Cryoablation-induced modulation of Treg cells and the TGF-β pathway in lung adenocarcinoma: implications for increased antitumor immunity

Background

Cryoablation plays a key role in the comprehensive management of lung adenocarcinoma, characterized by its ability to activate antitumor immunity. This study aimed to explore the impact of cryoablation on the local immune microenvironment, focusing on regulatory T cells (Tregs) and the TGF-β pathway.

Methods

Single-cell sequencing was employed to identify differences in immune cell populations and related pathway expression between lung adenocarcinoma tissues and adjacent noncancerous tissues. Prospective observations of changes in Tregs in the peripheral blood pre- and post-cryoablation for lung adenocarcinoma were conducted at Dongfang Hospital, Beijing University of Chinese Medicine. Bulk RNA-seq analysis of mouse tumor tissues was performed to predict the potential mechanisms underlying cryoablation-induced antitumor immunity. Finally, these predictions were validated through in vitro and in vivo experiments employing cell cryoablation and mouse subcutaneous tumor transplantation models.

Results

Single-cell RNA sequencing analysis revealed intricate interactions between Tregs subpopulations and the regulation of the immune response in lung adenocarcinoma, highlighting the involvement of the TGF-β pathway. A significant decrease in the level of Tregs was noted at 30 days post-cryoablation compared to pre-surgical and 3-day post-surgery levels. The cellular and murine cryoablation models validated the inhibitory effect of cryoablation on Tregs and its potential to stimulate antitumor immunity. Additionally, the results of bulk RNA-seq demonstrated the role of cryoablation in regulating postoperative immunity via the TGF-β pathway. Cryoablation decreased the expression levels of TGF-β1, suppressed the phosphorylation of Smad2 and Smad3, and downregulated the expression of FOXP3, thereby inhibiting the conversion of CD4 + T cell precursors into Tregs. Moreover, cryoablation enhanced the expression of interferon-gamma (IFN-γ), thereby promoting its antitumor activity.

Conclusions

This study revealed the effective modification of the lung adenocarcinoma microenvironment by cryoablation through the suppression of Tregs and activation of antitumor immunity via the TGF-β pathway. These findings hold implications for optimizing cryoablation-based therapies and guiding future clinical trials on lung adenocarcinoma treatment.

Trial registration

This trial was registered with the Chinese Clinical Trial Registry (Chictr.org.cn, ChiCTR2000038580, Sep 24, 2020).

Background

As is well documented, lung cancer is one of the most prevalent malignant tumors worldwide and threatens the health and life of humans. According to the Global Cancer Data 2022 report, lung cancer is currently the leading cause of cancer deaths globally, accounting for 18.7% of all cancer-related deaths [1]. Lung adenocarcinoma (LUAD) is the primary pathological type of lung cancer [2]. While radical surgery can yield positive outcomes in early-stage patients with a favorable prognosis, a considerable number of patients are diagnosed only at an advanced stage with a dismal prognosis [3, 4].

Cryoablation, a minimally invasive treatment for tumors, has long been used for the treatment of inoperable or unresectable LUAD patients. It offers distinct advantages over other treatment modalities, including a lower incidence of side effects and trauma, as well as improved visualization and targeting [5]. Cryoablation-mediated tumor cell death involves a multifaceted mechanism comprising rapid freezing and slow rewarming processes at ultra-low temperatures, leading to tumor cell demise and the induction of apoptosis [6, 7]. Furthermore, ablated tumors that persist within the patient’s body, being no longer viable, continually release tumor-associated antigens (TAAs) that are subsequently engulfed by antigen-presenting cells (APCs), thereby priming specific immune responses against residual tumor cells [8,9,10]. Notably, some studies have concluded that cryoablation can induce the regression of distant metastatic tumors, a phenomenon referred to as the abscopal effect [11].

The tumor microenvironment (TME) plays a pivotal role in the immune evasion of lung cancer. Comprised of tumor cells, immune cells, endothelial cells, fibroblasts, and extracellular matrix [12, 13], the TME is a complex interplay that collectively fosters conditions conducive to the growth and metastasis of lung cancer. Tregs, which characterized by the surface markers CD4 and CD25, are instrumental in maintaining immune suppression. High frequency of Tregs indicates worse outcome of NSCLC patients [14]. The differentiation and function of Tregs are crucially regulated by the TGF-β pathway, which promotes the expression of Forkhead box protein 3 (Foxp3). Foxp3, essential for the development and immunosuppressive function of Tregs, plays a pivotal role in their inhibitory activities. In addition to its role in Treg differentiation, TGF-β1 also contributes to the immunosuppressive tumor microenvironment by being secreted by Tregs themselves. This secretion of TGF-β1 by Tregs further suppresses antitumor immunity, which creates a negative feedback loop that allows the tumor to evade immune surveillance [15, 16]. In Foxp3-deficient mice, CD4 + CD25 + T cells have been shown to lack immune regulatory function. High frequency of Tregs indicates worse outcome of NSCLC patients [17].

Cryoablation of LUAD has demonstrated its effectiveness in not only killing cancer cells but also destroying the TME. Previous studies have shown that cryoablation can effectively reduce the number of Tregs in different types of solid tumors, such as kidney cancer, liver cancer, and prostate cancer [18,19,20]. However, there is currently no research investigating the alterations in Treg content and the underlying mechanisms following cryoablation, specifically in LUAD. In this study, multiomics approaches were employed to investigate the relationship between cryoablation and the local immune microenvironment in LUAD. Briefly, a cell cryoablation model and a mouse subcutaneous transplanted tumor model were established. In vitro and in vivo experiments validated that cryoablation inhibits Tregs and enhances antitumor immune responses by modulating the TGF-β pathway. Taken together, our findings highlight the dual role of cryoablation in eliminating lung adenocarcinoma cells and concomitantly stimulating antitumor immunity, offering potential benefits in preventing tumor progression and recurrence among LUAD patients.

Methods

scRNA-seq analysis

Tissue preparation and scRNA-seq library construction

Samples were obtained from the Dongfang Hospital, Beijing University of Chinese Medicine. Samples of carcinoma tissues were derived from three lung adenocarcinoma cases carrying non-metastasis, major, non-treated carcinomas receiving lung adenocarcinoma resecting. The adjacent normal tissues received the taking from over 5 mm in cancerous tissues’ negative margins. Tissues were preserved in sCelLive™ and processed within 30 min. Single-cell suspensions were prepared and scRNA-seq libraries were constructed using the GEXSCOPE® Single-cell RNA Library Kits protocol [21]. Libraries were sequenced on an Illumina NovaSeq 6000 platform to generate gene expression profiles.

Cell cluster analysis

Quality control was performed to exclude low-quality cells. Normalized and log-transformed data were subjected to variable gene selection and principal component analysis (PCA). The Louvain algorithm was used for cell clustering, and UMAP was used for visualization. Differential expression analysis identified DEGs, and pathway enrichment analysis was performed from the Kyoto Encyclopedia of Genes and Genomes (KEGG) to explore the functions of Treg subclusters. Cell type identification was determined based on the expression of canonical markers from the SynEcoSysTM database, and major cell types were further subtyped through reclustering.

Trajectory analysis

To explore cell differentiation and developmental potential, trajectory analysis was conducted using CytoTRACE and Monocle2. CytoTRACE v0.3.3 was used to predict the differentiation state of cell subpopulations based on gene counts and expression derived from scRNA-seq data. Monocle2 v2.22.0 was employed to reconstruct the cell differentiation trajectory of monocyte subtypes by selecting top highly variable genes using FindVairableFeatures in Seurat, performing dimension reduction using DDRTree, and visualizing the trajectory using the plot_cell_trajectory function of Monocle2. These analyses provided insights into cellular differentiation pathways and potential.

Sample collection and processing

Blood samples were collected from a group of stage IIIb/IV LUAD patients undergoing cryoablation at Dongfang Hospital, Beijing University of Chinese Medicine, to evaluate changes in TGF-β and Tregs levels before and after cryoablation. The cryoablation system was used for percutaneous probe insertion under CT guidance, with probe size and quantity tailored to the tumor’s characteristics. The procedure involved a dual freeze–thaw cycle: 20 min of ice ball formation, 10 min of thawing, and 15 min of refreezing, achieving cryoprobe temperatures of − 196 °C within 5 min (registration number: ChiCTR2000038580). Peripheral blood samples were collected from patients 1 day before, 3 days after, and 1 month after the procedure, and flow cytometry was performed to analyze the content of Tregs, whilst ELISA was used to analyze TGF-β1 levels. This prospective study was approved by the Ethics Committee of Dongfang Hospital, Beijing University of Chinese Medicine.

Mouse models and cryoablation

The animal experiments in this study were approved by the Laboratory Animal Ethics Committee of Dongfang Hospital, Beijing University of Chinese Medicine. Six to 8 weeks C57BL/6 male wild-type mice (n = 20) were purchased from Sibeifu Inc. (Beijing, China). A tumor‐bearing mouse model was constructed by subcutaneously injecting of Lewis lung adenocarcinoma cells (2 × 10⁶ cells/mouse) into the right flank of each mouse. Next, the mice were randomly divided into four groups, namely the control group (Con), TGF-β1 inhibitor group (SB431542), cryoablation group (Cry), and cryoablation + TGF-β1 inhibitor group (Cry + SB431542). After 7 days, the tumor was subjected to cryoablation using a cryoablation system for two cycles. The temperature of each cycle was reduced to − 120 °C for 10 s, followed by rewarming to 10 °C [22,23,24]. Mice in the control group underwent only surgical incision, and suturing under strict aseptic conditions. Mice in the TGF-β1 inhibitor group were only intraperitoneally injected with SB431542 (APExBIO, Houston, TX, USA) once daily. On the other hand, mice in the cryoablation + TGF-β1 inhibitor group were intraperitoneally injected with 1 μM SB431542 in DMSO once daily from postoperative day 0 to postoperative day 14. After the intervention, the surgical site was disinfected once daily to prevent infection. The size and volume of the tumors were recorded after the operation, and a growth curve was plotted for 14 days.

Histology and immunostaining

Tumor tissues from mice in each group were fixed and embedded in paraffin, and sections were prepared. Hematoxylin–eosin (HE) staining, Masson staining, and immunohistochemistry for Ki67 (1:200) (ab15580, Abcam, Cambridge, UK) were performed according to standard protocols [25, 26].

Enzyme‐linked immunosorbent assay

TGF-β1 and IFN-γ levels were measured in human and mouse serum samples using a commercial ELISA kit (Proteintech, Chicago, USA). ELISA was performed according to the manufacturer’s protocol.

