TFF3 as a Potential Prognostic Biomarker in Myelodysplastic Syndrome and Acute Myeloid Leukaemia

By Huali Hu^1,2, Fahua Deng^1,3, Siqi Wang³, Hai Huang^1,3, Tingting Lu^1,3, Sixi Wei^1,3

Affiliations

1. Center for Clinical Laboratories, The Affiliated Hospital of Guizhou Medical University, Guiyang, China
2. Department of Blood Transfusion, Guizhou Provincial People’s Hospital, Guiyang, China
3. Department of Clinical Biochemistry, School of Clinical Laboratory Science, Guizhou Medical University, Guiyang, China

doi: 10.29271/jcpsp.2025.06.735

ABSTRACT
Objective: To explore and validate the biomarker trefoil factor 3 (TFF3) of immune infiltration in myelodysplastic syndrome (MDS) and acute myeloid leukaemia (AML).
Study Design: Descriptive research.
Place and Duration of the Study: Center for Clinical Laboratories, The Affiliated Hospital of Guizhou Medical University, Guiyang, China, from April to October 2023.
Methodology: To screen for novel differentially expressed genes (DEGs) in MDS and AML, DEGs were obtained from the GEO database for bioinformatics analysis. Immune infiltration in MDS and AML datasets was investigated using the CIBERSORT algorithm. In addition, the PrognoScan database verified the correlation between immunity-associated DEGs and survival time and plotted the ROC curve. Finally, the expression of TFF3 in clinical peripheral blood samples and cell lines was verified by RT-qPCR, and the role of TFF3 in cell proliferation was analysed by CCK-8 assay.
Results: Thirty-two common DEGs were identified: Twenty-seven downregulated and five upregulated. The immune infiltration investigation revealed that the development of AML may include γδT cells, activated CD4 memory T cells, monocytes, and neutrophils. K‒M survival curves and ROC curves showed that the low expression of an infiltration-related gene named TFF3 in MDS and AML was associated with poor prognosis, and the ROC curve showed better predictive performance. Finally, verification results by RT‒qPCR also demonstrated that TFF3 expression was decreased in MDS and AML, and the CCK-8 assay revealed that si-TFF3 could promote Kasumi-1 cell proliferation.
Conclusion: Low expression of TFF3 is associated with poor prognosis and immune cell infiltration in MDS and AML, suggesting that TFF3 may serve as a potential biomarker in MDS and AML through immune regulation.

Key Words: Trefoil factor 3, Myelodysplastic syndrome, Acute myeloid leukaemia, Immune infiltration, Prognosis.

INTRODUCTION

Myelodysplastic syndrome (MDS) and acute myeloid leukaemia (AML) are haematological malignancies. Cytopenia, bone marrow dysplasia, and ineffective haematopoiesis are the characteristics of MDS.¹ The characteristics of AML are the accumulation of myeloid progenitors in blood or bone marrow due to a block in the differentiation of haematopoietic cells.² The prognosis of AML is influenced by multiple factors, such as the patients' physical condition, age, the process of prodromal MDS, chemotherapy regimens, common laboratory test indicators, the expression of immune molecules in leukaemia cells, the haematopoietic microenvironment of bone marrow, gene expression levels, and reactivity to chemotherapy medicines, etc.³

With the deepening of risk stratification diagnosis and treatment, the overall survival (OS) has been greatly improved. However, there are still a small number of AML patients who cannot tolerate strong chemotherapy or face relapse after remission, and the overall efficacy is still not optimistic.⁴ Many studies have shown that cytogenetics and molecular genetic changes are involved in the development of AML.⁵ Therefore, looking for AML potential molecular changes, to further explain the development of AML molecular mechanisms to improve the AML curative effect and prognosis have important significance.

Trefoil factor 3 (TFF3) belongs to a small peptide of the trefoil factor family, which is mainly secreted by intestinal goblet cells.⁶ This peptide has neuroprotective effects because it eliminates the activity of caspase-3 that damages microglia.⁷ TFF3 also has anti-apoptotic and proliferation-promoting functions and is considered to contribute to the progression of solid tumours.⁸ Studies have reported that TFF3 also activates myelid-derived suppressor cells (MDSCs) through the NF-κB/ COX2 signalling pathway,^6,9 and is negatively correlated with immune cell infiltration in papillary thyroid carcinoma.¹⁰ However, the relationship between TFF3 expression and MDS as well as AML remains poorly understood.

