Background
Around 50% of all cancer-related deaths worldwide can be attributed to a wasting condition known as “cachexia”. Cachexia is a complex syndrome that causes ongoing loss of adipose tissue and muscle, and which cannot be reversed with nutritional supplementation [
1‐
3]. The progressive wasting that occurs in cancer cachexia has been suggested to be mediated by circulating factors [
3] (e.g., proinflammatory cytokines, hormones, and metal ions), which can originate from various tissues and have different functions [
3]. It has also been suggested that interventions for cachexia could be developed by targeting inflammatory processes (
e.g., interleukin [IL]-6), although these efforts have not achieved the desired results [
4,
5]. Thus, advances in the basic understanding of cachexia and effective targeting strategies are needed.
The protein lipocalin 2 (LCN2) (also known as neutrophil gelatinase-associated lipocalin, siderocalin, or 24p3) functions as a mediator in several diseases associated with cachexia, including cancer, pneumonia, and kidney disease [
6‐
9]. LCN2 has been demonstrated in mechanistic studies to function as an iron-regulatory protein under physiological and inflammatory conditions. In prokaryotes, LCN2 inhibits bacterial siderophores from acquiring iron, thus inhibiting bacterial growth [
10]. In mammals, a study using a mouse model of leptomeningeal metastasis showed that cancer cells use lipocalin 2 to collect iron [
6]. However, LCN2 in the hypothalamus interacts with the melanocortin-4 receptor, which has been shown to mediate anorexia and promote the loss of lean (skeletal muscle) and fat (adipose tissue) mass in cancer cachexia [
11,
12].
Ferroptosis is a form of cell death. It is driven by iron-dependent phospholipid peroxidation and regulated by multiple metabolic and signaling pathways [
13,
14]. Since its discovery in 2012, diverse injuries to many organs and various malignant lesions have been pathogenically associated with ferroptosis [
13]. In recent years, multiple studies have reported links between ferroptosis and cancer progression. Egolf et al
. [
15] reported that knocking out the gene of an epigenetic regulator, myeloid/lymphoid or mixed-lineage leukemia 4, leads to ferroptosis inhibition and the development of pre-cancerous skin lesions in mice. Ma et al
. [
16] showed that cholesterol in the tumor microenvironment leads to increased cluster of differentiation (CD)36 expression in CD8
+ T cells and causes these cells to take up polyunsaturated fatty acids and to initiate ferroptosis. Liao et al. [
17] reported that CD8
+ T cells orchestrate tumor ferroptosis via acyl-coa synthetase long chain family member-4. Although a link between cachexia and ferroptosis has not been reported, many therapy-resistant cancer cells (especially those prone to metastasis) are highly susceptible to ferroptosis [
18]. Accordingly, it has been proposed that ferroptosis could be regulated by pharmacologic means to treat drug-resistant cancers [
19].
In the present study, we found that the ferroptosis of adipose tissue cells caused tissue wasting in experimental models of lung cancer cachexia. Mechanistically, we demonstrated that wasting tissue had an increased number of tissue-infiltrating neutrophils (TI-Neu), and showed that these cells promoted ferroptosis and tissue wasting through LCN2 secretion. Moreover, we showed that the chemical inhibition of ferroptosis inhibited tissue wasting in experimental models of lung cancer cachexia.
Methods
Mice
6–10 week-old male C57BL/6 (Cat# T002040), BALB/cJGpt-Foxn1nu/Gpt mice (Cat# D000521) were obtained from GemPharmatech (Nanjing, China). The Lcn2f/+ (Cat# NM-CKO-00134) and LysMcre mice (Cat# NM-KI-18018) were obtained from ShangHai Model Organisms. Lcn2f/f mice were crossed with LysMcre mice to obtain Lcn2f/f;LysMcre mice. All mice used were housed under specific pathogen-free conditions. All procedures involving experimental animals were approved by the Ethics Committee of the University of Science and Technology of China and were performed in accordance with the National Guidelines for Animal Usage in Research.
Cell lines
The Lewis lung carcinoma (LLC) and 3T3-L1 cell lines were obtained from The Cell Bank of Type Culture Collection of the Chinese Academy of Sciences (Cat# TCM 7 and Cat#SCSP-5038, respectively). Both cell lines were cultured in DMEM supplemented with 10% FBS (HyClone), 1% streptomycin and penicillin, and were maintained at 37 °C and 5% CO2. The cell lines were routinely tested for mycoplasma using the TransDetect PCR mycoplasma detection kit (Transgen, Cat#FM311).
2 days after the 3T3-L1 cells had reached 70% confluence, the cells were treated with 1 μM dexamethasone (Selleck, Cat#S1322), 5 μg/mL insulin (Novoprotein, Cat#P05019), 0.5 μM isobutylmethylxanthine (IBMX) (Selleck, Cat#S5836), and 1 μM rosiglitazone (Selleck, Cat#S2556). Another 2 days later, the cells were treated with 5 μg/mL insulin and 1 μM rosiglitazone. Starting on day 6, the cells were cultured in DMEM supplemented with 10% FBS, 1% streptomycin and penicillin, and treated with 200 ng/mL recombinant mouse LCN2 protein (Novoprotein, Cat#P11672) for 24 h.
Human samples
Serum samples from healthy donors and from lung cancer patients with and without cachexia were obtained from the First Affiliated Hospital of the University of Science and Technology of China. Patients were diagnosed with cachexia if they had lost > 5% of their body weight over the past 6 months or had a body mass index (BMI) < 20, plus had three of the following criteria: anorexia, decreased muscle strength, fatigue, low fat-free mass index, or abnormal biochemistry results, including increased levels of inflammatory markers (
e.g., CRP and IL-6), anemia, or low serum albumin. All samples were collected with the informed consent of the patients, and the experiments were approved by the Ethics Committee of the University of Science and Technology of China (2020-research-36). Details relating to the patients’ cancer type and cachexia are listed in Additional file
4: Table S5.
Tumor models
LLC-induced cachexia model was established as previous study described [
20]. In brief, the male C57BL/6 mice were subcutaneously inoculated with LLC cells (5 × 10
6 per mouse). Mice were randomly divided into treatment groups while ensuring that the average body weight in each group was roughly the same. Mice which showed significant loss (> 20%) in body weight were defined as having cachexia. Then, they were euthanized and WAT and muscle tissues were harvested to further confirm the cachexia symptom. We comprehensively compared the cachexia phenotype of mice at 1,2,3 weeks and finally determined the analysis of samples on day 21.
For lung cancer patient-derived xenograft (PDX)-induced cachexia, whether or not cachexia occurs is dependent upon the source of the tissue from the tumor patient. We established based on the 3#-Ade PDX of a lung adenocarcinoma. For specific steps and methods, PDX tumors in cold Dulbecco’s Modified Eagle’s Medium (DMEM) were minced into 30–50 mm3 fragments, and each fragment was subcutaneously transplanted into the dorsal flank of 6- to 10-week-old male BALB/cJGpt-Foxn1nu/Gpt mice. Body weight of these tumor-bearing mice were monitored regularly. Mice who showed significant loss (> 20%) in body weight were defined as having cachexia. Then, they were euthanized and WAT and muscle tissues were harvested to further confirm the cachexia symptom.
Animal treatment protocol
To deplete neutrophils, the male C57BL/6 mice were subcutaneously inoculated with LLC cells (5 × 106 per mouse) on day 0 and then intraperitoneally injected on days − 1, 1, 4, 6, 8, 11, 13, 15, 17, and 20 with a 100 μg dose of an anti-Ly6G (Bio X Cell, Cat# BE0075-1; RRID:AB_1107721) or an isotype IgG (Bio X Cell, Cat# BP0090, RRID: AB_2891360).
To deplete LCN2, male C57BL/6 mice were subcutaneously inoculated with LLC cells (5 × 106 per mouse) on day 0 and then were intraperitoneally injected on days 4, 7, 11, 14, 17, and 20 with a 50 μg dose of an anti-LCN2 (Novus Biologicals, Cat# AF1857, RRID:AB_355022) or an isotype IgG (Bio X Cell, Cat# BP0090, RRID: AB_2891360).
