In the current study, gene expression analysis of CYTH4 was carried out using various public datasets and online platforms. The Human Protein Atlas (HPA) database (https://www.proteinatlas.org/ENSG00000100055-CYTH4/tissue, accessed on 14 October 2022) (22) was used to analyze the expression of CYTH4 in different healthy human tissues and cancer cell lines. The Cancer Cell Line Encyclopedia (CCLE) database (depmap.org/portal/interactive/, accessed on 15 October 2022) (23) was used to compare the expression of CYTH4 among various cancer types.

The Cancer Genome Atlas (TCGA) database (TCGA-LAML, https://portal.gdc.cancer.gov/, accessed on 26 October 2022) is a large public database containing both genome and clinical information spanning 33 cancer types (24). The Tumor Immune Estimation Resource version 2.0 (TIMER2.0) web resource (http://timer.cistrome.org/, accessed on 20 October 2022) (25) was used in the present study to compare the expression of CYTH4 between tumor and adjacent normal tissues from TCGA. The University of Alabama at Birmingham CANcer (UALCAN) data analysis portal (http://ualcan.path.uab.edu, accessed on 24 October 2022) (26) was used to visualize the expression data among different AML French-American-British (FAB) subtypes in TCGA database. The Gene Expression Omnibus (GEO) dataset GSE30029 (27) was adopted to compare the expression of CYTH4 between AML and healthy BM CD34⁺ cells.

TCGA database was used to investigate the survival significance of CYTH4 (24). A total of 151 AML samples from TCGA-LAML dataset with intact RNA sequencing data and survival status were included in the current study (24). Receiver Operating Characteristic (ROC) analysis was conducted, and the Youden index was calculated as the sum of sensitivity and specificity minus one. The expression level that achieves the maximum of the Youden index is referred to as the cut-off value to divide patients into the low- and the high-CYTH4 groups. Overall survival (OS) and event-free survival (EFS) were calculated using the Kaplan-Meier method and comparisons were carried out using the log-rank test. Univariate and multivariate survival analyses were performed using the Cox regression model and described with the hazard ratio (HR) and 95% confidence interval (CI). A stepwise forward procedure was used in the multivariate analysis. The datasets GSE10358 (28) and GSE14468 (29) from the GEO database were also used in the survival analysis.

DESeq2 package (Version 1.42.0, github.com/thelovelab/DESeq2) (30) in R software was used to screen differentially expressed genes between the low- and the high-CYTH4 groups in AML. Significantly differentially expressed genes were defined using an adjusted P<0.05 and |fold change (FC)|>2. Gene association analysis was carried out using LinkedOmics (http://www.linkedomics.org/login.php, accessed on 1 November 2022) (31), a publicly available portal for analyzing multi-omics data based on TCGA dataset. Pearson's correlation coefficient was calculated to search for CYTH4-associated genes. Significantly-associated genes were determined based on the criteria of False Discovery Rate (FDR)<0.05 and |r|>0.5.

GO enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were carried out using the Database for Annotation, Visualization and Integrated Discovery online tool (Version v2023q3, https://david.ncifcrf.gov/, accessed on 20 October 2023) (32). GSEA was performed using the Molecular Signatures Database (Version 2023.1, https://www.gsea-msigdb.org/gsea/index.jsp, accessed on 21 October 2023) (33). Genes interacting with CYTH4 were also investigated using STRING (Version 11.5, https://string-db.org/, accessed on 19 November 2022). The CIBERSORTx tool (https://cibersortx.stanford.edu/, accessed on 22 October 2023) (34) was used to compare the difference in immune cell infiltration between the low- and high-CYTH4 groups.

