The Impact of Sodium Selenite and Seleno-L-Methionine on Stress Erythropoiesis in a Murine Model of Hemolytic Anemia

Abstract

Background

Selenium (Se) is an essential trace element that exerts most biological activities through selenoproteins. Dietary selenium is a key regulator of red cell homeostasis and stress erythropoiesis. However, it is unknown whether the form and increasing doses of Se supplementation in the diet impact stress erythropoiesis under anemic conditions.

Objectives

If inorganic (sodium selenite; Na₂SeO₃) or organic [seleno-L-methionine (Se-Met)] forms of Se in different amounts (deficient, adequate, supplemented, and supranutritional) support stress erythropoiesis in anemic mice.

Methods

Three-wk-old male C57BL/6 mice were subjected to graded amounts of Se in the form of <0.01 mg/kg Se [Se-deficiency (Se-D)], 0.1 mg/kg Na₂SeO₃ (adequacy), 0.4 mg/kg Na₂SeO₃ (supplemented), 3 mg/kg Na₂SeO₃ (supranutritional), 0.4 mg/kg Se-Met (supplemented), or 3 mg/kg Se-Met (supranutritional), for 10–12 wk before intraperitoneal phenylhydrazine administration to induce hemolytic anemia. Following 3 d of phenylhydrazine injection, spleen and blood samples were used to assess the impact of form and graded amounts of Se in the diet on stress erythropoiesis.

Results

Phenotypic parameters showed that supplementing the diet with Se in the form of Na₂SeO₃ or Se-Met alleviated hemolytic anemia and promoted stress erythropoiesis by supporting the formation of erythroblastic islands. Se-Met at 0.4 mg/kg enhanced erythroid progenitor differentiation by 2-fold compared with Se-D, while Na₂SeO₃ at 0.4 mg/kg and 3 mg/kg significantly (P < 0.05) aided monocyte recruitment and macrophage differentiation within erythroblastic islands. Additionally, 3 mg/kg of Se-Met triggered a stronger inflammatory response than the same dose of Na₂SeO_3.

Conclusions

While both Se-Met and Na₂SeO₃ effectively aid in stress erythropoiesis, Na₂SeO₃ supplementation effectively support stress erythropoiesis with a minimal inflammatory response, while Se-Met at supranutritional dosage lead to increased inflammation despite its support for stress erythropoiesis. These results indicate diverse mechanisms of action of Se on the alleviation of anemia by stress erythropoiesis, which should be considered for further studies to complement existing therapies.

Introduction

Selenium (Se) is a nutritionally essential trace element required for optimal physiologic functions. The recommended dietary allowance of 55 μg of Se for an adult per day [1,2] is derived from the diet. Generally, Se in the organic form, predominantly seleno-L-methionine (Se-Met), exists in plant foods, while animal foods are good sources of inorganic forms of Se in the form of sodium selenite (Na₂SeO₃) or selenate (Na₂SeO₄). Canonically, Se derived from the diet is incorporated into selenoproteins in the form of the 21st amino acid, selenocysteine [3], which is transported throughout the body by selenoprotein P [4]. The major difference between selenite and Se-Met lies in their absorption rates and downstream pathways of metabolism, distribution, and excretion. The absorption of organic Se is generally greater than the inorganic form of Se [5], but its utilization via the selenoprotein pool is less efficient when compared with that of selenite [6]. Selenite is absorbed in the small intestine, while Se-Met is absorbed as an amino acid and can be nonspecifically incorporated into proteins in competition with methionine [7], providing a Se reserve that can be mobilized when needed. However, the incorporation of Se-Met in place of methionine leads to a slow whole-body turnover time (∼363 d), indicating that long-term high-dose Se-Met intake may result in high amounts of Se accumulation and toxicity compared with long-term high-dose selenite intake, as selenite is more easily excreted [8]. While the debate regarding the ideal form of Se in supplements for optimal health benefits continues, it is abundantly clear that insufficient Se intake is associated with congestive cardiomyopathy (Keshan disease) [9], osteoarthritis (Kashin-Beck disease) [10], and some cancers [11]. Studies have shown that ∼55 μg of Se supplementation was sufficient to prevent and significantly reduce mortality in Keshan disease [12,13]. Laboratory studies have reported the ability of Se to inhibit the growth of cancer cells through its regulation of oxidative stress [14,15]. This underscores the importance of Se in the immune system as an essential trace element through its incorporation into selenoproteins [16]. Selenoproteins are critical in the regulation of oxidative stress [17], calcium homeostasis [18], thyroid hormone synthesis [19], skeletal muscle regeneration [20], and immune response [21,22]. Interestingly, low serum Se amount in humans was reported to be associated with anemia [23]. In fact, Se deficiency may explain the significant comorbidity in older adults who also present with iron deficiency [23]. While supplementation with Se could help mitigate anemia [24], the molecular underpinnings are still unclear. This report further highlights the importance of Se in erythrocyte development.