This study aimed to compare the samples and immune cells in the GEO database as a potential evidence of MDS and AML process to explore the proportion of different immune cells.

METHODOLOGY

This study was completed from April to October 2023 in the Centre for Clinical Laboratories, The Affiliated Hospital of Guizhou Medical University, Guiyang, China. Ethical approval was obtained from the Ethics Committee of the University [Approval Document: 2019(169) and 2023(655)].

The inclusion criteria for this study were 30 patients with primary MDS/AML diagnosed according to WHO criteria. The exclusion criteria were patients with mixed cell leukaemia and those preparing for bone marrow transplantation. In addition, patients with incomplete medical records were also excluded.

The gene expression omnibus (GEO) database (httpss://www. ncbi.nlm.nih.gov/geo/) was utilised to download gene expression profile data. Survival data for MDS and AML were acquired using the cancer genome atlas (TCGA) database (httpss://portal.gdc.cancer.gov/). The downloaded link is shown in Table I.

Table  I:  Dataset  link  of  the  gene  expression  profile.

Dataset	Link
GSE15061 (mRNA microarray of AML and MDS)	httpss://ftp.ncbi.nlm.nih.gov/geo/ series/GSE15nnn/GSE15061/ matrix/

Table  II:  Primers  for  qPCR  amplification.

Names	Sequences
GAPDH forward	TGGCGCTGAGTACGTCG
GAPDH reverse	ACACCCATGACGAACATG
TFF3 forward	TCCACCCGAGGACAGTTCTT
TFF3 reverse	TCCTAGGCTTTCCTGTGACG

The Bioconductor package limma of R 4.2.1 software was applied to screen DEGs from healthy cohorts, MDS and AML patients. To sort out functional categories of intersecting DEGs, Gene Ontology (GO) was performed.

The proportions of twenty-two subtypes of immune cells in a mixed cell population were quantitatively calculated from the GEO dataset by the CIBERSORT package. To explore the composition and correlation of 22 subtypes of immune cells from healthy cohorts and patients, the deconvolution algorithm was used to evaluate the scores.

ROC curve was plotted using the GSE15061 dataset. The value of the area under the ROC curve (AUC) was utilised to estimate the diagnostic effectiveness of genes in AML by pROC of R software and further validated in GSE15061. The 0.70.9 means that the gene has a high diagnostic value.

The MDS and AML cell lines SKM-1, MV4-11, Kasumi-1, and U937 were stored at laboratory. The negative control siRNA (si-NC) and siRNA TFF3 (si-TFF3) were transfected into Kasumi-1 cells.

Cell viability was measured by CCK-8 assay (Meilunbio, China). Transfected cells (si-NC and si-TFF3) were seeded in 96-well plates at a density of 2×10³ cells per wall. At 0, 24, 48, and 72 hours, 10 μl of CCK8 reagent was added to each well and incubated for 2 hours. The OD was measured at 450nm using a microplate reader.

Bone marrow samples from healthy donors and peripheral blood samples from healthy individuals and patients with MDS or AML were collected from the Affiliated Hospital of Guizhou Medical University. RT-qPCR amplification was conducted by Applied Biosystems 7500 Real-Time PCR System, and the expression of genes was analysed with the 2^−ΔΔCt method. The primers are shown in Table II.

All of the experiments have been repeated in triplicate at least. GraphPad Prism version 9.0 was used to analyse data and draw graphs. Student’s t-test was used for comparisons between the two groups and the Shapiro-Wilk’s test is used to assess normality before using the t-test. The one-way ANOVA was used for comparisons among more than two groups (including healthy controls, MDS, and AML. Healthy donors, SKM-1, Kasumi-1, MV4-11, and U937). A value of p <0.05 was considered statistically significant.