For deferoxamine (DFO) therapy, male C57BL/6 mice were subcutaneously inoculated with LLC cells (5 × 106 per mouse) on day 0 and then intraperitoneally injected with 15 mg/kg DFO (Selleck, Cat#S5742) on days 4, 7, 10, 13, 16, and 19.
For liproxtatin-1 therapy, male C57BL/6 mice were subcutaneously inoculated with LLC cells (5 × 106 per mouse) on day 0 and then intraperitoneally injected with 10 mg/kg liproxtatin-1 (Selleck, Cat#S7699) daily between days 1 and 20.
Lentivirus production and delivery
The pCDH-CMV-MCS-EF1-Lcn2 (Sangon Biotech) or pCDH-CMV-MCS-EF1 vectors (YouBio, Cat# VT1479) were extracted using the Endo-Free Plasmid DNA Mini Kit II (Omega, Cat#D6950) and co-transfected with the pRSV-Rev (YouBio, Cat# VT1445), pLP/VSVG (YouBio, Cat# VT1491), and pNL-GFP-RRE (YouBio) plasmids into 293 T cells using the lipofectamine (Invitrogen, Cat#11668019) transfection method. Supernatants containing the LCN2-expressing or control lentiviruses were collected 48 and 72 h later and centrifuged at 50,000 × g for 2 h at 4 °C to purify the virus. For overexpression of Lcn2, 2 × 109 plaque-forming units (PFUs) of LCN2-expressing lentivirus were injected intravenously into C57BL/6 mice once per week.
Flow cytometry
Leukocytes were isolated from the epididymal and inguinal white adipose tissue (eWAT and iWAT, respectively), gastrocnemius skeletal muscle (Gast), bone marrow, and blood, as previously described [
21,
22]. For intracellular staining, leukocytes were incubated with PMA (50 ng/mL), ionomycin (1 mg/mL), and monensin (10 ng/mL) for 4 h at 37 °C and 5% CO
2, followed by staining for surface markers for 30 min at 4 °C. Cells were fixed and then permeabilized with the Foxp3/Transcription Factor Staining Buffer and incubated with fluorescent antibodies for 30 min at 4 °C. Cells were acquired by flow cytometry (LSR II).
For analysis of human blood samples, blood from cachectic cancer patients was centrifuged and lysed using red blood cell lysis buffer. Cells were then stained for surface markers for 30 min at 4 °C. Cells were fixed and then permeabilized with the Foxp3/Transcription Factor Staining Buffer and incubated with the fluorescent antibodies for 30 min at 4 °C. Cells were acquired by flow cytometry (LSR II). Antibodies and related materials used in this study are listed in Additional file
1: Table S1. Data analysis was performed using FlowJo 10 software.
Isolation of adipocytes, neutrophils, and macrophages
eWAT samples from mice were cut into pieces and digested in DMEM with collagenase I (1 mg/mL) while shaking at 220 rpm for 30 min at 37 °C. The suspensions were filtered through sieves, and the filtrates were centrifuged at 500 × g for 5 min to separate the suspended adipocytes from pelleted leukocytes. Neutrophils were purified using the Neutrophil Isolation Kit (Miltenyi Biotec, Cat#130-097-658). Macrophages were purified using a PE-F4/80 antibody (eBioscience, Cat#12-4801-80; RRID:AB_465922) and anti-PE Microbeads (Miltenyi Biotec, Cat#130-048-801).
ELISA
The concentrations of human IL-6, human CRP, human LCN2, and mouse LCN2 in cell culture supernatants and serum were measured by ELISA, according to the manufacturers’ instructions; the kits used are listed in Additional file
1: Table S1.
Histological analysis
Tissues were fixed overnight with 10% neutral-buffered formalin, dehydrated, embedded in paraffin, and sectioned. The 4 μm slices were then stained with hematoxylin and eosin (H&E).
Immunofluorescence
Tissues were collected as described above. Tissues were fixed overnight with 10% neutral-buffered formalin, dehydrated, embedded in paraffin, and sectioned into 4 μm slices. The slides were dewaxed, rehydrated, and subjected to heat-induced epitope retrieval, followed by incubation with 5% goat serum for 1 h at room temperature to block non-specific antibody binding. Next, the sections were incubated with a primary anti-LCN2/NGAL antibody (Abcam, Cat# ab216462) overnight at 4 °C and then with an Alexa-Fluor-546-conjugated goat anti-rabbit IgG (5 μg/mL; Invitrogen, Cat# A-11010, RRID:AB_2534077) secondary antibody. Nuclear staining was performed using DAPI (5 min incubation). The stained sections were imaged using the LSM 880 Confocal Microscope (Zeiss, Jena, Germany) and analyzed with Image J software.
Iron assay
The concentrations of iron (Fe2+) in eWAT, iWAT, and Gast tissues were measured using the Iron Assay Kit (Sigma Aldrich, Cat#MAK025). eWAT, iWAT, and Gast tissues were homogenized in Iron Assay buffer by centrifugation at 16,000 × g for 10 min at 4 °C. The samples were then incubated with an iron reducer in a 96-well plate for 30 min at room temperature, and finally with an iron probe for 60 min at room temperature. Absorbance at 593 nm was measured using a microplate reader.
Lipid reactive oxygen species (ROS) analysis
Adipocyte lipid ROS production was measured using the Lipid Peroxidation Assay Kit (Abcam, Cat#ab243377). Adipocytes purified from eWAT or iWAT were stained with the Lipid Peroxidation Sensor for 30 min at 37 °C and analyzed immediately by flow cytometry (LSR II). The oxidized and non-oxidized lipids were detected on the FITC and PE channels, respectively. The FITC to PE mean fluorescence intensity (MFI) ratio was calculated for each sample. Data analysis was performed using FlowJo 10 software.
The malondialdehyde (MDA) lipid peroxidation assay
Lipid peroxidation was analyzed using the MDA Assay Kit (Sigma Aldrich, Cat#MAK085). Briefly, eWAT, iWAT, and Gast tissues were homogenized in MDA Lysis Buffer by centrifugation at 13,000 × g for 10 min. MDA in epididymal adipose tissue was mixed with TBA to generate the MDA-TBA adduct. Absorbance at 532 nm was measured using a microplate reader.
Quantitative (q)PCR assays
RNA was extracted from purified cells or frozen tissue samples using the TRIzol reagent (Invitrogen, Cat#15596018), glycogen (Thermofisher, Cat#AM9515), and sodium acetate (Thermofisher, Cat#R1181). RNA was reverse transcribed into cDNA using the M-MLV Reverse Transcriptase (Invitrogen, Cat#28025013). qPCR was then performed using the SYBR Green Premix Pro Taq HS qPCR Kit (Accurate biology, Cat#AG11701) on a LightCycler 96 instrument (Roche). PCR was performed using the 2 × TransTaq-T PCR SuperMix (Transgen, Cat#AS122). Relative mRNA levels were calculated using the 2
−ΔΔCt method and normalized to actin mRNA levels. Target gene primers are shown in Additional file
1: Table S2.
RNA sequencing (seq)
Total RNA was extracted from the eWAT of cachectic and control mice using the miRNeasy Mini Kit and treated with DEPC-treated water. A total amount of 1 µg RNA per sample was used as input material for the RNA sample preparations. RNA libraries were prepared for sequencing using the NEBNext UltraTM RNA Library Prep Kit (Illumina). Clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina), according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina Novaseq platform and 150 bp paired-end reads were generated. Raw data (raw reads in fastq format) were firstly processed through in-house perl scripts. Clean data were obtained by removing reads containing adapter and poly-N sequences, and low-quality reads. The reference genome and gene model annotation files were downloaded from the genome website directly. The reference genome index was built using Hisat2 software, and paired-end clean reads were aligned to the reference genome. FeatureCounts software was used to count the read numbers mapped to each gene. Fragments per kilobase of exon model per million reads mapped (FPKM) were calculated for each gene based on the length of the gene and the read counts mapped to this gene. Differential expression analysis of two groups was performed using the DESeq2 R package (1.16.1). The resulting P-values were adjusted using the Benjamini and Hochberg’s approach to control for the false discovery rate. Genes with a DESeq2-derived adjusted
P-value < 0.05 and log2fold change (FC) ≥ 2 were labelled as differentially expressed genes (DEGs; listed in Additional file
5: Table S6). The DEG heatmap was analyzed using MEV software Gene Ontology (GO) enrichment analysis of DEGs was implemented using the clusterProfiler R package. GO terms with a corrected
P-value < 0.05 were considered significantly enriched DEGs. We used clusterProfiler R package to test the statistical enrichment of DEGs in various KEGG enrichment pathways. For the Gene Set Enrichment Analysis (GSEA), the genes were ranked according to the degree of differential expression in the two samples being compared. The predefined Gene Sets were then tested to see if they were enriched at the top or bottom of the list. The RNA-seq data generated in this study were deposited in the GEO database repository and can be accessed using with the accession number GSE188479.
Serum protein array
The relative expression levels of human serum proteins in lung cancer patients with and without cachexia and in healthy donors were measured using the G-Series Human Cytokine Antibody Array 440 (RayBiotech). The relative expression levels of mouse serum proteins in cachectic and control mice were determined using the Quantibody® Mouse Cytokine Antibody Array 4000 (RayBiotech). Fluorescence signals were visualized in the Cy3 channel of a laser scanner. Data were extracted using the GAL file from
www.RayBiotech.com/Gal-Files.html. Heatmaps of differentially expressed proteins (DEPs) were analyzed using MEV software.
Western blotting
Purified adipocytes, neutrophils, and macrophages were lysed in RIPA buffer (Beyotime, Cat#P0013B) containing protease inhibitors on ice for 30 min, followed by centrifugation at 16,000 ×
g for 10 min at 4 °C. Protein concentrations in the supernatants were measured using the BCA Protein Assay Kit (Thermofisher, Cat#23225), and 50 μg aliquots of proteins were incubated at 95 °C for 10 min and separated on SDS-PAGE gels. The proteins were transferred to PVDF membranes, which were blocked in a solution of 5% milk in TBS buffer containing 0.1% Tween-20 for 1 h at room temperature. The membranes were then incubated with primary antibodies overnight at 4 °C and subsequently with HRP-conjugated secondary antibodies for 1 h at room temperature. The protein bands were detected by chemiluminescence autoradiography. The antibodies used are listed in Additional file
1: Table S1.
Statistical analysis
All statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA, USA). Results are presented as the mean ± standard error of the mean (SEM) and were compared using unpaired t-tests, one-way ANOVA, or two-way ANOVA. Mouse survival was estimated using the Kaplan–Meier method and compared using the log-rank test. Patient sample data were compared using the Wilcoxon signed rank test and Spearman’s rho was calculated as indicated.
Discussion and conclusions
The incidence of cachexia among cancer patients is relatively high, especially those with cancer of the pancreas, gastrointestinal tract, colon, or lung. Cachexia symptoms can appear early, even if the primary tumor is localized. These systemic changes affect many peripheral tissues that are not proximal to the tumor (
e.g., the muscles essential for breathing, moving, chewing, and swallowing food) and are often detrimental to the host [
34]. Furthermore, weakened muscle and adipose tissue reduce the tolerance of cancer cachexia patients to anti-tumor therapies. For instance, weakening of the heart muscles and diaphragm muscles often leads to premature death from heart failure or respiratory failure [
2,
35]. Frustratingly, efficacious treatment for cancer cachexia is lacking, despite > 100 clinical trials of mediators developed to treat this condition [
2]. In addition, many reported cachexia mediators target tumor metastasis-related cachexia but very few target cachexia arising from localized tumors or early tumors. This scenario is suboptimal given that the mediators of cachexia may differ between metastatic tumors and localized primary tumors [
2,
36].
We established, by experimental means, that the LCN2-induced ferroptosis of tissue parenchymal (
e.g., adipose) cells is one of the causes of tissue wasting in mice with non-metastatic lung cancer cachexia. Ferroptosis has been demonstrated to induce organ injury and various degenerative changes in diverse diseases [
19]. For example, there is strong evidence that ferroptosis contributes to ischemia–reperfusion injury, including stroke and ischemic disease of the heart, liver, and kidney [
37]. DFO (Desferal®) is a drug approved for the treatment of acute iron overdose and chronic iron overload resulting from repeated blood transfusions. It is an iron-chelating agent that binds excess free iron and forms a stable complex that inhibits ferroptosis [
38]. We found that the DFO-mediated inhibition of ferroptosis significantly and reduced wasting of adipose tissue in a mouse model of lung cancer cachexia. In addition, DFO treatment lengthened the survival of these mice significantly.
LCN2 is a mediator implicated in several diseases: cachexia, cancer, pneumonia, and kidney disease [
6‐
9]. There are two forms of LCN2 under physiological conditions. It is now clear that the function of the iron-free form of LCN2 is distinct from that of the iron-loaded from [
39]. Meier et al
. [
40] reported that iron-loaded LCN2 promoted the growth and progression of renal cell carcinoma, whereas iron-free LCN2 exerted anti-tumoral activity; however, the mechanistic details of cancer-related LCN2 signaling are not known. Iron-free LCN2 has been used as a marker for renal regeneration [
41]. Incidentally, the exogenous LCN2 used in our study was the iron-free form. Therefore, further studies assessing the impact of the iron-loaded form of LCN2 on ferroptosis are needed to deepen our understanding of the diverse functions of LCN2.
LCN2-related signaling varies among diseases affecting different organs. Liu et al
. [
42] reported that the stress-responsive transcription factor nuclear protein-1 transactivates LCN2 in pancreatic cancer cells which, in turn, induce ferroptosis resistance . Yao et al
. [
43] reported that an leukemia inhibitory factor receptor (LIFR)/nuclear factor-kappa B (NF-κB)/LCN2 axis controls liver tumorigenesis and vulnerability to ferroptosis. They showed that loss of LIFR activates NF-κB signaling, thereby leading to upregulation of expression of the iron-sequestering cytokine LCN2 which, in turn, depletes iron and renders liver cells insensitive to ferroptosis inducers.
LCN2 has been reported to mediate appetite suppression during pancreatic cancer cachexia [
12]. We found that LCN2 overexpression in healthy mice or LCN2 depletion in LLC tumor-bearing mice did not influence the food intake of animals significantly. We established that TI-Neu-derived LCN2 induced the ferroptosis of adipocytes directly, leading to the wasting of adipose tissues. A therapeutic antibody targeting LCN2 prolonged the survival of tumor-bearing mice effectively and prevented wasting of adipose tissue and muscle in these animals. In addition, LCN2 expression was upregulated in lung cancer patients and was associated with cachexia progression. Collectively, these findings indicate that LCN2 may represent a valuable target in the treatment of cachexia caused by lung cancer and other types of cancer.
Wang and colleagues suggested that LCN2 knockdown protected a lipopolysaccharide- induced model of acute respiratory distress syndrome via inhibition of ferroptosis-related inflammation and oxidative stress by inhibiting the mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) pathway [
44]. Huang et al
. [
45] found that signal transducer and activator of transcription-3 (STAT3)-mediated lysosomal cell death promoted ferroptosis in PDAC. They found the MEK-ERK pathway promotes STAT3 activation in ferroptosis, and that STAT3 contributes to erastin-induced ferroptosis. Wang and colleagues [
46] reported the crosstalk of Lcn2/Janus kinase 2 (JAK2)-STAT3 in neurotoxic microglia and astrocytes. We also found the JAK-STAT pathway to be enriched in the eWAT of cachectic mice (Fig.
2B). Thus, LCN2 may induce ferroptosis by activating the STAT3 pathway.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.