Erythropoiesis presents a particular problem to redox regulation as the presence of iron, heme, and unpaired globin chains lead to high amounts of free radical–mediated oxidative stress, which are detrimental to erythroid development and can lead to anemia. Essential cytoprotection against such prooxidant toxicity is required during the maturation of erythroblasts, especially during stress erythropoiesis [25]. In response to inflammation, bone marrow hematopoiesis is skewed toward myelopoiesis, which leads to a deficit in erythroid production. To compensate for this loss of output, extra-medullary stress erythropoiesis is induced, rapidly producing erythrocytes to maintain homeostasis [26]. The production of stress erythroid progenitors (SEPs) is coordinated by distinct cellular factors that are different from steady-state erythropoiesis, including bone morphogenetic protein 4, Hedgehog, and growth and differentiation factor 15. As in steady-state erythropoiesis, erythropoietin, a hormone generated from the kidney, stimulates the differentiation of SEPs into stress burst-forming unit erythroid cells (BFU-Es), followed by final erythroblast maturation [27]. The development of SEPs and erythroblasts depends on a specialized microenvironment, called erythroblastic islands (EBIs), which are formed by a central macrophage surrounded by erythroblasts. The central macrophages serve as nursing cells for the surrounding erythroblasts by providing survival and development signals, phagocytosing the expelled nucleus, and supplying nutrients (including heme) needed for SEPs and erythroblast maturation [28]. Thus, the importance of redox homeostasis becomes crucial in SEPs, erythroblasts, and the microenvironment, which together aid in ensuring efficient erythroid supply. However, it is unclear whether the SEPs and erythroblasts respond to specific chemical forms of exogenous Se supply to impact their expansion and differentiation to erythroblasts.

Previous studies in our laboratory showed that Se-deficient (Se-D) diet induced mild anemia and impaired stress erythropoiesis in mice and that a Se-adequate diet (0.08 mg/kg of sodium selenite) helped alleviate anemia, particularly when subjected to hemolytic agents [24,29]. Moreover, while it is known that Se and selenoproteins play a key role in stress erythropoiesis, whether the dose and the form of Se would affect stress erythropoiesis has not been systematically determined. In this study, we aimed to study the effect of organic or inorganic forms of Se in the diet of mice subjected to stress erythropoiesis via phenylhydrazine (PHZ)-induced hemolytic anemia. The effect of Se supplementation of mice with graded doses of organic or inorganic (Se-Met or selenite) Se in the diet on stress erythropoiesis, following challenge with PHZ to induce hemolytic anemia, was investigated with an emphasis on the frequency of SEPs, monocyte, macrophage, and EBI. Our studies suggest selenite and Se-Met impact distinct cell types to ultimately alleviate stress erythropoiesis.

Methods

Animal, diet, and PHZ treatment

Three-wk-old male C57BL/6 mice (purchased from Taconic) were maintained on a customized torula yeast-based Se-D TD.92163 diet (Inotiv) that contained <0.01 mg/kg of Se. Graded Se diets were prepared using the TD.92163 Se-D diet as the base diet to which various amounts of sodium selenite (Na₂SeO₃; Sigma) or selenomethionine (Se-Met; Sigma) were added as follows. 0.1 mg/kg of Na₂SeO₃; 0.4 mg/kg of Na₂SeO₃; 3 mg/kg of Na₂SeO₃; 0.4 mg/kg of Se-Met; and 3 mg/kg of Se-Met for 10–12 wk. Mice were provided with Milli-Q water and respective diets ad libitum. To induce hemolytic anemia, mice were injected intraperitoneally with 50 mg/kg body weight PHZ (Sigma). Mice were killed 3 d after PHZ treatment, and retro-orbital blood collection was used to measure hematologic parameters using a Hemavet blood analyzer (Drew Scientific).

Peripheral blood mononuclear cell isolation

Three days after PHZ treatment, peripheral blood was collected by cardiac puncture for peripheral blood mononuclear cell (PBMC) isolation. Briefly, peripheral blood mixed with 2 volumes of phosphate-buffered saline (PBS) was layered on Histopaque1077 (Sigma) and subjected to gradient centrifugation at 400g for 35 min at room temperature. PBMCs were isolated and washed 2 times with PBS to ensure they were free of granulocytes and contained minimal or no red blood cells (RBCs).

Splenocyte isolation and stress BFU-E colony assays

Spleen isolated from mice was gently smashed using the rear end of a 3-mL syringe plunger and filtered through a 70-μm cell strainer (Falcon). Splenocytes were collected following RBC lysis and washed once with sterile PBS. Isolated splenocytes were cultured in methylcellulose (StemCell Technology M3334) media containing erythropoietin at 2% oxygen to enumerate stress BFU-E colonies, as described earlier [29,30].

Splenic EBI isolation

The freshly isolated spleen was chopped into small pieces and incubated in RPMI1640 media containing 0.075% (m/v) Collagenase IV (Gibco 17104019) and 0.004% (m/v) DNase I (Invitrogen DN25) for 30 min at 37 °C with constant shaking. The suspension was then gently passed through an 18-gauge needle several times. After centrifugation at 500g for 5 min at 4 °C, the pellet was gently resuspended with 1 mL of RPMI1640 containing 0.004% DNase I and layered on top of 10 mL of 30% (vol:vol) heat-inactivated FBS (R&D systems) in IMDM (Gibco). Aggregates were enriched by gravitational sedimentation for 45 min at room temperature. Eight milliliters of supernatant was removed, and the aggregates were resuspended in 50% (vol:vol) Percoll (in IMDM) and layered on top of 100% (vol:vol) Percoll (Sigma P1644). Following centrifugation for 20 min at 400g, 25 °C, aggregates within the 50%–100% Percoll gradient were collected and washed with PBS until further use.

Flow cytometry

Freshly isolated splenocytes, PBMC, and EBI aggregates were stained with zombie yellow (BioLegend) for 15 min to examine viability. For SEP analysis, splenocytes and PBMC were stained with antibodies to Kit and CD133. For macrophage analysis, splenocytes were stained with antibodies to Ly6C, CD11b, F4/80, and vascular cell adhesion protein (VCAM)-1. For erythrocyte analysis, splenocytes were stained with antibodies to CD71 and Ter119. For monocyte analysis, PBMCs were stained with antibodies to Ly6C and CD11b. For the analyses of EBIs, cells were stained with antibodies to Ter119, Ly6C, CD11b, F4/80, and VCAM-1. Extracellular staining was performed at 4 °C for 30 min, followed by analysis using the Fortessa LSR (Becton Dickinson) flow cytometer. Supplemental Table 1 provides detailed antibody information.

qPCR and gene expression analysis

Total RNA was extracted from the spleen of mice using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. RNA reverse transcription was performed using the High-Capacity cDNA Reverse Transcriptase kit (Thermo Fisher). qRT-PCR was performed with Taqman probes (Thermo Fisher) and PerfeCTa qPCR SuperMix Master mix (Quanta Biosciences). Probes are listed in Supplemental Table 2. The relative mRNA expression level was quantified using the 2^−ΔΔCt method.

Western blotting

The total proteins of spleen were extracted using M-PER (Thermo Fisher) cell lysis buffers containing protease inhibitor cocktail (Roche#04693116001). The protein amount was measured using a bicinchoninic acid protein assay kit (Thermo Fisher). Equal amount of protein was loaded onto SDS-PAGE gels, followed by immunoblotting. All antibodies used are listed in Supplemental Table 3.

Statistical analysis

GraphPad Prism 7 software program was used for all statistical analysis. All quantitative data are reported as mean ± SEM of at least duplicate biological replicates. Kruskal–Wallis H test (1-way ANOVA on ranks) was used to analyze data with >2 groups using Dunn multiple comparisons tests. The significance level was set at α = 0.05. Data are representative of mean ± SEM (∗P < 0.05; ∗∗P < 0.001; ∗∗∗P < 0.005; ∗∗∗∗P < 0.0001.

Results

Diets containing organic or inorganic Se were equally effective in alleviating anemia

Three-wk-old male C57BL/6 mice maintained on custom Se diets (Se-D; 0.1 mg/kg Na₂SeO₃; 0.4 mg/kg Na₂SeO₃; 3 mg/kg Na₂SeO₃; 0.4 mg/kg Se-Met; 3 mg/kg Se-Met) for 10–12 weeks, following which they were treated with PHZ (50 mg/kg; 50% dose) by intraperitoneal injection to induce nonlethal hemolytic acute anemia (Figure 1A). Compared with the Se-D group, mice maintained on graded amounts of Na₂SeO₃ or Se-Met in the diet protected the mice from the severity of PHZ-induced anemia as indicated by mice weight, hematocrit, RBC count, hemoglobin, and platelet counts (Figure 1B–G). Se-D diet mice exhibited more severe anemia, which is consistent with our previous results [24]. Interestingly, irrespective of the form of dietary Se used, supplementation at adequate, supplemented, and supranutritional amounts of 0.1, 0.4, and 3.0 mg/kg, respectively, were protective. Mice on 0.4 mg/kg of Se-Met showed a nonsignificant decrease in hemoglobin compared with the 3-mg/kg Se-Met group but were both higher than those in the Se-D group. Taken together, it appears that Se supplementation, irrespective of the chemical form and dose, adequately facilitates stress erythropoiesis for further anemia recovery in murine models.

FIGURE 1.

Dietary supplementation in the form of organic and inorganic selenium are equally effective in alleviating the severity of hemolytic anemia. (A) Schematic of phenylhydrazine (PHZ) treatment and diet supplement on mice. Three days after PHZ (50 mg/kg) treatment, (B) hematocrit percentage, (C) spleen weight, (D) mouse weight change percentage, (E) red blood cells count, (F) hemoglobin, and (G) platelet were analyzed. Dotted line refers to average of each parameter in naïve untreated mice. n = 3–11 per group. ∗Significance difference between selenium-deficient (Se-D) and other diet groups.

Supplementation with Se-Met promoted differentiation of SEPs compared with that with Na₂SeO₃ in stress erythropoiesis

To further compare the efficacy of inorganic and organic forms of Se in the diet on stress erythropoiesis, we identified SEPs based on the differential expression of c-kit (stem cell factor receptor; CD117) and CD133 from the spleen of mice maintained on various Se diets as mentioned earlier. Flow cytometric analysis indicated that SEPs in the early expansion stage (Kit⁺CD133⁺) were not affected by increasing amounts and/or the form of Se used in the diet (Figure 2A). However, SEPs in the differentiation stage (Kit⁺CD133⁻) increased in 0.4-mg/kg Se-Met supplemented mice (Figure 2B, C), indicating that on day 3 following PHZ-induced anemia, Se-Met supplemented diet promoted the differentiation of SEPs compared with Se-D or Na₂SeO₃ group. To further investigate the effect of dietary Se on the differentiation of the erythroid lineage, we performed BFU-E cultures with SEPs from each of the Se-D and Se-supplemented groups. Surprisingly, all Na₂SeO₃-supplemented groups showed significantly decreased BFU-E colony formation compared with the Se-D group, and all Se-Met displayed a nonsignificant decrease in BFU-E colony formation compared with the Se-D group. PHZ induced a rapid stress erythropoiesis response, with stress BFU-E reaching maximal amounts at 3 h posttreatment. Se-D mice showed a profound delay in the development of stress BFU-E compared with mice fed on diets containing various amounts of Na₂SeO₃. In addition, Se-D mice being anemic at a steady state also impacted their response to PHZ (Figure 2D).

FIGURE 2.

Seleno-L-methionine (Se-Met) diet promotes the differentiation of stress erythroid progenitors (SEPs) following phenylhydrazine (PHZ) injection compared with Na₂SeO₃ dietary supplement. (A) Early expansion stage SEPs indicated by kit⁺CD133⁺ fluorescence frequency post-PHZ treatment. (B) Differentiation stage SEPs indicated by kit⁺CD133⁻ fluorescence frequency post-PHZ treatment and (C) representative flow cytometry diagrams. Red box indicates Kit⁺CD133⁻ in (B). (D) Burst-forming unit erythroid cell (BFU-E) colony culture post-PHZ treatment; n = 3–14 per diet group.

Se in the form of Na₂SeO₃ recruited monocytes into the spleen and accelerated the differentiation toward macrophage

Erythroid progenitors develop in the EBI niche, which contains erythroid progenitors in proximity to macrophages [28]. Our previous study showed that the central macrophage in stress erythropoiesis, which plays an important nursing role via the development and maturation of erythroid progenitors, is partially derived from monocytes in the spleen recruited from bone marrow [30]. As shown in Figure 3A, the process of monocyte maturation into macrophage is divided into 4 distinct populations as the cells advance from population I through population IV. Within the PBMCs, the infiltrating (Ly6C^hi) monocytes showed a nonsignificant decrease in all Se-added groups compared with the Se-D group, particularly at 0.1 mg/kg of Na₂SeO₃ (Figure 3B). In the spleen, Ly6C^hi monocytes were also reduced in all diet groups where Se was added, especially at supranutritional amounts (Figure 3C), indicating a defect in recruiting monocytes into the spleen from the bone marrow via systemic circulation. Compared with the Se-Met–supplemented mice, Na₂SeO₃-supplemented mice showed a higher trend in resident (Ly6C^loCD11b⁺) monocytes (Figure 3D) and higher F4/80⁺VCAM-1⁺ red pulp macrophages (RPMs) (Figure 3E), indicating that more monocytes were recruited from peripheral blood to mature into RPM located within the EBI.

FIGURE 3.

Dietary Na₂SeO₃ supplementation aids in monocyte recruitment and their differentiation to macrophages. (A) Monocyte and macrophage gating strategy. Splenic single cells are stained with antibodies to Ly6C, CD11b, F4/80 and vascular cell adhesion protein (VCAM)-1. Cells are gated into 4 major populations, representing Ly6C^hi monocytes (I), Ly6C^lo monocytes (II), pre–red pulp macrophages (RPMs) (III) (not shown), and RPMs (IV), respectively. Frequency of Ly6C^hi monocytes in peripheral blood mononuclear cell (PBMC) (B), Ly6C^hi monocytes in spleen (C), Ly6C^lo monocytes in spleen (D), and RPM [F4/80 and VCAM-1 as markers of RPM in erythroblastic island (EBI)] in spleen (E) following 3 d of phenylhydrazine (PHZ) treatment; n = 3–9 per diet group. ∗Significance difference between selenium-deficient (Se-D) and other diet groups.

Dose and form of Se were critical for the development of the EBI niche

EBIs provide a specialized microenvironment that regulates the proliferation and maturation of SEPs [28]. Ter119⁺F4/80⁺ aggregates in the spleen were identified as the EBI population (Figure 4A, B). As shown in Figure 4C, dietary Se supplementation at 0.4 mg/kg as Na₂SeO₃ or 0.4 mg/kg Se-Met significantly promoted the development of EBI compared with the Se-D group. The supplemented Na₂SeO₃ and Se-Met groups also showed a trend in improving EBI formation compared with adequate and supranutritional Na₂SeO₃ and Se-Met groups. Therefore, it appears that supplemented (0.4 mg/kg) organic and inorganic forms of Se are both equally effective in the restoration of the hematocrit through proper development of EBI.

FIGURE 4.

Effect of form and dose of selenium supplementation on the development of erythroblastic island (EBI) niche. EBI aggregates are isolated from mice spleen after 3 d of phenylhydrazine (PHZ) treatment. (A,B) Representative FlowSight image of EBIs. (C) F4/80⁺Ter119⁺ EBI aggregates frequency in total events by flow cytometry analysis; n = 3–9 per diet group. ∗Significance difference between selenium-deficient (Se-D) and other diet groups. Illustration was created in Biorender.com.

Supranutritional Se-Met supplementation triggered a stronger inflammatory response than supranutritional supplementation with Na₂SeO₃

To further explore the differences in cellular response to stress erythropoiesis as a function of dose and type of Se in the diet, total RNA and protein from the whole spleen were isolated from mice fed the abovementioned diets after 3 d of PHZ treatment. We examined the expression of proinflammatory (Il-1b, Tnfα, Ifnγ, Il6, Nox1, Cybb, and Hif1α) and anti-inflammatory mediators (Il-10, Sod1, Gpx4, Gpx1, and Hmox1) (Figure 5A, B). Interestingly, as the Se amounts in the diets increased, mice on both diets showed a U-shaped dose-response curve with regard to the expression of most of the genes in the 2 diet groups following PHZ treatment. Although the pattern of increase in the expression of Il-1b, Tnfα, Ifnγ, Il-6, Nox1, Cybb, and Sod1 with increase in dietary Se (compared with Se-D as a control) was seen in both diet groups, the magnitude of expression of these proinflammatory genes was different between the Na₂SeO₃ and Se-Met groups with an overall increased response in the Se-Met 3-mg/kg group compared with that in the Se-D group. In agreement with such a trend, the expression of Il-10, a major anti-inflammatory cytokine, as well as Gpx1 and Gpx4, followed a similar pattern. Expression of glutathione peroxidase (GPX) 1 protein was notably enhanced following supplementation with Se-Met, and GPX1 amounts increased in response to increasing amounts of both Na₂SeO₃ and Se-Met in supplemented diets (Figure 5C, D) compared with that to the Se-D diet. Interestingly, increased mRNA expression of Hif1α and Hmox1 in the 3-mg/kg Se-Met groups, but not in the corresponding Na₂SeO₃ diet group, as well as a consistently higher hypoxia-inducible factor (Hif) Hif1α and Hif2α protein expression in the 3-mg/kg Se-Met group, suggested hypoxic stress induced by supranutritional dose of Se-Met could increase erythroid turnover and erythrophagocytosis (Figure 5B–D).

FIGURE 5.

Protein and mRNA amounts of key proinflammatory cytokines, oxidative stress indicators, and selenoproteins. (A) Heatmap of mRNA expression. (B) mRNA expression level of Il1b, Il6, sod1, Cybb, Gpx4, Gpx1, Hif1α, and Hmox1. Normalized to the mean of selenium-deficient (Se-D) group; n = 3–5. (C) Western blot showing protein amounts of SelenoW, GPX4, GPX1, Hif1α, and Hif2α. (D) Quantification of protein expression level. Normalized to β-actin, then normalized to the mean of Se-D group; n = 2–5. Statistical analyses performed excluded the group with n = 2. ∗Significance difference between Se-D and other diet groups.

Discussion

Redox regulation in erythropoiesis is of paramount importance given the enormous oxidative stress that could impede erythroid development, leading to anemia [31]. Antioxidant protection is, therefore, critical to protect erythroid cells from oxidative stress for timely recovery to overcome anemia. Previous studies from our laboratory showed that Se deficiency not only precipitates anemia [29] but also severely affects stress erythropoiesis [24]. Moreover, GPX1 protein expression decreased during iron deficiency [32,33], suggesting Se supplementation could assist stress erythropoiesis under anemic conditions. However, supplementation with organic compared with inorganic forms of Se has not been investigated in the context of erythropoiesis, including stress erythropoiesis. Using a mouse model of hemolytic anemia that leads to stress erythropoiesis, our studies indicated equal effectiveness at 0.4 mg/kg of Na₂SeO₃ or Se-Met with regard to EBI development. While supplemented Se-Met promoted the development of SEPs, Na₂SeO₃ at supplementation and supranutritional amounts facilitated the maturation of recruited monocytes.

Between the 2 forms of Se used, Se-Met is reported to be better absorbed and better retained than the inorganic form of Se [[34], [35], [36]]. Although both Na₂SeO₃ and Se-Met are reported to incorporate Se as selenocysteine into the selenoprotein pool, albeit at different extents, the prospect of nonspecific incorporation of Se-Met into proteins always exists depending on the amount of methionine present in the diet [35,37]. Exposure to high amounts of Se-Met may cause increased Se retention in tissues, potentially leading to toxicity [38]. Our study showed an increase in Sod2, Cybb, and Nox1, with 3 mg/kg of Se-Met indicating potential toxicity of high dosage Se-Met rather than Na₂SeO₃. It is reported that low amounts of Se-Met supplementation alleviated the toxicity of fusarium T-2 toxin in rabbits by suppressing the inflammatory factors such as Il-1b, Il-6, and Tnfα, while high amounts of Se-Met worsened the toxic effect associated with a higher proinflammatory response [39], which is similar to what we observed in this study with 3 mg/kg of Se-Met. Such an inflammatory response triggered a stronger compensatory anti-inflammatory response.

In particular, GPX4, a Se-containing antioxidant enzyme that protects cells against endogenous and reactive lipid hydroperoxides, controls reticulocyte maturation and enucleation during erythropoiesis [40,41]. The increase in compensatory anti-inflammatory pathways along with hypoxic response as seen in the form of Hif1α and its downstream gene, Hmox1, further complements the increased amounts of inflammatory stress. Interestingly, metabolism studies suggest that Se-Met increases the RBC turnover compared with Na₂SeO₃ [42], therefore likely impacting erythrophagocytosis. In fact, inflammation that accompanies stress erythropoiesis as a cause and/or consequence triggers the burst in erythroblast development in the spleen to compensate for the loss of RBC. Although GPX4 protein amount does not show significant elevated regulation by Se in diet as GPX1, the relatively high mRNA amounts in the 3-mg/kg Se-Met group than those in other Se groups indicates that it is plausible that an increase in Gpx4 with high Se-Met supplementation leads to a cellular response in the form of increased erythrophagocytosis and erythroblast turnover rate indicated by higher Hif1α amounts, which perhaps assists the proliferation of SEP through promoting glycolysis during stress erythropoiesis [43]. These mechanisms remain to be further elucidated.

During Se deficiency, the absence or low expression of selenoproteins appeared to impair the expansion of SEP and maturation of erythroblasts [24], highlighting the importance of dietary Se. In this study, 0.4 mg/kg of Se-Met supplementation significantly promoted the differentiation of SEPs (Kit⁺CD133⁻) compared with Se-D or Na₂SeO₃-supplemented mice on day 3 post-PHZ treatment, which confirmed the dominant effect of Se-Met in SEP development under stress erythropoiesis. While Se-Met at supplementation amounts positively impacted the differentiation of SEP, our data also suggest that it was ineffective in rescuing the delay in macrophage development that we observed in the Se-D diet group. In contrast, Na₂SeO₃ supplementation significantly promoted the development of F4/80⁺VCAM-1⁺ RPMs. Interestingly, Ly6C^loCD11b^hi monocytes were significantly increased in all Se-Met–supplemented groups but not in Se-D or Na₂SeO₃-supplemented diet groups. The recruitment of these monocytes from peripheral blood is driven by Ccr2-dependent signaling, which responds to the inflammatory signal from stress erythropoiesis [44] to promote the development of F4/80⁺VCAM-1⁺ RPMs [30]. Together, regardless of the differential effects of the 2 forms of Se in the diet that affect SEPs compared with macrophages, our studies suggest that eventually, these differences lead to the near equal EBI development on day 3 post-PHZ treatment between 0.4-mg/kg Na₂SeO₃ and Se-Met supplementation groups. These results lend credence to the idea that Se, in the form of selenoproteins and/or selenometabolites, may aid in the resolution of the hematocrit via diverse mechanisms. Se-Met at supplemental amounts supported the development of SEPs, whereas Na₂SeO₃ at both supplemental and supranutritional amounts enhanced the maturation of recruited monocytes. The underlying mechanisms beg further detailed characterization (Figure 6).

FIGURE 6.

Schematic illustration of the effect of Na₂SeO₃ and seleno-L-methionine (Se-Met) on stress erythropoiesis. Under anemic conditions following phenylhydrazine (PHZ) treatment, Na₂SeO₃ and Se-Met supplementation both support the formation of EBIs. Na₂SeO₃ supplementation supports the migration of monocyte from bone marrow to spleen, as well as the maturation of monocyte into macrophage, while Se-Met supplementation predominately promotes the differentiation of stress erythroid progenitors (SEPs). Illustration was created in Biorender.com.

The curvilinear dose-response relationship in the form of a U-shaped behavior between the amounts of Se in the diet and observed adverse events have been previously reported in the form of DNA damage, apoptosis, diabetes, cardiovascular disease, and prostate cancer [8,23,45,46]. In stress erythropoiesis, Se-D or high doses of dietary Se also showed a U-shaped response in the form of higher proinflammatory responses, such as Il-1b and Il-6, as well as worsened stress erythropoiesis development. Se-Met supplementation at 3 mg/kg showed a decreased differentiation of SEPs compared with lower Se-Met–dose groups. In addition, 3 mg/kg of Na₂SeO₃ or Se-Met supplementation impaired EBI development. Such a U-shaped dose-response curve suggests that both Se amounts and the form need to be carefully considered for future clinical trials in conditions particularly related to anemia of inflammation [8].

Our studies highlight the importance of Se in efficient stress erythropoiesis. These results are consistent with our previous study, where Se-replete diets were able to offer protection from Se deficiency–dependent impairment of stress erythropoiesis in a murine model of anemia [24]. Se-D mice experienced lower hematocrit, body weight, RBC count, and hemoglobin compared with all Se-supplemented diet groups, regardless of the form used. To compensate for the defective stress erythropoiesis caused by Se deficiency, BFU-E was significantly increased, indicating an increased potential for erythroblast maturation.

In conclusion, our studies emphasize the critical role of Se supplementation in anemia as either Na₂SeO₃ or Se-Met at supplemental amounts (0.4 mg/kg). However, supranutritional (3 mg/kg) Se-Met clearly led to a strong inflammatory response in comparison with 3 mg/kg of Na₂SeO₃. Clearly, a U-shaped response was seen in our studies, suggesting providing anemia patients with supplemental amounts of Se could be useful. However, it is important to be cognizant of the inflammatory effects with Se-Met, particularly at supranutritional levels, which could potentially be harmful to anemic patients.

Author contributions

The authors’ responsibilities were as follows—HG, RFP, KSP: designed the research; HG, DR, YB: conducted the research; HG, RFP, KSP: wrote the paper; KSP: had primary responsibility for the final content; and all authors: analyzed data and read and approved the final manuscript. All illustrations were prepared using Biorender.com.

Funding

This work was funded, in part, by NIH grants, RO1 DK119865 (to KSP), and RO1 DK138865-04 and RO1 HL146528 (to RFP); USDA National Institute of Food and Agriculture and Hatch Appropriations under Projects #PEN04932/Accession #7006585 (to KSP) and #PEN04960/Accession #7006577 (to RFP).

Conflict of interest

The authors report no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: