An initial cohort of 35 patients with active ITP was screened. CTL induced platelet apoptosis was quantified as described in the Reference [20], and patients with values above the upper limit of the healthy volunteer range were classified as CTL-mediated ITP. Seventeen patients fulfilled this criterion and were included in all subsequent immunologic and mechanistic analyses. At enrollment, these 17 patients (11 females and 6 males; age range, 19–68 years; median age, 45 years), exhibited platelet counts ranging from 1 × 10⁹/L to 29 × 10⁹/L (median platelet count, 15 × 10⁹/L). Six ITP patients in remission with normal platelet counts and of treatment were enrolled as a comparator group. Baseline demographic and hematologic characteristics were provided in Table 1. Thirty-two age- and sex-matched healthy volunteers (18 females and 14 males; age range, 26–72 years; median age, 48 years) were recruited as healthy controls (HCs), with platelet counts ranging from 157 × 10⁹ to 294 × 10⁹/L (median, 254 × 10⁹/L). All participants gave written informed consent, and the study was approved by the ethics committees of the participating centers in accordance with the Declaration of Helsinki.

Table 1.

Clinical characteristics of enrolled ITP patients

Patients No.	Sex/age (years)	PLT (× 10⁹/L)	Bleeding symptoms	Anti-GPIIb/IIIa	Anti-GPIb/IX	Major previous drugs
CTL-mediated active ITP
1	M/21	29	PT + EC	−	−	Pred, IVIG, rhTPO
2	F/45	28	PT + EC	−	−	Pred, IVIG, rhTPO
3	F/27	5	ME + EC	+	−	Pred, rhTPO
4	F/34	1	EC + GH	−	−	DEX, rhTPO, IVIG
5	F/60	11	–	−	−	Pred, IVIG
6	F/19	22	PT + EC	−	−	Pred, TPO-RA
7	M/33	6	PT + EP	−	−	Pred
8	F/68	15	PT + EC	−	−	DEX
9	M/58	8	EC + GIH	−	−	Pred, IVIG
10	F/57	15	PT + GH	−	−	Pred, rhTPO
11	F/42	11	–	−	−	DEX
12	F/43	20	EC	−	−	Pred, IVIG
13	M/55	23	–	−	−	Pred, rhTPO
14	F/57	15	EC	−	−	DEX, IVIG
15	M/23	25	PT	−	−	Pred, IVIG
16	F/53	5	PT + GH + CH	−	−	DEX, rhTPO
17	M/51	13	–	−	+	Pred, IVIG, rhTPO
ITP in remission
1	F/32	180	–	ND	ND	Pred, CsA, RTX
2	F/67	210	–	ND	ND	DEX, IVIG, rhIL-11
3	F/78	250	–	ND	ND	Pred, rhTPO, Dana
4	M/67	310	–	ND	ND	DEX, IVIG, rhIL-11
5	M/51	270	–	ND	ND	Pred, rhTPO, RTX
6	M/32	240	–	ND	ND	Pred, rhTPO

Open in a new tab

CH, cerebral hemorrhage; CsA, ciclosporin; Dana, danazol; EC, ecchymoses; DEX, dexamethasone; EP, epistaxis; F, female; GH, gingival hemorrhage; GIH, gastrointestinal hemorrhage; ITP, primary immune thrombocytopenia; IVIG, intravenous gamma globulin; M, male; ME, menorrhagia; Pred, prednisone; PT, petechiae; rhIL-11, recombinant human interleukin-11; rhTPO, recombinant human thrombopoietin; RTX, rituximab; TPO-RA, thrombopoietin receptor agonist; ND, not determined; -, no bleeding symptom

To assess the potential impact of long-term ROCK2 inhibition on platelet function in vivo, peripheral blood samples were obtained from five cGVHD patients receiving active Belumosudil (KD025) treatment following allogeneic hematopoietic stem cell transplantation. Belumosudil was administered at 200 mg once daily, and treatment duration ranged from 4 to 18 months (median, 9 months). Platelet aggregation studies were performed using platelet-rich plasma isolated from these patients.

Active ITP mouse model

An active ITP mouse model was established as previously described [21]. Briefly, C57BL/6 Cd61-knockout (Cd61^−/−) mice were immunized weekly with platelets isolated from syngeneic wild-type (WT) C57BL/6 male mice (8–12 weeks old). Splenocytes from immunized mice were harvested, and 5 × 10⁴ cells per mouse were intravenously transferred into sublethally irradiated severe combined immunodeficient (Rag1^−/−) mice to construct the active ITP murine model. The strains and backgrounds of the mice are detailed in the Supplementary Methods.

At the time of splenocyte transfer, recipient mice were randomized to receive the following treatments: 2% DMSO control (intraperitoneal injection, once every 2 days and oral gavage, daily; Sigma-Aldrich), KD025 (100 mg/kg/day, oral gavage; MedChemExpress), IL-6 (2 mg/kg, intraperitoneal injection, once every 2 days; PeproTech), and the combination of IL-6 and KD025. Platelet counts were determined weekly. On day 28, mice were anesthetized and euthanized, and splenocytes were collected for flow cytometric (FCM) analysis.

To determine whether the therapeutic benefit of KD025 was mediated through direct modulation of CD8⁺ T cells, CD8⁺ T cells were purified from anti-CD61-sensitized splenocytes and cultured in vitro with KD025 (1 µM) or 0.1% DMSO for 72 h. These pretreated CD8⁺ T cells were then recombined with CD8⁺ T-cell-depleted splenocytes and adoptively transferred into irradiated Rag1^−/− mice to induce ITP.

All animal experiments were approved by the Animal Care and Use Committee of Qilu Hospital and conducted in accordance with institutional guidelines of Shandong University.

Isolation and culturing of CD8⁺ T Cells

Peripheral blood mononuclear cells (PBMCs) were isolated from peripheral venous blood of ITP patients and HCs using Ficoll-Hypaque density gradient centrifugation (Amersham Biosciences). CD8⁺ T cells were sorted from PBMCs by magnetic microbeads and MS separation columns (Miltenyi Biotec), with a purity of more than 90% as confirmed by FCM. For in vitro culture, CD8⁺ T cells were maintained in RPMI 1640 medium (Life Technologies) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco) and 1% penicillin and streptomycin (Solarbio). Cells were stimulated with anti-human CD3 antibodies (1 µg/mL; eBioscience), anti-human CD28 antibodies (1 µg/mL; eBioscience) and recombinant human interleukin-2 (rhIL-2, 10 ng/mL; PeproTech) for 72 h at 37 °C in a 5% CO₂ atmosphere, in the presence of KD025 (1 µM; MedChemExpress), IL-6 (50 ng/mL; PeproTech) or 0.1% DMSO (Sigma-Aldrich). After incubation, CD8⁺ T cells were collected for CTL proliferation, CTLs-mediated platelet apoptosis and activation assays by FCM. Culture supernatants from KD025- or DMSO-treated CD8⁺ T cells were collected for enzyme-linked immunosorbent assay (ELISA).

ROCK kinase activity assay

ROCK activity in PBMC lysates was measured using a 96-well ROCK Activity Assay Kit (Cell Biolabs) according to the manufacturer’s instructions. This enzyme immunoassay detects ROCK-dependent phosphorylation of recombinant MYPT1 pre-coated on a 96-well plate at the Thr696 site. Briefly, PBMC lysates were prepared in lysis buffer, and 90 µL of each sample was added to wells of the MYPT1-coated plate. The kinase reaction was initiated by adding 10 µL of 10× kinase reaction buffer containing ATP and DTT. According to the kit protocol, the reaction buffer consisted of Tris-based kinase buffer supplemented with MgCl₂, glycerol-2-phosphate, Na3VO4, ATP, and DTT. Samples were incubated at 30 °C for 30–60 min. After termination of the kinase reaction, wells were washed and incubated with anti-phospho-MYPT1 (Thr696) antibody, followed by HRP-conjugated secondary antibody. Signal was developed using substrate solution, stopped with stop solution, and absorbance was read at 450 nm using a microplate reader. ROCK activity was expressed on the basis of optical density after subtraction of blank/background values, according to the manufacturer’s protocol. The assay detects total ROCK kinase activity and is based on phosphorylation of the physiological substrate MYPT1.

Isolation of washed human platelets

Peripheral blood samples from HCs were collected into sodium citrate anticoagulant tubes. Samples were processed promptly after collection. Platelet-rich plasma (PRP) was prepared by centrifugation at 200 × g for 10 min at room temperature without brake. The upper PRP fraction was carefully aspirated and transferred to a fresh tube. Platelets were collected by centrifugation at 800 × g for 10 min, washed once with Tyrode’s buffer, and centrifuged again under the same conditions. The final platelet pellet was resuspended in medium for downstream assays. To minimize platelet activation, all aspiration and resuspension steps were performed gently. Platelet concentration was determined using an automated hematology analyzer before subsequent experiments.

CTL and platelet co-culture system

For experiments involving ITP patients, CD8⁺ T cells were isolated from PBMCs and cultured in complete RPMI 1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. To model the effector function of patient-derived CTLs, isolated CD8⁺ T cells were stimulated with anti-human CD3/CD28 antibodies and IL-2, followed by treatment with KD025, IL-6, or 0.1% DMSO for 72 h in vitro. IL-6 was used only during the CTL pretreatment stage to activate STAT3 signaling. After pretreatment, CTLs were thoroughly washed and resuspended in fresh medium before being added to the platelet co-culture system. Therefore, exogenous IL-6 was not intentionally maintained during the subsequent platelet co-culture assay.

This co-culture system was designed to evaluate disease-relevant CTL-mediated platelet injury. Based on the established concept that CD8⁺ CTLs in ITP patients may be primed in vivo against platelet-derived autoantigens, including platelet glycoproteins such as GPIIb/IIIa and GPIb/IX, patient-derived CD8⁺ T cells were used as effector cells [8, 22, 23]. Fresh platelets from healthy donors were used as standardized target cells because they express shared platelet antigenic determinants and avoid the confounding effects of severe thrombocytopenia, prior treatment, or pre-existing platelet activation in ITP patients. Thus, exposing ITP-derived CTLs to healthy donor platelets allows ex vivo assessment of the effector phase of platelet-directed CTL cytotoxicity.

To minimize alloimmune or blood group-related confounding effects, healthy donor platelets used in the co-culture assay were matched for ABO and Rh blood types with the corresponding ITP patient-derived CTLs. CD8⁺ T cells and platelets were adjusted to final concentrations of 1 × 10⁵ cells/mL and 1 × 10⁶ platelets/mL, respectively, and co-cultured for 4 h at 37 °C. No additional IL-6 was added during the co-culture stage. Platelet injury or activation after co-culture was then assessed using the indicated downstream assays.

For murine experiments, Cd61^−/− mice immunized with WT platelets weekly were concurrently treated with KD025 (100 mg/kg, oral gavage) or 2% DMSO (oral gavage) as a control, administered daily for 4 consecutive weeks. In the active ITP mouse model, mice were treated with 2% DMSO control (intraperitoneal injection, once every 2 days and oral gavage, daily), KD025 (100 mg/kg/day, oral gavage), IL-6 (2 mg/kg, intraperitoneal injection, once every 2 days), and the combination of IL-6 and KD025 for 4 consecutive weeks. After treatment, splenocytes were harvested from both immunized Cd61^−/− mice and active ITP mice, and CD8⁺ T cells were isolated using a negative-selection magnetic bead kit (Miltenyi Biotec). Platelets were isolated from WT mice. CD8⁺ T cells and platelets were adjusted to concentrations of 1 × 10⁵/mL and 1 × 10⁶/mL, respectively, and cocultured for 4 h at 37 °C.

Flow cytometry

For apoptosis assessment, PBMCs were stained with fluorescein isothiocyanate (FITC)-Annexin V and propidium iodide (PI) using a Cell Apoptosis Kit (Best Bio). Stained cells were subsequently analyzed using a Gallios Flow cytometer (Beckman Coulter). Data were processed using Kaluza (Beckman Coulter).

CTL proliferation was assessed by FCM. Proliferation was measured using the division index, defined as the total number of cell divisions divided by the number of cells before stimulation. Data analysis was performed using NovoExpress (Agilent).

For detection of phosphorylated STAT3 (p-STAT3), PBMCs were isolated and cultured in vitro. Cells were incubated with anti-CD3/CD28 antibodies and either KD025 or DMSO for 1 h, then stimulated with IL-6 for an additional 15 min. Next, cells were resuspended in cold PBS, and stained with FITC-conjugated anti-CD8 monoclonal antibodies (mAbs). Cells were then fixed with Phosflow Lyse/Fix buffer (BD Bioscience), permeabilized with 90% methanol on ice, and stained with APC-conjugated anti-p-STAT3 mAbs in PBS. For analysis, the cells were acquired on a NovoCyte Quanteon (Agilent), and data were analyzed using NovoExpress (Agilent).

To evaluate CTL-mediated platelet apoptosis, platelets were stained with JC-1 using a mitochondrial membrane potential assay kit (Beyotime). JC-1 was selected because platelets are anucleated cells in which apoptosis is closely associated with mitochondrial dysfunction. Although Annexin V/7-AAD is commonly used for apoptosis detection in nucleated cells, Annexin V positivity in platelets may also reflect activation-induced phosphatidylserine exposure rather than apoptosis. Therefore, JC-1 staining was used in this study to more specifically evaluate mitochondrial depolarization-associated platelet apoptosis and to avoid potential confounding from platelet activation in the CTL–platelet co-culture system. JC-1 aggregates (red fluorescence) indicated viable platelets, whereas JC-1 monomers (green fluorescence) indicated apoptotic platelets. Platelets were identified as APC-conjugated anti-CD61⁺ (eBioscience) events by FCM, and the percentage of JC-1 monomer-positive CD61⁺ platelets was defined as the platelet apoptosis rate. Spontaneous platelet apoptosis was determined in samples containing platelets cultured in medium alone. CTL-induced platelet apoptosis was calculated as the difference between apoptosis observed in coculture conditions and spontaneous apoptosis.

In parallel, platelet activation was assessed by staining with PE-conjugated anti-CD62P mAbs (eBioscience), and the geometric mean fluorescence intensity (MFI) of CD62P expression on platelets was quantified.

Flow cytometric analysis of CD8⁺ T-cell degranulation and CTL–platelet aggregate formation

To evaluate CD8⁺ T-cell degranulation and CTL–platelet aggregate formation, purified CD8⁺ T cells were first treated with DMSO or KD025 under the indicated culture conditions. After treatment, CD8⁺ T cells were washed to remove residual drug and then co-cultured with isolated platelets at the indicated ratio. For CD8⁺ T-cell degranulation analysis, APC-conjugated anti-CD107a mAbs was added during the co-culture period to capture surface mobilization of CD107a, a marker of cytotoxic granule exocytosis. After incubation, cells were harvested and stained with FITC-conjugated anti-CD3 mAbs and PE-conjugated anti- CD8 mAbs.

For CTL–platelet aggregate analysis, co-cultured samples were stained with PE-conjugated anti- CD8 mAbs and APC-conjugated anti-CD61 mAbs. CD8 was used to identify CTLs, whereas CD61 was used as a platelet-specific marker. CTL–platelet aggregates were defined as CD8⁺CD61⁺ double-positive events and were quantified as the proportion of CD8⁺CD61⁺ aggregates.

STAT3 inhibition (Stattic)

To evaluate the effect of STAT3 inhibition on CTL cytotoxic function, purified CD8⁺ T cells were pretreated with vehicle control, Stattic (MCE; 5 µM), KD025, or Stattic plus KD025. After drug treatment, CD8⁺ T cells were thoroughly washed to remove residual inhibitors and then co-cultured with platelets.

For CD107a degranulation analysis, APC-conjugated anti-CD107a mAbs was added during platelet–CTL co-culture. After incubation, cells were collected, washed, and stained with FITC-conjugated anti-CD3 mAbs and PE-conjugated anti-CD8 mAbs. CD107a expression was analyzed in the CD8⁺ T-cell population by FCM and presented as mean fluorescence intensity.

For intracellular granzyme B detection, cells were harvested after co-culture and stained with FITC-conjugated anti-CD3 mAbs and APC-conjugated anti-CD8 mAbs. After fixation and permeabilization, cells were incubated with PE-conjugated anti-granzyme B mAbs.

All FCM data were analyzed using FlowJo software (v10). A consistent gating strategy was applied across all experiments. First, lymphocytes were identified based on their characteristic Forward Scatter (FSC-A) and Side Scatter (SSC-A) properties. Second, doublets were excluded by gating on singlets using FSC-A versus FSC-H. Third, dead cells were excluded by gating on the negative population for a viability dye Zombie NIR™. From the live, singlet lymphocyte population, T cells were identified by gating on CD3⁺ cells. Finally, CTLs were identified as the CD8⁺ population within the CD3⁺ gate. All subsequent analyses of functional markers (e.g., p-STAT3, Granzyme B, CD107a) were performed on this final, stringently gated Live/Singlet/CD3⁺/CD8⁺ population.

In addition, we have further clarified the gating strategy used for platelet apoptosis and activation analyses. For platelet-related FCM experiments, isolated platelets were first identified according to their characteristic FSC/SSC distribution, and CD61 staining was then used to specifically define the platelet population. The CD61⁺ platelet gate was used as the parent population for all subsequent platelet analyses. Platelet apoptosis was assessed within the CD61⁺ platelet population by JC-1 staining, and the proportions of JC-1 monomers and aggregates were analyzed to evaluate mitochondrial membrane potential depolarization. Platelet activation was also analyzed within the same CD61⁺ platelet gate by measuring the expression of CD62P.

Reverse Transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from PBMCs derived from ITP patients and healthy controls, purified CD8⁺ T cells, and CD8⁺ T cells treated with DMSO or KD025 in vitro, using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. RNA quality and concentration were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Reverse transcription was performed using 1 µg of total RNA with a commercial cDNA synthesis kit (Vazyme) following the manufacturer’s protocol.

qPCR was performed using a SYBR Green PCR Master Mix on an ABI 7500 Real-Time PCR System (Applied Biosystems). Each 20 µL reaction mixture contained 10 µL of 2× SYBR Green Master Mix, 2 µL of cDNA template, 0.4 µM of each forward and reverse primer, and nuclease-free water. The PCR amplification program was as follows: 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 10 s and annealing/extension at 60 °C for 30 s. Melt curve analysis was performed at the end of the amplification procedure to confirm the specificity of the PCR products. Gene expression levels were normalized to β-Actin and calculated using the 2^-ΔΔCt method.

Primers were designed using NCBI Primer-BLAST to amplify fragments of 80–200 bp with similar melting temperatures and minimal secondary structures. Primers were preferentially designed across exon-exon junctions when possible. Primer specificity was verified by in silico BLAST analysis and experimentally confirmed by the presence of a single peak in the melt curve.

To assess assay performance, amplification efficiency for each primer pair was determined using serially diluted cDNA templates. The standard curve slope and correlation coefficient (R²) were used to calculate amplification efficiency with the formula E = (10^(−1/slope) − 1) × 100%. All primer efficiency data are presented in Supplementary Table 1.

Western blotting

Purified human peripheral blood CD8⁺ T cells and murine splenic CD8⁺ T cells were lysed in RIPA buffer (Beyotime) supplemented with protease and phosphatase inhibitor cocktails (Roche) on ice for 30 min. Cell lysates were centrifuged at 12,000 × g for 15 min at 4 °C, and the supernatants were collected. Protein concentrations were determined using a BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein (30 µg per lane) were separated by 10% SDS–PAGE and subsequently transferred onto PVDF membranes (Millipore).

After blocking with 5% non-fat milk in TBST for 1 h at room temperature, membranes were incubated overnight at 4 °C with the following primary antibodies: anti-ROCK2 antibody (CST), anti-STAT3 (CST) and p-STAT3 (Tyr705; CST) and anti-β-actin antibody (Proteintech) as an internal loading control. Membranes were then washed three times with TBST and incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL, Thermo Fisher Scientific) and detected using a ChemiDoc imaging system (Bio-Rad). Densitometric analysis was performed using ImageJ software (Imagej 1.54f). The relative abundance of ROCK2, STAT3 and p-STAT3 was determined by normalizing to β-actin as a loading control.

Seahorse metabolic flux analysis

CD8⁺ T cells were isolated from PBMCs of ITP patients and HCs. To compare the basal metabolic profiles between ITP patients and HCs, freshly isolated CD8⁺ T cells from both groups were directly seeded into Seahorse XF cell culture microplates and subjected to extracellular flux analysis without additional in vitro stimulation.

For the KD025 treatment experiments, CD8⁺ T cells isolated from ITP patients were cultured in complete RPMI-1640 medium and divided into the indicated treatment groups. Under unstimulated conditions, cells were treated with either DMSO or KD025 for 72 h. To evaluate whether ROCK2 inhibition directly regulated T-cell metabolism under preserved activation, another set of CD8⁺ T cells was stimulated with PMA and ionomycin in the presence of either DMSO or KD025 for 72 h. After treatment, cells were harvested, washed, and seeded into Seahorse XF cell culture microplates for subsequent glycolysis stress and mitochondrial stress assays.

For Seahorse assays, the sensor cartridge was hydrated overnight in Seahorse XF calibrant at 37 °C in a non-CO₂ incubator. On the day of the assay, Seahorse XF cell culture microplates were coated with Cell-Tak to facilitate adherence of suspension CD8⁺ T cells. CD8⁺ T cells were seeded into each well at the optimized density and gently centrifuged to attach the cells to the bottom of the plate. The culture medium was then replaced with Seahorse XF assay medium, and the cells were equilibrated at 37 °C in a non-CO₂ incubator before measurement.

Extracellular acidification rate (ECAR) was measured using the Seahorse glycolysis stress test. After baseline measurement, glucose (10 mM), oligomycin (1 µM), and 2-deoxy-D-glucose (50 mM) were sequentially injected to determine glycolysis, glycolytic capacity, and glycolytic reserve. Oxygen consumption rate (OCR) was measured using the Seahorse mitochondrial stress test. After baseline measurement, oligomycin (10 µM), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) (1.5 µM), and rotenone/antimycin A (1 µM) were sequentially injected to determine basal respiration, maximal respiration, ATP-linked respiration, and spare respiratory capacity. All experiments were performed as 3 independent biological replicates, and 5 technical replicates tested for each condition within each experiment. Data were normalized to cell number or protein content and analyzed using Seahorse Wave software (Agilent v.2.6.1).

Immunofluorescence staining

For immunofluorescence analysis, spleen tissue sections were prepared from splenectomy specimens obtained from ITP patients and healthy control donors undergoing splenectomy for traumatic injury, as well as from spleen tissues of active ITP mice and WT mice. Spleen tissues were fixed in 4% paraformaldehyde, embedded in OCT compound, and sectioned at 6–8 μm thickness. Tissue sections were permeabilized with 0.3% Triton X-100 for 15 min and blocked with 5% bovine serum albumin (BSA) for 1 h at room temperature. Sections were then incubated overnight at 4 °C with primary antibodies against CD8 (1:200, BioLegend) and ROCK2 (1:200, Abcam). After washing with PBS, sections were incubated with species-specific fluorescent secondary antibodies for 1 h at room temperature in the dark, and nuclei were counterstained with DAPI. Secondary antibody-only negative controls, in which primary antibodies were omitted, were included in each experiment to assess background and nonspecific fluorescence signals.