Bulk RNA-seq and data analysis

Three tumor tissue samples from mice in both the cryoablation group and the control group were selected for total RNA was extraction using the Trizol reagent kit (Invitrogen, Carlsbad, CA, USA). Next, the RNA samples were sequenced by Gene Denovo Biotechnology Co. (Guangzhou, China). Strand-specific libraries were constructed from the extracted RNA, followed by sequencing using the Illumina HiSeq 4000 platform. Differential expression analysis was performed using DESeq2 [27], comparing samples across different groups. Genes with a false discovery rate (FDR) < 0.05 and an absolute fold change ≥ 2 were considered differentially expressed genes and were further subjected to Gene Ontology (GO) and KEGG pathway enrichment analyses.

Real‐time quantitative polymerase chain reaction

Three tumor tissue samples were obtained from each group of mice. Total RNA was extracted using RNA extraction solution (Servicebio, Wuhan, China), while reverse transcription was performed using the Reveraid First Strand cDNA Synthesis Kit (Thermo Scientific, MA, USA), and PCR was performed using SYBR qPCR SuperMix Plus (Novoprotein, Shanghai, China). β-Actin served as an internal control for mRNA. The 2^−ΔΔCT method was used for relative quantification. The primers (5′–3′) used for gene amplification are listed in Additional file 1: Table S1.

Western blot (WB) analysis

WB analysis was performed as described previously [28]. The following primary antibodies were used: TGF-β1 (21,898–1-AP, Proteintech), SMAD2 (12,570–1-AP, Proteintech), p-SMAD2 (ab188334, Abcam), SMAD3 (66,516–1-Ig, Proteintech), p-SMAD3 (ab52903, Abcam), FOXP3 (ab215206, Abcam), Mouse anti-GAPDH (60,004–1-Ig, Proteintech), and Rabbit anti-GAPDH (10,494–1-AP, Proteintech).

Cell culture and treatment

Human lung adenocarcinoma cell lines A549 and H1299 were procured from Wuhan Pu-nuo-sai Life Technology Co. Ltd. (Wuhan, China); mouse lung adenocarcinoma cells (Lewis lung carcinoma cells, LLC) were sourced from Jiangsu Kaiji Bio-Technology Co. Ltd (Nanjing, China). Both A549 and LLC cells were incubated in high-glucose DMEM (HyClone, USA) medium supplemented with 10% FBS (Gibco, USA) and 1% penicillin/streptomycin (Gibco, USA), whereas H1299 cells were incubated in RPMI-1640 medium (HyClone, USA). All cells were incubated in an incubator at 37 °C and 5% CO₂.

To activate T cells within PBMCs obtained from healthy volunteers, a 24-well plate was coated with anti-CD3 antibody (clone: UCHT1, BioLegend, CA, USA) diluted to 5 μg/mL in PBS. A volume of 400 μL of the diluted antibody was added to each well and incubated overnight at 4 °C. Following incubation, the wells were aspirated and washed three times with PBS to remove any unbound antibody. The PBMCs were then added to these wells.

Construction of the cryoablation cell model

LUAD cells (5 × 10⁵ cells/mL, 5 mL) were sequentially collected from cryogenic tubes, frozen in a − 80 °C refrigerator for 7 min, and rewarmed in a 37 °C water bath. These cells remained partially viable and were characterized as sublethal cells outside the cryoprobe. The other cryogenic tube was frozen in liquid nitrogen for 5 min, resulting in cell lysis and necrosis. The supernatant was collected by centrifugation. The necrotic medium was prepared by combining the complete medium and supernatant at a ratio of 2:1 [29]. Experimental groups: Group A: healthy cells + complete medium (unfrozen area), Group B: healthy cells + necrotic medium (healthy cells located near the frozen area), Group C: sublethal cells + complete medium (away from the necrotic sublethal area), Group D: sublethal cells + necrotic medium (close to the necrotic sublethal area). After incubation for 24 h, the ensuing experiments were performed.

Co-culture assay

Co-culture experiments were performed in 6-well plates with inserts with a pore size of 0.4 μm (Corning, NY, USA). The LUAD cells (5 × 10⁵ cells, 1 ml), including the A549 and H1299 cell lines, were seeded and incubated in the outer wells of a 6-well plate. PBMCs (1 × 10⁶ cells, 2 mL) pre-treated with anti-CD3 antibody as described in section “Cell culture and treatment” were added to the inner wells of a Transwell system. All cells were incubated in a complete medium containing 5 µg/mL anti-CD28 (Clone: CD28.2, BioLegend, CA, USA) and 0.4 µL/mL IL-2 (Catalog No.: GMP-CD66, Novoprotein, Shanghai, China). After 72 h, cells in the inner wells were harvested to quantify the Tregs via flow cytometry.

Treatment of samples and flow cytometric analysis

The serum samples of patients, co-cultured cells in the inner wells, and spleen tissue of mice were collected to detect the content of Tregs. Splenic tissues were ground, filtered, and centrifuged into a single-cell suspension, respectively. The staining protocol was adapted based on the samples. Tregs were detected as either CD4, CD25^bright and Foxp3 staining, or CD4, CD25^bright, and CD127^low staining. Intracellular staining for Foxp3 was performed using a Foxp3 Staining Buffer Set (eBioscience, CA, USA), according to the manufacturer’s instructions. The antibodies are summarized in Additional file 1: Table S2.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8 (GraphPad, USA) and SPSS 20 (IBM, USA). P-value < 0.05 was considered statistically significant. Paired t-tests were employed for clinical patient samples, while independent t-tests were utilized for between-group comparisons. One-way ANOVA was conducted for comparisons across multiple groups.

Results

Human single-cell sequencing analysis reveals cellular heterogeneity in lung adenocarcinoma microenvironment

To gain a comprehensive understanding of the cellular diversity within the lung adenocarcinoma microenvironment, single-cell RNA sequencing (scRNA-Seq) was performed on six samples collected from three patients with lung adenocarcinoma. The samples comprised three cases of lung adenocarcinoma tissues and three adjacent nontumor tissues. Following stringent quality control measures, single-cell transcriptomes were acquired for 95,318 cells. As depicted in the UMAP plot, seven major cell clusters were identified based on marker gene expression profiles: epithelial cells, stromal cells, proliferating cells, B cells, T cells, NK cells, mast cells, and mononuclear phagocytes (Fig. 1A). Histograms illustrating the specificity of the expression of each cell cluster across different samples revealed that epithelial cells and stromal cells were significantly overexpressed in lung adenocarcinoma tissues than in adjacent nontumor tissues (Fig. 1B). Conversely, T and NK cells and mast cells exhibited higher expression levels in adjacent nontumor tissues. To further identify immune cells associated with lung adenocarcinoma, a sub-clustering analysis of T and NK cells was conducted according to marker identification and subsequent merger of the clusters, yielding 11 distinct subtypes (Fig. 1C, E). According to the histograms representing these immune cell subgroups, the expression levels of Treg-IL2RA, NKT-HSPA6, and NaïveT-RPL34 were higher in tumor tissues compared to adjacent nontumor tissues (Fig. 1D).

Fig. 1

Acquisition of scRNA-seq profiles of samples and data generation in LUAD. A, C UMAP illustrating the identified cell populations. B, D Histograms displaying the differences in expression levels of the cell populations in different samples. E Dot plots depicting markers of different T and NK cell populations

Detailed classification and developmental trajectory of Tregs in the human lung adenocarcinoma microenvironment

The UMAP plot revealed seven cell clusters identified which were labeled as follows: Treg-LTB, Treg-PELI1, Treg-CCL5, Treg-HSPA6, Treg-CXCL13, Treg-CCL20, and Treg-IFI27 (Fig. 2A). For instance, the Treg-CCL5 cell cluster primarily expressed CCL5, GZMB, and CCL4, while the Treg-CXCL13 cell cluster predominantly expressed CXCL13, DRAIC, and C19orf70. The Treg-HSPA6 cell cluster was characterized by overexpression of HSPE1, DNAJB1, and HSPA6 (Additional file 1: Fig. S1). To elucidate the differentiation and developmental process of Tregs, the monocle and cytoTRACE R packages were used for pseudotime trajectory analysis of Treg subpopulations.

Fig. 2

Analysis of Treg trajectories and pathway enrichment in LUAD. A UMAP plot illustrating the identified Treg populations. B, D The trajectory distribution of each Treg population over time. The top represents the starting point of development, whereas the bottom is the developmental endpoint. C Heatmap of differently expressed genes along pseudotime. The colorbar on the left denotes different clusters for each gene. Rows: expression of differentially expressed genes; columns: cells shown in pseudotime order. E Pathway analysis of Treg clusters. F Boxplot delineating differentiation potential across different types of Tregs

According to the pseudotime analysis, during the early branch-point stage, Treg-IFI27 cells dominated, possessing greater differentiation potential and gradually transitioning towards intermediate and late-stage Treg differentiation. In the late stages of differentiation, Tregs primarily expressed HSPA6 and PELI1. Conversely, Treg-CXCL13, Treg-CCL5, Treg-CCL20, and Treg-LTB cells were predominantly concentrated in the intermediate stages of differentiation (Fig. 2B–D). Interestingly, both early and late Treg cell types were primarily localized within cancer-adjacent tissues, whereas Tregs within lung cancer tissues exhibited a more mature phenotype (Additional file 1: Fig. S2). The gene clustering heatmap illustrated the temporal trends of critical genes throughout the pseudotime analysis (Fig. 2E). The present study specifically focused on TGF-β and simulated the expression trajectory of TGFB1, noting that its expression level was increased during Treg differentiation (Additional file 1: Fig. S3). Pathway analysis suggested that genes within the Treg population regulated PD-L1 expression, Th17 cell differentiation, and the TGF-β pathway, indicating close relationships with tumor immune responses (Fig. 2F).

Cryoablation-mediated regulation of Tregs in clinical and experimental models of LUAD

To investigate the impact of cryoablation on Tregs and its potential to enhance immune responses in LUAD, a total of 22 patients were recruited at the Dongfang Hospital, Beijing University of Chinese Medicine, between August 2021 and February 2022, with 17 completing the trial. CD4 + CD25^highCD127^low/− was employed as a marker for Treg expression. To assess Treg content, aggregates and debris were initially excluded based on cell size and granularity, then lymphocyte populations were selected based on CD45 expression, and CD4 + T lymphocytes were identified by CD3 + and CD4 + expression. Finally, Tregs were determined through CD25 + CD127^low/− expression. As anticipated, the proportion of CD4 + CD25^highCD127 ^low/− in the majority of lung adenocarcinoma patients initially increased and subsequently decreased after cryoablation (Treg frequencies were 7.893 ± 2.0744% pre-surgery, 8.624 ± 2.6687% 3 days post-surgery, and 6.612 ± 1.0948% 30 days post-surgery). Statistical analysis revealed a significant decrease in Treg content 30 days post-surgery compared to both pre-surgery levels and 3 days post-surgery (t = 2.793, p = 0.013; t = 2.923, p = 0.01) (Fig. 3A, B). We speculate that the transient increase in Tregs 3 days post-surgery may be linked to treatment response.

Fig. 3

Changes in the frequency of Tregs in the peripheral blood of LUAD patients before and after treatment. A In the 17 patients, Treg frequency initially increased, followed by a decrease. *Comparison between postoperative day 30 and pre-treatment, P < 0.05; #Comparison between postoperative day 30 and postoperative day 3, P < 0.05. B Variations in CD4⁺CD25^highCD127^low Tregs during the follow-up period are represented by the dot plots of patients

Furthermore, cellular and murine cryoablation models were constructed to examine the effects of cryoablation on Tregs both in vitro and in vivo. Lung adenocarcinoma A549 and H1299 cells were subjected to liquid nitrogen to simulate cell lysis during cryoablation. Considering the temperature gradient from the center to the periphery of the ice ball, an environment of − 80 °C was designed to simulate the freezing conditions of peripheral ice balls. To evaluate the influence of cryoablation on Treg differentiation in vitro, the cellular cryoablation model was co-cultured with PBMCs (previously activated by CD3/CD28 and IL2). The results demonstrated that groups B, C, and D influenced Treg differentiation to varying degrees compared to the control group A (Fig. 4A, B). Notably, after accounting for the influence of tumor cell heterogeneity, lung adenocarcinoma cell lines cultured in a necrotic medium were found to more effectively inhibit Treg differentiation, with minimal association with cryogenic interventions. Additionally, flow cytometry experiments on mouse spleens revealed a reduction in the number of Tregs 14 days post-cryoablation compared to the control group. Compared to the Cry + SB431542 group, the Cry group showed a trend towards higher Treg numbers, though not statistically significant (p > 0.05). This may result from the complex interactions between TGF-β signaling inhibition and the tumor microenvironment’s immune regulation, potentially masking the effects of cryoablation and SB431542 on Treg levels (Fig. 4C, D).

Fig. 4

Impact of cryoablation on Treg frequency in in vivo and in vitro experiments. A Effect of co-culture with PBMCs on Treg differentiation in the A549 and H1299 LUAD cell lines, wherein Group D exhibited the lowest Treg frequency. Representative plots are shown. A Healthy cells + complete medium, B healthy cells + necrotic medium, C sublethal cells + complete medium, D sublethal cells + necrotic medium). B Percentage of Tregs after co-culture with PBMCs. Data were pooled from 3 independent experiments. C Effect of cryoablation on Treg generation in the mouse spleen. Representative plots are shown. D Percentage of Tregs in the mouse spleen. n = 3 mice per group. Statistical significance between groups was assessed using the Brown-Forsythe and Welch ANOVA tests. *P < 0.05, **P < 0.01, and ***P < 0.001

Cryoablation enhances antitumor capacity in lung cancer mouse model

To comprehensively investigate the antitumorigenic effects of cryoablation, a mouse model of lung cancer treated with cryoablation was established. After treatment, a significant reduction in tumor volume was observed in the cryoablation group compared to the control group 14 days post-surgery (Fig. 5A, B). Moreover, cryoablation altered the size, morphology, and structure of lung cancer cells (Fig. 5C). Meanwhile, immunohistochemistry assays also that the expression level of Ki-67 was significantly lower in the cryoablation group compared to the control group (Fig. 5D, E). These findings collectively indicated that cryoablation effectively suppressed the proliferation of residual tumor cells after ablation, thereby inhibiting tumor progression.

Fig. 5

Cryoablation suppresses lung cancer cell proliferation in mouse models. A, B Tumor volume measurements in each group following cryoablation for 14 days, n = 5 mice per group. Statistical significance was assessed using one-way ANOVA. *P < 0.05, **P < 0.01. C HE staining results of tumor tissue 14 days post-cryoablation exposed changes in the Cry group, with cells appearing smaller and rounder, with condensed nuclei, dense chromatin, and displaying apoptotic morphology. D, E Cryoablation downregulated Ki-67 expression in lung cancer tissue. n = 3 mice per group. Statistical significance was assessed using one-way ANOVA. *P < 0.05, **P < 0.01

Interestingly, cryoablation induced changes in collagen fiber structure within the tumor tissue, rendering the post-ablation tissue loose and predominantly exhibiting curved structures surrounding the tumor nests. Moreover, the Cry + SB431542 group showed a more pronounced change in this regard when compared with the Cry group (Fig. 6A, B). Upon collagen fibers transitioning from a curled to a linear configuration traversing through the tumor foci, tumor cells rapidly migrated along these reconstructed linear collagen fibers, a phenomenon referred to as “highways” for tumor invasion by researchers [30, 31]. In the current study, cryoablation suppressed changes in collagen fibers in Lewis lung adenocarcinoma mouse xenografts, thereby inhibiting tumor metastasis. Additionally, a lower number of lung metastases were noted in mice in the cryoablation group (60% vs. 0%, Con vs. Cry) (Fig. 6C).

Fig. 6

Cryoablation suppressed lung cancer metastasis in mouse models. A, B Masson staining results displayed redeposition of collagen in the Con group, forming thick, dense, straight collagen fibers passing through tumor nests; collagen fibers in the Cry group appeared loose and curved structures surrounding tumor nests. n = 3 mice per group. Statistical significance was evaluated using one-way ANOVA. *P < 0.05, Con vs. Cry + SB431542. C Assessment of metastatic lesions in mouse lungs revealed minimal metastases in the cryoablation group, with a comparison showing 60% versus 0% metastatic incidence between the control and cryoablation groups, respectively. n = 5 mice per group

Bulk RNA-seq results indicate that cryoablation reverses the immunosuppressive microenvironment in mouse models

To further explore the molecular mechanisms underlying cryoablation-enhanced antitumor immunity, RNA seq was performed on tumors collected from mice 14 days after cryoablation. Tumors were snap-frozen in liquid nitrogen for 15 min and stored at − 80 °C until processing. DESeq analysis was carried out to identify DEGs, which were visualized using volcano plots (Fig. 7A). Our results identified 236 upregulated genes and 153 downregulation genes in the cryoablation group compared to the control group.

Fig. 7

Bulk RNA-seq analysis in mouse models post-cryoablation. A As depicted in the Volcano plot, 236 genes were upregulated, and 153 genes were downregulated (Con vs. Cry) (P < 0.05 and |log2FC|> 1.5). B, C Results of GO and KEGG enrichment analyses. GO function analysis histogram. BP is marked by dark cyan, CC is marked by sienna and MF is marked by steel blue. Histogram of the KEGG pathway enrichment analysis. The horizontal axis represents the enrichment rate of the input genes in the pathway, while the vertical axis represents the pathway name. The color scale indicates the thresholds of the p-value

GO and KEGG enrichment analyses were performed on the DEGs using ClusterProfiler to determine their primary biological functions. Of note, the DEGs were significantly enriched in biological processes (BP) terms related to the regulation of lymphocyte proliferation, regulation of mononuclear cell proliferation, regulation of inflammatory response, and leukocyte migration. They were significantly enriched in cellular components (CC) terms associated with the MHC protein complex and the receptor complex and in molecular functions (MF) terms linked to receptor ligand activity, immune receptor activity, and immunoglobulin binding (Fig. 7B). Moreover, KEGG enrichment analysis identified several signaling pathways associated with the differential gene expression profile resulting from cryoablation, including Th17 cell differentiation, HIF-1 signaling pathway, p53 signaling pathway, VEGF signaling pathway, TGF-beta signaling pathway, and JAK − STAT signaling pathway (Fig. 7C). These findings suggest the involvement of multiple signaling cascades in mediating the effects of cryoablation on immune regulation and antitumor responses, providing valuable insights into potential therapeutic targets for enhancing cryoablation efficacy in cancer treatment. Further research is warranted to unravel the intricate interplay between these pathways and their contributions to cryoablation-induced immune modulation.

Cryoablation inhibits the TGF-β/SMAD/FOXP3 pathway,suppressing immune evasion in LUAD

Our previous bulk RNA-seq unveiled that lung adenocarcinoma cryoablation influences the TGF-β signaling pathway. scRNA-seq demonstrated a strong correlation between the TGF-β pathway and Tregs. Concurrently, flow cytometry corroborated that cryoablation suppressed Treg differentiation. To further investigate the underlying mechanisms, TGF-β1 expression was detected in the serum of patients, revealing a post-treatment pattern of initial increase followed by a decrease, which paralleled the trend in Treg differentiation (Fig. 8A). Both in vivo mouse models and in vitro cell experiments substantiated the inhibitory effect of cryoablation on TGF-β1 expression (Fig. 8B, C).

Fig. 8

Effects of cryoablation on the TGF-β/SMAD/FOXP3 signaling pathway. A Changes in serum TGF-β1 levels before and after surgery in 17 patients. Statistical significance was determined using the paired t-test. ***P < 0.001. B Differences in serum TGF-β1 levels among mouse groups. n = 3 mice per group. Statistical significance was assessed using Brown-Forsythe and Welch ANOVA tests. *P < 0.05. C Expression levels of TGF-β1 in the supernatant after co-culture with a cell cryoablation model and PBMCs. Statistical significance was determined using the Brown-Forsythe and Welch ANOVA tests. *P < 0.05, ***P < 0.001. D, E, F, G Expression levels of key genes in the TGF-β/SMAD/FOXP3 signaling pathway in tumor tissues collected from mice in each group. n = 3 mice per group. Statistical significance was determined using one-way ANOVA. *P < 0.05, **P < 0.01. H Western blot analysis of proteins related to the TGF-β/SMAD/FOXP3 signaling pathway in tumor tissues harvested from mice in each group. I Cumulative data for quantification of the levels of proteins related to the TGF-β/SMAD/FOXP3 signaling pathway in blots shown in H. n = 3 mice per group. Statistical significance was determined using one-way ANOVA. *P < 0.05

qPCR and Western blot analyses determined that cryoablation downregulates TGF-β expression, thereby impairing the phosphorylation of Smad2 and Smad3 and subsequently suppressing downstream Foxp3 expression. The inhibitor SB431542 markedly suppressed the transcription levels of Smad2, Smad3, and Foxp3. Nevertheless, its effect on protein expression was less pronounced (Fig. 8D–I). These findings signaled that cryoablation inhibits the TGF-β/SMAD/FOXP3 pathway, hindering Treg differentiation and reshaping the immune microenvironment surrounding lung adenocarcinoma. This mechanistic understanding enhances our understanding of the mechanism by which cryoablation enhances antitumor immunity and provides potential therapeutic targets for optimizing cryoablation-based interventions in LUAD treatment.

Cryoablation enhances antitumor capacity via IFN-γ

To further investigate the potential of cryoablation to augment the killing of LUAD in addition to its effects on suppressing Tregs differentiation, we conducted a series of experiments. Our results demonstrated that the expression of IFN-γ progressively increased, whereas that of IL-2 decreased in lung cancer patients after cryoablation (Fig. 9A, D). Subsequently, ELISA was carried out to detect the levels of IFN-γ and IL-2 in mouse serum 2 weeks after cryoablation. Cryoablation was observed to produce a trend towards increased levels of IFN-γ, albeit non-statistically significant, which was higher when combined with a TGF-β1 inhibitor. In contrast, IL-2 expression was significantly lower in the treated group than in the control group (Fig. 9B, E). Moreover, a cell cryoablation model co-cultured with PBMCs was constructed, and ELISA was performed on the supernatant. Our results showed that, compared to the control group, the expression level of IFN-γ and IL-2 was increased and decreased in sublethal cells + necrotic medium (D), respectively, with a more prominent effect observed in the A549 cell line (Fig. 9C, F). These findings collectively suggest that cryoablation enhanced antitumor capacity through the modulation of IFN-γ and IL-2 expression. The upregulation of IFN-γ and concurrent downregulation of IL-2 may contribute to an improved immune response against lung adenocarcinoma cells, highlighting the potential of cryoablation as a complementary strategy in cancer immunotherapy.

Fig. 9

Cryoablation modulates IFN-γ and IL-2 expression to enhance antitumor immunity. A, D Changes in serum IFN-γ and IL-2 levels before and after surgery in 17 patients. Statistical significance was evaluated using the paired t-test. *P < 0.05, **P < 0.01, ***P < 0.001. B, E Serum IFN-γ and IL-2 levels in mouse models 2 weeks post-cryoablation. n = 3 mice per group. Statistical significance was evaluated using one-way ANOVA (B) and the Brown-Forsythe and Welch ANOVA tests (E). *P < 0.05, **P < 0.01. C, F Expression levels of IFN-γ and IL-2 in the supernatant after co-culture with a cell cryoablation model and PBMCs. Statistical significance was evaluated using one-way ANOVA (C) and the Brown-Forsythe and Welch ANOVA tests (F). *P < 0.05, ** P < 0.01

Discussion

Cryoablation is a well-established treatment modality for various solid tumors, including prostate cancer, breast cancer, kidney cancer, and lung cancer [32,33,34]. Using imaging techniques such as CT, cryogenic media (argon or liquid nitrogen) are delivered through probes into tumor tissue, achieving freezing of the tumor at ultra-low temperatures of − 175 °C, leading to cellular damage through repeated freeze–thaw cycles [35]. With the widespread clinical application of cryoablation, the post-procedure immune response has garnered increasing attention. Reports dating back to the 1970s have documented the regression of distant metastases following cryoablation [36]. Current research posits that cryoablation, by lysing tumor cells, exposes previously shielded tumor antigens to the immune system, offering therapeutic opportunities for immune cells [26].

Tumor cells can evade immune cell attack through multiple mechanisms, one of which involves the activation of Tregs to suppress immune responses. In this study, scRNA-seq was conducted on cells isolated from cancer-adjacent and lung adenocarcinoma tissues, yielding a total of 95,318 cells. Besides, distinct cell types were identified in the peripheral regions of LUAD, and differences in immune cell subtypes were examined by isolating T and NK cells. Particularly, we focused on Treg populations, elucidating their developmental trajectories and enriching pathways associated with DEGs. Seven Tregs subtypes were prominently enriched in the tumor microenvironment, namely Treg-LTB, Treg-PELI1, Treg-CCL5, Treg-HSPA6, Treg-CXCL13, Treg-CCL20, and Treg-IFI27. Pseudo-chronological analysis revealed the presence of both early-stage differentiating Treg-IFI27 cells and late-stage Treg-PELI1 cells in adjacent non-cancerous tissues, while mature Treg-LTB cells predominated in lung adenocarcinoma tissue. The immunosuppressive role of Tregs in the lung adenocarcinoma microenvironment is well-established [37]. Notably, prior investigations have concluded that Tregs can adapt to a hypoglycemic, high-lactate environment, allowing them to utilize energy sources inaccessible to other immune cells, thereby facilitating immune escape within the distinctive microenvironment of lung adenocarcinoma [38]. Modulating the lung adenocarcinoma microenvironment holds the potential to impact the survival status of Tregs.

Importantly, scRNA-seq identified a correlation between TGF-β1 and the developmental trajectory of Tregs. The TGF-β family comprises three isoforms, namely TGF-β1, TGF-β2, and TGF-β3, existing as homodimers within the biological system [39]. Noteworthily, TGF-β1, highly expressed in tumor tissues, plays a pivotal role in the tumor microenvironment (TME) [40]. Indeed, the TGF-β1 signaling pathway directly influences various cellular components within the TME, including cancer cells, immune cells, and other cell types, playing a decisive role in cancer progression. Consequently, inhibiting TGF-β signal transduction can potentially attenuate immune suppression, angiogenesis, and the activation of cancer-associated fibroblasts (CAFs) [41, 42], thereby enhancing the effectiveness of alternative therapeutic modalities. Considering the elevated TGF-β activity in the tumor vicinity, therapeutic strategies must concurrently target cancer cells and the immune and stromal components within the TME. This dual approach ensures sustained inhibition of TGF-β signal transduction, facilitating enduring regression of cancer. Overall, this research outlined the distinctive characteristics of cryoablation, emphasizing its targeted ablation capabilities. Our study explored the overarching impact of cryoablation, not only in completely eliminating of lung adenocarcinoma but also in altering the Tregs landscape within the tumor microenvironment. This alteration is speculated to enhance the host’s anti-tumor immune response. Clinical blood tests of patients undergoing cryoablation revealed a reduction in Treg frequency. Subsequent in vivo and in vitro experiments further corroborated this phenomenon. Importantly, the decrease in Tregs was correlated with changes in TGF-β1 levels. Pathological staining revealed cryoablation-induced alterations in the lung adenocarcinoma microenvironment in the murine lung cancer model. These alterations manifested as observable changes in tumor tissue structure, reduced invasiveness, and limited proliferation in LUAD.

Naïve precursor CD4 + T cells are guided by TGF-β1 to differentiate into Tregs, which suppress antitumor immunity. DEGs were analyzed in Treg subsets, and KEGG enrichment confirmed the association between Tregs and the TGF-β/Smad signaling pathway in LUAD. Furthermore, our transcriptomic study of mouse cryoablation revealed a correlation between the TGF-β signaling pathway and cryoablation. It is worthwhile emphasizing that the TGF-β signaling pathway promotes the progression of late-stage lung cancer. Upon activation, TGF-β forms a dimeric complex, phosphorylates Smad2 and Smad3, and binds to Smad4 to form the Smad complex. This complex subsequently translocates to the cell nucleus, activates Foxp3, and generates Tregs by promoting Foxp3 CNS1 site activation [37]. Our hypothesis was validated in the animal experiments, wherein increased TGF-β1 expression was observed in serum after cryoablation. Interestingly, variations in TGF-β1 expression were less pronounced in lung adenocarcinoma tissues. Observing the impact of TGF-β1 expression, downstream Smad2 and Smad3 exhibited increased phosphorylation, thereby activating FOXP3 and promoting the transformation of naïve precursor CD4 T cell precursors into Tregs. Furthermore, our study uncovered an additional facet of cryoablation, demonstrating its capability to upregulate the expression of IFN-γ, thereby amplifying its anti-tumorigenic effects. These findings not only shed light on the intricate interplay between the TGF-β/Smad signaling and cryoablation but also highlight the multifaceted impact of cryoablation on the immune landscape within the lung adenocarcinoma microenvironment.

Despite its proven clinical efficacy, cryoablation has limitations in achieving complete ablation due to factors such as tumor size and proximity to major blood vessels in lung adenocarcinoma cases. Leveraging the principle that cryoablation can activate the host’s antitumor immune response, we aimed to enhance the immunotherapeutic effects by combining cryoablation with Treg inhibitors. Regrettably, our experimental findings indicated that the use of Tregs inhibitors in isolation did not exert a significant effect. The combination drug regimen demonstrated only marginal improvements over the cryoablation-alone group. This suggests that the complex interplay between TGF-β signaling inhibition and immune modulation within the tumor microenvironment may have mitigated the direct effects of cryoablation and SB431542 on Treg levels. Such complexity highlights the need for a more detailed understanding of the interactions among cryoablation, Treg inhibitors, and the tumor microenvironment, to better elucidate these interactions and develop more effective strategies to enhance the immunotherapeutic benefits of cryoablation.

Nevertheless, some limitations of this study cannot be overlooked. To begin, while changes in the levels of IFN-γ and IL-2 were monitored post-cryoablation, our investigation did not detect the levels of serum cytokines or chemokines. Including these factors could provide a more holistic perspective on cryoablation-induced immunological changes. Secondly, the lack of data on tumor-infiltrating lymphocytes (TILs), additional chemokines, and the temporal evolution of immune cell populations following cryoablation restricted our understanding of the sustainability of the immune response. Thirdly, technical difficulties associated with acquiring post-cryoablation tissue samples impeded a spatial analysis of immune cells within the tumor microenvironment, which is crucial for elucidating the spatial dynamics of the immune response. To address these limitations, future research should aim to encompass a more exhaustive cytokine profile and conduct longitudinal analyses of immune cell populations. Such an approach will not only enhance our comprehension of the immunological effects of cryoablation but also contribute to the formulation of more precise and effective targeted cancer therapeutics.

Conclusions

Lung adenocarcinoma cryoablation emerges as a potent intervention capable of eradicating lung adenocarcinoma cells and reshaping the tumor microenvironment. Its influence on Treg differentiation mediated by the TGF-β/Smad signaling pathway represents a pivotal mechanism that fosters antitumor immune responses.