Differentiation of intestinal stem cells toward goblet cells under systemic iron overload stress are associated with inhibition of Notch signaling pathway and ferroptosis

Abstract

Iron overload can lead to oxidative stress and intestinal damage and happens frequently during blood transfusions and iron supplementation. However, how iron overload influences intestinal mucosa remains unknown. Here, the aim of current study was to investigate the effects of iron overload on the proliferation and differentiation of intestinal stem cells (ISCs). An iron overload mouse model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for a duration of 12 weeks, and an iron overload enteroid model was produced by treatment with 3 mM or 10 mM of ferric ammonium citrate for 24 h. We found that iron overload caused damage to intestinal morphology with a 64 % reduction in villus height/crypt depth ratio, and microvilli injury in the duodenum. Iron overload mediated epithelial function by inhibiting the expression of nutrient transporters and enhancing the expression of secretory factors in the duodenum. Meanwhile, iron overload inhibited the proliferation of ISCs and regulated their differentiation into secretory mature cells, such as goblet cells, through inhibiting Notch signaling pathway both in mice and enteroid. Furthermore, iron overload caused oxidative stress and ferroptosis in intestinal epithelial cells. In addition, ferroptosis could also inhibit Notch signaling pathway, and affected the proliferation and differentiation of ISCs. These findings reveal the regulatory role of iron overload on the proliferation and differentiation of ISCs, providing a new insight into the internal mechanism of iron overload affecting intestinal health, and offering important theoretical basis for the scientific application of iron nutrition regulation.

1. Introduction

Iron is a widely distributed trace element that is necessary for practically all biological tissues. It serves as a cofactor for numerous key enzymes involved in many metabolic activities, including DNA synthesis, oxidative phosphorylation, cell cycle progression and growth [1]. However, high quantities of intracellular free iron can produce a lot of free radicals via Fenton and Haber-Weiss processes. This can lead to oxidative damage to membrane lipids, proteins, and DNA, which can ultimately kill cells. As a result, it is essential for preserving intracellular iron homeostasis, which depends on the proper coordination of iron transport, absorption, and storage. The small intestine is crucial to iron metabolism since it is the only organ involved in dietary iron absorption.

Under physiological conditions, the body lacks specific mechanism for excreting iron, and iron metabolism is primarily influenced by dietary iron absorption (1–2 mg Fe/day), which serves as a compensatory measure for non-specific metal losses (primarily from cellular desquamation, menstrual bleeding, or other sporadic blood loss) [2]. The systemic hepcidin/ferroportin axis strongly regulates iron homeostasis. When the intricate homeostatic machinery is unable to maintain an iron equilibrium, an iron overload may happen. Iron overload can occur in patients with inherited diseases, such as hereditary haemochromatosis [3] and β thalassaemia [4], or secondary to iron overload during blood transfusion and haemolysis [5]. Iron accumulation in multiple organs causes extra clinical problems and oxidative damage in the liver [6], pancreas [7] and kidney [8]. The metabolism of non-transferrin-bound iron results in reactive oxygen species, which are linked to apoptosis, necrosis, and cellular malfunction [9]. The primary target of persistent oxidative stress resulting from the production of reactive oxygen species during aerobic metabolism is the mucosa of the gastrointestinal system. Studies have shown that elevated luminal iron levels are linked to a greater ability of healthy volunteers’ colon to produce free radicals [10]. Excessive intestinal iron in the gastrointestinal tract causes intestinal flora dysbiosis and intestinal injury, even increases the risk of colorectal cancer development [11,12]. Our previous studies also showed that iron overload caused intestinal inflammation, damaged villi morphology and reshaped the gut microbiota [13,14]. However, despite this strong correlation between iron overload and gut health, the role of iron in the mucosa of the gastrointestinal system and the method by which it operates are still unclear.

In this study, a mouse model of iron overload was established to investigate the effects of iron overload on intestinal morphology, epithelial function, proliferation and differentiation of ISCs and ferroptosis of intestinal cells. Furthermore, how iron overload affect the proliferation and differentiation of intestinal progenitors was explored in mouse enteroids and ferroptosis mouse models. This study helps to expand our understanding the negative effects of iron overload on intestinal health.

2. Materials and methods

2.1. Animals

Male C57BL/6 mice at the age of 6–8 weeks were purchased from the Chinese Academy of Sciences’ Animal Center in Shanghai, China. Lgr5^{-EGFP-IRES-CreERT2} transgenic mice were kindly provided by Professor Yeguang Chen (Beijing University). All mice were housed in standard plastic cages and received food and water ad libitum in a room with a 12 h dark-light cycle at 23 ± 1 °C, with a relative humidity of 50 % ± 10 %. For iron-overload treatment, the mice were administered an intraperitoneal injection of 120 mg/kg body weight (B.W.) of iron dextran (uniferon®, Pharmacosmos A/S, Holbaek, Denmark) as previously described once a fortnight for 12 weeks [15]. Control mice were given physiological saline (n = 6 per group). Blood was drawn via retro-orbital bleed at the end of the experiment. Following the cervical dislocation method of mice sacrifice, the duodenum was frozen in liquid nitrogen and kept at −80 °C. Lgr5^{-EGFP-IRES-CreERT2} transgenic mice were sacrificed for isolation of intestinal crypts and enteroids culture. For establishing a ferroptosis mouse model, mice were intraperitoneally injected with 25 mg/kg B W. of a classical inducer of ferroptosis, erastin (Selleck chemicals, n = 6) or physiological saline (control, n = 6) as previously described for 2 days at 12-h intervals [16]. The studies and usage of animals involved in the experiment were conducted in strict accordance with the guidelines of the Institute of Animal Research Committee of Zhejiang University and were approved by the Laboratory Animal Care Committee of Zhejiang University (ZJU2022-22522).

2.2. Measurement of serum iron content and transferrin saturation

Serum iron content was measured using serum iron assay kit (Jiancheng, Nanjing, China) according to the manufacturer's instructions. Total iron binding capacity was measured using total iron binding capacity assay kit (Jiancheng, Nanjing, China) according to the manufacturer's instructions. Optical density was measured absorbance at 520 nm by a microplate reader (Molecular Devices, Sunnyvale, USA). And transferrin saturation (TSAT, %) = serum iron (mg/L)/total iron binding capacity (mg/L) × 100.

2.3. Quantitative real-time PCR

The TRIzol reagent (Simgen, Hangzhou, China) was used to isolate total RNA. A NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, USA) was used to measure the quantity and purity of RNA. M-MLV reverse transcriptase (Thermo Fisher Scientific, Waltham, USA) was used to reverse transcribe 2 mg of RNA from each sample in order to create cDNA. The ABI 7500 StepOne Real-Time PCR System (Applied Biosystems, Foster City, USA) was then used to perform quantitative real-time PCR (qPCR) in triplicate using the Hieff UNICON® qPCR SYBR Green Master Mix (High Rox, YEASEN, Shanghai, China). An internal reference gene was the β-actin gene. The mRNA expression levels were ascertained using the 2^−ΔΔ cycle threshold approach. Table 1 lists the sequences of the particular primers.

Table 1.

List of real time PCR primers and sequences.

Gene	Forward primers	Reverse primers
Atoh1	GGGAAGCCCCGTGACAAATA	TGGACCATGAAACGATGCCA
BCAT2	CTCTGCCTGCCAGACTTTGA	CAGCCACAGTGGGTCCATAG
CCL9	GAGAAGGGTGTGAACTGCCA	TTCCCCTCTGTGACAAAGCC
ChrA	CCAAGGTGATGAAGTGCGTC	GGTGTCGCAGGATAGAGAGGA
Claudin1	TTTGCAGAGACCCCATCACC	CGATCCATCCCAGAGAAGCC
Dll1	GGTTTGTGTGTGACGAGCAC	ATCTTCTCCCCTCTGTCCCC
Dll4	CCAGCAACCCCTGTCGAAAT	ACAGTGCTGGCCATAGTAGC
DMT1	TCTGGGCAGTGGGGATCCTG	GACGAGCAGGGTGGGGATGA
FABP2	GACGGCACGTGGAAAGTAGA	TTTCCCTCAATGGTCCAGGC
FtH	TCAGTCACTACTGGAACTGC	CGTGGTCACCCAGTTCTTTA
Fpn1	ATGGGAACTGTGGCCTTCAC	TCCAGGCATGAATACGGAGA
GLUT2	GCCCAGCAGTTCTCAGGAAT	ACATGCCAATCATCCCGGTT
Gapdh	TGCGACTTCAACAGCAACTC	GCCTCTCTTGCTCAGTGTCC
Hes1	TCAACACGACACCGGACAAA	ATTCTTGCCCTTCGCCTCTT
IL-13	ACAAGACCAGACTCCCCTGT	TCTGGGTCCTGTAGATGGCA
Ki67	ACCATCATTGACCGCTCCTT	TCACTCTTGTCAGGGTCAGC
Lgr5	GAGTCAACCCAAGCCTTAGTATCC	CATGGGACAAATGCAACTGAA
Lf	GGGCTTTAAGGTGTCTGGCT	ACAGGTTTCCGGGTGTCATC
Lyz	ACTGCAGCCATACAATGTGC	GGGACAGATCTCGGTTTTGA
Muc2	CTACAATGGCTGCACCAAGA	CGGACACTGGTCTTCTCCTC
Notch1	TGGACTGTTCTGTGCATCCC	TGGGGATCAGAGGCCACATA
Occludin	GTGAATGGGTCACCGAGGG	AGATAAGCGAACCTGCCGAG
Ptgs2	ATACGCTGAGTGTGGTTTGC	CTCACTAAACCATCCAATCGG
RBP-Jk	TGTGGTTTTCTGTGCCACCT	TCAAAGACGATGATGCACGC
Relm-β	ATGTGCTTGTGGCTATGGCT	GGTCGAGACCGTGGTTTCAT
SGLT1	ATGTCTCACGTGAAGGCTGG	TGGTGTGCCGCAGTATTTCT
SLC7A5	AAGGGCAGGGATTCATGGTG	GTAGGGGTGTCTTTCAGGGC
TFF3	GTATGGTGCCGGCAAATGTC	AGAGGTTTGAAGCACCAGGG
TfR	TGCGGAAGGAAGTGACGTAG	CTTGCCGAGCAAGGCTAAAC
Vil1	TCAAAGGCTCTCTCAACATCAC	AGCAGTCACCATCGAAGAAGC
Zg16	GTGATCCAGGTGTCTGGCAA	ACAGCGTTGAAACTTGTGCC
ZO-1	CAGCAGGAATGCCTTACCCT	TCAGATGCACAGCCCAAAGA
β-actin	GCCACTGCCGCATCCTCTTC	AGCCTCAGGGCATCGGAACC

2.4. Prussian blue staining

The intestinal tissues were preserved with 4 % paraformaldehyde and embedded in paraffin. Perls’ Prussian blue was used to stain sections of a thickness of 5 μm by Servicebio (Wuhan, China). In short, slices of intestinal tissue were serially rehydrated in alcohols after being dewaxed. After that, the tissue was incubated for 30 min in a 1:1 mixture of 2 % potassium ferrocyanide and 2 % hydrochloric acid. The slides were dehydrated, mounted, and counterstained with eosin for 20 s after being cleaned with tap water. The Olympus NP-26 (Beijing, China) was used to shoot the photos.

2.5. Western blot (WB) analysis

After homogenizing tissues in lysis buffer (KeyGEN BioTECH, Nanjing, China) containing a protease inhibitor cocktail (Thermo Fisher Scientific, Waltham, USA), the precipitate was removed from the tissues by centrifuging the mixture at 3000 g for 5 min at 4 °C. The BCA Protein Assay Kit (KeyGEN BioTECH, Nanjing, China) was utilized to quantify the total protein content. The primary antibodies were used to measure the expression levels of protein samples separated by SDS-PAGE, which are listed in Table 2. HRP-conjugated anti-rabbit and anti-mouse IgG (Biosharp, Hefei, China) served as secondary antibodies. The ECL Plus detection equipment from Clinx Science Instruments in China was used to detect the signals, and ImageJ software (Rawak Software, Stuttgart, Germany) was used to quantify them.

Table 2.

List of antibodies used in Western blot analysis and immunofluorescence microscopy.

Antibodies	Source	Identifier	Application
Anti-Atoh1	Proteintech	21215-1-AP	WB
Anti-BCAT2	HuaBio	EM1707-98	WB
Anti-Claudin1	HuaBio	ER1906-37	WB
Anti-ChrA	Abcam	ab15160	WB
Anti-COX-2	HuaBio	ET1610-23	WB
Anti-DMT1	Abcam	ab55735	WB
Anti-FABP2	HuaBio	ER1911-53	WB
Anti-Fpn1	Bioss	bs-4906R	WB
Anti-FtH	Bioss	bs-5907R	WB
Anti-GLUT2	Bioss	bs-0351R	WB
Anti-GPX4	HuaBio	ER1803-15	WB
Anti–HO–1	Bioss	bs-2075R	WB
Anti-Keap1	Abcam	ab119403	WB
Anti-Ki67	Abcam	ab15580	IF
Anti-Lf	Abcam	ab166803	WB
Anti-Lgr5	Abcam	ab75732	WB
Anti-Lyz	Abcam	ab108508	WB
Anti-Muc2	Santa Cruz	(F-2): sc-515032	WB
Anti-NICD	Cell Signaling Technology	4147	WB, IF
Anti-Nrf2	Abcam	ab62352	WB
Anti-Notch1	HuaBio	D6F11	WB, IF
Anti-Occludin	HuaBio	R1510-33	WB
Anti-OLFM4	Cell Signaling Technology	39141	IF
Anti-RBP-Jk	Cell Signaling Technology	5313	WB
Anti-SGLT1	HuaBio	ER1916-51	WB
Anti-SLC7A5	HuaBio	ER1916-84	WB
Anti-SLC7A11	HuaBio	ER1905-55	WB
Anti-TfR	Bioss	bs-0988R	WB
Anti-Vil1	Proteintech	16488-1-AP	WB, IF
Anti-β-actin	Cell Signaling Technology	3700	WB
Alexa Fluor 488 goat anti-rabbit	Abcam	ab150077	IF
Alexa Fluor 594 goat anti-rabbit	Abcam	ab150080	IF

2.6. Measurement of intestinal permeability

Fluorescein isothiocyanate (FITC)-dextran permeability assay was used to measure the intestinal permeability according to previously published methods [17]. Briefly, C57BL/6 mice were orally gavaged with 50 μL of 10 mg/kg B W. FITC-dextran (Sigma, Shanghai, China). After 4 h, blood was harvested by retro-orbital bleed and stored away from light. FITC fluorescence signal was quickly measured in serum samples using a fluorescent plate reader (Molecular Devices, Sunnyvale, USA).

2.7. Histological examination of tissues

Tissue samples were fixed with 4 % paraformaldehyde overnight before being cleaned in PBS, dehydrated with alcohol and embedded in paraffin blocks. Sections of 5-μm were deparaffinized and hydrated, and then stained with hematoxylin and eosin (H&E). Images were taken by Olympus NP-26 (Beijing, China). The villus height (from the tip of the villi to the villus-crypt junction) and crypt depth (depth of the invagination between adjacent villi) were measured and recorded using Image-Pro Plus 7.0 (Media Cybernetics, MD, USA). A total of 6 intact, well-organized crypt villus units were selected for each intestinal cross-section per mouse. The villus height/crypt depth was calculated as the villus height relative to the crypt depth.

2.8. Measurement of Malonyldialdehyde and glutathione content in tissues

Malonyldialdehyde (MDA) and glutathione (GSH) contents were measured using MDA and GSH assay kit (Solarbio, Beijing, China) according to the manufacturer's instructions. Optical density was measured absorbance at 450 nm, 532 nm and 600 nm for MDA content or 412 nm for GSH content by a microplate reader (Molecular Devices, Sunnyvale, USA).

2.9. Electron microscopy

The duodenum was preserved for a night at 4 °C with 2.5 % glutaraldehyde. Analyses using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) were carried out in compliance with published procedures [18]. The SEM observation was carried out utilizing a SU-8010 SEM (Hitachi, Tokyo, Japan). A Hitachi H-7650 TEM (Tokyo, Japan) was used for TEM observation. As for quantification of mitochondrial length, we randomly selected one mouse per group for TEM and take 6 images per mouse. The quantification of mitochondrial length was performed using at least one mitochondria per image. Mitochondrial length was measured using Image-Pro Plus software (Media Cybernetics, MD, USA).

2.10. Immunofluorescence (IF) microscopy

It was standard procedure to deparaffinize and rehydrate duodenum cross-sections. Following antigen retrieval, the sections were blocked, penetrated, and incubated with the primary antibodies and florescent secondary antibodies (Table 2). For immunostaining in enteroids, the enteroids in matrigel were fixed in 4 % PFA, then permeabilized, blocked and incubated with anti-rabbit Ki67 (1: 100, Abcam, Shanghai, China) overnight, and then incubated with anti-rabbit Alexa Fluor 594 (1: 200, Abcam, Shanghai, China). All the samples were counterstained with DAPI (1: 5000, Sangon Biotech, Shanghai, China). A confocal fluorescent microscope (Zeiss, Jena, Germany) was used to obtain the images. Cell numbers were measured using Image-Pro Plus software (Media Cybernetics, MD, USA). For Ki67-positive cell and Lgr5-positive cell counting, a total of 6 crypts were selected for each intestinal cross-section per mouse, respectively.

2.11. Periodic acid-schiff (PAS) staining

Histological slices embedded in paraffin, measuring 5 μm in thickness, were obtained following a series of alcohol washings (from 70 % to 100 %) and deparaffinization. Sections were deparaffinized, hydrated, and then exposed to 1 % periodic acid for 15 min. Next, they were cleaned with water and Schiff's reagent, and finally stained with Gills hematoxylin. Photos were captured with Olympus NP-26 (Beijing, China). Cross sections of PAS staining were quantified using Image-Pro Plus software (Media Cybernetics, MD, USA). For goblet cell counting, a total of 6 field were selected for each intestinal cross-section per mouse.