Fluorescence images were acquired using a confocal laser scanning microscope (Leica SP8) under identical acquisition settings across all comparative groups, including matched laser intensity, detector gain, exposure time, and offset parameters. For each sample, at least five non-overlapping fields were captured, and representative images were selected from fields without signal saturation or overexposure. Fluorescence images were acquired using a confocal laser scanning microscope under identical acquisition settings across all comparative groups, including matched laser intensity, detector gain, exposure time, and offset parameters. For each sample, at least five non-overlapping fields without signal saturation or overexposure were analyzed using ImageJ/Fiji. CD8⁺ cells were identified based on CD8 fluorescence together with DAPI nuclear staining, and ROCK2 positivity was determined using a consistent threshold defined according to secondary antibody-only controls. The percentage of ROCK2⁺CD8⁺ cells was calculated as the number of ROCK2⁺CD8⁺ double-positive cells divided by the total number of CD8⁺ cells × 100%. The mean value from multiple fields was used as one biological replicate for statistical analysis. Quantification was performed in a blinded manner by two independent investigators.

Enzyme-linked immunosorbent assay

Levels of granzyme A, granzyme B, and perforin in culture supernatants were determined using commercial sandwich ELISA kits (Human Granzyme A ELISA Kit, Human Granzyme B ELISA Kit, and Human Perforin ELISA Kit; Multi sciences). All assays were performed according to the manufacturer’s instructions. Briefly, standards and appropriately diluted samples were added in duplicate to 96-well plates pre-coated with specific capture antibodies and incubated at room temperature for the recommended time. After washing, horseradish peroxidase (HRP)-conjugated detection antibodies were added, followed by incubation and additional washes. Color development was initiated by adding tetramethylbenzidine (TMB) substrate solution and stopped with 1–2 N sulfuric acid. Absorbance was measured at 450 nm with a reference wavelength of 570–620 nm using a microplate reader. Analyte concentrations were calculated from standard curves generated using serial dilutions of recombinant standards. The low detection limit for granzyme A was 0.31 pg/ml, for granzyme B was 3.54 pg/ml, and for perforin was 19.5 pg/ml.

IFN-γ ELISPOT assay

To assess platelet antigen-induced CTL activation at the single-cell level, an IFN-γ ELISPOT assay was performed using purified CD8⁺ T cells from patients with CTL-mediated ITP. Briefly, CD8⁺ T cells were isolated from peripheral blood mononuclear cells using magnetic bead-based separation and cultured in the presence of DMSO or KD025. After treatment, CD8⁺ T cells were washed and co-cultured with isolated platelets in IFN-γ ELISPOT plates. Wells containing medium alone, platelets alone, or CD8⁺ T cells alone were included as background and negative controls. After incubation at 37 °C in a humidified 5% CO₂ incubator, the plates were washed and sequentially incubated with biotinylated anti-IFN-γ detection antibody, enzyme-conjugated streptavidin, and substrate solution. IFN-γ-secreting cells were visualized as individual spots, and the number of spot-forming units was quantified using an automated ELISPOT reader. The results were expressed as the number of IFN-γ spot-forming units per well.

Lentiviral transduction of primary human CD8⁺ T cells

Primary human CD8⁺ T cells were isolated from PBMCs of ITP patients and pre-activated with anti-human CD3/CD28 antibodies for 24 h in complete RPMI 1640 medium. For ROCK2 knockdown, activated CD8⁺ T cells were transduced with GFP-labeled lentiviral vectors carrying either ROCK2-specific shRNA or non-targeting control shRNA (shNC).To optimize ROCK2 knockdown efficiency, candidate ROCK2 shRNA sequences were first evaluated, and the sequence showing the highest knockdown efficiency was selected for subsequent experiments. Lentiviral particles were packaged from the corresponding shRNA plasmid by GeneChem and used according to the manufacturer’s recommended transduction procedure. Based on viral titer, the manufacturer’s recommended protocol, and preliminary optimization, shNC and shROCK2 lentiviruses were used at the same MOI of 10. To further improve lentiviral transduction efficiency, activated CD8⁺ T cells were subjected to spin infection by centrifugation for 1 h after lentivirus addition. After transduction, cells were cultured in complete RPMI 1640 medium for 48 h before downstream analyses.

Transduction efficiency was assessed by detecting GFP-positive CD8⁺ T cells using FCM. ROCK2 knockdown was confirmed by western blotting and RT-qPCR. The downstream effects of ROCK2 silencing were evaluated by detecting p-STAT3 and total STAT3 by western blotting and by measuring GZMB and PRF1 mRNA expression using RT-qPCR. All analyses were performed by comparing shROCK2-transduced cells with matched shNC-transduced control cells.

RNA sequencing and bioinformatic analysis

CD8⁺ T cells were isolated from ITP patients and cultured in vitro in the presence of either 0.1% DMSO or KD025. Three independent biological replicates were included in each group. Total RNA was extracted from CD8⁺ T cells using TRIzol reagent according to the manufacturer’s instructions. RNA concentration and purity were assessed using a NanoDrop spectrophotometer, and RNA integrity was evaluated before library construction.

RNA-seq libraries were prepared and sequenced on an Illumina sequencing platform. After quality control and removal of low-quality reads, clean reads were aligned to the human reference genome. Gene expression levels were normalized and used for subsequent differential expression analysis. Differentially expressed genes between the DMSO- and KD025-treated groups were identified using an adjusted P value < 0.05 and |log₂(fold change)| > 1 as the cutoff criteria. Heatmaps were generated using normalized gene expression values to visualize cytotoxicity-related and glycolysis-related genes. KEGG and Gene Ontology enrichment analyses were performed to identify biological pathways and functional categories enriched among differentially expressed genes. Gene set enrichment analysis was used to further evaluate pathway-level transcriptional changes between the two groups.

Upstream transcription factor (TF) enrichment analysis was performed on the list of differentially expressed genes (DEGs) using the ChEA3 web server (https://maayanlab.cloud/chea3/). TF enrichment based on the “Literature ChIP-seq” library, with results filtered by a False Discovery Rate (FDR) < 0.05.

Platelet aggregation assay

Platelet aggregation was assessed by light transmission aggregometry (LTA) using a Chrono-Log platelet aggregometer with matched aggregation cuvettes and magnetic stir bars. Platelet-rich plasma (PRP) was used for aggregation testing, whereas platelet-poor plasma (PPP) from the same sample was used only as the blank reference for calibration. Aggregation was induced with ADP (10 µM), collagen (2 µg/ml), arachidonic acid (AA, 0.5 mM), or thrombin (1 U/ml), and the maximal aggregation percentage was calculated from the aggregation tracing.

For the in vitro human platelet assay, PRP from HCs was preincubated with KD025 at 37 °C for 15 min before stimulation with the indicated agonists. For the murine ex vivo assay, ITP mice were treated with KD025 or DMSO for 7 consecutive days at 100 mg/kg, after which PRP was prepared and subjected to LTA. For the clinical validation cohort, PRP from cGVHD patients receiving Belumosudil/KD025 (200 mg once daily) and HCs was analyzed ex vivo using the same procedure. Aggregation responses were continuously recorded under stirring conditions, and representative tracings and maximal aggregation values were used for analysis.

Statistical analysis

Statistical analyses were performed using SPSS version 29.0 or GraphPad Prism version 10.1.2. Continuous variables were presented as mean ± standard deviation (SD) or median (range). Descriptive statistics were used to summarize demographic and baseline clinical characteristics of the enrolled patients. For comparisons among more than two groups, one-way analysis of variance (ANOVA) followed by the Student–Newman–Keuls test was used for normally distributed data, whereas the Kruskal–Wallis H test followed by the Nemenyi test was used for non-normally distributed data. Differences between two independent groups were analyzed using the independent-samples t-test or Mann–Whitney U test, as appropriate. The paired t-test was used to compare parameters in the same specimens with or without KD025 treatment. Each dot in the quantitative graphs represents one biological replicate, and the number of dots indicates the corresponding sample size; the n value for each experiment was also indicated in the figure legends. P values < 0.05 were considered statistically significant.

For RT-qPCR analysis, relative gene expression was calculated using the 2^−ΔΔCt method and presented as fold-change values with individual data points. Statistical analyses were performed using log-transformed fold-change values rather than untransformed 2^−ΔΔCt values. When normality could not be reliably assumed, especially for small-sample datasets, non-parametric tests were preferentially applied. The Mann–Whitney U test was used for two independent groups, the Wilcoxon matched-pairs signed-rank test for paired comparisons, and the Kruskal–Wallis H test followed by the Nemenyi post hoc test for multiple-group comparisons involving non-normally distributed data or small samples.

Results

ROCK2 expression was elevated in ITP patients and mice

Using an ELISA-based assay, we observed that ROCK kinase activity was significantly higher in PBMCs from patients with CTL-mediated ITP than from HCs, and was markedly reduced in ITP patients in remission (Fig. 1A). In purified CD8⁺ T cells, ITP patients showed higher ROCK2 mRNA and protein expression levels than HCs (Fig. 1B and C). Moreover, immunofluorescence staining of human spleen sections showed stronger ROCK2 signals within CD8⁺ T-cell-enriched areas in ITP patients than in HCs, and quantitative analysis confirmed a significantly higher proportion of ROCK2⁺CD8⁺ cells among total CD8⁺ cells in ITP spleens (Fig. 1D).

Fig. 1 — ROCK2 expression was elevated in ITP patients and in an active ITP mouse model. A ROCK kinase activity in PBMCs from healthy controls (HCs; n = 16), patients with CTL-mediated ITP (n = 8), patients with non-CTL-mediated ITP (n = 8), and ITP patients in remission (n = 6). B *ROCK2* mRNA in purified CD8⁺ T cells from HCs and ITP patients, as determined by RT-qPCR. Expression of each target gene was normalized to β-actin. Data are presented as the means ± SD from three independent experiments (n = 6). C Representative western blots and quantification of ROCK2 protein expression in purified CD8⁺ T cells from HCs and ITP patients (n = 5). D Representative immunofluorescence staining of ROCK2 (green) and CD8 (red) in spleen sections and secondary antibody-only controls from active ITP patients and HCs. Nuclei were counterstained with DAPI (blue). Quantification shows the percentage of ROCK2⁺CD8⁺ cells among total CD8⁺ cells (n = 6). Scale bars, 30 μm. E Schematic of the active ITP mouse model: irradiated *Rag1*^-/- mice were infused with 5 × 10⁴ splenocytes from *Cd61*^-/- mice immunized with WT C57BL/6 platelets. F *Rock2* mRNA expression in splenic CD8⁺ T cells from WT and ITP mice (n = 6). G Representative western blots and quantification of ROCK2 protein expression in splenic CD8⁺ T cells from WT and ITP mice (n = 5). H Representative immunofluorescence staining of spleen sections from WT and ITP mice, with quantification of ROCK2⁺CD8⁺ cells among CD8⁺ cells (n = 6). Scale bars, 30 μm. Data were presented as mean ± SEM or median (range). *P < 0.05, **P < 0.01, ***P < 0.001

To further evaluate the role of ROCK2 in the pathogenesis of ITP, we next assessed ROCK2 expression in spleen tissues from a murine model of active ITP. In line with human data, splenic CD8⁺ T cells from ITP mice exhibited elevated ROCK2 mRNA and protein levels compared with WT mice (Fig. 1F and G). Furthermore, spleen sections from active ITP mice exhibited increased ROCK2 expression in CD8⁺ T-cell regions compared with WT mice, with a significantly elevated percentage of ROCK2⁺CD8⁺ cells among total CD8⁺ cells (Fig. 1H). These observations suggested that ROCK2 expression was elevated within CD8⁺ T cells in both human and murine ITP.

KD025 suppressed CD8⁺ T-cell proliferation, degranulation, cytotoxicity, and CTL cell-platelet interactions in ITP patients

The optimal concentration of KD025, a selective ROCK2 inhibitor, was determined using a range of gradient doses. Early apoptotic PBMCs were identified as Annexin V⁺/PI⁻ cells by FCM. To determine the working concentration of KD025, PBMCs from ITP patients were cultured with graded doses, and early apoptosis (Annexin V⁺/PI⁻) was assessed by FCM. KD025 at 5 and 10 µM significantly increased PBMC apoptosis, whereas 0.5 and 1 µM had no pro-apoptotic effect (Supplemental Fig. 1A). Therefore, 1 µM KD025 was selected for subsequent in vitro experiments.

A representative gating strategy was used to sequentially identify live singlet CD3⁺CD8⁺ CTLs and CD61⁺ platelets, which served as the parent populations for subsequent analyses of platelet apoptosis, platelet activation, CTL degranulation, and CTL–platelet aggregate formation (Fig. 2A).

Fig. 2 — KD025 attenuated CTL cytotoxicity in ITP patients in vitro. A Representative flow cytometry gating strategy for identifying live singlet CD3⁺CD8⁺ CTLs and CD61⁺ platelets, which were used as the parent populations for subsequent analyses of platelet apoptosis, platelet activation, CTL degranulation, and CTL–platelet aggregate formation. B Representative JC-1 flow cytometry plots and quantitative analysis of platelet apoptosis in platelets cultured alone, platelets co-cultured with CTLs, and platelets co-cultured with KD025-treated CTLs. Platelet apoptosis was defined by mitochondrial membrane depolarization, reflected by increased JC-1 monomer signal (n = 7). C Representative histograms and quantification of platelet CD62P expression under the indicated co-culture conditions (n = 5). D Representative histograms and quantification of CD107a expression on CTLs alone, CTL+ Platelet and KD025-CTL + Platelet (n = 5). E CTL–platelet aggregate formation was analyzed by flow cytometry as CD8⁺CD61⁺ events and further visualized by confocal microscopy. DAPI indicates nuclei, CD8 indicates CTLs, and CD61 indicates platelets. Scale bars, 10 μm (n = 7). F Relative mRNA expression of cytotoxicity-associated genes, including *GZMA*, *GZMB*, *PRF1*, *TBX21*, *EOMES*, and *IFNG*, in DMSO- or KD025-treated CTLs. Expression of each target gene was normalized to β-actin (n = 10). G Concentrations of granzyme A, granzyme B, and perforin in culture supernatants after DMSO or KD025 treatment (n = 8). H Representative IFN-γ ELISPOT images and quantification of IFN-γ spot-forming units under the indicated stimulation conditions (n = 8). CTL–platelet co-culture enhanced platelet apoptosis, platelet activation, CTL degranulation, CTL–platelet adhesion, cytotoxic gene expression, granule release, and IFN-γ secretion, whereas KD025 markedly attenuated these CTL-mediated responses. Data were presented as mean ± SEM or median (range). *P < 0.05, **P < 0.01, ***P < 0.001

To explore the functional effects of KD025 on CTLs, 5(6)-carboxyfluorescein diacetate N-succinimidyl ester (CFSE)-labeled CD8⁺ T cells stimulated with anti-CD3/CD28 antibodies were cultured in the presence or absence of KD025. As shown in Supplemental Fig. 1B, the division index in the KD025-treated group was significantly lower than that in the DMSO control group, indicating that KD025 suppresses CTL proliferation.

As shown in Supplementary Fig. 3, baseline platelet apoptosis remained low in co-cultures with CD8⁺ T cells from healthy controls and non-CTL-ITP patients, but was markedly increased in co-cultures with CD8⁺ T cells from CTL-ITP patients. These data support our patient stratification strategy and provide a baseline for evaluating the inhibitory effect of KD025 on CTL-mediated platelet injury.

Next, CTL cytotoxicity was assessed by analyzing platelet apoptosis after 72 h of co-culture, as determined by mitochondrial membrane depolarization using JC-1 staining (loss of JC-1 aggregate signal and increase in JC-1 monomer signal). Platelets cocultured with CTLs exhibited higher levels of apoptosis than platelets cultured alone. In contrast, the addition of KD025 to CTL-platelet cocultures significantly reduced platelet apoptosis compared with coculture with CTLs alone (Fig. 2B). Additionally, platelet CD62P expression, as one of the marker for platelet activation, was markedly increased after co-culture with CTLs, and this increase was significantly attenuated when KD025 was present during the co-culture (Fig. 2C), suggesting that KD025 could reduce CTL-mediated platelet activation.

CD107a is normally expressed on granule membranes of CTLs and becomes exposed on the cell surface following degranulation, thereby serving as a standard marker of CTL degranulation [8]. Membranous CD107a expression on CD8⁺ T cells was significantly elevated after co-culture with platelets compared with CTLs cultured alone, and was significantly reduced when KD025 was present during the co-culture (Fig. 2D). By contrast, ex vivo assays on CD8⁺ T cells from HCs showed minimal baseline platelet apoptosis, and KD025 neither increased apoptosis nor produced a significant reduction in CD107a degranulation (Supplementary Fig. 4A and B).

To assess whether ROCK2 inhibition affects the ability of CTLs to physically engage with their targets, CD8⁺ T cells from ITP patients were co-cultured with platelets from HCs for 4 h. KD025 treatment substantially reduced CTL-platelet aggregate formation compared to DMSO controls, as quantified by CD8⁺CD61⁺ aggregates (Fig. 2E). Consistently, confocal microscopy further confirmed reduced CTL-platelet aggregate formation in KD025-treated co-culture compared with the DMSO controls (Fig. 2E). These observations suggested that ROCK2 activity was involved in the stable adhesion between CTLs and platelets, a critical prerequisite for targeted cytotoxicity.

We next assessed cytotoxic granule secretion by CTLs. At the transcriptional level, KD025-treated CTLs exhibited significantly lower mRNA expression of GZMA, GZMB, PRF1, TBX21, eomesodermin (EOMES), and interferon-γ (IFNG; Fig. 2F). Besides, levels of granzyme B (GZMB) and perforin (PRF1) in culture supernatants were significantly decreased after KD025 treatment, whereas granzyme A (GZMA) remained unchanged (Fig. 2G). In parallel, IFN-γ ELISpot assays showed that platelet stimulation induced robust antigen-specific IFN-γ responses in CD8⁺ T cells from CTL-mediated ITP patients, and KD025 markedly reduced the number of IFN-γ-secreting spots (Fig. 2H). Thus, KD025 attenuated the cytotoxic machinery of CTLs at both protein and transcriptional levels and suppressed proliferation, degranulation, and CTL-platelet interactions in ITP.

RNA sequencing revealed substantial transcriptional reprogramming between DMSO- and KD025-treated CD8⁺ T cells. The volcano plot demonstrated that 397 genes were upregulated and 624 were downregulated after KD025 treatment, while 15,322 genes remained unchanged (Fig. 3A). Multiple cytotoxic effector molecules, including GZMA, GZMB, GZMK, GZMH, FASLG, and IFNG, were consistently reduced by KD025 (Fig. 3B). RNA-sequencing further showed that multiple glycolysis-related genes, including HK2, PFKM, PKM, ENO1, LDHA, and PFKFB3, were downregulated after KD025 treatment (Fig. 3C). KEGG pathway and GO enrichment analyses of differentially expressed genes (DEGs) indicated that KD025 predominantly downregulated cell-cycle- and proliferation-related programs, and modulated cytokine-driven signaling, including JAK-STAT pathways, while GSEA revealed decreased enrichment of MYC targets and mTORC1 signaling (Fig. 3D–F). An unbiased transcription-factor enrichment analysis of the RNA-seq data identified STAT3 among the top transcription factors predicted to be inhibited by KD025 in CTLs (Fig. 3G), providing an independent, data-driven rationale for focusing on the ROCK2–STAT3 axis. Together, these results indicated that KD025 induced coordinated transcriptional changes that reshaped CTL effector function.

Fig. 3 — Transcriptome sequencing of CD8⁺ T cells from ITP patients cultured in vitro with DMSO or KD025. A Volcano plot showing DEGs between DMSO- and KD025-treated CD8⁺ T cells from 3 ITP patients. KD025 treatment resulted in 397 upregulated genes and 624 downregulated genes, whereas 15,322 genes remained unchanged. B Heatmap of selected cytotoxicity-related genes, demonstrating reduced expression of *GZMA*, *GZMB*, *GZMK*, *GZMH*, *FASLG* and *IFNG* after KD025 treatment. C RNA-sequencing analysis of glycolysis-related genes in CD8⁺ T cells from ITP patients treated with KD025 versus DMSO, showing reduced expression of *HK2*, *PFKM*, *PKM*, *ENO1*, *LDHA*, and *PFKFB3* after KD025 treatment. D KEGG pathway enrichment analysis of DEGs, highlighting significant enrichment of pathways involved in DNA replication and repair and cytokine-mediated signaling, including the JAK-STAT pathway. E GO enrichment analysis indicated that attenuated cell-cycle associated processes, including DNA replication, chromosomal segregation, and mitotic phase transition. F GSEA enrichment analysis showed decreased enrichment of MYC targets and mTORC1 signaling, alongside increased activation of immune- and stress-response pathways such as allograft rejection, IL2-STAT5 signaling, TNFα-NFκB signaling, and the unfolded protein response. G Upstream transcription factor enrichment analysis using ChEA3 identified several T-cell regulatory transcription factors, including STAT3, among the top predicted regulators affected by KD025. Red bars indicate key T-cell master regulators, and blue bars indicate other significant or top-ranked transcription factors. Together, these transcriptomic data indicate that KD025 reshapes CTL effector programs by suppressing cytotoxicity-, glycolysis-, proliferation-, and STAT3-associated transcriptional signatures. Each column in the heatmaps represents one biological replicate

KD025 attenuated CD61-driven activation and cytotoxicity of CD8⁺ T cells in Cd61^−/− mice

To investigate the in vivo impact of KD025 on platelet antigen-specific CD8⁺ T-cell responses, Cd61^−/− mice were intravenous transfused weekly with platelets from WT mice and concurrently administered KD025 (100 mg/kg) or a 2% DMSO daily for 4 consecutive weeks. At the end of treatment, splenocytes were harvested, CD8⁺ T cells were isolated, and then cocultured with WT platelets for 4 h (Fig. 4A). Consistent with the in vitro findings, CD8⁺ T cells from KD025-treated Cd61^−/− mice induced significantly less platelet apoptosis (Fig. 4B) and activation (Fig. 4C) than those from DMSO-treated controls.

Functionally, KD025 also restrained CD8⁺ T-cell degranulation, as evidenced by reduced surface CD107a expression (Fig. 4D). In parallel, KD025-treated mice displayed diminished CTL-platelet aggregates, reflected by a lower proportion of CD8⁺CD61⁺ population after coculture (Fig. 4E). These results suggested that KD025 mitigated CD61-induced activation and effector responses of CD8⁺ T cells in Cd61^−/− mice.

Metabolic reprogramming in CD8⁺ T cells from ITP patients and its regulation by ROCK2 inhibition

Metabolic reprogramming is a hallmark of CD8⁺ T-cell activation and proliferation, characterized by enhanced glycolysis, amino acid turnover, and protein synthesis, concomitant with decreased aerobic glucose oxidation [7]. To evaluate metabolic activity in CD8⁺ T cells, we performed a Seahorse glycolysis and mitochondrial stress assay, measuring the extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) as indicators of glycolytic flux and mitochondrial respiration, respectively. Following glucose injection, CTLs from ITP patients displayed a markedly greater increase in ECAR compared to HCs, indicating hyperactivated glycolytic metabolism. Quantitative analysis confirmed that basal glycolysis, glycolytic capacity, and glycolytic reserve were all significantly elevated in ITP-derived CTLs (Fig. 5A).

Fig. 5 — Metabolic reprogramming in CD8⁺ T cells from ITP patients and its regulation by ROCK2 inhibition. A Representative glycolytic stress assay of CD8⁺ T cells from ITP patients and HCs, showing ECAR changes after sequential injection of glucose (10 mM), oligomycin (1 µM), and 2-deoxyglucose (2-DG, 50 mM). Comparison of glycolysis, glycolytic capacity, and glycolytic reserve in CD8⁺ T cells from ITP patients and HCs (n = 5). B Representative mitochondrial stress test (OCR) of CD8⁺ T cells from ITP patients and HCs after oligomycin (10 µM), FCCP (1.5 µM), and rotenone/antimycin A (1 µM). Quantification of mitochondrial respiration parameters in CD8⁺ T cells from ITP patients and HCs, including basal respiration, maximal respiration, ATP-linked respiration, and spare respiratory capacity (n = 5). C RT-qPCR analysis of glycolysis-related genes in CD8⁺ T cells, showing increased mRNA expression of *GLUT1*, *HK2*, *PKM2*, and *LDHA* in ITP patients compared to HCs (n = 5). D ECAR analysis of CD8⁺ T cells from ITP patients treated with DMSO or KD025, showing that KD025 reduced glycolysis, glycolytic capacity, and glycolytic reserve (n = 5). E OCR analysis showing that KD025 increased basal respiration, maximal respiration, ATP production, and spare respiratory capacity (n = 5). F RT-qPCR analysis showing decreased expression of *GLUT1*, *HK2*, *PKM2*, and *LDHA* in CD8⁺ T cells after KD025 treatment (n = 5). G ECAR analysis of unstimulated or PMA/ionomycin-stimulated CD8⁺ T cells treated with DMSO or KD025. Under matched high-activation conditions, KD025 still reduced glycolytic flux, including glycolysis, glycolytic capacity, and glycolytic reserve (n = 5). H OCR analysis of unstimulated or PMA/ionomycin-stimulated CD8⁺ T cells treated with DMSO or KD025. KD025 enhanced mitochondrial respiration, including basal respiration, maximal respiration, ATP production, and spare respiratory capacity (n = 5). Data were presented as mean ± SEM or median (range) from 3 independent biological replicates, and that within each experiment, each condition was tested with 5 technical replicates (wells). *P < 0.05; **P < 0.01; ***P < 0.001

In contrast, basal OCR prior to oligomycin injection was significantly lower in CTLs from the ITP patients than from HCs. Consistently, basal respiration, maximal respiration, ATP-linked respiration and spare respiratory capacity were all markedly reduced in ITP CTLs (Fig. 5B), indicating impaired mitochondrial oxidative phosphorylation (OXPHOS). At the transcriptional level, the expression of key glycolytic genes—including glucose transporter 1 (GLUT1), hexokinase 2 (HK2), pyruvate kinase M2 (PKM2), and lactate dehydrogenase (LDHA)—was significantly upregulated in CD8⁺ T cells from ITP patients compared with HCs (Fig. 5C), further supporting a glycolytic metabolic bias.

To delineate the underlying mechanisms by which ROCK2 inhibition regulates CD8⁺ T cells, we next evaluated the bioenergetic status of cultured CD8⁺ T cells from ITP patients. KD025 treatment significantly decreased ECAR, including reductions in basal glycolysis, glycolytic capacity, and glycolytic reserve (Fig. 5D). By contrast, OCR was markedly increased, accompanied by enhanced basal respiration, maximal respiration, ATP-linked respiration and spare respiratory capacity (Fig. 5E). Consistent with these functional metabolic changes, RT-qPCR analysis revealed that KD025 treatment downregulated glycolysis-related genes, including GLUT1, HK2, PKM2, and LDHA, in CD8⁺ T cells from ITP patients (Fig. 5F).

To dissociate metabolic reprogramming from reduced upstream activation, we activated CD8⁺ T cells with phorbol 12-myristate 13-acetate and ionomycin (PMA/ionomycin) to bypass proximal TCR signaling and maintain comparable activation between conditions. Under these matched high-activation conditions, KD025-treated CD8⁺ T cells still exhibited a significant reduction in glycolytic flux and an increase in OXPHOS relative to vehicle controls. Specifically, PMA/ionomycin-stimulated KD025 cells showed lower ECAR and higher OCR (Fig. 5G and H). These results indicated that the metabolic effects of KD025 were not solely secondary to reduced T-cell activation and supported a more direct role for ROCK2 in regulating CTL metabolic programming.

ROCK2 inhibition suppressed STAT3 phosphorylation to attenuate CTL-mediated platelet destruction

Having established that ROCK2 inhibition profoundly suppresses CTL effector functions, we next sought to identify the upstream signaling mechanism. Pathway enrichment analysis of our RNA sequencing data revealed the JAK-STAT signaling pathway as a top hit. Given the pivotal role of STAT3 in controlling T-cell cytotoxicity, proliferation, and metabolism [24–26], we specifically hypothesized that STAT3 is the critical signaling node downstream of ROCK2 that orchestrates CTL pathogenicity. To test this, we evaluated whether ROCK2 inhibition could restrain STAT3 activation in vitro. PBMCs were treated with KD025, IL-6, or DMSO, and p-STAT3 expression in CD8⁺ T cells were quantified by FCM. IL-6 markedly enhanced p-STAT3 expression in CD8⁺ T cells, whereas KD025 significantly reduced p-STAT3 levels. Importantly, IL-6 partially restored p-STAT3 expression in KD025-treated CD8⁺ T cells, indicating that ROCK2 inhibition restrains STAT3 activation but can be counteracted by IL-6 stimulation (Fig. 6A). In contrast to patient-derived CD8⁺ T cells, CD8⁺ T cells from HCs showed minimal baseline p-STAT3, and KD025 produced only minimal changes in p-STAT3 levels (Supplementary Fig. 4C). In ITP mice, KD025 administration reduced p-STAT3 expression in splenic CD8⁺ T cells, while IL-6 treatment increased STAT3 phosphorylation and partially reversed the inhibitory effect of KD025 (Fig. 6B). Together, these data demonstrated that ROCK2 activity was required for robust STAT3 phosphorylation in CTLs both in vitro and in vivo.

Fig. 6 — ROCK2 inhibition suppressed STAT3 phosphorylation to attenuate CTL-mediated platelet destruction. A Representative flow cytometry histograms and quantification of p-STAT3 expression in CD8⁺ T cells from ITP patients treated with DMSO, IL-6, KD025, or KD025 plus IL-6 (n = 5). B Representative histograms and quantification of p-STAT3 expression in splenic CD8⁺ T cells from ITP mice after the indicated treatments (n = 5). C Representative CFSE histograms and division index of CD8⁺ T cells from ITP patients treated with DMSO, IL-6, KD025, or KD025 plus IL-6 (n = 8). D Representative JC-1 plots and quantification of platelet apoptosis after co-culture with CTLs under the indicated conditions. IL-6 enhanced CTL-mediated platelet apoptosis, whereas KD025 reduced platelet apoptosis; this protective effect was partially reversed by IL-6 and abrogated by anti-IL-6 neutralization (n = 5). E MFI of CD62P on platelets after co-culture with CD8⁺ T cells pretreated with DMSO, IL-6, KD025, or KD025 plus IL-6 in vitro. F MFI of CD107a on CD8⁺ T cells pretreated with DMSO, IL-6, KD025, or KD025 plus IL-6. (n = 5). G Levels of CD8⁺ T cell-platelet aggregates in the DMSO, IL-6, KD025, and KD025 plus IL-6 groups (n = 5). H Representative histograms and quantification showing that Stattic-mediated STAT3 inhibition reduced CD107a expression on CTLs (n = 5). I Flow cytometry plots and quantification showing that Stattic reduced the proportion of GZMB⁺CD8⁺ T cells, indicating impaired cytotoxic effector molecule expression (n = 5). J GFP expression showing comparable lentiviral transduction efficiency between shNC- and shROCK2-transduced CD8⁺ T cells (n = 5). K Western blot analysis showing that ROCK2 knockdown reduced ROCK2 protein expression and STAT3 phosphorylation (n = 5). L RT-qPCR analysis showing decreased *ROCK2*, *GZMB*, and *PRF1* expression after ROCK2 knockdown (n = 5). These data indicate that ROCK2 promotes CTL effector function through STAT3 phosphorylation. Data were presented as mean ± SEM or median (range). *P < 0.05; **P < 0.01; ***P < 0.001

To determine whether STAT3 activation is sufficient to restore CTL effector function in the presence of ROCK2 inhibition, we treated KD025-exposed CD8⁺ T cells with IL-6. Indeed, co-treatment with IL-6 not only enhanced CTL proliferation (Fig. 6C), but also restored the overall cytotoxic capacity of CTLs suppressed by KD025. To determine if this restoration translated to the ultimate pathological outcome, we assessed direct CTL-mediated platelet damage. In these assays, the protective effect of KD025 on platelet apoptosis was significantly reversed by the addition of IL-6 (Fig. 6D). This hierarchical response was mirrored across a range of parallel functional readouts. Specifically, platelet activation (CD62P), CTL degranulation (CD107a), and CTL-platelet aggregate formation were all maximal in the IL-6 group, minimal in the KD025 group, and displayed an intermediate phenotype with the combination treatment. (Fig. 6E - G). We also performed neutralization experiments using a monoclonal IL-6 neutralizing antibody (anti-IL-6 mAb; 10 µg/mL), which completely abrogated IL-6-induced enhancement in platelet apoptosis, returning the cytotoxicity levels to baseline (Fig. 6D).

To further validate the involvement of STAT3 signaling in CTL effector responses, patient-derived CD8⁺ T cells were incubated with the selective STAT3 inhibitor Stattic. Specifically, Stattic significantly decreased CD107a surface exposure and reduced Granzyme B release compared with vehicle treated cells (Fig. 6H and I). These results confirmed that STAT3 activity was necessary for the spectrum of ROCK2-dependent CTL effector functions.

To provide genetic validation of ROCK2’s role in CTL programming, we performed transient ROCK2 knockdown in primary human CD8⁺ T cells using a high-efficiency lentiviral shRNA approach. GFP-based flow cytometry showed comparable transduction efficiency between shNC- and shROCK2-transduced cells (Fig. 6J). Lentiviral transduction achieved robust reduction of ROCK2 protein levels compared with a non-targeting control shRNA (Fig. 6K). In consistence with the pharmacologic inhibition, ROCK2 knockdown in CD8⁺ T cells led to a significant decrease in STAT3 phosphorylation and in the expression of cytotoxic effector transcripts including GZMB and PRF1 (Fig. 6L). These genetic data corroborated the pharmacologic findings and supported a cell-intrinsic requirement for ROCK2 in maintaining STAT3 activity and the cytotoxic transcriptional program in pathogenic CTLs.

IL-6/STAT3 signaling counteracted ROCK2 inhibition-mediated metabolic reprogramming in CD8⁺ T cells

Given that ROCK2 is a critical upstream regulator of STAT3 activation, we next examined whether IL-6-STAT3 signaling could interfere with KD025-induced metabolic reprogramming in CD8⁺ T cells from ITP patients. In the glycolysis stress test (Fig. 7A), KD025 treatment markedly reduced ECAR over time, with glycolysis, glycolytic capacity, and glycolytic reserve all significantly decreased compared with DMSO (Fig. 7B), indicating a broad suppression of glycolytic function. IL-6 co-stimulation largely restored ECAR in KD025-treated cells, significantly increasing glycolysis, glycolytic capacity, and glycolytic reserve relative to KD025 alone, thereby re-establishing a glycolytic phenotype.

Conversely, in the mitochondrial stress test (Fig. 7C), KD025 significantly enhanced OCR, with increased basal respiration, maximal respiration, ATP production and spare respiratory capacity compared with DMSO (Fig. 7D), consistent with a shift toward OXPHOS. IL-6 partially reversed these effects: all 4 parameters were significantly reduced in the IL-6 plus KD025 group versus KD025 alone, trending back toward the IL-6-driven glycolytic profile. These results indicated that ROCK2 inhibition ameliorated abnormal CTL immunometabolism in ITP patients through STAT3 suppression.

KD025 ameliorated thrombocytopenia in active ITP mice

To determine whether the in vitro findings of KD025 could be recapitulated in vivo, an active murine model of ITP was established (Fig. 8A). Following irradiation and adoptive transfer of anti-CD61-immunized splenocytes into recombination-activating gene 1 (Rag1^−/−) mice, platelet counts gradually declined, reaching a nadir by day 14. KD025 treatment significantly ameliorated thrombocytopenia, with platelet counts remaining markedly higher than control on day 21 and day 28. In contrast, IL-6 administration induced the most profound platelet suppression on day 28. Notably, co-administration of IL-6 with KD025 abolished the therapeutic benefit of KD025, as platelet counts in the IL-6 plus KD025 group failed to increase and were markedly lower than those in the KD025 group, indicating that IL-6 abolished the therapeutic benefit of KD025 (Fig. 8B).

To further verify whether the platelet-protective effect of KD025 was maintained at the effector-cell level, splenic CD8⁺ T cells were purified from each treatment group and co-cultured with WT platelets for 4 h. CD8⁺ T cells from KD025-treated group induced significantly less platelet apoptosis than those from DMSO control group, confirming an attenuation of platelet-destructive activity. In contrast, IL-6 treatment significantly increased platelet apoptosis, and IL-6 plus KD025 conditioning completely abrogated the protective effect of KD025, yielding apoptosis levels comparable to those observed in the DMSO control group (Fig. 8C and Supplemental Fig. 2A). Platelet CD62P expression showed a consistent pattern, with KD025 markedly reducing platelet activation compared with control, whereas IL-6 co-treatment significantly weakened this inhibitory effect of KD025 on platelet activation (Fig. 8C and Supplemental Fig. 2B).

Consistent with these findings, KD025 dampened CD8⁺ T-cell cytotoxic function in vivo. Compared to DMSO controls, KD025 significantly reduced the MFI of CD107a on CTLs (Fig. 8C and Supplemental Fig. 2C) and decreased the frequency of CTL-platelet aggregates (Fig. 8C and Supplemental Fig. 2D). Importantly, concurrent IL-6 administration fully reversed these inhibitory effects, restoring CD107a expression and CTL-platelet conjugate formation to levels comparable to those in controls (Fig. 8C).

To determine whether the therapeutic benefit of KD025 was mediated through intrinsic reprogramming of CD8⁺ T cells, CD8⁺ T cells were purified from anti-CD61-sensitized splenocytes and pretreated in vitro with KD025 or 0.1% DMSO for 72 h. These pretreated CD8⁺ T cells were then recombined with CD8⁺ T-cell-depleted splenocytes and adoptively transferred into irradiated Rag1^−/− mice (Fig. 8D). This design preserved the antigen-presenting and helper cell compartments required to reconstitute the active ITP phenotype while ensuring that the only variable between groups was KD025 exposure of the CD8 compartment. Mice receiving KD025-treated CD8⁺ T cells exhibited significantly higher platelet counts on day 21 (Fig. 8E) and markedly reduced platelet apoptosis (Fig. 8F) compared with mice receiving DMSO-treated CD8⁺ T cells. These results suggested that KD025-treated CD8⁺ T cells exhibited diminished platelet-destructive activity in vivo.

Because ROCK2 has been implicated in platelet activation, we directly assessed whether KD025 could impair platelet function. At the KD025 concentrations that suppressed CTL effector functions, the ITP mice showed only minimal changes in ex vivo platelet aggregation induced by adenosine diphosphate (ADP), collagen, arachidonic acid (AA), or thrombin (Fig. 8G). Complementing these in vivo data, washed human platelets from HCs were pre-incubated with KD025 showed only minimal reductions in maximum aggregation versus control (Fig. 8H). We measured whole-blood aggregation in patients with cGVHD receiving KD025 (200 mg once daily), and observed negligible changes in aggregation responses to ADP, collagen, AA, and thrombin (Fig. 8I).

Discussion

ITP has long been regarded as an autoantibody-mediated disorder, while accumulating evidence supports an important contribution of CTL-mediated platelet destruction [27–29]. Effector CD8⁺ T cells induce platelet apoptosis via exocytosis of cytotoxic granules, forming stable conjugates with platelets and delivering polarized granule secretion that triggers caspase activation and programmed platelet death [30, 31]. In parallel, T-cell receptor (TCR) engagement triggers a metabolic switch from mitochondrial OXPHOS toward aerobic glycolysis to support clonal expansion and acquisition of effector function [32]. Our study extended these concepts to ITP by demonstrating that CD8⁺ T cells from patients underwent pathological immunometabolic reprogramming, characterized by enhanced glycolysis, reduced OXPHOS, and upregulation of glycolysis-related genes, consistent with a glycolysis-addicted, bioenergetically inflexible state that sustained pathogenic CTL activation. In this context, we showed that KD025, a selective ROCK2 inhibitor, restored metabolic balance by suppressing glycolytic flux and enhancing OXPHOS, thereby attenuating CTL-mediated platelet destruction.

The therapeutic efficacy of KD025 in ITP appeared to be multifaceted, integrating effects on both the physical interaction and the intrinsic cytotoxic programming of CTLs. The observed reduction in CTL-platelet conjugate formation was consistent with the canonical function of ROCK2 as a master regulator of the actin cytoskeleton, which was essential for stable immune synapse assembly and cell adhesion [33, 34]. While this modulation of cell mechanics likely represented an important layer of regulation, our data revealed a more profound impact on the core pathogenic identity of CTLs. We found that KD025 could dismantle the STAT3-glycosis signaling axis, a central metabolic and transcriptional engine for T-cell effector function. This aligned with the emerging paradigm of ROCK2 acting as a non-canonical signal transducer that directly influenced transcription factor activity in the nucleus [16, 35]. Therefore, we proposed that KD025’s potent suppressive effect stemmed from a powerful synergy: it not only disrupted the physical engagement of CTLs with platelets but, more critically, it de-energized and transcriptionally rewired the CTLs, stripping them of their pathogenic capabilities.

STAT3 phosphorylation is a central regulator of CD8⁺ T-cell biology [25, 36]. It promotes survival and clonal expansion, upregulates perforin and granzyme B, and augments production of inflammatory cytokines such as IFN-γ and TNF-α [26, 37]. Activating STAT3 mutations in CD8⁺ T cells correlate with heightened cytotoxicity in rheumatoid arthritis [38], and STAT3 also contributes to the shift from OXPHOS to glycolysis during CD8⁺ T-cell activation by enhancing glycolytic enzymes and glucose transporters [38, 39]. ROCK2 has recently emerged as an upstream node that couples cellular metabolism, inflammation, and cytoskeletal dynamics [40, 41]. Meanwhile, ROCK2 signaling can drive glycolysis-dependent pathogenic programs in multiple tissues and disease settings [42–44], and pharmacologic ROCK2 inhibition with KD025 has shown therapeutic efficacy in immune-mediated and fibrotic conditions [35, 45, 46]. These observations led us to hypothesize that aberrant ROCK2 activity might sustain a STAT3-dependent, glycolysis-biased program in CTLs from ITP patients.

In line with this, we observed increased ROCK kinase activity and elevated ROCK2 mRNA expression in PBMCs from ITP patients, as well as in spleens from active ITP mice. Previous studies have reported that ROCK2 could phosphorylate and activate downstream transcription factor STAT3, thereby promoting pathogenic Th17 polarization and impairing regulatory T-cell function in autoimmune settings [18, 43]. Consistent with a functional association between ROCK2 and STAT3 signaling in CTLs, we found that ROCK2 inhibition with KD025 significantly reduced STAT3 phosphorylation in vitro and in vivo. This suppression of STAT3 signaling coincided with normalization of CTL metabolism and attenuation of their cytotoxic program, suggesting that aberrant ROCK2-STAT3 activity sustains both the effector and metabolic phenotypes of pathogenic CTLs in ITP.

Functionally, ROCK2 inhibition with KD025 curtailed multiple steps in the CTL-mediated platelet destruction cascade. In human ITP CTLs, KD025 reduced proliferation, degranulation (CD107a exposure), and secretion of granzyme B and perforin, and it diminished CTL-platelet conjugate formation. As a result, platelets co-cultured with KD025-treated CTLs exhibited lower apoptosis and decreased CD62P expression, indicating reduced activation and damage. Similar inhibitory effects were observed in Cd61^−/− mice immunized with WT platelets, where KD025 treatment reduced platelet apoptosis, CD62P upregulation, CTL degranulation, and CTL-platelet aggregates. These findings supported the notion that ROCK2 inhibition attenuated antigen-specific CTL responses against platelet antigens.

A key mechanistic insight from our study was that KD025 corrected abnormal CTL immunometabolism in ITP via STAT3 suppression. In CD8⁺ T cells from cytotoxic ITP patients, KD025 inhibited glycolysis and enhanced mitochondrial respiration, accompanied by downregulation of glycolytic genes. Conversely, IL-6, a canonical STAT3 activator, restored glycolytic activity and reversed KD025-mediated increases in OXPHOS. Functionally, IL-6 also rescued CTL proliferation, degranulation, cytotoxicity, and CTL-platelet aggregate formation that had been suppressed by KD025. While other cytokines or pathways might cooperate with IL-6, the combined gain- and loss-of-function results supported a central role for ROCK2-STAT3 signaling: this axis integrated metabolic and transcriptional programs to shift CTLs between a glycolysis-dominated, highly cytotoxic state and an OXPHOS-supported, less cytotoxic phenotype.

Our in vivo data provided proof-of-concept that targeting this axis could ameliorate disease. In an active ITP model that recapitulates both humoral and cellular anti-platelet immunity, KD025 treatment significantly increased platelet counts and reduced CTL-mediated platelet destruction. Phenotypically, CD8⁺ T cells from KD025-treated mice displayed less degranulation and fewer interactions with platelets. These findings indicated that ROCK2 inhibition could attenuate ongoing platelet damage in a complex autoimmune milieu. Importantly, at the therapeutic/immunomodulatory doses used here had minimal effects on platelet aggregation. However, because transferred splenocytes contained heterogeneous lymphocyte populations, including CD4⁺ T cells and B cells, we could not fully exclude KD025 effects on other immune subsets that contributed to disease modulation.

To more interrogate CD8⁺ T cells, we performed adoptive transfer experiments using ex vivo KD025-pretreated CTLs recombined with CD8⁺ T cell-depleted splenocytes. Mice receiving KD025-conditioned CD8⁺ T cells exhibited higher platelet counts and lower platelet apoptosis on day 21 than mice receiving DMSO-treated CD8⁺ T cells, although this benefit diminished by day 28 as thrombocytopenia naturally improved in both groups. These results indicated that intrinsic reprogramming of CD8⁺ T cells by KD025 was sufficient to transiently mitigate platelet destruction even in the presence of humoral autoimmunity.

Our study had several limitations. The co-culture assays provided a functional readout consistent with disease-relevant, antigen-driven CTL killing, since patient-derived CD8⁺ T cells were primed in vivo against platelet antigens shared by healthy donor platelets; however, definitive assignment to a single peptide-MHC I epitope was not performed. Notably, classifying patients solely by platelet apoptosis might overlook mixed humoral/cellular mechanisms and other CTL features (e.g., clonal expansion, CD107a), so future studies should adopt multi-parameter “CTL activity” scores. Besides, the number of patients with cytotoxicity-dominant ITP included in the detailed metabolic profiling was relatively modest, and larger cohorts are needed to validate whether ROCK2-associated STAT3 signaling-related metabolic features may serve as potential biomarkers. We focused mainly on CD8⁺ T cells in the study. However, the effects of ROCK2-STAT3 signaling on CD4⁺ T cells, B cells, and myeloid cells in ITP were also worth further evaluation. Potential off-target effects of KD025 were not systematically assessed, and murine ITP models only partially recapitulated the chronic and heterogeneous nature of human disease.

In conclusion, our study supports the involvement of ROCK2-STAT3 signaling in the immunometabolic regulation of ITP. ROCK2 inhibition with KD025 downregulated STAT3 phosphorylation, shifted CD8⁺ T-cell metabolism from pathological glycolysis toward OXPHOS, and attenuated CTL-mediated platelet destruction in vitro and in vivo. These findings supported ROCK2 inhibition as a strategy to restore immune homeostasis and reduce platelet destruction, particularly in CTL-driven disease.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1^{(8.6MB, docx)}

Author contributions

Contribution: J.S., A.X., Q.L., and B.D. performed the experiments, M.L., X.S., and F.W. analyzed the results and created the figures, J.H., L.L., and J.S. designed the research and wrote the paper; X.L., and H.Z. funded the research and edited the manuscript.

Funding

This work was supported by the National Key Research and Development Program of China (2024YFC2510500), National Natural Science Foundation of China (No. 82570171 and No. 82170123), Young Taishan Scholar Foundation of Shandong Province (Grant/Award No. tsqn202312325), Natural Science Foundation of Henan Province (No. 252300421374) and Henan Province Young and Middle-aged Health Science and Technology Innovation Leading Talent Training Program (No. YXKC2022008).

Data availability

No datasets were generated or analysed during the current study. For original data, please contact liuxingrant@email.sdu.edu.cn.

Declarations

Ethics approval and consent to participate

Our study was approved by the ethics committees of Qilu Hospital of Shandong University according to the Helsinki Principles. Written informed consent was obtained from all participants included in the study.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Juanjuan Song and Bingjie Ding have contributed equally to this work.

Contributor Information

Xinguang Liu, Email: liuxingrant@163.com.

Hu Zhou, Email: tigerzhoupumc@163.com.

References

1.Kuen DS, Park M, Ryu H, et al. Critical regulation of follicular helper T cell differentiation and function by Gα(13) signaling. Proc Natl Acad Sci U S A. 2021;118:43. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Lozano ML. Expanding therapeutic options for patients with primary immune thrombocytopenia. Lancet Haematol. 2023;10(6):e390–1. [DOI] [PubMed] [Google Scholar]
3.Yu X, Menard M, Prechl J, Bhakta V, Sheffield WP, Lazarus AH. Monovalent Fc receptor blockade by an anti-Fcγ receptor/albumin fusion protein ameliorates murine ITP with abrogated toxicity. Blood. 2016;127(1):132–8. [DOI] [PubMed] [Google Scholar]
4.Bolton-Maggs PHB, George JN. Immune thrombocytopenia treatment. N Engl J Med. 2021;385(10):948–50. [DOI] [PubMed] [Google Scholar]
5.Liu XG, Hou Y, Hou M. How we treat primary immune thrombocytopenia in adults. J Hematol Oncol. 2023;16(1):4. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Guo L, Kapur R, Aslam R, et al. CD20 + B-cell depletion therapy suppresses murine CD8⁺ T-cell-mediated immune thrombocytopenia. Blood. 2016;127(6):735–8. [DOI] [PubMed] [Google Scholar]
7.Cao J, Liao S, Zeng F, Liao Q, Luo G, Zhou Y. Effects of altered glycolysis levels on CD8⁺ T cell activation and function. Cell Death Dis. 2023;14(7):407. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Malik A, Sayed AA, Han P, et al. The role of CD8⁺ T-cell clones in immune thrombocytopenia. Blood. 2023;141(20):2417–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Provan D, Semple JW. Recent advances in the mechanisms and treatment of immune thrombocytopenia. EBioMedicine. 2022;76:103820. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Zhou Q, Mei Y, Shoji T, et al. Rho-associated coiled-coil-containing kinase 2 deficiency in bone marrow-derived cells leads to increased cholesterol efflux and decreased atherosclerosis. Circulation. 2012;126(18):2236–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Nalkurthi C, Schroder WA, Melino M, et al. ROCK2 inhibition attenuates profibrogenic immune cell function to reverse thioacetamide-induced liver fibrosis. JHEP Rep. 2022;4(1):100386. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Zanin-Zhorov A, Chen W, Moretti J, et al. Selectivity matters: selective ROCK2 inhibitor ameliorates established liver fibrosis via targeting inflammation, fibrosis, and metabolism. Commun Biol. 2023;6(1):1176. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Li Y, Yarahmadov T, Jahnke L, et al. Rho-kinase inhibition reduces subretinal fibrosis. Cell Death Discov. 2025;11(1):428. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Zanin-Zhorov A, Weiss JM, Nyuydzefe MS, et al. Selective oral ROCK2 inhibitor down-regulates IL-21 and IL-17 secretion in human T cells via STAT3-dependent mechanism. Proc Natl Acad Sci U S A. 2014;111(47):16814–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Wang X, Liu Y, Chen T, Yao H, Song Q, Zhang X. ROCK 2 inhibition with belumosudil for the treatment of chronic graft-versus-host disease: a narrative review. Ther Adv Hematol. 2025;16:20406207251367462. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Chen W, Nyuydzefe MS, Weiss JM, Zhang J, Waksal SD, Zanin-Zhorov A. ROCK2, but not ROCK1 interacts with phosphorylated STAT3 and co-occupies TH17/TFH gene promoters in TH17-activated human T cells. Sci Rep. 2018;8(1):16636. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Zanin-Zhorov A, Weiss JM, Trzeciak A, et al. Cutting edge: selective oral ROCK2 inhibitor reduces clinical scores in patients with psoriasis vulgaris and normalizes skin pathology via concurrent regulation of IL-17 and IL-10. J Immunol. 2017;198(10):3809–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Jagasia M, Lazaryan A, Bachier CR, et al. ROCK2 Inhibition with belumosudil (KD025) for the treatment of chronic graft-versus-host disease. J Clin Oncol. 2021;39(17):1888–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Neunert C, Terrell DR, Arnold DM, et al. American society of hematology 2019 guidelines for immune thrombocytopenia. Blood Adv. 2019;3(23):3829–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Zhou H, Qiu JH, Wang T, et al. Interleukin 27 inhibits cytotoxic T-lymphocyte-mediated platelet destruction in primary immune thrombocytopenia. Blood. 2014;124(22):3316–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Chow L, Aslam R, Speck ER, et al. A murine model of severe immune thrombocytopenia is induced by antibody- and CD8⁺ T cell-mediated responses that are differentially sensitive to therapy. Blood. 2010;115(6):1247–53. [DOI] [PubMed] [Google Scholar]
22.Vrbensky JR, Nazy I, Clare R, Larché M, Arnold DM. T cell-mediated autoimmunity in immune thrombocytopenia. Eur J Haematol. 2022;108(1):18–27. [DOI] [PubMed] [Google Scholar]
23.Bu S, Liu M, Yang L, et al. The function of T cells in immune thrombocytopenia. Front Immunol. 2025;16:1499014. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Semenzato G, Calabretto G, Teramo A, et al. The constitutive activation of STAT3 gene and its mutations are at the crossroad between LGL leukemia and autoimmune disorders. Blood Cancer J. 2024;14(1):13. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Sun Q, Zhao X, Li R, et al. STAT3 regulates CD8+ T cell differentiation and functions in cancer and acute infection. J Exp Med. 2023;220(4):e20220686. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Yu CR, Dambuza IM, Lee YJ, Frank GM, Egwuagu CE. STAT3 regulates proliferation and survival of CD8⁺ T cells: enhances effector responses to HSV-1 infection, and inhibits IL-10 + regulatory CD8⁺ T cells in autoimmune uveitis. Mediat Inflamm. 2013;2013:359674. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Badovinac VP, Haring JS, Harty JT. Initial T cell receptor transgenic cell precursor frequency dictates critical aspects of the CD8(+) T cell response to infection. Immunity. 2007;26(6):827–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Restifo NP, Dudley ME, Rosenberg SA. Adoptive immunotherapy for cancer: harnessing the T cell response. Nat Rev Immunol. 2012;12(4):269–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Callahan MK, Postow MA, Wolchok JD. Targeting T cell co-receptors for cancer therapy. Immunity. 2016;44(5):1069–78. [DOI] [PubMed] [Google Scholar]
30.Henning AN, Roychoudhuri R, Restifo NP. Epigenetic control of CD8(+) T cell differentiation. Nat Rev Immunol. 2018;18(5):340–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Zhang N, Bevan MJ. CD8(+) T cells: foot soldiers of the immune system. Immunity. 2011;35(2):161–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Jaccard A, Wyss T, Maldonado-Pérez N, et al. Reductive carboxylation epigenetically instructs T cell differentiation. Nature. 2023;621(7980):849–56. [DOI] [PubMed] [Google Scholar]
33.Suzuki J, Yamasaki S, Wu J, Koretzky GA, Saito T. The actin cloud induced by LFA-1-mediated outside-in signals lowers the threshold for T-cell activation. Blood. 2007;109(1):168–75. [DOI] [PubMed] [Google Scholar]
34.Yamada S, Nelson WJ. Synapses: sites of cell recognition, adhesion, and functional specification. Annu Rev Biochem. 2007;76:267–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Biswas PS, Gupta S, Chang E, et al. Phosphorylation of IRF4 by ROCK2 regulates IL-17 and IL-21 production and the development of autoimmunity in mice. J Clin Invest. 2010;120(9):3280–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Liang CL, Wei YY, Chen Y, et al. Zhen-Wu-Tang ameliorates lupus nephritis by diminishing renal tissue-resident memory CD8(+) T cells via suppressing IL-15/STAT3 pathway. Biomed Pharmacother. 2024;174:116597. [DOI] [PubMed] [Google Scholar]
37.Wang D, Yu W, Lian J, et al. Th17 cells inhibit CD8(+) T cell migration by systematically downregulating CXCR3 expression via IL-17A/STAT3 in advanced-stage colorectal cancer patients. J Hematol Oncol. 2020;13(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Moosic KB, Olson TL, Freijat M, et al. Activating STAT3 mutations in CD8⁺ T-cells correlate to serological positivity in rheumatoid arthritis. Front Immunol. 2024;15:1466276. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Zhou Y, Peng X, Fang C, et al. Histones methyltransferase NSD3 inhibits lung adenocarcinoma glycolysis through interacting with PPP1CB to decrease STAT3 signaling pathway. Adv Sci (Weinh). 2024;11(38):e2400381. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Zhang X, Chen T, Li Z, et al. NORAD exacerbates metabolic dysfunction-associated steatotic liver disease development via the miR-511-3p/Rock2 axis and inhibits ubiquitin-mediated degradation of ROCK2. Metabolism. 2025;164:156111. [DOI] [PubMed] [Google Scholar]
41.Wei L, Shi J. Insight into rho kinase isoforms in obesity and energy homeostasis. Front Endocrinol (Lausanne). 2022;13:886534. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Lee JY, Stevens RP, Kash M, et al. KD025 Shifts pulmonary endothelial cell bioenergetics and decreases baseline lung permeability. Am J Respir Cell Mol Biol. 2020;63(4):519–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Matoba K, Takeda Y, Nagai Y, et al. ROCK2-induced metabolic rewiring in diabetic podocytopathy. Commun Biol. 2022;5(1):341. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Cheng X, Gao Z, Zhang J, Zheng H, Shan S, Zhou J. TAGLN-RhoA/ROCK2-SLC2A3-mediated mechano-metabolic axis promotes skin fibrosis. Int J Biol Sci. 2025;21(2):658–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Weiss JM, Chen W, Nyuydzefe MS, et al. ROCK2 signaling is required to induce a subset of T follicular helper cells through opposing effects on STATs in autoimmune settings. Sci Signal. 2016;9(437):ra73. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Zanin-Zhorov A, Blazar BR. ROCK2, a critical regulator of immune modulation and fibrosis has emerged as a therapeutic target in chronic graft-versus-host disease. Clin Immunol. 2021;230:108823. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1^{(8.6MB, docx)}

Data Availability Statement

No datasets were generated or analysed during the current study. For original data, please contact liuxingrant@email.sdu.edu.cn.

[CR1] 1.Kuen DS, Park M, Ryu H, et al. Critical regulation of follicular helper T cell differentiation and function by Gα(13) signaling. Proc Natl Acad Sci U S A. 2021;118:43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Lozano ML. Expanding therapeutic options for patients with primary immune thrombocytopenia. Lancet Haematol. 2023;10(6):e390–1. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Yu X, Menard M, Prechl J, Bhakta V, Sheffield WP, Lazarus AH. Monovalent Fc receptor blockade by an anti-Fcγ receptor/albumin fusion protein ameliorates murine ITP with abrogated toxicity. Blood. 2016;127(1):132–8. [DOI] [PubMed] [Google Scholar]

[CR4] 4.Bolton-Maggs PHB, George JN. Immune thrombocytopenia treatment. N Engl J Med. 2021;385(10):948–50. [DOI] [PubMed] [Google Scholar]

[CR5] 5.Liu XG, Hou Y, Hou M. How we treat primary immune thrombocytopenia in adults. J Hematol Oncol. 2023;16(1):4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Guo L, Kapur R, Aslam R, et al. CD20 + B-cell depletion therapy suppresses murine CD8⁺ T-cell-mediated immune thrombocytopenia. Blood. 2016;127(6):735–8. [DOI] [PubMed] [Google Scholar]

[CR7] 7.Cao J, Liao S, Zeng F, Liao Q, Luo G, Zhou Y. Effects of altered glycolysis levels on CD8⁺ T cell activation and function. Cell Death Dis. 2023;14(7):407. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Malik A, Sayed AA, Han P, et al. The role of CD8⁺ T-cell clones in immune thrombocytopenia. Blood. 2023;141(20):2417–29. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Provan D, Semple JW. Recent advances in the mechanisms and treatment of immune thrombocytopenia. EBioMedicine. 2022;76:103820. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Zhou Q, Mei Y, Shoji T, et al. Rho-associated coiled-coil-containing kinase 2 deficiency in bone marrow-derived cells leads to increased cholesterol efflux and decreased atherosclerosis. Circulation. 2012;126(18):2236–47. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Nalkurthi C, Schroder WA, Melino M, et al. ROCK2 inhibition attenuates profibrogenic immune cell function to reverse thioacetamide-induced liver fibrosis. JHEP Rep. 2022;4(1):100386. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Zanin-Zhorov A, Chen W, Moretti J, et al. Selectivity matters: selective ROCK2 inhibitor ameliorates established liver fibrosis via targeting inflammation, fibrosis, and metabolism. Commun Biol. 2023;6(1):1176. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Li Y, Yarahmadov T, Jahnke L, et al. Rho-kinase inhibition reduces subretinal fibrosis. Cell Death Discov. 2025;11(1):428. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Zanin-Zhorov A, Weiss JM, Nyuydzefe MS, et al. Selective oral ROCK2 inhibitor down-regulates IL-21 and IL-17 secretion in human T cells via STAT3-dependent mechanism. Proc Natl Acad Sci U S A. 2014;111(47):16814–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Wang X, Liu Y, Chen T, Yao H, Song Q, Zhang X. ROCK 2 inhibition with belumosudil for the treatment of chronic graft-versus-host disease: a narrative review. Ther Adv Hematol. 2025;16:20406207251367462. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Chen W, Nyuydzefe MS, Weiss JM, Zhang J, Waksal SD, Zanin-Zhorov A. ROCK2, but not ROCK1 interacts with phosphorylated STAT3 and co-occupies TH17/TFH gene promoters in TH17-activated human T cells. Sci Rep. 2018;8(1):16636. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Zanin-Zhorov A, Weiss JM, Trzeciak A, et al. Cutting edge: selective oral ROCK2 inhibitor reduces clinical scores in patients with psoriasis vulgaris and normalizes skin pathology via concurrent regulation of IL-17 and IL-10. J Immunol. 2017;198(10):3809–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Jagasia M, Lazaryan A, Bachier CR, et al. ROCK2 Inhibition with belumosudil (KD025) for the treatment of chronic graft-versus-host disease. J Clin Oncol. 2021;39(17):1888–98. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Neunert C, Terrell DR, Arnold DM, et al. American society of hematology 2019 guidelines for immune thrombocytopenia. Blood Adv. 2019;3(23):3829–66. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Zhou H, Qiu JH, Wang T, et al. Interleukin 27 inhibits cytotoxic T-lymphocyte-mediated platelet destruction in primary immune thrombocytopenia. Blood. 2014;124(22):3316–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Chow L, Aslam R, Speck ER, et al. A murine model of severe immune thrombocytopenia is induced by antibody- and CD8⁺ T cell-mediated responses that are differentially sensitive to therapy. Blood. 2010;115(6):1247–53. [DOI] [PubMed] [Google Scholar]

[CR22] 22.Vrbensky JR, Nazy I, Clare R, Larché M, Arnold DM. T cell-mediated autoimmunity in immune thrombocytopenia. Eur J Haematol. 2022;108(1):18–27. [DOI] [PubMed] [Google Scholar]

[CR23] 23.Bu S, Liu M, Yang L, et al. The function of T cells in immune thrombocytopenia. Front Immunol. 2025;16:1499014. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Semenzato G, Calabretto G, Teramo A, et al. The constitutive activation of STAT3 gene and its mutations are at the crossroad between LGL leukemia and autoimmune disorders. Blood Cancer J. 2024;14(1):13. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Sun Q, Zhao X, Li R, et al. STAT3 regulates CD8+ T cell differentiation and functions in cancer and acute infection. J Exp Med. 2023;220(4):e20220686. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Yu CR, Dambuza IM, Lee YJ, Frank GM, Egwuagu CE. STAT3 regulates proliferation and survival of CD8⁺ T cells: enhances effector responses to HSV-1 infection, and inhibits IL-10 + regulatory CD8⁺ T cells in autoimmune uveitis. Mediat Inflamm. 2013;2013:359674. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Badovinac VP, Haring JS, Harty JT. Initial T cell receptor transgenic cell precursor frequency dictates critical aspects of the CD8(+) T cell response to infection. Immunity. 2007;26(6):827–41. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Restifo NP, Dudley ME, Rosenberg SA. Adoptive immunotherapy for cancer: harnessing the T cell response. Nat Rev Immunol. 2012;12(4):269–81. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Callahan MK, Postow MA, Wolchok JD. Targeting T cell co-receptors for cancer therapy. Immunity. 2016;44(5):1069–78. [DOI] [PubMed] [Google Scholar]

[CR30] 30.Henning AN, Roychoudhuri R, Restifo NP. Epigenetic control of CD8(+) T cell differentiation. Nat Rev Immunol. 2018;18(5):340–56. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Zhang N, Bevan MJ. CD8(+) T cells: foot soldiers of the immune system. Immunity. 2011;35(2):161–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Jaccard A, Wyss T, Maldonado-Pérez N, et al. Reductive carboxylation epigenetically instructs T cell differentiation. Nature. 2023;621(7980):849–56. [DOI] [PubMed] [Google Scholar]

[CR33] 33.Suzuki J, Yamasaki S, Wu J, Koretzky GA, Saito T. The actin cloud induced by LFA-1-mediated outside-in signals lowers the threshold for T-cell activation. Blood. 2007;109(1):168–75. [DOI] [PubMed] [Google Scholar]

[CR34] 34.Yamada S, Nelson WJ. Synapses: sites of cell recognition, adhesion, and functional specification. Annu Rev Biochem. 2007;76:267–94. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Biswas PS, Gupta S, Chang E, et al. Phosphorylation of IRF4 by ROCK2 regulates IL-17 and IL-21 production and the development of autoimmunity in mice. J Clin Invest. 2010;120(9):3280–95. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Liang CL, Wei YY, Chen Y, et al. Zhen-Wu-Tang ameliorates lupus nephritis by diminishing renal tissue-resident memory CD8(+) T cells via suppressing IL-15/STAT3 pathway. Biomed Pharmacother. 2024;174:116597. [DOI] [PubMed] [Google Scholar]

[CR37] 37.Wang D, Yu W, Lian J, et al. Th17 cells inhibit CD8(+) T cell migration by systematically downregulating CXCR3 expression via IL-17A/STAT3 in advanced-stage colorectal cancer patients. J Hematol Oncol. 2020;13(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Moosic KB, Olson TL, Freijat M, et al. Activating STAT3 mutations in CD8⁺ T-cells correlate to serological positivity in rheumatoid arthritis. Front Immunol. 2024;15:1466276. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Zhou Y, Peng X, Fang C, et al. Histones methyltransferase NSD3 inhibits lung adenocarcinoma glycolysis through interacting with PPP1CB to decrease STAT3 signaling pathway. Adv Sci (Weinh). 2024;11(38):e2400381. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Zhang X, Chen T, Li Z, et al. NORAD exacerbates metabolic dysfunction-associated steatotic liver disease development via the miR-511-3p/Rock2 axis and inhibits ubiquitin-mediated degradation of ROCK2. Metabolism. 2025;164:156111. [DOI] [PubMed] [Google Scholar]

[CR41] 41.Wei L, Shi J. Insight into rho kinase isoforms in obesity and energy homeostasis. Front Endocrinol (Lausanne). 2022;13:886534. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Lee JY, Stevens RP, Kash M, et al. KD025 Shifts pulmonary endothelial cell bioenergetics and decreases baseline lung permeability. Am J Respir Cell Mol Biol. 2020;63(4):519–30. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Matoba K, Takeda Y, Nagai Y, et al. ROCK2-induced metabolic rewiring in diabetic podocytopathy. Commun Biol. 2022;5(1):341. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Cheng X, Gao Z, Zhang J, Zheng H, Shan S, Zhou J. TAGLN-RhoA/ROCK2-SLC2A3-mediated mechano-metabolic axis promotes skin fibrosis. Int J Biol Sci. 2025;21(2):658–70. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Weiss JM, Chen W, Nyuydzefe MS, et al. ROCK2 signaling is required to induce a subset of T follicular helper cells through opposing effects on STATs in autoimmune settings. Sci Signal. 2016;9(437):ra73. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Zanin-Zhorov A, Blazar BR. ROCK2, a critical regulator of immune modulation and fibrosis has emerged as a therapeutic target in chronic graft-versus-host disease. Clin Immunol. 2021;230:108823. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

ROCK2 inhibition suppresses cytotoxic T lymphocyte-mediated platelet destruction in primary immune thrombocytopenia

Juanjuan Song

Bingjie Ding

Qian Liu

Ao Xie

Mengjuan Li

Xuewen Song

Fengdi Wang

Junxia Hu

Liu Liu

Xinguang Liu

Hu Zhou

Abstract

Graphical abstract

Supplementary Information

Introduction

Materials and methods

Patients and healthy controls

Table 1.

Active ITP mouse model

Isolation and culturing of CD8+ T Cells

ROCK kinase activity assay

Isolation of washed human platelets

CTL and platelet co-culture system

Flow cytometry

Flow cytometric analysis of CD8+ T-cell degranulation and CTL–platelet aggregate formation

STAT3 inhibition (Stattic)

Reverse Transcription quantitative polymerase chain reaction (RT-qPCR)

Western blotting

Seahorse metabolic flux analysis

Immunofluorescence staining

Enzyme-linked immunosorbent assay

IFN-γ ELISPOT assay

Lentiviral transduction of primary human CD8+ T cells

RNA sequencing and bioinformatic analysis

Platelet aggregation assay

Statistical analysis

Results

ROCK2 expression was elevated in ITP patients and mice

Fig. 1.

KD025 suppressed CD8+ T-cell proliferation, degranulation, cytotoxicity, and CTL cell-platelet interactions in ITP patients

Fig. 2.

Fig. 3.

KD025 attenuated CD61-driven activation and cytotoxicity of CD8+ T cells in Cd61−/− mice

Fig. 4.

Metabolic reprogramming in CD8+ T cells from ITP patients and its regulation by ROCK2 inhibition

Fig. 5.

ROCK2 inhibition suppressed STAT3 phosphorylation to attenuate CTL-mediated platelet destruction

Fig. 6.

IL-6/STAT3 signaling counteracted ROCK2 inhibition-mediated metabolic reprogramming in CD8+ T cells

Fig. 7.

KD025 ameliorated thrombocytopenia in active ITP mice

Fig. 8.

Discussion

Supplementary Information

Author contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Isolation and culturing of CD8⁺ T Cells

Flow cytometric analysis of CD8⁺ T-cell degranulation and CTL–platelet aggregate formation

Lentiviral transduction of primary human CD8⁺ T cells

KD025 suppressed CD8⁺ T-cell proliferation, degranulation, cytotoxicity, and CTL cell-platelet interactions in ITP patients

KD025 attenuated CD61-driven activation and cytotoxicity of CD8⁺ T cells in Cd61^−/− mice

Metabolic reprogramming in CD8⁺ T cells from ITP patients and its regulation by ROCK2 inhibition

IL-6/STAT3 signaling counteracted ROCK2 inhibition-mediated metabolic reprogramming in CD8⁺ T cells