Structurally, the mammalian ferritin complex is a hollow protein sphere or a nanocage composed of 24 subunits from two protein chains, light (FtL: 183 amino acids) and heavy (FtH: 182 amino acids). A ferritin nanocage can accommodate about 4500 Fe(III) atoms, mostly in the form of bioavailable mineral ferrihydrite, and a significant amount is present in the liver, heart, spleen, and kidneys (1, 2, 3).
Several genetic studies reported a more vital role of FtH than of FtL, as its silencing causes early embryonic death, whereas mice without FtL survive. Although both chains are structured as intracellular nanocages, FtH also promotes ferroxidase activity, which mediates oxidation of the ferrous [Fe (II)] into a less reactive ferric [Fe (III)] ion (1, 2, 3, 4, 5, 6, 7, 8).
Previous sex and aging studies have shown that ferritin expression reflects iron accumulation and that ferritin is higher in female than male and in older than younger experimental animals (9, 10, 11, 12, 13, 14, 15, 16, 17, 18). These sex and age differences may be owed to the liver hormone hepcidin as the main regulator of systemic iron availability. Hepcidin expression is higher in female rats and depends on steroid hormones (7, 18).
At the moment, however, we do not know how sex and age differences affect the expression of either ferritin chain in the kidney and liver and whether they affect them differently, as, to the best of our knowledge, such studies have not yet been carried out (9, 10). The aim of this study was, therefore, to determine liver and kidney expression of both ferritin chains in adult and old male and female animals and to establish the effect of hormone loss (through gonadectomy of adult animals) on their individual expression.
Sex and age differences were studied in three-month and two-year-old male and female Wistar rats from the colony bred at the Institute for Medical Research and Occupational Health, Zagreb, Croatia. Animals had free access to tap water and standard pelleted food (4RF21, Mucedola, Settimo Milanese, Italy) and were treated in accordance with the Directive 2010/63/EU (19) on the protection of animals used for scientific experiments. The gonadectomised group consisted of four 10-week old males castrated via scrotal route and four females ovariectomised via dorsal route two weeks before sacrifice. For control we used four respective male and female sham-gonadectomised animals, which underwent operation without removing their reproductive organs (20). The sacrifice and operations were performed under general anaesthesia with intraperitoneal ketamine (80 mg/kg body mass) and xylazine (12 mg/kg body mass) (Chassot AG, Bern, Switzerland). The studies were approved by the Institute’s Ethics Committee and by the Croatian Ministry of Agriculture under the AGEMETAR project.
Monoclonal antibodies against target proteins FtL (sc-390558, D-1) and FtH (sc-376594, B-12) or housekeeping proteins alphatubulin (sc-23948, aT), beta-actin (sc-47773, bA), and glyceraldehyde-3-phosphate dehydrogenase (sc-365062, GAPDH, G) were obtained from Santa Cruz Biotechnology (Dallas, TX, USA). Secondary antibodies were bought from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). The alkaline phosphatase-labelled goat anti-mouse IgG (GAMAP; Cat. No. 115-056-068) was used for western blotting and CY3-labeled donkey anti-mouse IgG (DAMCY3; Cat. No. 715-165-151) for immunohistochemical analysis.
We removed part of the liver and kidney from each sacrificed animal and dissected it manually to prepare 10 % tissue homogenates in an ice cold pH 7.4 buffer containing 300 mmol/L mannitol, 5 mmol/L EGTA, 12 mmol/L Tris/HCl, and protease inhibitors (SIGMA
For sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE), liver or kidney tissue homogenates were mixed with Laemmli sample buffer (20, 21) with or without 2-mercaptoethanol and denatured at various temperatures. Proteins (40 or 60 μg per lane) were separated with 18 % or 20 % SDS-PAGE gels using a Mini-PROTEAN 3 system and then wet-transferred using Mini Trans-Blot Electrophoretic Transfer Cell (both from Bio-Rad Laboratories, Hercules, CA, USA) to the poly(vinylidene fluoride) (PVDF) membrane (Immobilon®, Millipore, Bedford, MA, USA). Some PVDF membranes were stained with Coomassie blue (CBS) (Figure 1B) to check that we had the same load of proteins for all samples and the rest were immediately incubated with 1 % glutaraldehyde (GA) for 30 min as described elsewhere in detail (20, 21) to fixate small proteins on the membrane. We did this because FtL (antibody D-1) did not stain without GA fixation, whereas FtH (antibody B-12) staining was not affected by GA, which indicates different hydrophilic properties of the two subunits (data not shown).
PVDF membranes were incubated with target and housekeeping antibodies diluted in blotto milk – D-1 at the 1:1000 ratio, and B-12, aT, bA, and GAPDH at the 1:500 ratio. As we could not detect housekeeping proteins in some liver samples with colourimetry because of high gel density, we relied on the nonspecific band (NB) stained with secondary antibodies, as it appeared uniform below the ferritin protein band (Figures 1B, 4A, and 5A). After colourimetric detection, we evaluated protein band densities relative to the density of the strongest protein band in control samples (1 arbitrary unit) using the open source ImageJ program (US National Institutes of Health, MA, USA).
As described in detail in our previous work (20, 21), liver and kidneys were perfused with 4 % paraformaldehyde (PFA)
The slices were soaked overnight in 30 % sucrose (in PBS), embedded in an optimal cutting temperature medium (Tissue-Tek, Sakura, Japan), frozen to -25 °C, cut to 4 μm thick cryosections with a Leica CM 1850 cryotome (Leica instruments, Nussloch, Germany), and placed on Superfrost/Plus microscope slides (Fischer Scientific, Pittsburgh, PA, USA). For immunostaining we followed an optimised antigen retrieval procedure for both target antibodies described in detail elsewhere (20, 21). Immunohistochemical staining was examined under an Opton III RS fluorescence microscope (Opton Feintechnik, Oberkochen, Germany) and images captured and processed with a Spot RT Slider camera and software (Diagnostic Instruments, Sterling Heights, MI, USA). We then used Adobe Photoshop 6.0 (Adobe, San Jose, CA, USA) to obtain black and white images, label them, and arrange in panels.
Immunohistochemical and western blot data were obtained from four animals per experimental group. Group means ± standard errors of the mean (SEM) were compared with the
Considering previous research of age and sex differences in ferritin expression (9, 10, 11, 12, 13, 14, 15, 16, 17, 18), our study has confirmed its dependence on age, sex, and steroid hormones for both chains in the liver and kidney of 3-month and 2-year-old male and female Wistar rats.
Immunochemical findings in the liver are given in Figure 1. Figure 2 shows kidney panels with parallel ferritin staining in the female kidney based on consecutive cryosections. Detailed distribution of ferritin chain expression in the kidneys is shown in Figure 3.
Figures 4 and 5 show immunochemistry findings of light and heavy ferritin chains in the liver and Figures 6 and 7 in the kidney. The effect of gonadectomy on both ferritin chains in the liver is detailed in Figure 8 and in the kidney in Figure 9.
We have confirmed sex and age-dependent expression of both ferritin chains in the liver and kidneys (Figures 4–7), as female rats had more than two times as high expression as males. Similar findings in the liver and kidney were reported for iron accumulation (11, 12, 13, 14, 20, 22).
We have also demonstrated that the expression of both ferritin chains depends on sex hormones, as it was higher after castration and lower after ovariectomy (Figures 8 and 9), which is in line with studies of iron accumulation conducted a few decades earlier (9, 10, 14) and our more recent report of iron turnover after castration and ovariectomy (20).
Even though both ferritin chains are responsive to sex hormones, they do not have oestrogen or androgen response elements in their promoters. This suggests that iron regulates their expression through iron response proteins (IRP) and elements (IRE) in conformations of mRNA (1, 2, 3, 4, 5, 6, 7).
Some details we observed in the staining of both ferritin chains in three-month and two-year-old kidneys warrant more rigorous investigation that would include colocalisation analysis of specific proteins to confirm the presence of both in same cellular locations, as indicated by unexpected localisation of ferritin chains in intercalated cells and
The role of iron in kidney diseases has recently been emphasised, and more precise localisation of ferritin throughout the kidney may tell us more about its functions in (pato)physiology (5, 25, 26, 27, 28). However, protein expression in organs detected by western blotting in our study has yet to be confirmed by
Our findings have shown strong ferritin staining of interstitial cells irregularly present in all kidney zones. We assume that it could be associated with macrophages (Figures 6 and 7, c and d), as macrophages have recently been recognised as a new source of serum ferritin (25, 28). This could be relevant for our positive findings of ferritin in aged rats.
Another interesting finding in our study is intracellular abundance of ferritin-positive vesicles, which may reflect ferritin/ iron lysosomal turnover. Lysosomal iron may be protected with higher expression of inducible metallothioneins (MT) in some proximal tubules, which may attenuate reactive oxygen species (ROS) or directly sequester iron in organelles (29, 30, 31, 32, 33). Also, these vesicles can stay in the cell and form lipofuscin as time goes by (aging pigment of incompletely degraded proteins) (32). Our previous study (21) reports that expelled vesicles to the lumen of proximal tubules through exocytosis may correlate with MT-positive vesicles. It is possible that some reabsorption mechanism takes place in the form of megalin/cubilin scavenger receptors (21), which are present on the apical membrane of the S3 segment of proximal tubules. This is visible in Figure 3 (a and b, bottom).
Interactions between essential and toxic transition metals (TM) and mechanisms that protect or damage the cell/organism have been studied for many years (30). For iron as an endogenous source of ROS, scavenger interactions with ferritin and MT are of particular interest (31, 32, 33, 34, 35, 36, 37).
The significance of ferritin is also visible at the molecular level through regulation of iron accumulation and autophagocytosis and recycling of material within the cell (38, 39). Sex differences in autophagy and some diseases call for a revision of current knowledge about the metabolism of iron and other TMs and of implications related to aging, sterile inflammation, and degenerative diseases (40, 41, 42, 43, 44, 45, 46). Essential TMs, iron in particular, are involved in all cellular processes, and our study shows that it is important to establish whether sex differences or their absence play an important role in research models and human diseases, as suggested earlier (5, 26, 40, 41, 42).
Some novel markers, such as nuclear receptor coactivator 4 (NCOA4) are believed to drive ferritinophagy – autophagic degradation of ferritin that can lead to ferroptosis as regulated cell death independent of apoptosis, with iron lipid oxidation and GSH depletion (38, 39).
One of our aims was to establish the role of aging on ferritin expression. Until now, such an association between aging and iron accumulation in organs or between expressions of the two ferritin chains was studied only individually, like in female liver along with methionine-centred redox cycle changes (11) or in male heart and liver with oral iron chelator that decreased apoptosis protein expression (12), or in the male muscle in relation to protein oxidative damage (47). The last study suggested that accumulated iron and ferritin caused sarcopaenia. However, molecular pathways in aging and the role of ROS-producing iron and both ferritin chains remain elusive (5, 26–28, 30).
Our work contributes to current research with new data on sex and age differences in the expression and distribution of both ferritin chains and provides the background of all physiological processes.
Our study returns to the basic questions about the importance of sex, sex hormones, and age in the expression of both ferritin chains as the main iron storage cell structures by giving new insights into their distribution in the liver and kidney. These insights call for new (patho)physiological and toxicological investigations, which should clarify how the sex and age-dependent relationship between iron and ferritin affects the system, especially in oxidative stress.