2.12. Isolation of intestinal crypts and enteroids culture

Enteroids were isolated from fresh intestinal crypts obtained from wild-type and Lgr5^{-EGFP-IRES-CreERT2} transgenic mice as previously described [19] with some modifications. The isolated small intestines were dissected longitudinally. After washing, the tissue fragments about 5 mm were incubated on ice with 2 mM ethylene diamine tetra acetic acid (EDTA) plus PBS for 30 min. After EDTA medium was removed, tissue fragments were suspended with cold PBS. The supernatant (villus fraction) was discarded and the sediment was re-suspended with PBS. After further vigorous suspension and centrifugation, crypts were enriched in the supernatant. Residual villous material were removed through a 40-μm cell strainer (Corning, New York, USA). Isolated crypts were centrifuged at 150–200 g for 3 min to separate the crypts from single cell, and then were embed in a 24-well plate with 50 μL Matrigel® (Corning, New York, USA). After polymerizing at 37 °C for 10 min, fresh IntestiCult™ Organoid Growth Medium (Stemcell technology, Toronto, Canada) were added, and changed every 2 days. The enteroids were passaged every 4–6 days. Iron overloaded enteroid model were established by treatment of 3 mM and 10 mM ferric ammonium citrate (FAC, Sigma-Aldrich, Missouri, USA) on 2 d after passage for 24 h, respectively. The concentrations of FAC were determined by our preliminary experiment. Enteroids in control group were treated with equal volume solvent (PBS) for 24 h. As for Lgr5 positive cell counting, three organoids were established from three Lgr5^{-EGFP-IRES-CreERT2} transgenic mice, and two intact crypts (shown as buds in enteroid image) were randomly selected per organoid.

2.13. Statistical analysis

All the statistical analyses were performed using GraphPad Prism version 8.0 (GraphPad Software, CA, USA) as mean ± SEM. Unpaired two-tailed Student t tests were used to compare statistical differences between the two groups. Tukey's post hoc test was used after one-way ANOVA to examine statistical differences between several groups. P < 0.05 was used as the significance threshold. The symbols *P < 0.05 and **P < 0.01 indicate statistically significant differences.

3. Results

3.1. Iron overload reduced growth performance and damaged duodenal epithelium

In order to explore the role of iron overload in the intestine, we first established an iron overload mice model by intraperitoneal injection of 120 mg/kg B W. iron dextran once a fortnight for 12 weeks. Iron-overloaded mouse model was successfully established by 5.8-fold increased serum iron concentration, 1.7-fold elevated transferrin saturation, excessive iron deposition in the duodenum, and characteristic changes in expression of key iron metabolism associated proteins (Fig. S1). Iron-overloaded mice showed evident growth retardation, with 37 % less body weight gain (P < 0.05) than control mice (Fig. 1A). After 12-week treatment, the average body weight of iron-overloaded mice was 24.6 g, while that of control mice was 27.3 g. However, no significant difference in feed consumption was exist between two groups (Fig. 1B). In addition, the concentration of FITC-dextran in serum of iron-overloaded mice was significantly (P < 0.05) increased (Fig. 1C) compared with that of control mice, which indicated that intestinal permeability was increased after iron treatment. H&E straining images of duodenum tissues of iron-overloaded mice showed slight intestinal injury, including tiny cracks in villus tip and large lumen (Fig. 1D). In iron-overloaded mice, villus height of duodenum was significantly (P < 0.01) decreased and crypt depth was significant (P < 0.05) increased, resulting to a considerable (P < 0.01) decreased by 64 % in the villus height/crypt depth ratio (Fig. 1E–G). TEM analysis confirmed that iron overload induced intestinal damage by causing more scattered and sparser epithelial microvilli in the duodenum (Fig. 1H). Furthermore, as iron is an element inducing oxidative stress in a variety of organs, we also detected reactive oxygen species (ROS)-related proteins in the duodenum. As expected, the protein expression levels of ROS-related proteins including heme oxygenase (HO)-1, cyclooxygenase (COX)-2, nuclear respiratory factor 2 (Nrf2) and Kelch-like ECH-associated protein-1 (Keap1) were all significantly (P < 0.05) increased in the duodenum of iron-overloaded mice (Fig. 1I), which implied oxidative stress was exist in the intestine of iron overload mice. Taking together, iron overload inhibited growth performance, damaged duodenal villus and caused oxidative stress in mice.

Fig. 1.

Iron overload damaged duodenal epithelium. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for 12 weeks. (A) Body weight. (B) Food intake. (C) Duodenal epithelial permeability were measured by fluorescein isothiocyanate (FITC)-dextran concentrations in the serum 4 h after gavage of FITC-4 kDa dextran solution. (D) Representative images for duodenal epithelial morphology by H&E staining (Upper panel: × 200 magnification. Lower panel: × 400 magnification). (E–G) Quantitative analysis of villus height, crypt depth and villus height/crypt depth ratio of duodenum from H&E staining. (H) Representative images for duodenal microvilli by SEM analysis. (I) Representative immunoblot and quantitative analysis showing the expression level of HO-1, COX-2, Nrf2 and Keap1 in the duodenum. n = 5 for each treatment. *P < 0.05 and **P < 0.01 compared with the control group.

3.2. Iron overload mediated duodenal epithelial function

Since villus height and villus height/crypt depth ratio are regarded as important indicators of intestinal absorption function, we assumed that iron overload would impair duodenal absorption with the reduction of villus height and villus height/crypt depth ratio. Transporter proteins on the brush border or basolateral membranes of the intestinal epithelium mediate the absorption of nutrients. Compared with control group, the protein expressions of nutrient transporters, including amino acid transporters (BCAT2 and SLC7A5), fatty acid transporter (FABP2), and glucose transporters (SGLT1 and GLUT2) were all significantly (P < 0.05) decreased in the duodenum of iron-overloaded mice (Fig. 2A and B). Similar trends in their mRNA expression patterns were evident (Fig. 2B). In addition to facilitating the absorption of nutrients, the intestinal epithelium serves as a barrier to keep out dangerous chemicals. The expression levels of intracellular linker protein (ZO-1) and transmembrane proteins (claudin 1 and occludin) can well reflect intestinal barrier integrity. However, there is no significant difference in the expression of ZO-1, claudin 1 and occludin between two groups (Fig. S2). Furthermore, enterocytes contribute intestinal immunity by secreting immune components to resist the invasion of pathogenic bacteria, which is the embodiment of intestinal barrier. In the case of duodenal secretion, the mRNA expression levels of secretory factors, such as CCL9, IL-13 and TFF3 were all significantly (P < 0.05) elevated in iron-overloaded group compared with control group (Fig. 2C). Taking together, iron overload inhibited the expression of nutrient transporters, but enhanced that of secretory factors of the duodenum, indicating that iron overload might affect duodenal absorptive and secretary functions of mice.

Fig. 2.

Iron overload mediated duodenal epithelial function. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for 12 weeks. (A) Representative immunoblot and quantitative analysis showing the expression level of amino acid transporters (BCAT2 and SLC7A5), fatty acid transporter (FABP2) and glucose transporters (SGLT1 and GLUT2) in the duodenum. (B) The mRNA expression levels of BCAT2, SLC7A5, FABP2, SGLT1 and GLUT2 in the duodenum. (C) The mRNA expression levels of secretory factors (CCL9, IL-13, Relm-β, TFF3 and ZG16) in the duodenum. n = 5 for each treatment. *P < 0.05 and **P < 0.01 compared with the control group.

3.3. Iron overload inhibited the proliferation of duodenal ISCs and regulated their differentiation

Since intestinal absorption and secretion functions are performed by enterocytes and secretory epithelial cells respectively, we speculated that iron overload might regulate the differentiation of duodenal intestinal stem cells. Compared with control mice, the mRNA and protein expression levels of ChrA (marker of enteroendocrine cells), Lyz (marker of Paneth cells) and Muc2 (marker of goblet cells) were all significantly (P < 0.05) increased in the duodenum of iron-overloaded mice, while Vil1 (marker of enterocytes) were significantly reduced (Fig. 3A and B). Immunofluorescence and AB-PAS staining showed that Vil1 protein level was obviously reduced and the number of AB-PAS-positive goblet cells was 2-fold increased (P < 0.01) in the duodenum of iron-overloaded mice (Fig. 3C and D). These data inferred that iron overload did influence the differentiation of ISCs. In addition, ISCs can not only differentiate into mature cells, but also self-renew and proliferate to maintain sufficient number of intestinal epithelial cells. The number of OLFM4-positive ISCs was decreased by 88 % (P < 0.01) in the duodenum of iron-overloaded mice compared with that of control mice (Fig. 4A and B). The mRNA expression of Lgr5 (marker of ISCs) showed the likewise reduction with extra iron treatment (Fig. 4C). Similarly, the number of Ki67-positive proliferative cells and the transcriptional level of Ki67 were both significantly (P < 0.05) lower in iron-overloaded group than those in the control group (Fig. 4D–F). Taking together, our results suggested that iron overload could inhibit the proliferation of duodenal ISCs, promote their differentiation into secretory mature cells and prevent their differentiation into absorptive mature cells.

Fig. 3.

Iron overload regulated the differentiation of duodenal ISCs. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for 12 weeks. (A) The mRNA expression levels of markers for enteroendocrine cells (ChrA), Paneth cells (Lyz), goblet cells (Muc2) and enterocytes (Vil1) in the duodenum. (B) Western blot analysis showing the expression level of markers for enteroendocrine cells (ChrA), Paneth cells (Lyz), goblet cells (Muc2), enterocytes (Vil1) and ISCs (Lgr5) in the duodenum. (C) Representative images for Vil1 by immunofluorescence (Upper panel: × 100 magnification. Lower panel: × 400 magnification). (D) Representative images and quantitative analysis for Goblet cells by AB-PAS staining (Upper panel: × 200 magnification. Lower panel: × 400 magnification). AB-PAS-positive goblet cells were counted under × 400 magnification. n = 5 for each treatment. *P < 0.05 and **P < 0.01 compared with the control group.

Fig. 4.

Iron overload inhibited the proliferation of duodenal ISCs. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for 12 weeks. (A–B) Representative images and quantitative analysis for OLFM4 (marker of intestinal stem cells) by immunofluorescence (Upper panel: × 200 magnification. Lower panel: × 400 magnification). OLFM4-positive duodenal stem cells were counted under × 400 magnification. (C) The mRNA expression levels of markers for ISCs (Lgr5) in the duodenum. (D–E) Representative images and quantitative analysis for Ki67 (marker of proliferative cells) by immunofluorescence (Upper panel: × 100 magnification. Lower panel: × 400 magnification). Ki67-positive duodenal proliferative cells were counted under × 400 magnification. (F) The mRNA expression levels of Ki67 in the duodenum. n = 5 for each treatment. *P < 0.05 and **P < 0.01 compared with the control group.

3.4. Iron overload inhibited the proliferation of ISCs and regulated their differentiation in enteroids

In order to verify the effect of iron overload on intestinal cells in vivo, we treated mouse enteroids by FAC to establish an iron-overloaded enteroid model of mice. FAC treatment could clearly affect the growth of mouse enteroids in a dose-dependent manner (Fig. S3A). Compared with control group, mouse enteroids showed complete structure in 3 mM FAC group, while collapsed structure and death when treated with 9 mM or higher concentrations of FAC (Fig. S3A and 5A). Thus, we selected 3 mM and 10 mM FAC as mild and severe degree for the modeling of iron overload mouse enteroids. The Fe²⁺ probe, FerroOrange, was used to monitor the Fe²⁺ concentration in mouse enteroids, and it showed that orange deepened significantly in the 3 mM FAC and 10 mM FAC groups when compared with the control group (Fig. S3B). Furthermore, an increase of FtH mRNA expression was related with a decrease of DMT1, Fpn1, and TfR mRNA expression (Fig. S3C). These findings suggested that the iron-overloaded mouse enteroid model had been successfully established. We then investigated the effect of iron overload on proliferation and differentiation of ISCs in vitro using these enteroids. The growth retardation of enteroids, which caused by FAC treatment, could be a sign of proliferation damage of ISCs (Fig. 5A). Therefore, the proliferation of ISCs was firstly determined. Compared with control group, the numbers of Lgr5-positive ISCs were significantly (P < 0.01) lower in 3 mM FAC group, and they were almost lost in 10 mM FAC group (Fig. 5B and C). The mRNA expression of Lgr5 was also both reduced significantly (P < 0.01) in iron-treated enteroids (Fig. 5D). Similarly, the number of Ki67-positive proliferative cells (Fig. 5E and F) and the mRNA expression of Ki67 (Fig. 5G) were significantly (P < 0.01) decreased both in 3 Mm FAC and 10 mM AFC group compared with control group. In addition, the transcriptional level of Muc2 were significantly (P < 0.01) increased, while those of ChrA, Lyz and Vil1 were significantly (P < 0.01) reduced in the enteroids with iron-overloaded treatment (Fig. 5H). The finding was consistent with the results in vivo, confirming that iron overload can inhibit the proliferation of ISCs, regulate their differentiation into goblet cells and prevent their differentiation into absorptive mature cells.

Fig. 5.

Iron overload inhibited the proliferation of ISCs and regulated their differentiation in enteroids. The mediums of mouse enteroids were added same amount of normal saline (Control), 3 mM FAC or 10 mM FAC. (A) Representative bright-field images of mouse enteroids grown for 5 days (Left panel: × 100 magnification. Right panel: × 400 magnification). (B–C) Representative images and quantitative analysis for Lgr5-EGFP expressing ISCs by immunofluorescence ( × 400 magnification) in mouse enteroids. (D) The mRNA expression levels of Lgr5 in mouse enteroids. (E–F) Representative images and quantitative analysis for Ki67 by immunofluorescence ( × 400 magnification) in mouse enteroids. (G) The mRNA expression levels of markers for enteroendocrine cells (ChrA), Paneth cells (Lyz), goblet cells (Muc2) and enterocytes (Vil1) in mouse enteroids. n = 3 independent experiments. *P < 0.05 and **P < 0.01 compared with the control group.

3.5. Iron overload induced ferroptosis of intestinal epithelial cells

Our previous study revealed that iron overload resulted in ferroptosis in the liver of mice [15]. Therefore, we hypothesized that iron overload may also cause ferroptosis in the duodenum and enteroids of mice. Compared to control mice, the mRNA expression of Ptgs2 (marker of ferroptosis) were significantly (P < 0.05) increased in the iron-overloaded mice either in vivo or in vitro (Fig. 6A and B). MDA, a lipid peroxidation intermediate, was 2.5-fold (P < 0.05) higher in the duodenum of iron-overloaded mice, whereas GSH, a major antioxidant, was 35 % (P < 0.05) lower (Fig. 6C and D). Moreover, compared with the control group, the protein levels of key ferroptosis associated proteins, SLC7A11 and GPX4, were both significantly (P < 0.05) decreased in the duodenum of iron-overloaded mice (Fig. 6E). In addition, TEM analysis revealed that duodenal epithelial cells from iron-overloaded mice had greatly reduced mitochondrial crista and shrunken mitochondria (Fig. 6F), which was quantitated by the mitochondria's reduced long axis length (Fig. 6G). All these findings corresponded to the molecular and subcellular characteristics of ferroptosis, suggesting that iron overload could cause ferroptosis of intestinal epithelial cells of mice.

Fig. 6.

Iron overload induced ferroptosis of ISCs. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight iron dextran once a fortnight for 12 weeks. The mediums of mouse enteroids were added same amount of normal saline (Control), 3 mM FAC or 10 mM FAC. (A) The mRNA expression levels of marker for ferroptosis (Ptgs2) in the duodenum. n = 5 for each treatment. (B) The mRNA expression levels of Ptgs2 in mouse enteroids. n = 3 independent experiments. (C) MDA contents. (D) GSH contents. (E) Representative immunoblot and quantitative analysis showing the expression level of ferroptosis related proteins (SLC7A11 and GPX4) in the duodenum. n = 3 for each treatment. (F–G) Representative images and quantitative analysis of mitochondrial morphology by TEM analysis (Scale bars = 0.2 μm) in the duodenum. Blue arrows indicate mitochondria crista and red arrows indicate shrunken mitochondria. n = 5 for each treatment. *P < 0.05 and **P < 0.01 compared with the control group. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.6. Iron overload inhibited the Notch signaling pathway

Previous studies have reported that the inhibition of Notch signaling pathway would direct intestinal progenitors to secretory cell rather than absorptive cell fate [20], which was similar with our findings. So, we determined the expression levels of Notch signaling pathway related proteins. The mRNA levels of Dll1 and Dll4 (ligands of Notch signaling pathway), Notch1 (a receptor of Notch signaling pathway) and RBP-Jκ (a DNA binding factor of Notch signaling pathway) were all significantly (P < 0.05) reduced in iron-overloaded group compared with control group, either in vivo or in vitro (Fig. 7A and B). Compared with the control enteroids, the mRNA expression of Atoh1 (the negative regulator of Notch signaling pathway) was significantly (P < 0.01) increased in 3 mM FAC treated enteroids (Fig. 7B). The protein expression levels showed the similar trend, with Notch1 and RBP-Jκ greatly reduced and Atoh1 dramatically raised in iron-treated duodenum (Fig. 7C). IF staining confirmed that Notch1 protein level was decreased with iron-overloaded treatment (Fig. 7D). Notch intracellular domain (NICD) is known as the activated form of Notch1 [21]. Both WB analysis (Fig. 7C) and IF staining (Fig. 7E) revealed that the NICD protein level was considerably lower in the iron-overloaded group's duodenum than in the control group. Above results suggested that iron overload inhibited Notch signaling pathway in the duodenum and enteroids of mice.

Fig. 7.

Iron overload inhibited the Notch signaling pathway. Iron overload mice model was established by intraperitoneal injection of 120 mg/kg body weight of iron dextran once a fortnight for 12 weeks. The mediums of mouse enteroids were added same amount of normal saline (Control), 3 mM FAC or 10 mM FAC. (A–B) The mRNA expression levels of Notch related proteins (Dll1, Dll4, Notch1, RBP-Jκ, Hes1 and Atoh1) in the duodenum (n = 5) and mouse enteroids (n = 3). (C) Representative immunoblot and quantitative analysis showing the expression level of Notch1, NICD, RBP-Jκ and Atoh1 in the duodenum (n = 3). (D–E) Representative images of Notch1 and NICD by immunofluorescence (Upper panel: × 100 magnification. Lower panel: × 400 magnification). *P < 0.05 and **P < 0.01 compared with the control group.

3.7. Ferroptosis affected the differentiation of ISCs and inhibited Notch signaling pathway

To further investigate the crosstalk between ferroptosis and Notch signaling pathway in intestinal epithelium, we established a ferroptosis mouse model by intraperitoneal injection of erastin (a classical inducer of ferroptosis) as we previously reported [16]. IF staining showed that the number of OLFM4-positive duodenal stem cells and Ki67-positive proliferative cells were significantly (P < 0.01) decreased by 70 % and 33 %, respectively, in the erastin group compared with the control group (Fig. 8A–D). Moreover, PAS staining results revealed that the number of PAS-positive goblet cells was 2-fold increased (P < 0.01) in the duodenum of ferroptosis mice (Fig. 8C and D). These findings matched the intestinal phenotype of iron-overloaded mice, implying that ferroptosis may also affect the Notch signaling pathway in the duodenum. The mRNA expression of Dll4, Notch1, and RBP-J were found to be significantly (P < 0.05) lower in the erastin group compared to the control group, whereas the mRNA level of Atoh1 was found to be significantly (P < 0.05) higher (Fig. 8E). IF staining showed that the protein level of NICD was significantly lower in the duodenum of ferroptosis mice than that in the control mice (Fig. 8F).

Fig. 8.

Ferroptosis affected the differentiation of ISCs and inhibited Notch signaling pathway. C57BL/6J mice were intraperitoneally injected with saline (Control) or 25 mg/kg body weight of erastin, a classical inducer of ferroptosis, respectively for 2 d at 12-h intervals. (A–B) Representative images of OLFM4 and Ki67 by immunofluorescence (Upper panel: × 200 magnification. Lower panel: × 400 magnification) in the duodenum. (C) Representative images of Goblet cells by PAS staining (Upper panel: × 200 magnification. Lower panel: × 400 magnification) in the duodenum. (D) Quantitative analysis for OLFM4-positive ISCs, Ki67-positive cells, and PAS-positive goblet cells in the duodenum (n = 6). (E) The mRNA expression levels of Notch related proteins (Dll1, Dll4, Notch1, RBP-Jκ and Atoh1) in the duodenum (n = 6). (F) Representative images of NICD by immunofluorescence (Upper panel: × 200 magnification. Lower panel: × 400 magnification) in the duodenum. *P < 0.05 and **P < 0.01 compared with the control group.

4. Discussion

Our study demonstrated that iron overload inhibited the growth performance, decreased the expression of nutrient transporters and enhanced those of secretory factors in the duodenum, damaged duodenal villus and caused oxidative stress of mice. Meanwhile, iron overload inhibited the proliferation of ISCs and regulated their differentiation into secretory mature cells, such as goblet cells. Furthermore, iron overload promoted the ferroptosis of intestinal epithelial cells and inhibited Notch signaling pathway (Fig. 9). In addition, ferroptosis could also inhibit Notch signaling pathway and affect the proliferation and differentiation of ISCs.

Fig. 9.

Schematic overview of the putative mechanism by which iron overload affect the proliferation and differentiation of ISCs. Schematic intestinal epithelia and mesenchymal cells are depicted top left. Schematic organoid is depicted top right. The schemata was created with BioRender.com.

Intestine is the primary site for the digestion and absorption of nutrients, and the proper function of intestinal epithelium is essential for their digestive capacity. We found that iron overload could damage the intestinal epithelium, which is manifested as cracks at the top of villi, messy microvilli, weakened absorption function and exacerbated oxidative stress, thus harming the growth performance of mice. These results were consistent with our previous study [13] and other publications [22]. However, iron overload caused no significant changes in the expression of tight junction proteins in the duodenum. This is similar to the findings that there was no significant difference in the expression of tight junction proteins in the duodenum of rats with high iron diet, but they were decreased significantly in the colon [23] or in the jejunum [22]. These results suggest that the effect of iron overload on intestinal epithelial tight junctions is likely to be intestinal segment specific. Furthermore, iron overload inhibited the expression of intestinal nutrient transporters (BCAT2, SLC7A5, FABP2, SGLT1, and GLUT2). BCAT2 is a key enzyme in branched-chain amino acids catabolism, reversibly catalyzing the first stage of degradation to branched-chain acyl-CoA [24]. SLC7A5 plays a key role in sustaining the intracellular amino acid pool, which is essential for proper mTORC1 signaling and effective protein synthesis [25]. FABP2, also known as intestinal fatty acid binding protein, is specifically expressed in enterocytes of small intestine [26]. The majority of luminal glucose transport across the intestinal epithelium is carried out by SGLT1, and enterocyte metabolism in the intestinal glucose absorption system is facilitated by GLUT2's glucose channel [27]. Consequently, the impairment to intestinal digestion and absorption function may be reflected in the reduced expression levels of these transporters.

In normal physiology, iron is necessary for cellular proliferation. However, we found that the proliferation of duodenal ISCs was inhibited in the duodenum of iron overload mice. Moreover, iron overload changed the differentiation progress of ISCs. ISCs undergo self-renewal and generate enterocytes and secretory progenitors, which then develop into Paneth cells, goblet cells, enteroendocrine cells, tuft cells, and enterocytes [28]. In this study, we found that iron overload promoted the differentiation of ISCs into secretory mature cells, such as Paneth cells, goblet cells and enterondocrine cells, while preventing their differentiation into absorptive mature cells, such as enterocytes. Paneth cells are the lowest-lying cells in the crypt and are well-known for secreting bactericidal substances including defensins and lysozyme. Goblet cells are found on villi and in crypts, and they release mucus and a variety of hormones [29]. The increase of secretory mature cells provided the information that the intestine might be under some stress. The absorptive enterocyte, a highly polarized columnar cell with an intricate lumenal brush border, is the most abundant cell on the villus. The damage to the intestinal villi and the decline in absorption capacity may be intimately linked to the decrease in enterocyte. There are several signaling pathways that control the progression of differentiation, including BMP, Hedgehog, and Notch [30]. The Notch signaling pathway plays a key dual role in the proliferation and differentiation of ISCs. Our results revealed that iron overload inhibited the Notch signaling pathway both in the duodenum and enteroids of mice. The activation of Notch signaling pathway promotes the proliferation and self-renewal of ISCs, and the stem cell area in the crypt is the strongest area of Notch signal in the whole crypt-villous axis [31]. It could explain how iron overload inhibited the proliferation of ISCs. Moreover, ISCs will differentiate more readily into secretory mature cells if the Notch signaling pathway is inhibited, and more quickly into absorptive mature cells if the Notch signaling pathway is activated [30]. This is completely consistent with our results that iron overload inhibited the Notch signaling pathway and regulated the differentiation direction of ISCs to goblet cells. It's the first time to link the relationship between iron overload and Notch signaling pathway in intestinal epithelium. However, in the results of enteroid experiments, the expression of Atoh1 gene expression level was significantly increased when treated with 3 mM FAC but decreased when treated with 10 mM FAC. We speculated that the concentration of 10 mM FAC may be too high, which has an excessive impact on intestinal epithelial cells and broke the cell homeostasis. The cells themselves have changes that cannot be self-regulated, so there is no increase in Atoh1 gene expression. However, the exact mechanism still needs further study.

Ferroptosis is a new Fe²⁺-dependent and nonapoptotic type of cell death. There are two typical characteristics of ferroptosis: the accumulation of Fe²⁺ and the increase of lipid peroxidation which manifested as the increase of MDA content [32]. As of right now, ferroptosis can be detected using three biomarkers: elevated lipid peroxidation, elevated Ptgs2 gene expression, and lowered glutathione peroxidase 4 (GPX4) expression. In the duodenum and enteroids of iron-overloaded animals, we observed classic features of ferroptosis, such as elevated iron levels and increased lipid peroxidation. GPX4 and SLC7A11, two potential molecular indicators of ferroptosis, significantly decreased in response to iron overload, while MDA and Ptgs2 mRNA expression increased markedly. Iron overload inhibited the expression of the amino acid transporter BCAT2, which was used as an inhibitor of ferroptosis [33]. Iron overload increased COX-2 level (an oxidase encoding by Ptgs2 gene that can lead to inflammation and tissue damage), which is not only an important indicator of oxidative stress, but also an important biomarker of ferroptosis [34]. There are also some literatures about iron overload inducing ferroptosis in the liver [35], dopaminergic neurons [36], and motor neurons [37]. Now we innovatively confirmed that ferroptosis occurred in iron-overloaded intestinal epithelial cells. However, the underlying mechanism of ferroptosis in iron-overloaded intestinal epithelium still needs to be further explored.

In addition, a ferroptosis mouse model was established by intraperitoneal injection of erastin to investigate the crosstalk between ferroptosis and Notch signaling pathway in the intestinal epithelium. The classic ferroptosis inducer erastin was found in 2003 to induce cancer cell death, which inhibits the expression of SLC7A11 (a component of the cysteine glutamate reverse transport receptor) on the cell membrane [38]. The cystine glutamate reverse transport receptor on the cell membrane is responsible for the transport of cystine from the extracellular to the intracellular to synthesize the antioxidant GSH, and the conversion of GSH between the reduced state and the oxidized state can help the synthesis of GPX4 [39]. When erastin inhibits SLC7A11, the uptake of cysteine by cells decreases, and the synthesized GSH is continuously consumed and cannot continue to be synthesized, then the synthesis of GPX4 is also hindered, which weakens the ability of cells to scavenge ROS, leading to the accumulation of ROS, thereby triggering ferroptosis [40]. We found that Notch signaling pathway was indeed inhibited in ferroptosis mice. There are some other researches noticing the interaction between ferroptosis and Notch signaling pathway. In endometrial cancer cells, ferroptosis-related regulators affect the occurrence and development of endometrial cancer by regulating the Notch signaling pathway [41]. The potential mechanism of heme-induced ferroptosis in human nucleus pulposus cells may be related to the Notch signaling pathway [42]. Further investigation is needed to determine the precise relationship between ferroptosis and the Notch signaling pathway.

In conclusion, this study demonstrated that iron overload could inhibit the proliferation of ISCs, regulate the differentiation of ISCs into goblet cells, promote ferroptosis of intestinal epithelial cells and block Notch signaling pathway. Ferroptosis also could inhibit the proliferation of ISCs, regulate their differentiation into goblet cells and inhibit the Notch signaling pathway in intestinal epithelium, suggesting that iron overload might inhibit the Notch signaling pathway through ferroptosis. These findings reveal the regulatory role of iron overload on the proliferation and differentiation of ISCs, providing a new insight into the internal mechanism of iron overload affecting intestinal health, and offering important theoretical basis for the scientific application of iron nutrition regulation.

Funding statement

This work was supported by National Key Research and Development Program of China (2023YFD1301104), China; and National Natural Science Foundation of China (82270670, 31872363), China.

CRediT authorship contribution statement

Jing Zhao: Writing – review & editing, Writing – original draft, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Wan Ma: Methodology, Formal analysis, Data curation. Sisi Wang: Methodology, Formal analysis, Data curation. Kang Zhang: Data curation. Qingqing Xiong: Data curation. Yunqin Li: Methodology. Hong Yu: Writing – review & editing, Resources, Methodology, Funding acquisition. Huahua Du: Writing – review & editing, Supervision, Project administration, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that there are no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.