RESULTS

To screen for novel DEGs in MDS and AML, the GEO database was used to search for gene expression microarrays in MDS and AML. In the GEO datasets, the gene expression in four cohorts was compared (Cohort 1: Healthy people vs. patients with MDS and AML; Cohort 2: healthy people vs. AML patients; Cohort 3: Healthy people vs. the MDS patients; Cohort 4: MDS patients vs. AML patients) and obtained heatmaps of differential expression across cohorts (Figure 1A-D). Subsequently, the Venn diagram showed that 32 DEGs were shared between healthy people and patients with AML and MDS and between patients with MDS and AML (Figure 1E). The volcanic plot results showed that 32 DEGs contained 27 downregulated genes and 5 upregulated genes (Figure 1F). Therefore, the 32 DEGs were mainly analysed in the following studies.

To study the biological functions of DEGs between MDS and AML, the GO and KEGG pathways were analysed in the R software. For biological processes (BP), the DEGs might regulate the defense response to bacteria and the activation of myeloid leucocytes, neutrophils, and granulocytes. For cellular components (CC), the DEGs could regulate various immune responses, antimicrobial peptides mediated antibacterial humoral immune response. Moreover, for molecular functions (MF), the DEGs could regulate the progression of AML by regulating the activity of immune receptors (Figure 2A). The infiltration of immune cells in MDS and AML was analysed, the types and amounts are shown on the heatmap (Figure 2B).

Figure 1: Identification of DEGs of MDS and AML. (A-D) Heatmap of DEGs between MDS and AML patients and the healthy cohort. (E) Venn plot of DEGs between MDS and AML patients and the healthy cohort. (F) Volcano plot of DEGs between the healthy cohort, MDS and AML patients. Figure 2: Evaluation of immune cell proportions in DEGs. (A) Significant GO biological functional analyses of the 32 DEGs. (B, C) Heatmap and correlation heatmap of 22 subtypes of immune cell infiltration in MDS and AML. (D) Violin plots of immune cell expression in MDS and AML.
Red indicates AML, and blue indicates MDS.  Figure 3: Validation of the function of TFF3. (A) K–M survival curves of TFF3 with significant prognosis obtained from the TCGA analysis. (B) ROC curves of TFF3. (C-D) The expression of TFF3 in healthy controls (n = 30), MDS patients, AML patients (n = 30), and in healthy donors (n = 3), the MDS line SKM-1, and the AML cell lines Kasumi-1, MV4-11 and U937 were detected by RT‒qPCR. (E) The knockdown efficiency of si-TFF3 in Kasumi-1 cells was detected by RT‒qPCR. si-NC: Negative-control siRNA, si-TFF3: TFF3 knockdown siRNAs. (F) Cell proliferation was detected by the CCK8 assay. Student’s t-test was used for comparisons between the two groups (E, F) and one-way ANOVA was used for comparison amongst more than two groups (C, D).

The Spearman’s correlation test was used to calculate the correlation coefficient, M0 macrophages were positively correlated with neutrophils (r = 0.53), and resting CD8 T cells were positively correlated with resting NK cells (r = 0.44). However, M2 macrophages and B-cell memory were negatively correlated with M0 macrophages (Figure 2C). Nevertheless, tumour-promoting immune cell, macrophages M2 (p <0.001), was increased in AML (Figure 2D). Collectively, the above results suggest that these immune cells may play a crucial role in MDS and AML.    

The poor prognosis of AML is influenced by multiple factors, such as the patients' age, the process of prodromal MDS, chemotherapy regimens, the expression of immune molecules in leukaemia cells, the haematopoietic microenvironment of bone marrow, gene expression levels, and reactivity to chemotherapy medicines, etc.³

To explore the potential value of the 32 candidate DEGs in predicting the OS in AML patients, the survival data from the TCGA-LAML dataset were analysed, and the Kaplan-Meier survival curve was constructed. Finally, a common gene - TFF3 was obtained, which had low expression in MDS and AML and negatively correlated with OS (Figure 3A). The K‒M curve and the AUC value of TFF3 was (0.895) suggested that this gene could be used as an indicator of the progression and prognosis in AML patients (Figure 3B). RT‒qPCR result revealed that TFF3 expression was low in both MDS and AML patients compared with healthy controls (Figure 3C). In addition, the examination of TFF3 gene levels in MDS line SKM-1, AML cell line Kasumi-1, MV4-11 and U937, and samples of healthy donors verified that TTF expression in MDS and AML cell lines was notably decreased compared to healthy donors, particularly in Kasumi-1 cells (Figure 3D). Therefore, Kasumi-1 cells were selected for further study. Subsequently, specific siRNA (si-TFF3) was constructed to knock down the TFF3 gene in Kasumi-1 cells. The knockdown efficiency data are shown in Figure 3E. Finally, the CCK-8 assay indicated that cell proliferation was increased after TFF3 knockdown (p = 0.0031; Figure 3F).

In summary, the results of this study suggest that TFF3 is expressed at low levels in MDS and AML and plays an essential role in promoting the growth and proliferation of AML cells.

DISCUSSION

MDS and AML are haematological malignancies with poor prognosis and high mortality.^1,2 Accumulating evidence suggests that molecular biomarkers have huge potential in the diagnosis, treatment, and prognosis of tumours. Therefore, it is essential to find molecular indicators to help predict, diagnose, and treat MDS and AML.

This study obtained the DEGs between healthy people and MDS/AML from the public database. GO analysis of these DEGs revealed that DEGs could regulate various immune responses, antimicrobial peptides mediated antibacterial humoral immune response. Subsequently, analysis of the distribution of 22 immune cells in the GEO dataset using CIBERSORT also indicated an imbalance of immune cells in MDS and AML, which is consistent with MDS/AML pathogenesis.^11,12 Previous studies have indicated that immune cells, as the primary factors affecting haematologic tumour immunity, their proportion in the bone marrow microenvironment and the expression of different genes in immune cells may perform an essential function in MDS and AML disease status.¹³ Moreover, stromal cells and immune cells are important parts of the marrow stromal environment affect the AML cells proliferation, survival, and treatment of medicine resistance.¹⁴ It is believed that with further development and research, the relationship between immune cells and AML will gradually clarify.

The prognosis of AML is influenced by multiple factors, such as the patients' age, the process of prodromal MDS, the expression of immune molecules in leukaemia cells, chemotherapy regimens, common laboratory test indicators, the haematopoietic microenvironment of bone marrow, gene expression levels, and reactivity to chemotherapy medicines, etc.³ The expression of immune molecules in leukaemic cells is also one of the key factors affecting the prognosis of AML. According to previous studies, it has been found that TFF3 is involved in the development and inflammatory process of tumours.^9,15 The expression of TFF3 in gastric cancer tissues is related to the stage and prognosis of local lymph node metastasis.¹⁶ TFF3 may be a new diagnostic and prognostic indicator for endometrial cancer and ovarian cancer.¹⁷ Low expression of TFF3 in thyroid carcinoma may be associated with poor prognosis, but the rate of immune cell infiltration is high.¹⁰ Similarly, in this study, combined with the survival data downloaded from the TCGA database, it was found that TFF3 was expressed at a low level in MDS and AML and negatively correlated with OS. ROC analysis was used to further explore the clinical efficacy of TFF3, the AUC of TFF3 was 0.895, showing good diagnostic performance. The results may suggest that TFF3 has great potential as a biomarker in the prognosis of MDS and AML.

Studies have shown that cell proliferation has been regarded as one of the main reasons for the poor prognosis of tumour patients.¹⁸ As a growth factor-like small molecule peptide, TFF3 can activate ERK1/2, which is an important component of the MAPK signal transduction pathway and can promote the proliferation of tumour cells.¹⁹ Knockdown of TFF3 can inhibit the proliferation of lung cancer cells and promote cell apoptosis.²⁰ However, some authors have reported that TFFs can inhibit the proliferation of tumour cells. Studies have found that overexpression of TFF3 inhibited EGFR phosphorylation and cell proliferation in human colon cancer cell lines.²¹ TFF3 could reduce the phosphorylation of MAPK/ERK, thereby significantly inhibiting the proliferation, growth, differentiation and apoptosis of cells.²² The CCK-8 experiment used in this study also indicated that the knockdown of TFF3 could promote the growth and proliferation of AML cells, suggesting that its low expression might have played a role in promoting the proliferation of AML cells.

