In transfusion-dependent thalassemia, increased iron overload is associated with lower serum alpha-klotho, which is strongly associated with lower total and ionized calcium concentrations

a Clinical Analysis Department, College of Pharmacy, Hawler Medical University, Havalan City, Erbil, Iraq. E-mail: shatha003@yahoo.com. b Department of Chemistry, College of Science, University of Kufa, Iraq. E-mail : headm2010@yahoo.com. c Department of Chemistry, College of Science, University of Kufa, Iraq. E-mail : zainab.alhillawi@uokufa.edu.iq. d Department of Psychiatry, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand. e Department of Psychiatry, Medical University of Plovdiv, Plovdiv, Bulgaria. f IMPACT Strategic Research Centre, Deakin University, PO Box 281, Geelong, VIC, 3220, Australia .

Zainab Hussein Alhillawi has no financial conflict of interests.
Michael Maes has no financial conflict of interests.

Introduction
Beta-thalassemia major (β-TM) is a hematologic disorder caused by absent or severely reduced synthesis of the β-globin chain in the hemoglobin A molecule resulting in damage to the erythrocyte membrane and subsequent anemia 1,2 . Patients with β-TM require lifelong blood transfusions to increase hemoglobin levels and minimize the detrimental effects of inefficient erythropoiesis 3 . Patients with the latter condition, denoted as transfusion-dependent thalassemia (TDT), are prone to many complications due to the frequent blood transfusions 4 . Chronic blood transfusions may cause severe iron overload, which may cause toxicity to various organs including the liver, heart, endocrine organs 5-7 , bones and joints 8 . The latter may be associated with severe consequences including spontaneous pathological fractures, osteoporosis, osteopenia, skeletal deformities, and bone pain [9][10][11] .
Other biomarkers which play a role in bone disorders and are elevated in patients with TDT are calcyphosin (CAPS1) and fibroblast growth factor receptor 2 (FGFR2) 21 .
Calcyphosin is a calcium-binding protein involved in both Ca 2+ -phosphatidylinositol and cAMP signal cascades 22 . FGFR2 is expressed on preosteoblasts and osteoblasts during the later phase of bone formation 23 . Dysregulation of FGFR2 results in a spectrum of bone pathologies 24,25 . Other authors examined serum soluble α-Klotho in TDT, but could not find a difference between TDT and control groups 26 . Klotho is a β-glucosidase-like membrane-bound protein that displays a secreted splice form 27,28 . The α-Klotho gene is predominantly expressed in tissues that are involved in calcium homeostasis including the parathyroid glands, kidney and the choroid plexus 29,30 . α-Klotho regulates calcium and phosphate reabsorption in the kidney and indirectly, as a cofactor for FGF23, regulates vitamin D metabolism 31 . Moreover, α-Klotho promotes endothelial nitric oxide production and inhibits Wnt signaling and oxidative stress pathways 32,33 , and inhibits intracellular insulin and IGF-1 signaling, which is an evolutionarily conserved pathway associated with an extended life span 34 . Also, in animal models, α-Klotho may delay the ageing process in association with suppressing insulin and IGF-1 signaling and oxidative stress toxicity 34,35 .
However, there are no data whether α-Klotho in TDT is associated with aberrations in calcium homeostasis and iron overload.
Hence, the present study aims to examine whether TDT in children is accompanied by lowered serum α-Klotho and whether there are significant associations between serum α-Klotho and calcium concentrations or calcium-related biomarkers (Vitamin D3, PTH, calcyphosin, FGFR2, phosphate) and iron overload biomarkers (iron, ferritin, transferrin saturation). This study recruited 90 participants, namely 30 healthy controls and 60 TDT   children, aged 3-12 years old and of both sexes. The TDT patients were recruited at the   Thalassemia Unit at Al-Zahra'a Teaching Hospital severe anemia, hepatosplenomegaly, and abnormal bone growth), hematological tests including hemoglobin <7g/dl and hypochromic microcytic RBCs with anisopoikilocytosis and high reticulocyte percentage, and by elevated HbA2 levels as assayed using HPLC (VARIANT TM β-Thalassemia Short Program). Thirty apparently healthy children were recruited as the control group. None of the controls was anemic or had an immuneinflammatory or systemic disease. We excluded any subject with splenectomy, systemic diseases such as renal failure, diabetes mellitus, or subjects with overt inflammation defined as serum C-reactive protein (CRP) levels > 6mg/l. The latter exclusion criterion was used to ascertain that the change in ferritin or other acute-phase reactant proteins is due to iron overload rather than to an acute phase response.

Participants
The frequency of administration of blood transfusions with packed RBCs at 2 or 4week intervals was based on Hb levels that should be kept above 9 g/dL. Moreover, patients were on an iron-chelating therapy (3-5 times weekly) with deferoxamine mesylate USP (Desferal ® ) infusion at a dose range between 25-50 mg/kg/day over 8 hours/day depending on the ferritin levels. Folic acid was also given to most patients to reduce ineffective erythropoiesis. TDT patients were treated with vitamin C to assist the chelation of iron with deferoxamine through stimulation of iron release from the reticuloendothelial system.
Written informed consent was obtained from the patient's first-degree relatives (mother or father) after appropriate oral explanation according to the Declaration of Helsinki. The study was approved by the IRB of the University of Kufa number 419/2018.

Measurements
Five mL of venous blood were drawn from all participants after an overnight fast.
The patients' samples were collected just before their blood transfusion session. Blood was left at room temperature for 10 minutes for clotting, centrifuged 3000 rpm for 5 minutes, and then serum was separated and transported into Eppendorf tubes. Serum albumin, calcium, and phosphate were measured using a ready for use kit supplied by Biolabo ® Co (Maizy France). Ionized calcium was calculated from the following formula: I.Ca 2+ = 0.813 × T.Ca 0.5 -0.006 × Albumin 0.75 + 0.079 36 , which give the best approximate result.
The amount of iron in sera was determined by colorimetric kits supplied by Spectrum ® (Cairo, Egypt). Transferrin saturation percentage (TS%) was calculated from the following equation: TS% = Iron * 100/TIBC 37 . TIBC was measured by saturation of serum transferrin with iron, and the unbound iron portion is precipitated with magnesium carbonate, and then the iron was remeasured in the supernatant. Serum PTH and soluble α-Klotho levels were measured using ELISA kits supplied by MyBioSource ® (San Diego, USA). Serum ferritin levels were measured by using ELISA kit supplied by Elabscience ® (Wuhan, China). Serum calcyphosin and FGFR2 were measured using an enzyme-linked immunosorbent assay (ELISA) using kits supplied by Bioassay Technology Laboratory (Shangai, China). These kits were designed for human samples depending on the biotin double antibody sandwich technology. Hematological parameters were measured by a fivepart differential Mindray BC-5000 hematology analyzer (Mindray Medical Electronics Co., Shenzhen, China). Vitamin D was determined by a fluorescence immunoassay (FIA) using kits designed for the I-Chroma™ instrument (BioLabs Diagnostics, Italy) to estimate total 25(OH)D2/D3 level in human serum.
The inter-assay CV% of ferritin, PTH, and soluble α-Klotho kits were <15%, <10%, and <10%, respectively, and the sensitivities of the ferritin, PTH and α-Klotho assays were 10.0 ng/ml, 15.6 pg/ml, and <56.25 pg/ml, respectively. The inter-assay CV% of iron was <2.19%. The inter-assay CV of calcyphosin was <10%, and sensitivity= 0.026 mM, while the inter-assay CV% of FGFR2 was <10%, and sensitivity=0.09 ng/ml. For samples with highly concentrated analytes, we employed sample dilutions. We computed a z unit-weighted composite score which reflects iron overload as z iron + z transferrin saturation % + z ferritin (IO index). CRP was measured using a kit supplied by Spinreact ® , Spain, which is based on latex agglutination.

Statistical analysis
Analysis of contingency tables (χ 2 test) was employed to assess associations between nominal variables while analysis of variance (ANOVAs) was used to assess differences in continuous variables among diagnostic groups. Associations between scale variables were computed using Pearson's product-moment correlation coefficients. Multivariate general linear model (GLM) analysis followed by tests of between-subject effects and pairwise comparisons among treatment groups were used to examine the associations between TDT (versus controls) and the biomarkers. A false-discovery rate (FDR) procedure was employed to control for type I errors when performing multiple comparisons 38 . Simple boxplots with the minimum, Q1, median, Q3, and maximum values, and out-and far-out values were employed to display the results of α-Klotho assays.
All tests were two-tailed, and a p-value of 0.05 was used for statistical significance. We used IBM SPSS 25 windows version to analyze the data .
Partial Least Squares (PLS) structural equation modelling was employed using the Smart PLS software 39

Demographic and Clinical data
The socio-demographic and clinical data in TDT and healthy control children are presented in Table 1. The patient group was further divided into two groups, namely those with normal α-Klotho (n=30) concentrations and those with low α-Klotho (n=30) levels using the median split method (median=350.3 pg/mL). There were no significant differences in age, sex ratio, and rural/urban ratio between the three study groups.

Biomarkers and diagnostic groups
Univariate GLM analysis showed that (after controlling for age and sex) α-Klotho was significantly (F=8.24, df=1/86, p=0.005) lower in TDT (mean ±SE=344.9 ±32.7 pg/mL) than in normal control children (508.9 ±46.5 pg/mL). show increases in serum PTH, FGFR2, and calcyphosin as compared with control children, while vitamin D3, RBCs and Hb were decreased in TDT patients as compared with the control group. No significant differences in serum phosphate were detected between the study groups. A multivariate GLM analysis with age and sex as covariates did not change these results and showed no significant effects of these covariates on the biomarkers except phosphate (F=8.45, df=1/85, p=0.005), which was higher in girls than in boys. Table 2 shows the intercorrelations between α-klotho, number of blood transfusion, the iron overload index, and the other biomarkers. In the whole study group, serum α-Klotho was significantly correlated with total and ionized calcium and negatively with PTH. There was a significant inverse correlation between α-Klotho levels and iron, TS%, and ferritin, and the iron overload index. α-Klotho levels were also significantly and positively correlated with Hb and negatively with the number of blood transfusions. In the control group, α-Klotho was significantly correlated with total calcium (r=0.380, p=0.039) and calcyphosin (r=0.523, p=0.003). In TDT patients, α-Klotho was significantly and positively correlated with total calcium (r=0.617, p<0.001) and ionized calcium (r=0.610, p<0.001), and inversely with Hb (r=-0.301, p=0.020), whereas no significant correlations with calcyphosin could be found (r=-0.066, p=0.617). In the whole study group, the number of blood transfusion and iron overload were strongly intercorrelated and showed similar correlations with the other biomarkers. Table 3 shows the results of different multiple regression analyses with total and ionized calcium levels as dependent variables and other biomarkers as explanatory variables while allowing for the effects of age and sex. Regression #1 shows that 40.1 % of the variance in total calcium could be explained by α-Klotho, vitamin D (both positively), and calcyphosin (inversely). Figure 2 shows the partial regression of total calcium on α-Klotho after adjusting for the variables listed in Table 3, regression #1. We found that 42.5% of the variance in ionized calcium (Regression #2) was explained by α-Klotho, vitamin D (both positively), and calcyphosin (inversely). In the healthy children control group, 14.9% of the variance of total calcium could be explained by serum albumin (regression #3). In TDT patients, 38.1% of the variance in total calcium was explained by α-Klotho, and vitamin D. Figure 3 shows the partial regression of total calcium on α-Klotho in TDT after adjusting for the variables listed in Table 3.  Figure 4 shows the results of the PLS analysis. calcium. All other paths were non-significant and thus deleted from the study, e.g. between calcium and PTH, and between α-Klotho and vitamin D3, PTH, FGFR2, and calcyphosin.

Discussion
The first major finding of this study is that TDT patients have lower α-Klotho levels than controls and that a meaningful part (around 50%) of TDT patients show low α-Klotho levels. In one study, serum α-Klotho levels tended to be lower in TDT patients as compared with controls, although the difference was not statistically significant 26 . Our PLS analysis showed that the number of blood transfusions significantly predicted lowered α-Klotho and that this effect was mediated by iron overload. Previously, it was shown that serum iron overload is accompanied by decreased expression of α-Klotho in the kidneys and that iron chelation may attenuate the angiotensin-II-associated decreases in α-Klotho expression 41 .
It is interesting to note that, in patients with chronic kidney disease, iron deficiency may lead to increased α-Klotho expression 41 . α-Klotho deficiency may cause activation of hypoxia-inducible factors (HIF) which regulate serum iron, which in turn negatively affect α-Klotho levels 42 . Nevertheless, the associations established in our study between α-Klotho and iron overload may, in theory, also be explained by the consequences of iron overload including chelation treatment, activated immune-inflammatory and oxidative stress pathways 40 . In this respect, it was shown that the type of chelation treatment did not affect α-Klotho levels 26 . In animal studies, iron overload may trigger down-regulation of α-Klotho expression while iron chelation may reverse this down-regulation, suggesting that abnormal iron metabolism is implicated 43 . TDT is associated with inflammation and oxidative stress toxicity as a direct consequence of iron toxicity 44 . Su and Yang concluded that α-Klotho might behave as an acute phase response since restraint stress is accompanied by a downregulation of α-Klotho mRNA and increased serum α-Klotho protein 45 .
Importantly, α-Klotho acts as an anti-inflammatory modulator through regulation of the production of nuclear factor-κB associated inflammatory proteins thereby reducing the production of several pro-inflammatory cytokines and oxidative stress toxicity 34 . At the cellular and organismal level, α-Klotho confers protection against oxidative stress [46][47][48] whereby α-Klotho attenuates superoxide production, oxidative damage, and apoptosis through the cAMP/PKA pathway 49  The second major finding of this study is that α-Klotho levels are strongly associated with total/ionized calcium levels and that TDT children belonging to the low α-Klotho group show deficient calcium levels. Previous studies showed that, in β-TM patients, α-Klotho correlated with serum and urine calcium 54 . α-Klotho participates in the regulation of calcium homeostasis in cerebrospinal fluid and blood by effects in the choroid plexus, parathyroid glands, and distal tubules 55,56 . In this regard, α-Klotho is a critical player that integrates "a multi-step regulatory system of calcium homeostasis", which continually adjusts calcium concentrations and maintains calcium within a narrow physiological range 57 . Reabsorption of calcium in the distal tubule of the kidney is facilitated by specific channels 58 which are activated by α-Klotho 59 . As such, α-Klotho expression responds to Ca 2+ concentration through Na + , K + -ATPase in the order of seconds, indicating that α-Klotho is a fast regulator of Ca 2+ absorption 60 . Moreover, α-Klotho regulates vitamin D3 production, which is a major regulator of intestinal calcium absorption 55 . The third major finding of our study is that TDT is accompanied by lower total and ionized calcium, and vitamin D3, but increased PTH, FGFR2, and calcyphosin levels while there are no significant differences in phosphate levels. These results extend those of previous papers which reported reduced levels of serum calcium and vitamin D3 and increased levels of calcyphosin, FGFR2 and PTH in thalassemia 15,65,66 . One hypothesis is that some of those changes could be induced by the effects of lower α-Klotho since a deficiency in α-Klotho was proposed to induce high serum PTH, phosphate, and FGF23 levels [67][68][69][70][71][72][73] . In addition, α-Klotho is a significant regulator of vitamin D biosynthesis 56 .
Nevertheless, in our study no significant associations between α-Klotho, on the one hand, and PTH, FGFR2 and vitamin D3, on the other hand, could be detected after considering the effects of iron overload. The latter was significantly associated with PTH, FGFR2, calcyphosin, (positively) and vitamin D3 (negatively), suggesting that mechanism related to iron overload may be involved. Previously, higher PTH levels were detected in β-TM patients, and these were positively associated with increased ferritin, one of the indicants of iron overload 74 . Chronic inflammation with increased levels of IL-1β and iron deficiency increase ferritin and FGF23 cleavage levels 75