Background
The incidence of thyroid cancer has augmented dramatically worldwide [
1,
2], and it recently ranks third of all malignant tumors among Chinese women [
3]. Thyroid cancers are classified into differentiated and undifferentiated thyroid cancers [
4]. The former accounts for the vast majority of thyroid cancers, which can be efficiently cured by thyroidectomy combined with postoperative thyroid hormone suppression and radioactive iodine treatment [
5]. However, there are still minority patients who relapse and develop into undifferentiated thyroid cancer, which are resistant to most conventional therapy, resulting in poor prognosis and fatal outcomes [
6]. Thus, there is an urgent need to develop effective and safe therapy for this disease.
Mannose, a natural hexose existing in daily food, takes part in several physiological processes, such as energy metabolism, protein glycosylation and immune reaction [
7‐
9]. As mannose is not detrimental to human health, some studies have explored its clinical function as a drug. For example, mannose has been demonstrated to treat urinary tract infections, especially in female patients who suffer from recurrent infections [
10,
11]. There is also study showing that mannose can alleviate the progression of type 1 diabetes by activating Treg cells and suppressing immunity [
12,
13]. Besides, a previous study proved that mannose could change the intestinal flora, thereby improving energy metabolism. Thus, it is considered as a potential weight-loss medicine [
14]. In recent years, more attention has been paid to its antitumor role. For example, a previous study showed that higher concentration of plasma mannose predicted better prognosis for patients with esophageal adenocarcinoma [
15]. It has been revealed that mannose preferentially killed cancer cells which expressed lower phosphate mannose isomerase (PMI) [
16].
PMI is ubiquitous in prokaryotic and eukaryotic cells as a housekeeping enzyme, which reversibly catalyze the conversion of mannose-6-phosphate (M-6-P) and fructose- 6-phosphate (F-6-P) to participate in energy metabolism [
17]. Mannose is transported into tumor cells through glucose transporter GLUTs and converts into M-6-P by hexokinase. When PMI is lowly expressed in tumor cells, M-6-P cannot be transformed into F-6-P, leading to accumulation of M-6-P and inhibition of glycolysis. In this way, the response of tumor cells to mannose is related to PMI protein expression [
16]. It is clear that the function of PMI relies on zinc ions (Zn
2+) which participate in the structure and activity of several proteins [
17,
18].
The metal ion transporter families SLC39 and SLC30 regulate import and export of Zn
2+ to meet steady state of cell environment. The SLC30 family (ZnT family) contains ten members which are responsible for transporting Zn
2+ from the cytoplasm to extracellular area or organelles, while the SLC39 family (alias ZIP family) consists of 14 members that shift Zn
2+ from extracellular area or organelles to the cytoplasm [
19,
20]. ZIP10, an important member of the SLC39 family, exerts various physiological and pathological functions by affecting Zn
2+ concentration and changing enzyme activity. For example, there is evidence showing that ZIP10 influences histone acetyltransferase by controlling intracellular Zn
2+ concentration, keeping skin healthy [
21]. Meanwhile, it has been proved that ZIP10 knockdown induces apoptosis in early B-cell as ZIP10 regulates caspase activity by Zn
2+ balance [
22].
In this study, our data showed that mannose could selectively kill thyroid cancer cells by a series of in vitro and in vivo studies, and this effect was highly dependent on ZIP10 expression levels, but not PMI expression levels.
Materials and methods
Cell culture and drug treatment
Human thyroid cancer cell lines TPC-1, BCPAP, FTC133, IHH4, 8305C, 8505C and K1 were kindly provided by Dr. Haixia Guan (Guangdong Provincial People’s Hospital, Guangzhou, P.R. China). These cell lines were authenticated by analyzing short tandem repeat at Genesky Co. Ltd., (Shanghai, P.R. China), and the results (Additional file
1: Table S1) were consistent with a previous study [
23]. We routinely cultured these cells at 37 °C in RPMI-1640 (Gibco) or DMEM/Ham’s F-12 (Gibco) medium with 10% fetal bovine serum (FBS), and treated cells with D-Mannose (Aladdin) or TPEN (Sigma) at the indicated concentrations and time points.
Cells (800 to 2000/well) were seeded in 96-well plates. After cell attachment to the plate, different doses of D-mannose were added into culture medium at the time points. Next, we performed the MTT assay to determine the effect of mannose on cell proliferation, and then calculate IC
50 value of each cell line as described previously [
24].
Cells (3000 to 5000/well) were seeded on 12-well plates and cultured with medium containing gradually increasing concentrations of D-mannose for 7 days. After cells were fixed, washed and stained, colony number was counted under an inverted microscope. We defined more than 50 cells as a colony. Each assay was carried out in triplicate.
Cell cycle assay
After attachment to cell plate, cells were cultured in serum-free medium for 12 h and treated with mannose or not for 24 h. Next, we fixed cells with 66% cold methanol for at least 2 h. Cells were then stained with PI and subjected to flow cytometry to analyze cell cycle distributions.
Short interfering RNAs (siRNAs), lentivirus short hairpin RNAs (shRNAs) and expression plasmids
Oligonucleotides of siRNAs targeting PMI or ZIP10 and control siRNAs were purchased from Ruibobio (Guangzhou, China), and the sequences were showed in Additional file
1: Table S2. One day before transfection, cells were seeded on a 12-well plate to achieve 50% confluence overnight. Next, we transfected these cells with the above siRNAs at a final concentration of 50 nM using X-treme GENE siRNA Transfection Reagent (Roche Diagnostics).
Lentivirus encoding PMI-shRNA, ZIP10-shRNA, control shRNA, PHBLV-ZIP10 and PHBLV-vector were purchased from HanBio Biotechnology Co., Ltd. The sequences of shRNAs were shown in Additional file
1: Table S3. Cells were cultured to achieve 50% confluence and transfected with different constructs or a final lentivirus multiplicity of infection (MOI) of 10–100.
RNA extraction and quantitative RT-PCR (qRT-PCR)
RNA extraction, cDNA synthesis and qRT-PCR assays were performed as described previously [
25]. The mRNA expression was normalized to
18S rRNA. The primer sequences were shown in Additional file
1: Table S4. Each sample was analyzed in triplicate.
Western blot analysis
Cells were cultured and treated with the indicated conditions. After cells were washed and lysed, equal amounts of protein lysates were subjected to 10% SDS-PAGE electrophoresis, and transferred onto polyvinylidene fluoride membranes (Roche Diagnostics GmbH, Mannheim, Germany). Next, we incubated the membranes with primary antibodies at 4 °C overnight as follow: anti-PMI (Abcam), anti-ZIP10 (Novus Biologicals), anti-cyclin D (Abcam), anti-cyclin E (Santa Cruz), anti-phospho-CDK2 (pCDK2, CST), anti-CDK2 (ABclonal), anti-p53(Santa Cruz), and anti-β-actin (Abcam). After being immunoblotted with corresponding secondary antibodies, immunoblotting signals were collected using the Western Bright ECL detection system (Advansta, Menlo Park, CA).
Measurement of PMI enzyme activity
Cysteine carbazole sulfuric acid method was used to measure enzyme activity of PMI as described previously [
26]. In brief, cells were washed, and then lysed by three freeze–thaw cycles and Ultrasonic cracker on ice. Next, the reactions were initiated by the addition of equal protein samples into the reaction buffer containing 40 mM Tris-HCl pH 7.4, 6 mM MgCl
2, 5 mM Na
2HPO
4/KH
2PO
4 and 20 mM mannose-6-phosphate. After a 2-h incubation, the reactions were stopped by adding 1.5% cysteine hydrochloride and 0.12% alcoholic solution of carbazole with concentrated sulfuric acid. After shaking this reaction buffer, the amount of complex formed was estimated spectrophotometrically in a spectrophotometer at 560 nm at room temperature. In parallel, β-actin was detected by western blot analysis to prove protein amount consistency.
Measurement of intracellular Zn2+
Cells were seeded in a 6-well plate and treated with TPEN or the indicated siRNAs. After a 48 h-treatment, cells were washed and incubated with FluoZin™-3 (Invitrogen) at concentration of 1 μM for 30 min in a dark 37 °C carbon dioxide incubator. Next, we washed and incubated cells with PBS at 37 °C for 20 min in a carbon dioxide incubator to de-esterify. The concentration of intracellular Zn2+ was then measured by flow cytometer.
Animal studies
Eight-week-old female BALB/c athymic mice were purchased from SLAC laboratory Animal Co., Ltd. (Shanghai, China). Next, we subcutaneously injected 8305C/sh-NC cells (1 × 107), 8305C/sh-ZIP10 cells (1 × 107) and FTC133 cells (1 × 107) into armpit region of these mice to establish xenograft mouse model, and randomly divided them into two groups (vehicle control vs. mannose; five mice/group). Normal water was then replaced by 15% mannose for mannose group. Meanwhile, mice received 20% mannose water or sterilized water by oral gavage (150 μL) four times per week. We weighted mice every 2 days, measured length and width of tumors using digital caliper, and calculated tumor volumes according to the formula: length × width2 × 0.5. In the end of the experiments, the mice were given final oral gavage of mannose and sacrificed 5 h later. Tumors were gathered and weighted. Part of xenograft tumors was used for western blot analysis, while the others were used for IHC assay. The antibody used for IHC assays were as follows: Ki-67 (Abcam), cyclin D (Abcam), cyclin E (Santa Cruz) and pCDK2 (ABclonal). Image J software was used to analyze the staining levels of Ki-67, cyclin D, cyclin E and pCDK2. The staining point was then scored 0, 1, 2, 3 representing negative, weak positive, positive and strong positive. Next, we quantified staining levels using the score multiply relating proportion and plus together. The staining levels of each sample was presented relative to the deepest staining of control group.
TPO-Cre and
BrafCA mice were kindly provided by Drs. Kimura Shioko (National Institutes of Health, USA) and Martin McMahon (University of California, USA), respectively. These two strains of mice mated and produced mice with
BrafV600E-driven thyroid cancer according to a previous study [
27]. Mice were randomly divided into two groups, and then received 20% mannose water or sterilized water by oral gavage (200 μL) every day, respectively. Next, we detected tumor burden by Vevo1100 ultrasound imaging system every week. The area of largest thyroid cross-section was quantified and normalized to the size at week 4 as described previously [
28]. At the end of the experiments, all mice were sacrificed, and thyroids were collected and weighted. The tissues were then fixed, embedded and sectioned for IHC assays. The above experiments were approved by the Laboratory Animal Center of Xi’an Jiaotong University.
Seahorse glycolytic stress test
The extracellular acidification rate (ECAR) which indicates for glycolysis function was measured by using seahorse XF glycolysis stress kit and XF96 analyzer (seahorse biosciences) as described previously [
29]. First, cells (20000–40,000 cells/well) were seeded in seahorse XF96 microplates. After attachment to the plate, cells were treated with 20 mM D-mannose or not for 36 h, washed twice with base medium (seahorse bioscience) containing 2 mM glutamate, and incubated in a non-CO
2 incubator for 1 h. after calibration, a final concentration of 10 mM glucose, 1 mM oligomycin, and 50 mM 2-deoxy-D-glucose were injected sequentially into cells. Changes in extracellular acid rate (ECAR) after the injection of glucose (activating glycolysis), oligomycin (suppressing mitochondrial adenosine triphosphate synthase) and 2-deoxy-D-glucose (suppressing glycolysis) represent cellular glycolysis and its maximum capacity, respectively.
Statistical analysis
All data were analyzed using GraphPad Prism 8.3.0 software. Reed-Muench method was used to calculate IC50 values. Student’s t-tests were used to compare the control and treatment groups. Data were represented as mean ± SD. A difference of P < 0.05 indicated statistical significance.
Discussion
Some ingredients existing in natural foods have been proved to kill cancer cells as drugs, such as Vitamin C and curcumin [
27,
38]. These compounds not only effectively suppress the growth of malignancies, but also exert no harmful effect on normal tissues. Mannose, which exists in various vegetables and fruits, selectively inhibits tumor progression. This selectivity is highly related to expression status of PMI that converts M-6-P to F-6-P. When PMI expression is extremely low, M-6-P cannot convert into F-6-P and accumulates in tumor cells, thereby killing cancer cells by inhibiting cellular glycolysis [
16]. However, its role in thyroid cancer still remains elusive.
In the present study, we demonstrated that mannose selectively killed thyroid cancer cells, which was similar to a previous study [
16]. Surprisingly, we did not find that there was statistical difference in PMI expression between mannose-sensitive cells and mannose-insensitive cells. However, mannose-insensitive cells became sensitive to mannose when PMI was knocked down in these cells. There was study showing that enzyme activity of PMI dictates the response of LPS-activated macrophage to mannose [
39]. Thus, we tested enzyme activity of PMI in a panel of thyroid cancer cells, and found that enzyme activity of PMI was higher in mannose-insensitive cells than mannose-sensitive cells, indicating that high enzyme activity of PMI impairs the response of thyroid cancer cells to mannose.
It has been disclosed that Zn
2+ plays an essential role in maintaining enzyme activity of PMI by binding to its catalytic core area [
17,
18]. Meanwhile, after extracting PMI proteins in vitro and chelating Zn
2+ with EDTA, its catalytic activity significantly reduced [
26]. Previous study implied that Zn
2+ did not change the structure of PMI protein [
40]. Thus, we attempted to determine whether there was a similar mechanism existing in thyroid cancer cells. Our results demonstrated that Zn
2+ chelator TPEN treatment significantly reduced intracellular Zn
2+ concentration and inhibited enzyme activity of PMI, while did not change PMI expression. Besides, we expectedly found that TPEN treatment obviously improved the response of mannose-insensitive thyroid cancer cells to mannose, and Zn
2+ supplement could effectively reverse this effect. It is well-known that zinc transporter ZIP10 is required for the entry of Zn
2+ into the cell [
41]. Next, we investigated ZIP10 expression in thyroid cancer cells, and found that ZIP10 expression was clearly higher in mannose-insensitive cells than mannose-sensitive cells. Functional studies demonstrated oncogenic role of ZIP10 in thyroid cancer, which was consistent with a previous study showing that ZIP10 promoted zinc-triggered mitosis [
42]. Importantly, ZIP10 knockdown decreased intracellular Zn
2+ concentration and enzyme activity of PMI, enhancing the response of mannose-insensitive cells to mannose. Conversely, ectopic expression of ZIP10 in mannose-sensitive cells decrease their response to mannose. These findings indicate that expression status of ZIP10 is a major determinant for anti-tumor activity of mannose in thyroid cancer cells, and may be a potential therapeutic target to sensitize the response of cancer cells to mannose.
Warburg effect is a special metabolism found in cancer cells which tend to obtain energy through anaerobic respiration in spite of sufficient oxygen, thus the inhibition of cellular glycolysis will be an effective strategy for cancer therapy [
43,
44]. M-6-P has been demonstrated to impede cellular glycolysis by suppressing hexokinase, phosphoglucose isomerase and glucose-6-phosphate dehydrogenase [
36]. Thus, it is reasonable to speculate that mannose kills thyroid cancer cells by this mechanism. When mannose enters into the cell, it transforms into M-6-P which accumulates when PMI enzyme activity is low, thereby decreasing cellular glycolytic levels [
16,
45]. As supported, our data showed that mannose treatment significantly suppressed the glycolysis of mannose-sensitive cells, while did not change that of mannose-insensitive cells. However, PMI or ZIP10 knockdown in mannose-insensitive cells could effectively inhibit enzyme activity of PMI, thereby promoting the inhibitory effect of mannose on cellular glycolysis.
Evidently, mannose can enhance the response of cancer cells to chemotherapy due to the accumulation of mannose-6-phosphate and subsequent inhibition of glycolysis [
16]. Besides, a previous study showed that PKM2 promoted chemotherapy resistance by enhancing the glycolysis in ER-positive breast cancer [
46]. However, whether mannose can improve the response of thyroid cancer cells to chemotherapy needs to be explored in the near further. In addition to chemotherapy, the radiotherapy is another therapeutic strategy for thyroid cancer. There is evidence indicating that targeting HIF1 enhanced the radiosensitivity of breast cancer cells by reducing cellular glycolytic levels and the content of lactate acid [
47], suggesting that mannose may improve the response of thyroid cancer cells to ionizing radiation by inhibiting cellular glycolysis.
In recent years, immunotherapy has become a highly promising strategy for cancer treatment [
48]. There is increasing evidence hinting that inhibition of glycolysis could effectively improve the response of cancer cells to CTLA-4 blocker by impairing the stability of Treg cells [
49]. Besides, lactic acid, the production of glycolysis, has been proved to facilitate the infiltration of Treg cells in tumor, interrupting the lethal function of effector T cells [
50]. Thus, we speculate that mannose may improve the efficacy of immune checkpoint inhibitors by suppressing cellular glycolysis. These observations support that mannose has potential clinical use in thyroid cancer therapy when combined with chemotherapy, radiotherapy or immunotherapy.
Conclusion
In summary, by a series of in vitro and in vivo experiments, we demonstrate that mannose selectively kills thyroid cancer cells, and this effect is highly dependent on enzyme activity of PMI rather than its expression. Further studies reveal that PMI can be activated by zinc transport protein ZIP10 through promoting Zn2+ influx, thereby decreasing the response of thyroid cancer cells to mannose. Thus, our data highlight a crucial role of expression status of ZIP10 in affecting the response of thyroid cancer cells to mannose, and offer a mechanistic rationale for exploring clinical use of mannose in thyroid cancer therapy, especially combining with chemotherapy, radiotherapy or immunotherapy.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.