In general, the study on immune infiltration of TFF3 is still lacking. In this study, immune cell infiltration was analysed based on the bioinformatics database, and the correlation between them still needs further experimental verification, which may provide a novel understanding of the pathogenesis of MDS and AML. However, the study also has limitations, the cell proliferation experiment of TFF3 has only been verified in Kasumi-1 cells, and more experiments are needed in the future to verify the in-depth mechanism of TFF3.

CONCLUSION

This study identified an immune infiltration-related gene, TFF3, which is associated with the prognosis of MDS and AML and is expected to become a new diagnostic marker for AML.

FUNDING:
The research was supported by research grants from the NSFC to Sixi Wei (Grant No: 81960031, No: 82260033, No: 81660027), the Science and Technology Project of Guizhou Province to Sixi Wei (Grant No: 20185779-70), the Research Funds for Guizhou Provincial Innovative Talents Team for 2019 (Grant No: 5610), the Doctoral Research Start-up Fund of Affiliated Hospital of Guizhou Medical University (Grant No: gyfybsky-2021-29) and the Regional Fund Cultivation Project of National Natural Science Foundation of Affiliated Hospital of Guizhou Medical University (Grant No: gyfynsfc-2021-48).

ETHICAL  APPROVAL:
Ethical approval was obtained from the Ethics Committee of the Affiliated Hospital of Guizhou Medical University, Guiyang, China [Approval Document: 2019(169) and 2023(655)]. As public databases, GEO and TCGA can be used for data mining and analysis. Ethical approval has been obtained for the patients involved in the database, and the data can be freely downloaded and used for related research. The sequencing data involved in this study are open-source data, there are no ethical problems or other conflicts of interest.

PATIENTS’  CONSENT:
Signed informed consent was obtained from study participants prior to their enrolment in the study.

COMPETING  INTEREST:
The authors declared no conflict of interest.

AUTHORS’  CONTRIBUTION:
HH: Study conception, design, and preparation of the manu-script.
FD: Data collection and performing the experiment.
SW: Data and specimen collection.
HH: Revision of the article.
TL: Project administration.
SW: Facilitation of the conception of this research and revision of the article.
All authors approved the final version of the manuscript to be published.

REFERENCES

1. Cazzola M. Myelodysplastic syndromes. N Engl J Med 2020; 383(14):1358-74. doi: 10.1056/NEJMra1904794.
2. Dohner H, Estey E, Grimwade D, Amadori S, Appelbaum FR, Buchner T, et al. Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood 2017; 129(4):424-47. doi: 10.1182/ blood-2016-08-733196.
3. Liersch R, Muller-Tidow C, Berdel WE, Krug U. Prognostic factors for acute myeloid leukaemia in adults-biological significance and clinical use. Br J Haematol 2014; 165(1): 17-38. doi: 10.1111/bjh.12750.
4. Minciacchi VR, Kumar R, Krause DS. Chronic myeloid leukaemia: A model disease of the past, present and future. Cells 2021; 10(1):117. doi: 10.3390/cells10010117.
5. Quessada J, Cuccuini W, Saultier P, Loosveld M, Harrison CJ, Lafage-Pochitaloff M. Cytogenetics of pediatric acute myeloid leukaemia: A review of the current knowledge. Genes 2021; 12(6):924. doi: 10.3390/genes12060924.
6. Yang Y, Lin Z, Lin Q, Bei W, Guo J. Pathological and therapeutic roles of bioactive peptide trefoil factor 3 in diverse diseases: Recent progress and perspective. Cell Death Dis 2022; 13(1):62. doi: 10.1038/s41419-022-04504-6.
7. Liu SQ, Roberts D, Zhang B, Ren Y, Zhang LQ, Wu YH. Trefoil factor 3 as an endocrine neuroprotective factor from the liver in experimental cerebral ischemia/reperfusion injury. PloS One 2013; 8(10):e77732. doi: 10.1371/journal.pone.00 77732.
8. Diao S, Zheng Q, Gao J, Yao Y, Ren S, Liu Y, et al. Trefoil factor 3 contributes to the malignancy of glioma via regulating HIF-1α. Oncotarget 2017; 8(44):76770-82. doi: 10. 18632/oncotarget.20010.
9. Liu J, Yang Q, Chen Z, Lv S, Tang J, Xing Z, et al. TFF3 mediates the NF-κB/COX2 pathway to regulate PMN-MDSCs activation and protect against necrotizing enterocolitis. Eur J Immunol 2021; 51(5):1110-25. doi: 10.1002/eji.202048768.
    
10. Yang L, Zhang X, Zhang J, Liu Y, Ji T, Mou J, et al. Low expression of TFF3 in papillary thyroid carcinoma may correlate with poor prognosis but high immune cell infiltration. Future Oncol 2022; 18(3):333-48. doi: 10.2217/fon-2020-1183.
11. Trowbridge JJ, Starczynowski DT. Innate immune pathways and inflammation in haematopoietic aging, clonal hematopoiesis, and MDS. J Exp Med 2021; 218(7):e20201544. doi: 10.1084/jem.20201544.
12. Vadakekolathu J, Minden MD, Hood T, Church SE, Reeder S, Altmann H, et al. Immune landscapes predict chemotherapy resistance and immunotherapy response in acute myeloid leukaemia. Sci Transl Med 2020; 12(546):eaaz0463. doi: 10.1126/scitranslmed.aaz0463.
13. Lambert C, Wu Y, Aanei C. Bone marrow immunity and myelodysplasia. Front Oncol 2016; 6:172. doi: 10.3389/fonc. 2016.00172.
14. Tabe Y, Konopleva M. Role of microenvironment in resistance to therapy in AML. Curr Hematol Malign Rep 2015; 10(2):96-103. doi: 10.1007/s11899-015-0253-6.
15. Hoffmann W. Trefoil factor family (TFF) peptides and their links to inflammation: A re-evaluation and new medical perspectives. Int J Mol Sci 2021; 22(9):4909. doi: 10.3390/ijms22094909.
16. Guleng B, Han J, Yang JQ, Huang QW, Huang JK, Yang XN, et al. TFF3 mediated induction of VEGF via hypoxia in human gastric cancer SGC-7901 cells. Mol Biol Rep 2012; 39(4): 4127-34. doi: 10.1007/s11033-011-1195-2.
17. Turan H, Vitale SG, Kahramanoglu I, Corte LD, Giampaolino P, Azemi A, et al. Diagnostic and prognostic role of TFF3, Romo-1, NF-кB and SFRP4 as biomarkers for endometrial and ovarian cancers: A prospective observational translational study. Arch Gynecol Obstet 2022; 306(6):2105-14. doi: 10.1007/s00404-022-06563-8.
18. Jayaraman S, Pazhani J, Veeraraghavan VP, Raj AT, Somasundaram DB, Patil S. PCNA and Ki67: Prognostic proliferation markers for oral cancer. Oral Oncol 2022; 130:105943. doi: 10.1016/j.oraloncology.2022.105943.
19. Storesund T, Schenck K, Osmundsen H, Roed A, Helgeland K, Kolltveit KM. Signal transduction and gene transcription induced by TFF3 in oral keratinocytes. Eur J Oral Sci 2009; 117(5):511-7. doi: 10.1111/j.1600-0722.2009.00652.x.
20. Liu M, Xiao Y. Trefoil factor 3 silencing can inhibit the proliferation and apoptosis of lung cancer cells. J BUON 2021; 26(5):1842-9.
21. Uchino H, Kataoka H, Itoh H, Hamasuna R, Koono M. Overexpression of intestinal trefoil factor in human colon carcinoma cells reduces cellular growth in vitro and in vivo. Gastroenterology 2000; 118(1):60-9. doi: 10.1016/s0016-5085(00) 70414-8.
22. Kosriwong K, Menheniott TR, Giraud AS, Jearanaikoon P, Sripa B, Limpaiboon T. Trefoil factors: Tumour progression markers and mitogens via EGFR/MAPK activation in cholangiocarcinoma. World J Gastroenterol 2011; 17(12):1631-41. doi: 10.3748/wjg.v17.i12.1631.