MV4-11, HL-60, THP-1, U-937, Kasumi-1, K-562, RS4;11 and 293T cells were purchased from American Type Culture Collection. HEL, Reh and MOLT-4 cells were purchased from the Chinese National Collection of Authenticated Cell Cultures. NOMO-1, MOLM-13, NB4, BALL-1, NALM-6, and SUP-B15 cells were purchased from Procell Life Science & Technology Co., Ltd. MV4-11, HL-60, THP-1, U-937, Kasumi-1, K-562, RS4;11, HEL, Reh, MOLT-4, NOMO-1, MOLM-13, NB4, BALL-1, NALM-6, and SUP-B15 leukemia cell lines were maintained in RPMI-1640 medium (VivaCell BIOSCIENCES), supplemented with 10% fetal bovine serum (FBS, TransGen Biotech Co., Ltd) and 1% penicillin-streptomycin (VivaCell BIOSCIENCES). 293T cells were maintained in DMEM medium (VivaCell BIOSCIENCES) supplemented with 10% FBS and 1% penicillin-streptomycin. Cells were cultured in a humidified incubator (Esco Lifesciences) with 5% CO₂ at a temperature of 37°C. All the cell lines were tested and authenticated by using short tandem repeat matching analysis. No mycoplasma contamination was detected.

Human CYTH4 was amplified from cDNA and cloned into the pLV3-EF1α-MCS-puro lentiviral construct (Wuhan MiaoLing Biotech Science Co., Ltd.). shRNA-targeting CYTH4 and non-targeting control were constructed using synthesized shRNA-encoded DNA oligos and cloned into the pLKO.1-puro vector (Addgene, Inc.). The designed target sequences were as follows: Scramble (TGAGGAAATTGCGGCTTATTT), shCYTH4 #1 (TRCN0000242587, CCGCCAAGGGTATCCAGTATT), shCYTH4 #2 (TRCN0000242586, TTGCACGGTTCCTGTATAAAG). The lentivirus was produced in 293T cells by transfecting the designed plasmid together with the packing vectors pLP/VSVG and psPAX2 (Addgene, Inc.). Cells were subsequently infected with lentiviral particles via two rounds of ‘spinoculation’ with 8 µg/ml polybrene.

The detection of mRNA expression level was carried out on day 2 after lentiviral infection to assess the overexpression and knockdown of CYTH4. Total RNA was isolated from cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. Isolated RNA was converted into cDNA using the TransScript^® All-in-One First-Strand cDNA Synthesis SuperMix for qPCR kit (One-Step gDNA Removal; cat. no. AT341; TransGen Biotech Co., Ltd.). The reaction was carried out by incubating the mixture at 42°C for 15 min, followed by inactivation at 85°C for 5 sec. The expression of CYTH4 was detected by qPCR using the KAPA SYBR^® Fast Universal kit (cat. no. KK4601; Sigma-Aldrich; Merch KGaA) on the ABI Prism 7500 sequence detection system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The process included 3 parts: initial denaturation at 95°C for 2 min, cycling stage (35 cycles) with denaturation at 95°C for 15 sec and annealing plus extension at 60°C for 1 min, and melt curve stage with 95°C for 15 sec, 60°C for 1 min, 95°C for 30 sec and 60°C for 15 sec. Expression of CYTH4 was determined by the comparative Cq method (35) using GAPDH for normalization. The following CYTH4 primer sequences were used: Forward, ATTGGGCGCAAGAAGTTCAAC; Reverse, TTTATACAGGAACCGTGCAATGT. The following GAPDH primer sequences were used: Forward, CTCTGCTCCTCCTGTTCGAC; Reverse, GCCCAATACGACCAAATCC.

Western blotting was performed on day 2 after lentiviral infection. Cells were lysed using RIPA buffer supplemented with 1 mM phenylmethane sulfonyl fluoride (both Beijing Solarbio LIFE SCIENCES). Total protein concentration was measured using a BCA assay kit (Solarbio LIFE SCIENCES). Equal amounts of protein (~30 µg) were separated by 12% SDS-PAGE and transferred onto polyvinylidene fluoride membranes. The membrane was blocked with 5% non-fat milk at room temperature for 1 h, then incubated with primary antibody CYTH4 (cat. no. H00027128-B01P; Novus Biologicals, LLC; Bio-Techne) at 4°C overnight for about 12 h. The incubation of HRP-linked secondary antibody (cat. no. 7076; Cell Signaling Technology, Inc.) was carried out at room temperature for 1 h. After detecting CYTH4, the membrane was washed with stripping buffer (Solarbio LIFE SCIENCES) at room temperature for 30 min and blocked again. It was then incubated with primary β-Actin (HRP conjugate; cat. no. 5125; Cell Signaling Technology, Inc.) antibody for 2 h at room temperature. The primary antibody was diluted at 1:1,000 and the secondary antibody 1:3,000. TBST buffer with 0.05% Tween 20 (Solarbio LIFE SCIENCES) was used for washing. The immobilon western chemiluminescent HRP substrate (cat. no. WBKLS0100; MilliporeSigma) was added to the membrane, and blot signals were detected using a ChemiDoc XRS+ System (Bio-Rad Laboratories, Inc.).

Cell proliferation was assessed using the Cell Counting Kit-8 (CCK8; MedChemExpress) assay according to the manufacturer's instructions. At 72 h after infection, 5,000 viable cells counted by Trypan blue staining were seeded into 96-well plates. After incubating the media with CCK-8 reagent for 3 h, the absorbance was measured at 450 nm using a Multiskan FC Microplate Photometer (Thermo Fisher Scientific, Inc.). The CCK-8 assay was performed at the same time for 5 consecutive days. At 72 h after infection, cells were harvested and fixed with 75% ice-cold ethanol at 4°C for about 12 h. Cell cycle analysis was conducted using a Fluorescence-activated cell sorting (FACS) flow cytometer (Beckman Coulter, Inc.) after staining the samples with propidium iodide (PI) for 30 min. At 96 h after infection, apoptosis was detected using Annexin V-fluorescein isothiocyanate (FITC)/PI apoptosis detection kit (Dojindo Molecular Technologies, Inc.) according to the manufacturer's instructions. Briefly, cells were stained with Annexin V-FITC and PI at room temperature for 15 min and analyzed with a FACS flow cytometer (Beckman Coulter, Inc.). In FACS analysis, 10,000 cells were gated for cell cycle and apoptosis detection. Regarding the colony formation assay, cells were harvest at 72 h after infection with scramble or shCYTH4 lentivirus. Variable cells were seeded in methylcellulose medium (MethoCult™ H4434, Stemcell Technologies, Inc.) at a density of 1,000 cells/ml. Colonies (≥50 cells) were counted manually after 10 days.

SPSS (version 22.0; IBM Corp.) and GraphPad Prism (version 7; Dotmatics) were used for statistical analyses. Clinical features between the two groups were compared using the χ² test for categorical variables and the Fisher's exact test for the expected frequency of an event (<5 in any cell of 2×2 tables). Continuous variables were compared using the non-parametric Mann-Whitney U test. One-way ANOVA was used to compare scramble cells and CYTH4-knockdown cells, with scramble cells serving as the control. Dunnett's multiple comparison test was used for testing. Two-way ANOVA followed by Sídák's multiple comparisons test were used to compare the cell viability between different groups on each day. Data are presented as mean ± standard deviation. P<0.05 was considered to indicate a statistically significant difference.

The present study focused on CYTH4 because its expression was much higher than other cytohesins in AML (Fig. S1). The HPA dataset was first explored to examine the RNA tissue specificity of CYTH4 in healthy humans. Results showed that the expression of CYTH4 was enhanced in the BM and lymphoid tissues, while being expressed at low levels in other tissues (Fig. 1A).

The level of CYTH4 expression was then analyzed in cell lines based on the latest next-generation sequencing data from the CCLE dataset (Table SI). Results showed that CYTH4 was expressed at high levels in lymphoma and leukemia cell lines, followed by thyroid cancer (Fig. 1B). The analysis of the HPA dataset also revealed that CYTH4 was expressed at high levels in myeloid cancer cells compared with other cancer cell lines such as brain, liver and kidney cancer cell lines (Fig. S2). In addition, the data from the CCLE leukemia cell lines were used to compare the expression level of CYTH4 in different types of leukemia, and it was found that CYTH4 was expressed at high levels in AML compared with acute lymphoblastic leukemia (ALL) and chronic myeloid leukemia (CML; Fig. 1C). The AML cell lines NOMO-1, MV4-11, HL-60, THP-1, MOLM-13, Kasumi-1, HEL, NB4 and U-937, the ALL cell lines RS4;11, BALL-1, Reh, NALM-6, MOLT-4 and SUP-B15, and the CML cell line K-562 were used in the present study to examine the expression of CYTH4. RNA was extracted from these cell lines and qPCR analysis was performed. Results showed that CYTH4 was expressed at high levels in NOMO-1, MV4-11, HL-60, THP-1 and MOLM-13 AML cell lines (Fig. 1D).

To investigate CYTH4 expression in human cancers, TCGA-LAML dataset was analyzed. Fig. 2A displays an overview of the different expression of CYTH4 between tumors and adjacent normal tissues across TCGA dataset, suggesting that CYTH4 expression was higher in patients with AML compared with that in patients with other tumors. These results for the AML samples matched those obtained from the cell lines (Fig. 1B). Analysis of the GSE30029 dataset showed CYTH4 expression was significantly upregulated in AML BM CD34⁺ cells compared with that in normal BM CD34⁺ cells (Fig. 2B). The French-American-British (FAB) classification system divides AML into 8 subtypes, designated Myeloid 0–7 (M0-M7), based on the morphology and the appearance of the leukemia cells. We compared the CYTH4 expression among different AML subtypes and patients with M3-AML had the lowest expression of CYTH4 (Fig. 2C).

To investigate the significance of CYTH4 expression in AML prognosis, survival was compared between the high- and low-CYTH4 expression groups in the different datasets. ROC analysis was performed to determine the cut-off value between the low- and high-CYTH4 expression groups (Fig. 3A). As shown in Fig. 3A and B, high CYTH4 expression was significantly associated with unfavorable OS (high vs. low; HR=2.19; 95% CI, 1.40–3.44; P=0.0006) and EFS (high vs. low; HR=2.32; 95% CI, 1.43–3.75; P=0.0006). This conclusion was validated in the GSE10358 dataset (Fig. 3C) and in the GSE14468 dataset (Fig. 3D). Next, survival was compared between the low- and high-CYTH4 expression groups by treatment using TCGA dataset. The results showed that high CYTH4 expression was associated with poor OS (Fig. 3E; high vs. low; HR=3.12; 95% CI, 1.82–5.34; P<0.0001) in patients treated with chemotherapy alone. However, in cases of patients who received both chemotherapy and transplantation, no significant difference was found (Fig. 3F; P=0.398). Survival was then compared between patients treated with chemotherapy alone and patients treated with chemotherapy plus transplantation grouped by CYTH4 expression level. Transplantation did not show a significant difference in the OS of the low-CYTH4 group (Fig. 3G; P=0.974), but significantly improved OS in the high-CYTH4 expression group compared with chemotherapy alone (Fig. 3H; transplantation vs. chemotherapy; HR=0.29; 95% CI, 0.18–0.47; P<0.0001). These findings suggest that transplantation may attenuate the adverse effect of high CYTH4 expression on patient survival in AML.

Based on the cut-off value in Fig. 3A, the characteristics of patients in the low- and the high-CYTH4 group were analyzed (Table SII). Both clinical features and gene mutations were listed in Table I. It was observed that the white blood cell count (WBC) varied significantly between the two groups, with patients in the high-CYTH4 group exhibiting higher WBC than those in the low-CYTH4 group (median WBC, 25.9 vs. 9.7; P=0.004). Moreover, in the M4-AML subtype, there were more patients with high CYTH4 expression than patients with low expression (P=0.001), while all patients with M3-AML were in the low-CYTH4 expression group (P<0.0001). This discovery is in line with the previous result that patients with M3-AML had the lowest CYTH4 expression (Fig. 2C). The low-CYTH4 group had higher percentages of cases with PML-RARA and RUNX1-RUNX1T1 karyotypes (P<0.0001). Regarding risk status, low expression of CYTH4 was significantly associated with a good-risk status (low vs. high, 58.8 vs. 9.4; P<0.0001). Besides, patients in the high-CYTH4 group tended to be older than those in the low-CYTH4 group (57 vs. 51 years; P=0.071). Patients in the high-CYTH4 group had a higher percentage of RUNX1 mutation (12% vs. 0; P=0.034). The percentage of patients with more than one mutation did not differ significantly between the low- and high-CYTH4 groups. No significant difference was found between the low- and high-CYTH4 group concerning sex, and BM and peripheral blood (PB) blasts.

To further explore the prognostic effect of CYTH4 in AML, univariate and multivariate survival analyses were performed using the Cox regression model (Table II). In the univariate analysis, high CYTH4, older age, high WBC, poor cytogenetics risk, FLT3, DNMT3A and TP53 mutations, and non-transplantation were associated with poor OS. High CYTH4, WBC, PB blasts, poor cytogenetics risk and DNMT3A mutation were identified as inferior prognostic factors for EFS. When age, WBC, cytogenetics risk, FLT3, DNMT3A and TP53 mutations, and transplantation were combined in the multivariate Cox regression analysis, results confirmed that high expression of CYTH4 was independently associated with inferior OS (HR=2.49; 95% CI, 1.28–4.83; P=0.007) and EFS (HR=2.56; 95% CI, 1.48–4.42; P=0.001).

To better understand the role of CYTH4 in AML, the transcriptomes were compared between the high- and the low-CYTH4 groups in TCGA dataset. A total of 552 genes showed significantly different expression (adjusted P<0.05; |FC|>2) including 394 and 158 genes significantly upregulated and downregulated in the high-CYTH4 group, respectively (Fig. 4A; Table SIII). LinkedOmics tools were then used to perform correlation analysis, and a total of 451 significantly co-expressed genes with a cut-off value of FDR<0.05 and |r|>0.5 were identified (Fig. 4B; Table SIV). Of these co-expressed genes, 326 and 125 genes were positively and negatively correlated with CYTH4 expression, respectively (Table SIV). By integrating the results of these two analyses, 159 genes were identified that were upregulated in the high-CYTH4 group and positively correlated with CYTH4 expression (Fig. 4C). By contrast, only 17 genes were found to be both downregulated in the high-CYTH4 group and negatively correlated with CYTH4 expression (Fig. 4D). The overlapping genes were further analyzed for their biological functions.

Next. the possible biological function of CYTH4 was explored. GO and KEGG pathway enrichment analyses were performed (Fig. 5A and B). These overlapping genes were significantly associated with ‘innate immune response’, ‘inflammatory response’, ‘signal transduction’, ‘apoptotic process’, ‘phagosome’ and ‘allograft rejection’. GSEA and the enrichment plot showed that these overlapping CYTH4-associated genes were significantly enriched in the gene set related to the immune response (Fig. 5C). Additionally, STRING analysis was used to investigate the genes that interacted with CYTH4 (Fig. 5D). Then, the fractions of 22 distinct immune cell types were estimated using the CIBERSORTx algorithm. Results showed that high CYTH4 expression was significantly correlated with CD4⁺ memory T cells resting (P<0.01), monocytes (P<0.0001) and mast cells resting (P<0.0001; Fig. 5E).

To further investigate the function of CYTH4 in AML, in vitro validation was carried out. CYTH4 knockdown was conducted in the AML cell lines MOLM-13, NOMO-1 and THP-1. These cell lines were chosen because they exhibit relatively high expression of CYHT4 (Fig. 1D). Lentivirus-expressing shRNA significantly reduced the mRNA and protein expression levels of CYTH4 (Fig. 6A; Fig. S3). Cell proliferation analysis showed that CYTH4 knockdown significantly suppressed the cell growth of AML cells (Fig. 6B). Cell cycle assays revealed a significant G0/G1 phase arrest in all three AML cell lines (Fig. 6C). The results of the colony-forming assays demonstrated that the silencing of CYTH4 significantly impaired the clonogenic potential of the three leukemia cell lines (Fig. 6D). Increased apoptosis was also recorded in the leukemia cell lines following transfection with CYTH4 shRNA (Fig. 6E). Furthermore, CYTH4 was overexpressed by lentivirus in the U-937 (acute monocytic leukemia) and Kasumi-1 (acute myeloblastic leukemia with maturation) AML cell lines (Fig. S4A and B), in which the CYTH4 expression was low (Fig. 1D). Results showed that the overexpression of CYTH4 enhanced the cell growth of the U-937 and Kasumi-1 cell lines (Fig. S4C). Taken together, these results indicated that CYTH4 plays an oncogenic role in AML cells.