Abstract
Clinical management of acute myeloid leukemia (AML) poses significant challenges due to its poor prognosis and heterogeneous nature. Discovering new biomarkers is crucial for improving risk assessment and customizing treatment approaches. While leukocyte-specific transcript 1 (LST1) is implicated in inflammation and immune regulation, its function in AML remains ambiguous. In this investigation, we conduct a comprehensive investigation into LST1 expression profiles, clinical implications, functional pathways, and immune interactions in AML, leveraging multi-omics data and experimental validations. Our examination shows increased levels of LST1 expression in AML when compared to regular hematopoietic tissues, a discovery validated by RT-qPCR and Western blot analyses in a separate group. Elevated LST1 levels correlate with distinct clinicopathological features, including increased white blood cell counts, non-M3 FAB subtype, and intermediate/poor cytogenetic risk. Importantly, heightened LST1 levels predict unfavorable overall survival outcomes across various subgroups, independently of age and cytogenetic risk. We develop an integrative nomogram incorporating LST1 expression, demonstrating robust prognostic efficacy for patient survival. Transcriptomic profiling identifies 275 differentially expressed genes between LST1-high and -low AML cases, enriched in cytokine signaling, immune modulation, cell adhesion, and oncogenic pathways. Furthermore, LST1 exhibits significant associations with the infiltration of diverse immune cell subsets within the AML microenvironment, particularly myeloid cells and regulatory T cells (Tregs). In conclusion, our study establishes LST1 as a novel prognostic indicator with immunological relevance in AML, emphasizing its potential therapeutic implications. Further mechanistic elucidation of LST1 in AML pathogenesis is crucial for its clinical translation.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12672-025-02082-5.
Keywords: LST1, AML, Biomarker, Immune infiltration, Prognosis
Introduction
Acute myeloid leukemia (AML) is marked by the existence of abnormal myeloid progenitor cell proliferation, which gives rise to a wide variety of hematologic malignancies, impairing the differentiation and uncontrolled proliferation [1, 2]. Despite progress in treatment and support, the prognosis for AML patients stays bleak, with a mere 29.5% survival rate after 5 years [3]. This grim reality is attributed to the considerable genetic and biological heterogeneity intrinsic to AML, leading to variable treatment responses and increased susceptibility to relapse[4, 5].While existing risk stratification methods incorporate cytogenetic and molecular markers, their prognostic accuracy is limited, underscoring the urgent need for innovative biomarkers capable of more precisely predicting clinical outcomes and guiding personalized therapeutic approaches.
Leukocyte-specific transcript 1 (LST1), also called B144, encodes a transmembrane protein predominantly expressed in hematopoietic lineage cells [6]. Recent investigations highlight the crucial function of LST1 in controlling the immune cell migration and stimulation, thereby shaping the tumor microenvironment [7]. Aberrant LST1 expression has been implicated in various cancers, encompassing bladder cancer, germline tumor, and B-cell acute lymphoblastic leukemia, where it correlates with disease progression and unfavorable outcomes [8–10]. However, the association between LST1 expression and AML prognosis remains largely unexplored. Exploring the intersection of LST1 expression and immune cell infiltration could offer a valuable understanding of the pathobiology of AML and aid in pinpointing potential therapeutic targets, considering the substantial impact of the immune system on AML development and response to treatment.
This research suggests that the levels of LST1 expression could be connected with the immunological cell presence in the AML tumor microenvironment, potentially acting as a dependable predictive marker for patients. Using transcriptomic data from The Cancer Genome Atlas (TCGA) and additional datasets, we aim to ascertain the link between LST1 expression and the composition of immune cells, along with its predictive value for overall survival (OS) and other clinicopathological factors in patients with AML. Our plan is to use a thorough bioinformatics strategy, including analyzing functional enrichment and examining protein–protein interaction (PPI) networks, to unveil the molecular mechanisms involved in the LST1 impact on AML development and immune regulation. Our study holds promise for developing a novel predictive model incorporating LST1 levels, thereby enhancing risk stratification and facilitating the design of personalized immunotherapeutic strategies for AML patients.
Materials and methods
RNA seq data acquisition and bioinformatics analysis
Utilizing the UCSC XENA platform [11], the Pan-cancer RNA-seq datasets from TCGA and Genotype-Tissue Expression (GTEx) were obtained and standardized with the Toil help. The Level 3 HTSeq-FPKM and HTSeq-Count data for AML specimens were acquired from the TCGA data repository for further examination. All procedures conducted in this study adhered to the established guidelines outlined by TCGA and GTEx. Bone marrow specimens were collected from 5 newly diagnosed acute myeloid leukemia (AML) patients and 5 healthy donors at the Department of Hematology, Anqing Municipal Hospital. This study was approved by the ethics committee of the Anqing Municipal Hospital (ID: No. 2024-72). All the enrolled patients signed an informed consent form. The study was conducted in accordance with the Declaration of Helsinki.
Differential gene expression analysis
The DESeq 2 R package was used to evaluate differences in LST1 expression patterns in AML samples (HTSeq-Count), with a 50% threshold used to distinguish between low and high levels of expression [12]. This analysis aided in identifying genes that exhibited differential expression (DEGs). Following this, a visual representation in the form of a heat map was generated to visually illustrate the top 10 differently expressed genes (DEGs).
Functional enrichment analysis
Genes that met the criteria of having an absolute value of logFC > 2 and a padj less than 0.05 were chosen for analysis of functional enrichment. The ClusteProfiler package in R was employed to execute an examination of cellular component (CC), molecular function (MF), and biological process (BP) categories in Gene Ontology (GO), along with pathway analysis from the Kyoto Encyclopedia of Genes and Genomes (KEGG) [13].
The process of gene set enrichment analysis (GSEA)
GSEA involves the analysis of gene sets to determine enrichment levels. We used the R software ClusteProfiler (v3.14.3) to analyze the variations in functions and pathways between LST1 high and low-expression groups employing GSEA [13]. Each analysis included permuting the gene set 1,000 times. Significant outcomes were determined based on adjusted P < 0.05 and False Discovery Rate (FDR) q < 0.25.
Immune cell infiltration analysis
The infiltration patterns of immune cells were evaluated using single-sample gene set enrichment analysis (ssGSEA) through the GSVA R package. Following the methodology outlined by Bindea et al. [14], we calculated enrichment scores for 24 distinct immune cell populations based on their characteristic gene expression profiles. The correlation between LST1 expression levels and immune cell enrichment scores was determined using Spearman's correlation analysis.
Assessment of using single-sample gene set enrichment analysis (ssGSEA)
The connection between immune cell infiltration and LST1 expression was evaluated by performing ssGSEA in R (version 3.6.3) with the GSVA package. We utilized a detailed array of 24 unique infiltrating immune cell types, as outlined in a previous publication [14]. To ascertain the connection between LST1 expression levels and the enrichment scores of 24 different immune cells types, a Spearman correlation analysis was applied. Enrichment scores were contrasted between groups with high and low LST1 expression levels using the Wilcoxon rank-sum test.
Analyzing PPI networks
The Search Tool for the Retrieval of Interacting Genes (STRING) database was utilized to predict the interaction network of genes with altered expression levels [15]. A threshold of 0.4 for the interaction score was used as the cutoff point. Afterward, the PPI network was displayed with Cytoscape (version 3.7.1) [16]. The PPI network was analyzed employing MCODE (version 1.6.1) to detect crucial components [17], considering MCODE scores higher than 5, a minimum degree of 2, a node score cutoff of 0.2, a maximum depth of 100, and a k-score of 2. Pathway and process enrichment analysis of the identified modules were conducted employing Metascape.
Predictive model development and prediction
With the utilization of the RMS R package (version 5.1-3), a nomogram was created to customize the forecast of OS in individuals diagnosed with AML. This nomogram incorporated essential clinical characteristics and was accompanied by calibration plots to assess its performance. Calibration curves were generated through plotting the nomogram-projected likelihood versus actual rates, with the ideal predictive values depicted by the 45° diagonal line. We assessed the nomogram's ability to discriminate by calculating the C-index with 1000 bootstrap samples. Additionally, we compared the nomogram's predictive performance with single prognostic factors through C-index and receiver operating characteristic (ROC) analysis. Statistical analyses were executed employing a two-tailed method, with a significance level set at 0.05.
RT-qPCR for quantitative reverse transcription
Cells were employed to isolate total RNA with Trizol reagents depending on the manufacturer’s recommendations. Following this, 1 µg of mRNA underwent reverse transcription to cDNA with a high-capacity cDNA kit as per the manufacturer's recommendations. The recovered cDNA was subsequently amplified employing qPCR with the SYBR Premix Ex Taq kit following the manufacturer's recommendations and examined on the ABI7300 Sequence Detection System. The qPCR procedure commenced with a denaturation phase lasting 10 min at 95 °C, followed by 40 cycles of 15-s denaturation at 95 °C and 1-min annealing/extension at 60 °C. Every test was done three times, and the data was analyzed using the 2^ − ΔΔCT technique. For qRT-PCR analysis, the following primers were utilized: human LST1 (Forward: 5′- TCAGAGCAGGAACTCCACTAT -3′, Reverse: 5′- CAGCAATGCAGGCATAGTC -3′) and human GAPDH (Forward: 5′- TCAAGAAGGTGGTGAAGCAGG -3′, Reverse: 5′- TCAAAGGTGGAGGAGTGGGT—3′).
Western blot analysis
Cellular proteins were extracted utilizing a cold RIPA lysis buffer, and their quantities were estimated employing the BCA Protein Assay Kit. Then, 30 µg of proteins were extracted employing 10% SDS-PAGE gels and subsequently put onto 0.45-mm PVDF membranes (Millipore, USA). After a 2-h period, during which 5% non-fat milk at the ambient temperature was employed, the membranes were then incubated overnight at 4 °C with primary antibodies. Subsequent to washing the membranes three times with TBST buffer, they were then exposed to secondary antibodies linked to HRP for 1 h at ambient temperature. The ECL detection system was utilized to visualize particular bands. The primary antibodies utilized were anti-LST1 (Abcam, 21361-1-AP, 1:2000) and anti-GAPDH (Proteintech, AB-P-R 001, 1:1000).
Statistical analysis
The statistical analyses and figures were applied using R software, specifically version 3.6.2. The expression levels of LST1 in unpaired specimens were examined employing the Wilcoxon rank-sum test, whereas, for paired specimens, the Wilcoxon signed-rank test was deployed. The connection between clinical and cytogenetic features and LST1 expression was assessed using statistical methods, including the Kruskal–Wallis and Wilcoxon signed-rank tests and logistic regression analysis. Cox regression analysis and the Kaplan–Meier technique were employed to assess predictive factors like the LST1 expression level, with multivariate Cox analysis comparing its impact on survival to different clinical features. The median LST1 expression level served as the cut-off value. All statistical analyses were applied with a significance level set at P < 0.05. Furthermore, the pROC tool was employed for ROC analysis to evaluate how well the expression of the LST1 gene can distinguish between samples from individuals with AML and those who are healthy. The AUC score, which falls within the range of 0.5 to 1.0, reflects the capacity to differentiate between 50 and 100%.
Results
Evaluation of LST1 expression across diverse cancer types and AML
RNA-seq information was obtained from the UCSC XENA repository in TCGA and GTEx variations and standardized through the Toil framework. Analyzing LST1 expression in normal specimens from TCGA and GTEx databases in contrast to cancer samples from TCGA showed a notable rise in LST1 levels across 23 various cancer types, such as AML (Fig. 1A). LST1 expression was found to be significantly upregulated in AML compared to normal hematopoietic tissues (Fig. 1B).
Fig. 1.
Expression of LST1 in acute myeloid leukemia (AML) patients. A Comparative examination of LST1 expression levels in various cancer tissues of AML patients versus normal tissues sourced from TCGA. B AML patients exhibit increased levels of LST1 in comparison to normal tissues. C Validation of LST1 overexpression in AML using the GSE65409 dataset. D Quantitative LST1 mRNA expression via qRT-PCR in AML patient specimens (n = 5) versus controls (n = 5). E, F Western blot was executed to measure the LST1 protein levels in individuals with AML. Significance levels: *p < 0.05; **p < 0.01; ***p < 0.001
The association between LST1 and AML was further supported by an analysis of LST1 expression and its prognostic implications, utilizing data from GSE65409 within the Gene Expression Omnibus (GEO) repository. Consistent with TCGA observations, AML specimens exhibited elevated LST1 expression, with heightened LST1 levels significantly correlating with adverse prognosis among AML patients (refer to Fig. 1C). To further validate the upregulation of LST1 in AML, we analyzed bone marrow specimens from five healthy donors and five AML patients, revealing consistently elevated expression at both transcriptional (RT-qPCR, Fig. 1D) and translational levels (Western blot analysis, Fig. 1E, F).
Key findings from transcriptomic and immune infiltration analyses
Differential gene expression analysis identified 275 genes exhibiting significant expression changes between LST1-high and -low AML cases, with 142 genes upregulated and 133 genes downregulated (Fig. 2A, B). Functional enrichment analyses revealed that these differentially expressed genes were involved in pathways related to cytokine signaling, immune modulation, cell adhesion, and oncogenesis (Fig. 3A, B, Supplementary Table 1).
Fig. 2.
Detection of genes with differential expression (DEGs) in high and low LST1 expression categories. A Volcano plot displaying DEGs, including 142 elevated and 133 mitigated genes. Normalized expression levels are shown on a scale ranging from green to red. B Heat map showing 10 DEGs, with 5 genes being elevated and 5 genes being declined. Specimens are represented on the X-axis, with DEGs indicated on the Y-axis. Green and red tones indicate declined and raised genes, respectively
Fig. 3.
Enrichment analysis of DEGs between high and low LST1 expression in AML patients from TCGA. Examining differentially expressed genes (DEGs) employing the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) approaches. Performing Gene Set Enrichment Analysis (GSEA) on DEGs
Examination of the relationship between LST1 expression and immune cell infiltration in the AML microenvironment using ssGSEA showed significant positive correlations with myeloid cells (e.g., MDSCs) and regulatory T cells (Tregs), while negative correlations were observed with cytotoxic immune cells such as NK and CD8 + T cells (Fig. 4A, B). These findings suggest that LST1 overexpression may contribute to shaping an immunosuppressive milieu in AML, potentially promoting leukemia cell survival and immune evasion.
Fig. 4.
Association of LST1 Expression with Immune Infiltration in the AML Microenvironment. A The relationship between the enrichment score of various tumor-infiltrating immune cells and the LST1 expression level is measured in Transcripts Per Million (TPM). B The connection between LST1 gene expression and the percentage of 24 distinct immune cell types. Each dot represents the absolute value of Spearman's correlation coefficient in terms of its magnitude
STRING analysis was employed to explore protein–protein interactions between LST1 and its co-expressed genes, identifying 275 significant DEGs (|logFC|> 2, P < 0.05). The resulting PPI network, visualized using Cytoscape-MCODE, contained 210 nodes and 247 edges (Fig. 5A). The most significant module showed an MCODE score of 8.133, with 12 nodes and 52 edges (Fig. 5B), suggesting a densely interconnected functional network involving LST1 in AML.
Fig. 5.
The main module of the Protein–Protein Interaction (PPI) Network involving DEGs associated with LST1. A The PPI network of differentially expressed genes (DEGs) was assembled employing Cytoscape. B The predominant module extracted from the PPI network comprises 12 nodes and 52 edges
Correlation of LST1 gene expression with clinical characteristics and cytogenetic risk factors
The primary clinical features of AML in TCGA are presented in Table 1. The 151 instances (68 females and 83 males) examined in this investigation had a mean age of 56.7 years. Out of the AML patients, 75 cases (49.3%) had decreased LST1 expression levels, with the remaining 76 cases (50.3%) exhibiting elevated levels. Strong correlations were found between LST1 levels and cytogenetics, FAB classification, and white blood cell (WBC) count through correlation analysis with a significance level of less than 0.001. Additionally, LST1 expression demonstrated significant associations with different factors, encompassing cytogenetic risk (P = 0.008) and NPM1 mutation (P = 0.014).
Table 1.
Evaluation of the correlation between LST1 gene expression and clinical characteristics in acute myeloid leukemia (AML) specimens from the TCGA database
| Characteristic | Low expression of LST1 | High expression of LST1 | p |
|---|---|---|---|
| n | 75 | 76 | |
| Gender, n (%) | 0.457 | ||
| Female | 31 (20.5%) | 37 (24.5%) | |
| Male | 44 (29.1%) | 39 (25.8%) | |
| Race, n (%) | 0.154 | ||
| Asian | 0 (0%) | 1 (0.7%) | |
| Black or African American | 9 (6%) | 4 (2.7%) | |
| White | 64 (43%) | 71 (47.7%) | |
| Age, n (%) | 0.114 | ||
| ≤ 60 | 49 (32.5%) | 39 (25.8%) | |
| > 60 | 26 (17.2%) | 37 (24.5%) | |
| WBC count(× 10^9/L), n (%) | < 0.001 | ||
| ≤ 20 | 49 (32.7%) | 28 (18.7%) | |
| > 20 | 25 (16.7%) | 48 (32%) | |
| BM blasts (%), n (%) | 0.665 | ||
| ≤ 20 | 28 (18.5%) | 32 (21.2%) | |
| > 20 | 47 (31.1%) | 44 (29.1%) | |
| PB blasts (%), n (%) | 1.000 | ||
| ≤ 70 | 36 (23.8%) | 36 (23.8%) | |
| > 70 | 39 (25.8%) | 40 (26.5%) | |
| Cytogenetic risk, n (%) | 0.018 | ||
| Favorable | 21 (14.1%) | 10 (6.7%) | |
| Intermediate | 33 (22.1%) | 49 (32.9%) | |
| Poor | 21 (14.1%) | 15 (10.1%) | |
| FAB classifications, n (%) | < 0.001 | ||
| M0 | 10 (6.7%) | 5 (3.3%) | |
| M1 | 18 (12%) | 17 (11.3%) | |
| M2 | 22 (14.7%) | 16 (10.7%) | |
| M3 | 15 (10%) | 0 (0%) | |
| M4 | 4 (2.7%) | 25 (16.7%) | |
| M5 | 2 (1.3%) | 13 (8.7%) | |
| M6 | 2 (1.3%) | 0 (0%) | |
| M7 | 1 (0.7%) | 0 (0%) | |
| Cytogenetics, n (%) | < 0.001 | ||
| Normal | 29 (21.5%) | 40 (29.6%) | |
| + 8 | 8 (5.9%) | 0 (0%) | |
| del(5) | 0 (0%) | 1 (0.7%) | |
| del(7) | 4 (3%) | 2 (1.5%) | |
| inv(16) | 0 (0%) | 8 (5.9%) | |
| t (15;17) | 11 (8.1%) | 0 (0%) | |
| t (8;21) | 7 (5.2%) | 0 (0%) | |
| t (9;11) | 0 (0%) | 1 (0.7%) | |
| Complex | 12 (8.9%) | 12 (8.9%) | |
| FLT3 mutation, n (%) | 0.247 | ||
| Negative | 53 (36.1%) | 49 (33.3%) | |
| Positive | 18 (12.2%) | 27 (18.4%) | |
| IDH1 R132 mutation, n (%) | 0.077 | ||
| Negative | 64 (43%) | 72 (48.3%) | |
| Positive | 10 (6.7%) | 3 (2%) | |
| IDH1 R140 mutation, n (%) | 0.819 | ||
| Negative | 68 (45.6%) | 69 (46.3%) | |
| Positive | 5 (3.4%) | 7 (4.7%) | |
| IDH1 R172 mutation, n (%) | 0.238 | ||
| Negative | 71 (47.7%) | 76 (51%) | |
| Positive | 2 (1.3%) | 0 (0%) | |
| RAS mutation, n (%) | 0.063 | ||
| Negative | 73 (48.7%) | 69 (46%) | |
| Positive | 1 (0.7%) | 7 (4.7%) | |
| NPM1 mutation, n (%) | 0.008 | ||
| Negative | 65 (43.3%) | 52 (34.7%) | |
| Positive | 9 (6%) | 24 (16%) |
Logistic regression was deployed to ascertain the connection between AML clinicopathological factors and the binary expression levels of LST1. As a result, there was a clear correlation between elevated levels of LST1 and an increase in WBC count over 20 × 10^9/L (with an odds ratio [OR] of 3.099; P < 0.001) and intermediate/normal cytogenetic risk (with an OR of 2.824; P = 0.020), as seen in Table 2. The Wilcoxon Rank Sum test was deployed to ascertain variations in LST1 expression levels among subjects exhibiting distinct clinicopathological features. The findings revealed that LST1 was significantly overexpressed in patients aged over 60, with WBC counts exceeding 20 × 10^9/L, non-M3 FAB classification, intermediate and poor cytogenetic risk, positive RAS mutation, and positive NPM1 mutation (refer to Fig. 6A–F). However, LST1 expression showed no significant associations with bone marrow blast percentage, peripheral blood blast percentage, FLT3 mutation status, IDH1 R132 mutation status, gender, or race (Supplementary Fig. 1). These comprehensive analyses suggest that LST1 expression is selectively associated with specific clinical and molecular features in AML.
Table 2.
Logistic regression analysis investigating the association between clinicopathological factors of AML and the expression of LST1
| Characteristics | Total (N) | OR (95% CI) | P value |
|---|---|---|---|
| Gender (Male vs. Female) | 150 | 0.685 (0.359–1.307) | 0.251 |
| Race (White vs. Asian&Black or African American) | 149 | 1.938 (0.617–6.086) | 0.257 |
| Age (> 60 vs. ≤ 60) | 150 | 1.835 (0.952–3.538) | 0.047 |
| WBC count(× 10^9/L) (> 20 vs. ≤ 20) | 149 | 3.099 (1.588–6.046) | < 0.001 |
| BM blasts (%) (> 20 vs. ≤ 20) | 150 | 0.756 (0.392–1.458) | 0.404 |
| PB blasts (%) (> 70 vs. ≤ 70) | 150 | 0.948 (0.499–1.800) | 0.870 |
| Cytogenetic risk (Intermediate/normal vs. Favorable) | 112 | 2.824 (1.175–6.787) | 0.020 |
| FLT3 mutation (Positive vs. Negative) | 146 | 1.452 (0.715–2.950) | 0.302 |
| IDH1 R132 mutation (Positive vs. Negative) | 148 | 0.270 (0.071–1.026) | 0.055 |
| IDH1 R140 mutation (Positive vs. Negative) | 148 | 1.400 (0.423–4.629) | 0.581 |
| RAS mutation (Positive vs. Negative) | 149 | 7.515 (0.901–62.679) | 0.043 |
| NPM1 mutation (Positive vs. Negative) | 149 | 2.831 (1.237–6.477) | 0.014 |
Bold values indicate statistical significance (P < 0.05)
Fig. 6.
Correlation between LST1 expression and clinical characteristics. Information is provided on A Age; B White Blood Cell (WBC) counts (20 × 10^9); C French-American-British (FAB) classification; D Cytogenetics risk; E RAS mutation; F NPM1 mutation. Significance levels: *p < 0.05; **p < 0.01; ***p < 0.001
Effect of elevated LST1 levels on AML outlook in individuals with varied clinicopathological conditions
Kaplan–Meier analysis was applied to ascertain the connection between LST1 expression and prognosis in AML patients. Those who had high LST1 expression possessed a considerably poorer prognosis contrasted with those with low LST1 expression (Fig. 7A). The hazard ratio (HR) was 2.029 (1.35–3.22), indicating a higher risk. The p-value was 0.001, indicating statistical significance. Subgroup analyses showed that higher levels of LST1 expression were linked to worse outcomes in specific subgroups defined by Age ≤ 60 (P = 0.011), Caucasian ancestry (P = 0.004), WBC counts ≤ 20 × 10^9/L (P < 0.001), PB blasts > 20% (P = 0.007), BM blasts > 70% (P = 0.002), FAB classification-M4 (P = 0.018), Intermediate Cytogenetic risk (P = 0.016), FLT3 mutation-positive (P = 0.019), NPM1 mutation-negative (P = 0.003), and RAS mutation-negative (P = 0.001) (see Fig. 7B–K).
Fig. 7.
Elevated LST1 expression linked to inferior overall survival (OS) in AML patients. Kaplan–Meier survival curves showing OS in A all acute myeloid leukemia (AML) patients and in patients who are B 60 years old or younger, C of Caucasian descent, D have a white blood cell count of less than or equal to 20 × 10^9/L, E bone marrow blasts exceeding 20%, F peripheral blood blasts exceeding 70%, G classified as FAB-M4, H with intermediate cytogenetic risk, I positive for FLT3 mutation, J negative for NPM1 mutation, and K negative for RAS mutation
Furthermore, the forest plot depicted the predictive significance of LST1 across different AML subcategories through univariate and multivariate Cox regression analyses, corroborating the aforementioned findings (refer to Fig. 8A, B).
Fig. 8.
Univariate and multivariate analysis of clinicopathological factors correlating with OS in AML patients. A Forest plot according to univariate Cox analysis for OS. B Forest plot according to multivariate Cox analysis for OS
Further scrutiny via univariate Cox proportional hazards regression revealed that high LST1 expression, poor cytogenetic risk, intermediate cytogenetic risk, and age over 60 were all significant predictors of worse OS. A multivariate Cox regression analysis was conducted to include cytogenetic risk, age, and LST1. The findings indicated that people above the age of 60 (P < 0.001), with unfavorable or moderate cytogenetic risk (P = 0.006 and P = 0.03, respectively), and elevated LST1 levels (P = 0.028) were autonomous prognostic elements associated with poorer survival rates (P < 0.05) (as shown in Table 3).
Table 3.
Univariate and multivariate cox regression analysis of factors influencing overall survival (OS) in AML patients
| Characteristics | Total(N) | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|---|
| Hazard ratio (95% CI) | P value | Hazard ratio (95% CI) | P value | ||
| Age | 140 | ||||
| ≤ 60 | 79 | Reference | |||
| > 60 | 61 | 3.333 (2.164–5.134) | < 0.001 | 2.791 (1.712–4.550) | < 0.001 |
| Cytogenetic risk | 138 | ||||
| Favorable | 31 | Reference | |||
| Intermediate | 76 | 2.957 (1.498–5.836) | 0.002 | 2.162(1.079–4.335) | 0.03 |
| Poor | 31 | 4.157 (1.944–8.893) | < 0.001 | 2.993(1.365–6.563) | 0.006 |
| LST1 | 140 | ||||
| Low | 70 | Reference | |||
| High | 70 | 2.038 (1.323–3.139) | 0.001 | 1.763 (1.063–2.923) | 0.028 |
| FLT3 mutation | 136 | ||||
| Negative | 97 | Reference | |||
| Positive | 39 | 1.271 (0.801–2.016) | 0.309 | ||
| NPM1 mutation | 139 | ||||
| Negative | 106 | Reference | |||
| Positive | 33 | 1.137 (0.706–1.832) | 0.596 | ||
| Gender | 140 | ||||
| Female | 63 | Reference | |||
| Male | 77 | 1.030 (0.674–1.572) | 0.892 | ||
| RAS mutation | 139 | ||||
| Negative | 131 | Reference | |||
| Positive | 8 | 0.643 (0.235–1.760) | 0.390 | ||
| WBC count(× 10^9/L) | 139 | ||||
| ≤ 20 | 75 | Reference | |||
| > 20 | 64 | 1.161 (0.760–1.772) | 0.490 | ||
| PB blasts (%) | 140 | ||||
| ≤ 70 | 66 | Reference | |||
| > 70 | 74 | 1.230 (0.806–1.878) | 0.338 | ||
| Race | 137 | ||||
| Black or African American | 10 | Reference | |||
| White | 127 | 1.383 (0.506–3.780) | 0.527 | ||
| IDH1 R132 mutation | 138 | ||||
| Negative | 126 | Reference | |||
| Positive | 12 | 0.588 (0.238–1.452) | 0.249 | ||
| IDH1 R140 mutation | 138 | ||||
| Negative | 127 | Reference | |||
| Positive | 11 | 1.131 (0.565–2.264) | 0.727 | ||
| IDH1 R172 mutation | 138 | ||||
| Negative | 136 | Reference | |||
| Positive | 2 | 0.610 (0.085–4.385) | 0.623 | ||
Bold values indicate statistical significance (P < 0.05)
Prognostic model of LST1 in AML
A nomogram was developed using the RMS R program (Fig. 9A) to predict prognosis in AML patients, based on Cox regression analysis results. The model incorporated three prognostic indicators (age, cytogenetic risk, and LST1 expression), selected using a statistical significance threshold of 0.2, with total variable scores normalized to a 0–100 scale. Performance validation through calibration plots showed strong concordance between predicted and observed survival probabilities at 1, 3, and 5 years across the AML cohort (Fig. 9B). The bootstrap-adjusted C-index of 0.728 (95% CI 0.695–0.768) indicated moderate predictive accuracy for overall survival. Analysis revealed survival probabilities of less than 50% at one year, declining to below 20% and 10% at three and five years, respectively.
Fig. 9.
AML prognostic predictive model integrating LST1. A Nomogram forecasting the likelihood of an OS of 1, 3, and 5 years for AML. B Calibration graph displaying the nomogram's accuracy in anticipating the likelihood of OS at 1, 3, and 5 years
Discussion
This investigation provides a comprehensive examination of the expression pattern, clinical relevance, functional pathways, and immune associations of LST1 in AML, leveraging multi-omics data alongside preliminary experimental validation. The main discoveries can be outlined as follows: (1) Expression Pattern Our research revealed a notable increase in LST1 concentrations in AML in contrast to healthy blood samples, validated through RT-qPCR and Western blot methods. (2) Clinical Importance: High LST1 expression was linked to aggressive clinical characteristics like high WBC counts, non-M3 FAB subtype, and intermediate/poor cytogenetic risk, indicating its potential as a predictive biomarker. (3) Predictive Value: Increased LST1 levels were an independent predictor of worse OS in both the entire AML group and various subgroups, but prospective cohort studies are required for confirmation. (4) Functional Pathways and Immune Connections: LST1 expression was linked to specific gene expression profiles and immune infiltration patterns in the AML environment, suggesting its potential role in AML development and immune response regulation. In conclusion, these results emphasize the diverse functions of LST1 in AML, showcasing its usefulness as a prognostic tool and its potential implications for understanding AML pathophysiology and immune regulation. Further research into the mechanistic underpinnings of LST1 dysregulation in AML is warranted, potentially opening avenues for targeted therapeutic interventions.
These outcomes underscore the clinical relevance of LST1 as a new predictive biomarker in AML. While previous investigations have elucidated the predictive significance of various molecular markers in AML [18–21], including NPM1, FLT3-ITD, CEBPA, and WT1, these markers often possess limitations, primarily restricted to specific AML subtypes and fail to encompass the disease's intricate heterogeneity. In contrast, our investigation demonstrates that LST1 overexpression is a prevalent characteristic across diverse AML subgroups and independently predicts poor prognosis. Notably, LST1 retains its significant prognostic importance even after adjustment for established risk factors encompassing age and cytogenetics, indicating its additional prognostic utility. Furthermore, the integration of LST1 expression into a nomogram model shows promising performance in predicting patient survival. The incorporation of this system could aid in risk stratification and impact treatment decisions for AML.
To elucidate LST1's role in AML pathogenesis, our integrative analysis revealed a coordinated network of inflammatory and immune regulatory pathways. As demonstrated by Bakhtiyaridovvombaygi et al. [22], cytokine-induced memory-like NK cells represent a promising immunotherapeutic strategy in AML, aligning with our findings on LST1's immunomodulatory effects. Patterson et al. [23] showed that the MYC-NFATC2 axis maintains cell cycle and mitochondrial function in AML cells, providing mechanistic insights into LST1's cellular effects. Our GO enrichment analysis identified significant differential expression in immune-related processes, consistent with Wang et al. [24] characterization of IFNγ signaling in AML. Allert et al. [25] demonstrated the hematopoietic niche's role in therapy resistance, supporting LST1's involvement in microenvironmental interactions. Our GSEA revealed distinct pathway signatures, complementing Grenier et al. [26] findings on adhesion molecules and Luciano et al. [27] characterization of the cytokine network. LST1-high cases showed enrichment in chemokine signaling and Toll-like receptor pathways, supported by Korbecki et al. [28] work on CXCR ligands and Baakhlagh et al. [29] evidence linking Toll-like receptor 4 signaling to AML pathophysiology.
The immunological impact of LST1 was further validated through ssGSEA results. Building on Tsuchiya and Shiota's [30] work on immune evasion mechanisms, our findings show parallels with LST1 overexpression patterns. Ferrell and Kordasti [31] documented Treg expansion in IFNγ-rich AML microenvironments, supporting our observation of increased Treg correlation with high LST1 expression. We found positive correlations with MDSCs, Tregs, and exhausted T cells, patterns characterized by Zhigarev et al. [32] in AML patients. Integrating these findings with Kim and Choi's[33] insights on NK cell function, we propose that LST1 overexpression orchestrates inflammatory signaling through chemokine and Toll-like receptor pathways. This model, supported by Zhong et al. [34] characterization of CD8 + T cell-based subtypes, creates a self-perpetuating immunosuppressive microenvironment. As shown by Al-Kahiry et al. [35], such immune pathway dysregulation significantly impacts patient outcomes, consistent with our LST1 expression findings.
However, our investigations also have certain limitations that warrant acknowledgment. Despite these promising findings, several methodological considerations merit discussion. Although we analyzed a substantial number of AML cases from public datasets, no independent cohort validation was performed, highlighting the need for external validation studies. Furthermore, while our bioinformatic analyses revealed compelling associations between LST1 and various functional pathways, experimental validation of these computational findings would strengthen our conclusions.The immune cell infiltration analysis in myeloid malignancies requires careful interpretation due to potential overlap with normal blood cell signatures. Moreover, the lack of detailed chemotherapy data in the TCGA cohort limits our understanding of LST1's relationship with treatment response.
In summary, our investigation unveils the widespread occurrence of LST1 overexpression in AML, serving as an independent predictor of unfavorable outcomes. We explore the relationship between LST1 levels and unique gene expression patterns, as well as immune cell infiltration in the AML environment, which could impact disease advancement and immune system avoidance. The results highlight the possibility of LST1 as a new predictive biomarker and target for treating AML. Further comprehensive exploration, both functionally and clinically, is imperative to delineate the precise role of LST1 in AML pathogenesis and to translate these insights into enhanced patient care.
Supplementary Information
Acknowledgements
We would like to extend our appreciation to all researchers and institutions for their efforts in making the original datasets employed in this work accessible to the public.
Author contributions
"Haitao Xu originated and formulated the research. Dangui Chen, Lihong Wang, and Long Zhong performed the experiments and bioinformatics analyses. Fei Chen and Jia Lu contributed to data acquisition, analysis, and interpretation. Haitao Xu drafted the manuscript."
Funding
Funding for this project was provided by The Anhui Health and Medical Research Foundation (Grant No. AHWJ2022c002).
Data availability
The research data produced and examined can be accessed in the TCGA and GEO repositories through the following URLs: [TCGA](https://portal.gdc.cancer.gov/); [GEO](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi).
Declarations
Ethics approval and consent to participate
This study was approved by the ethics committee of the Anqing Municipal Hospital (ID: No. 2024–72). All the enrolled patients signed an informed consent form. The study was conducted in accordance with the Declaration of Helsinki.
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.
References
- 1.Bornhauser M, Schliemann C, Schetelig J, Rollig C, Kramer M, Glass B, Platzbecker U, Burchert A, Hanel M, Muller LP, Klein S, Bug G, Beelen D, et al. Allogeneic hematopoietic cell transplantation vs standard consolidation chemotherapy in patients with intermediate-risk acute myeloid leukemia: a randomized clinical trial. JAMA Oncol. 2023;9(4):519–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dohner H, Weisdorf DJ, Bloomfield CD. Acute myeloid leukemia. N Engl J Med. 2015;373(12):1136–52. [DOI] [PubMed] [Google Scholar]
- 3.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70(1):7–30. [DOI] [PubMed] [Google Scholar]
- 4.Dohner H, Wei AH, Appelbaum FR, Craddock C, DiNardo CD, Dombret H, Ebert BL, Fenaux P, Godley LA, Hasserjian RP, Larson RA, Levine RL, Miyazaki Y, et al. Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood. 2022;140(12):1345–77. [DOI] [PubMed] [Google Scholar]
- 5.Papaemmanuil E, Gerstung M, Bullinger L, Gaidzik VI, Paschka P, Roberts ND, Potter NE, Heuser M, Thol F, Bolli N, Gundem G, Van Loo P, Martincorena I, et al. Genomic classification and prognosis in acute myeloid leukemia. N Engl J Med. 2016;374(23):2209–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Neville MJ, Campbell RD. A new member of the Ig superfamily and a V-ATPase G subunit are among the predicted products of novel genes close to the TNF locus in the human MHC. J Immunol. 1999;162(8):4745–54. [PubMed] [Google Scholar]
- 7.Fabisik M, Tureckova J, Pavliuchenko N, Kralova J, Balounova J, Vicikova K, Skopcova T, Spoutil F, Pokorna J, Angelisova P, Malissen B, Prochazka J, Sedlacek R, et al. Regulation of inflammatory response by transmembrane adaptor protein LST1. Front Immunol. 2021;12:618332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.D’Aloia A, Berruti G, Costa B, Schiller C, Ambrosini R, Pastori V, Martegani E, Ceriani M. RalGPS2 is involved in tunneling nanotubes formation in 5637 bladder cancer cells. Exp Cell Res. 2018;362(2):349–61. [DOI] [PubMed] [Google Scholar]
- 9.Shin H, Haupt KA, Kershner AM, Kroll-Conner P, Wickens M, Kimble J. SYGL-1 and LST-1 link niche signaling to PUF RNA repression for stem cell maintenance in Caenorhabditis elegans. PLoS Genet. 2017;13(12):e1007121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yang SY, Hayer KE, Fazelinia H, Spruce LA, Asnani M, Black KL, Naqvi AS, Pillai V, Barash Y, Elenitoba-Johnson KSJ, Thomas-Tikhonenko A. FBXW7beta isoform drives transcriptional activation of the proinflammatory TNF cluster in human pro-B cells. Blood Adv. 2023;7(7):1077–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Goldman MJ, Craft B, Hastie M, Repecka K, McDade F, Kamath A, Banerjee A, Luo Y, Rogers D, Brooks AN, Zhu J, Haussler D. Visualizing and interpreting cancer genomics data via the Xena platform. Nat Biotechnol. 2020;38(6):675–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bindea G, Mlecnik B, Tosolini M, Kirilovsky A, Waldner M, Obenauf AC, Angell H, Fredriksen T, Lafontaine L, Berger A, Bruneval P, Fridman WH, Becker C, et al. Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer. Immunity. 2013;39(4):782–95. [DOI] [PubMed] [Google Scholar]
- 15.Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Simonovic M, Doncheva NT, Morris JH, Bork P, Jensen LJ, Mering CV. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47(D1):D607–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Demchak B, Hull T, Reich M, Liefeld T, Smoot M, Ideker T, Mesirov JP. Cytoscape: the network visualization tool for GenomeSpace workflows. F1000Research. 2014;3:151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bandettini WP, Kellman P, Mancini C, Booker OJ, Vasu S, Leung SW, Wilson JR, Shanbhag SM, Chen MY, Arai AE. MultiContrast Delayed Enhancement (MCODE) improves detection of subendocardial myocardial infarction by late gadolinium enhancement cardiovascular magnetic resonance: a clinical validation study. J Cardiovasc Magn Resonance. 2012;14(1):83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Falini B. NPM1-mutated acute myeloid leukemia: new pathogenetic and therapeutic insights and open questions. Am J Hematol. 2023;98(9):1452–64. [DOI] [PubMed] [Google Scholar]
- 19.Loo S, Roberts AW, Anstee NS, Kennedy GA, He S, Schwarer AP, Enjeti AK, D’Rozario J, Marlton P, Bilmon IA, Taper J, Cull G, Tiley C, et al. Sorafenib plus intensive chemotherapy in newly diagnosed FLT3-ITD AML: a randomized, placebo-controlled study by the ALLG. Blood. 2023;142(23):1960–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Georgi JA, Stasik S, Kramer M, Meggendorfer M, Rollig C, Haferlach T, Valk P, Linch D, Herold T, Duployez N, Taube F, Middeke JM, Platzbecker U, et al. Prognostic impact of CEBPA mutational subgroups in adult AML. Leukemia. 2024;38(2):281–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schwede M, Jahn K, Kuipers J, Miles LA, Bowman RL, Robinson T, Furudate K, Uryu H, Tanaka T, Sasaki Y, Ediriwickrema A, Benard B, Gentles AJ, et al. Mutation order in acute myeloid leukemia identifies uncommon patterns of evolution and illuminates phenotypic heterogeneity. Leukemia. 2024. [DOI] [PMC free article] [PubMed]
- 22.Bakhtiyaridovvombaygi M, Yazdanparast S, Mikanik F, Izadpanah A, Parkhideh S, ShahbazGhasabeh A, Roshandel E, Hajifathali A, Gharehbaghian A. Cytokine-induced memory-Like NK cells: emerging strategy for AML immunotherapy. Biomed Pharmacother Biomed Pharmacother. 2023;168:115718. [DOI] [PubMed] [Google Scholar]
- 23.Patterson SD, Massett ME, Huang X, Jorgensen HG, Michie AM. The MYC-NFATC2 axis maintains the cell cycle and mitochondrial function in acute myeloid leukaemia cells. Mol Oncol. 2024;18(9):2234–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wang B, Reville PK, Yassouf MY, Jelloul FZ, Ly C, Desai PN, Wang Z, Borges P, Veletic I, Dasdemir E, Burks JK, Tang G, Guo S, et al. Comprehensive characterization of IFNgamma signaling in acute myeloid leukemia reveals prognostic and therapeutic strategies. Nat Commun. 2024;15(1):1821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Allert C, Muller-Tidow C, Blank MF. The relevance of the hematopoietic niche for therapy resistance in acute myeloid leukemia. Int J Cancer. 2024;154(2):197–209. [DOI] [PubMed] [Google Scholar]
- 26.Grenier JMP, Testut C, Fauriat C, Mancini SJC, Aurrand-Lions M. Adhesion molecules involved in stem cell niche retention during normal haematopoiesis and in acute myeloid leukaemia. Front Immunol. 2021;12:756231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Luciano M, Krenn PW, Horejs-Hoeck J. The cytokine network in acute myeloid leukemia. Front Immunol. 2022;13:1000996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Korbecki J, Kupnicka P, Barczak K, Bosiacki M, Zietek P, Chlubek D, Baranowska-Bosiacka I. The role of CXCR1, CXCR2, CXCR3, CXCR5, and CXCR6 ligands in molecular cancer processes and clinical aspects of acute myeloid leukemia (AML). Cancers. 2023;15(18):4555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Baakhlagh S, Kashani B, Zandi Z, Bashash D, Moradkhani M, Nasrollahzadeh A, Yaghmaei M, Mousavi SA, Ghaffari SH. Toll-like receptor 4 signaling pathway is correlated with pathophysiological characteristics of AML patients and its inhibition using TAK-242 suppresses AML cell proliferation. Int Immunopharmacol. 2021;90:107202. [DOI] [PubMed] [Google Scholar]
- 30.Tsuchiya H, Shiota G. Immune evasion by cancer stem cells. Regen Ther. 2021;17:20–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ferrell PB, Kordasti S. Hostile takeover: tregs expand in IFNgamma-rich AML microenvironment. Clin Cancer Res. 2022;28(14):2986–8. [DOI] [PubMed] [Google Scholar]
- 32.DiNardo CD, Erba HP, Freeman SD, Wei AH. Acute myeloid leukaemia. Lancet. 2023;401(10393):2073–86. [DOI] [PubMed] [Google Scholar]
- 33.Kim S, Choi J. Restoring NK cell functions in AML relapse. Blood. 2022;140(26):2765–6. [DOI] [PubMed] [Google Scholar]
- 34.Zhong F, Yao F, Jiang J, Yu X, Liu J, Huang B, Wang X. CD8 + T cell-based molecular subtypes with heterogeneous immune landscapes and clinical significance in acute myeloid leukemia. Inflamm Res. 2024;73(3):329–44. [DOI] [PubMed] [Google Scholar]
- 35.Al-Kahiry WMA, Dammag EAM, Abdelsalam HST, Fadlallah HK, Owais MS. Toll-like receptor 9 negatively related to clinical outcome of AML patients. J Egypt Natl Canc Inst. 2020;32(1):15. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The research data produced and examined can be accessed in the TCGA and GEO repositories through the following URLs: [TCGA](https://portal.gdc.cancer.gov/); [GEO](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi).