Comparison of Apical and Basolateral Cu Treatment for Iron-related Gene Regulation During Deferoxamine Induced Iron Deficiency Anemia


 Background: Intestinal copper transporter (Atp7a) mutant brindled mice with systemic Cu deficiency had elevated Cu levels in enterocyte cells without any perturbation of iron regulating genes, suggesting that blood Cu level might be important for intestinal iron homeostasis during iron deficiency anemia (IDA). We hypothesized that the blood Cu level and polarization (apical and basolateral) of enterocyte cells might be important regulators for the compensatory response on the regulation of genes in enterocyte cells during IDA. Methods: We grew Caco-2 cells on a bicameral cell culture plate to mimic the human intestine system and on a regular tissue culture plate. IDA was induced by Deferoxamine (DFO). The cells were treated with Cu and Cu with Fe following mRNA expressions of DMT1, FPN, TFR, and ANKRD37 were analyzed. Results: Our main finding was that basolateral treatment of Cu significantly reduced mRNA expressions of iron-regulated genes, including DMT1, FPN, TFR, and ANKRD37, compared to DFO treated and DFO with apical Cu treated groups in both bicameral and regular tissue culture plates. Conclusions: Cu level in the basolateral side of Caco-2 cells significantly influenced the intracellular gene regulation in DFO induced iron-deficient condition, and polarization of the cells might be important factor gene regulation in enterocyte cells.


Background
Iron is an essential and vital nutrient for life to sustain the functions of living organisms. In the human body, iron is essential to carry oxygen to tissues since it is the main component of hemoglobin. Moreover, it is necessary to control cell division and differentiation, mitochondrial energy production, DNA replication and repair, and immune response against pathogenic microorganisms 1,2 . Iron levels in the human body are tightly controlled via intestinal absorption since mammalians do not have a speci c active iron excretory mechanism. Therefore, intestinal enterocyte cells maintain adequate levels of iron in the body. Inadequate dietary iron absorption through enterocyte cells of the intestine leads to iron de ciency anemia, one of the major global nutritional de ciencies, especially in women, infants and children. Iron de ciency causes elevated copper mineral levels in the intestinal mucosa 3 , liver, and serum 4 in many mammalian species, including humans 5 . Previous investigations have implicated that copper homeostasis perturbations in uence iron metabolism 1,5−8 .
Divalent metal transporter (DMT1), the iron exporter ferroportin 1 (FPN1), can be a potential connection between iron and copper metabolisms since there are studies that suggest iron amount is regulated by copper in intestine [9][10][11] . Intestinal hypoxia-inducible factor 2α (HIF-2α) is essential for iron absorption during iron de ciency since it regulates apical and basolateral iron transporters 12 . Copper in uences the DNA-binding activity of HIF-2α, illustrating another connection of copper in uencing iron homeostasis.
DMT1 and FPN were shown as direct target genes related to HIF-2α 13 . mRNA regulation of Ankyrin repeat domain 37 (ANKRD37), prolyl 4-hydroxylase (P4ha1), and HIF prolyl hydroxylase 3 (EGLN3) are the most well-known marker genes for the hypoxic signal under iron de ciency.
Recent studies have shown that the regulation of genes involved in iron and copper homeostasis were altered in brindled mice 9 . However, when intracellular Cu level increased during IDA, Cu did not affect the genes' regulation, which plays a role in enterocyte iron metabolism. In the same study, enterocyte cells of mice intestine were treated with high intracellular and low blood Cu levels. Thus, blood Cu level might be a more dominant regulator of iron metabolism-related genes during IDA. The apical side of the enterocyte is in contact with dietary nutrients, whereas the basolateral part is in contact with the blood. Considering this apical and basolateral polarization, different sides of enterocytes which in relation to different environmental conditions (diet or blood) may be necessary for controlling the molecular and genetic regulation of iron metabolism in enterocyte cells. However, to the best of our knowledge, no research exists related to the dependency of the apical versus basolateral Cu for enterocyte cell iron metabolism during IDA. Given this background, the current study investigated the effects of dietary and blood copper treatments separately on iron and iron-dependent hypoxic gene regulations of anemic enterocyte cells by in vitro modeling of the human intestine system.

Modeling of Human Intestine System and TEER Measurements of Monolayer Caco-2 Cells
The apical and basolateral polarization of Caco-2 cells can behave like the human small intestine 15 .
Therefore, Caco-2 cells were grown on the bicameral cell culture system for 21 days and then DFO and mineral treatments were performed. TEER was measured from all experimental groups at the end of 21 days and after treatments to control the stability of the monolayer integrity of the cells. As shown in the Fig. 1 we did not observe any signi cant change in TEER values between the experimental groups before and after Cu and Fe treatments. Furthermore, TEER values that are much higher than 250 ohm/cm 2 re ect the experimentally polarized cells (as apical and basolateral) in the human intestinal system.

Apical and Basolateral Treatments of Cu in DFO Induced Anemic Monolayer Caco-2 Cells
When the Caco-2 cells are grown on the bicameral cell culture system, it allows us to study the effects of apical and basolateral Cu treatment on regulation of iron metabolism-related genes in enterocyte cells during DFO induced IDA. DFO treatment induced DMT1, TFR, and ANKRD37, and EGLN3 mRNA levels; however, FPN and MT1a were not affected ( Figure-2a). When Cu was given to the apical side of the cells, it did not affect mRNA levels of genes, including DMT1, TFR, and ANKRD37, and EGLN3 that were induced by DFO. Cu treatment of the basolateral side of cells signi cantly reduced DMT1, FPN and TFR, and ANKRD37 mRNA levels under the IDA condition. Moreover, DMT1, FPN, and TFR gene expressions were not changed by apical side treatments of the cells by Cu and Fe, whereas ANKRD37 and EGLN3 mRNA levels were slightly reduced. Furthermore, when both Cu and Fe were given to the basolateral side of the cells, mRNA levels of iron regulating genes (DMT1, FPN, and TFR), and hypoxia-related genes (ANKRD37, and EGLN3) were signi cantly lower than all treatment groups. MT1a mRNA levels were signi cantly increased by basolateral Cu treatment of cells, and this induction was higher when both Cu and Fe were given to the basolateral side of the cells ( Fig. 2a and 2b).
The Effect of Cu on Iron De ciency Anemia in Caco-2 Cells Grown on the 12-well Tissue Culture Plate Copper and iron interaction in enterocyte cells has attracted interest since their homeostasis is controlled by intestinal absorption. In this study, we grew Caco-2 cells on 12-well tissue culture plates for 21 days and then treated by DFO to induce IDA. We treated the cells Cu with and without Fe. Then we analyzed the mRNA levels of iron regulating and iron-dependent hypoxia regulating genes to see the effects of Cu on iron metabolism during IDA ( Fig. 3a and 3b). DFO treatment signi cantly induced DMT1, TFR, and ANKRD37, and EGLN3 mRNA expression levels, whereas FPN and MT1a mRNA levels were not affected. The Cu treatment did not reduce DFO induced gene mRNA expressions and did not affect FPN and MT1a mRNA levels. However, Cu with Fe treatment signi cantly reduced DMT1, TFR, ANKRD37, and EGLN3 mRNA levels. Moreover, we observed that Cu and Fe together signi cantly upregulated MT1a mRNA expression.

Discussion
Enterocyte cells are unique cell types in the intestine due to the interaction between diet and the internal circulatory system of the body. Thus, the dietary and blood-derived stimulus might play a role in enterocyte nutrient metabolism. Copper and iron have physiological interactions in enterocyte cells during IDA. In many mammalian species, iron de ciency leads to elevated Cu levels in enterocyte 16 , serum 17 , and liver 18 , indicating that copper might in uence intestinal iron metabolism. Iron and copper interactions have been investigated in brindled mice (Mo Br/y ) and researchers observed no perturbation in iron absorption during anemia 9 . In this study, the enterocyte copper level of mutant mice was high due to nonfunctional ATP7A protein. In contrast, these mice were systemic copper-de cient; thus, enterocyte cells were exposed to two different copper levels. Furthermore, it was observed that dietary-induced iron de ciency upregulated mRNA expressions of iron-regulated genes, including DMT1, TFR, and ATP7A, but high enterocyte copper level did not affect the gene regulations suggesting that blood Cu level might be an important dietary factor for gene regulations in enterocyte cells during IDA. Thus, we grew Caco-2 cells on a bicameral cell culture plate to mimic the human intestine system to test whether dietary (apical side) or blood copper (basolateral side) are involved in iron-dependent genes regulation during IDA.
In the current study, we selected marker genes that were regulated by iron de ciency anemia including iron transporters (DMT1, TFR and FPN) and iron-dependent hypoxic genes (ANKRD37 and EGLN3). Furthermore, MT1A mRNA level was used to control intracellular mineral uptake since MT1A is regulated by intracellular Cu and Zn, respectively 19,20 . Our results showed that Cu and Cu with Fe in the basolateral side of the cells reduced marker genes mRNA levels (DMT1, FPN, TFR1, and ANKR37) compared to the apical Cu and Cu with Fe treatment group during IDA. This suggests that Cu sensing of the basolateral side of the cells might be different from Cu sensing of the apical side. Cu regulates the MT1A mRNA levels and its level is correlated by elevated intracellular Cu 20 . We observed that MT1A mRNA levels were signi cantly induced when Cu and Cu with Fe were given to the basolateral side of the cells suggesting that the level of Cu in the blood might be a compelling factor for the regulation of iron de ciency-related genes in the intestine. Furthermore, the basolateral side of the enterocyte cells may play a role in Cu sensing. The importance of the molecular and genetic interactions between basolateral and apical sides of polarized enterocyte cells has been indicated in different studies. It was shown that glucose treatment in the basolateral side of the Caco-2 cells induced apical cholesterol uptake and the mRNA levels of the cholesterol transporter gene 21 , indicating that basolateral signaling in uences fatty acid metabolism through apical side of enterocyte cells. Moreover, Han et al. (2002) 22 showed that both apical and basolateral Cu treatment increased Fe uptake in non-anemic Caco-2 cells. Moreover, Cu treatment of apical and basolateral sides also induced DMT1, TFR, and FPN mRNA levels compared to the control group. All together results suggest that basolateral and apical sides of enterocyte cells might have a different physiological response regarding the same nutritional stimulus.
Caco-2 cells can differentiate on the regular tissue culture plate when they are grown 21 post-con uent days. Differentiation of Caco-2 cells gives different genetic responses for a variety of cellular physiological pathways 15 . However, polarization of those cells on the insert system might be another factor that can in uence genetic regulation. Thus, we also grew Caco-2 cells on regular tissue culture plates at the same time that we performed bicameral cell culture experiments. We observed similar results for marker genes regulation between apical side Cu treatment groups and Cu treatment on Caco-2 cells that were grown on tissue culture plates. However, basolateral Cu treatment affected regulations of DMT1, FPN, MT1A, ANKRD37 genes compared to results from the cells that were grown on tissue culture plates. Our results suggest that the polarization of the Caco-2 cells might be an important factor for gene regulation in terms of Cu treatment in our study. It might be better to account polarization of the Caco-2 cells for gene regulation or whole-genome array studies.

Conclusion
The nutrient-dependent regulation of enterocyte cells is central to the intestinal nutrient-sensing mechanism. The basolateral and apical sides of the enterocyte cells are the primary targets to understand nutrient sensing in terms of nutrient overload or de ciency. Furthermore, polarization of Caco-2 cells also might in uence gene regulation. Our results suggest that intracellular gene regulation was mainly affected by Cu treatment in the basolateral side of enterocyte cells during IDA, indicating that blood copper level might have ability to control the enterocyte iron metabolism at molecular and genetic levels during iron de ciency anemia. The blood Cu level might be an important regulator for intestinal iron metabolism during iron de ciency anemia.

Modeling of the Human Intestinal System and Cu Treatments of Apical and Basolateral Sides of the Cells During Iron De ciency Anemia
To test the effect of the dietary and blood copper on iron de ciency anemia, we mimicked the human intestines. The Caco-2 cells were seeded into bicameral collagen-coated polytetra uoroethylene membrane with 0.4-µm pore size and 1.12 cm 2 diameter (12-well inserts) (Corning, Cat. No.: 3493). After three days of con uent culture, the cells were grown for an additional 21 days to form a monolayer and

Statistical Analyses
The results were expressed as the mean values ± standard deviation of three independent experiments with at least two parallels of each experimental group. Statistical analysis for more than two groups was carried out by one-way analysis of variance (ANOVA) with Tukey's post hoc test by using GraphPad Prism 6 (GraphPad Software Inc., CA, USA). Data were considered signi cant for p ≤ 0.05. Availability of data and materials

List Of Abbreviations
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interest.

Funding
This work was supported by the Scienti c and Technological Research Council of Turkey (grant number:215Z041).
Author Contributions