Disease-modifying effects of iron deficiency in mouse models of chronic renal failure Open Access

Chronic kidney disease (CKD) is a global health challenge affecting >800 million individuals in both developed and developing nations.^1,2 With no effective treatments to cure this disease, clinical management prioritizes the mitigation of factors that contribute to CKD progression.^3,4 CKD is often exacerbated by systemic complications, such as anemia, inflammation, organ fibrosis, and vascular calcification.^5-9 These conditions drive cardiovascular disease, the leading cause of premature death in CKD.¹⁰ 

A prominent aspect of CKD is altered iron homeostasis, with a high prevalence of iron deficiency (ID) and anemia.^11-13 Anemia in CKD is associated with poor outcomes, including cognitive impairment and mortality.^14,15 The etiology of this anemia is driven by a complex interplay of reduced red blood cell (RBC) life span, relative erythropoietin deficiency, systemic inflammation, and both absolute and functional ID.¹⁶ Absolute ID occurs when total body iron stores are depleted, whereas functional ID arises when iron stores are adequate but insufficiently mobilized to support erythropoiesis.¹⁷ This iron restriction is driven by elevated hepcidin levels, which are increased by inflammatory cytokines, mainly interleukin-6 (IL-6).^18,19 Hepcidin is a key regulator of iron homeostasis, controlling flow of iron into circulation by inhibiting ferroportin, the sole known cellular iron exporter.²⁰ This inhibition restricts iron efflux from iron recycling macrophages and duodenal enterocytes, leading to reduced serum iron levels (hypoferremia) and erythropoiesis. Chronically inadequate intestinal iron absorption combined with increased blood loss contributes to development of absolute ID.^21,22 

Anemia worsens quality of life and outcomes in CKD.^16,23-25 Notably, ID and anemia might accelerate kidney injury progression, organ fibrosis, and cardiomyopathy.^26-30 Beyond these effects, emerging evidence links ID to vascular calcification.^31-33 A recent study reported increased coronary artery calcification scores in patients with CKD with reduced transferrin saturation.³⁴ Corroborating these findings, mice with metabolic syndrome on an iron-deficient diet exhibit enhanced cardiac calcification compared with those on a standard iron diet.³⁵ Despite these observations, the impact of ID severity on outcomes remains unclear. This study evaluates moderate and severe ID in 2 mouse models of CKD.

Male mice were used for a consistent CKD phenotype.^36,37 Mice were housed in ventilated systems (12 hours dark/light, 23 ± 1°C), fed ad libitum, and maintained on a standard diet (PicoLab Rodent Diet 20) before switching to a specialized diet. C57BL/6J mice were from Jackson Laboratory (JAX 000664). Col4a3^–/– knockout (Alport) mice were on a Sv129 background in heterozygous breeding.

Two models of CKD were used, as follows: adenine-induced nephropathy and genetic Col4a3^–/– mice. Assessing the impact of ID on CKD in adenine nephropathy, we established moderate and severe ID models (Figure 1A-B; supplemental Tables 1 and 2). Moderate ID was generated by feeding 8-week-old C57BL/6J mice a customized iron-deficient (4 ppm iron) 0.2% adenine-rich diet (TD.130826, Envigo) for 8 weeks. Iron-replete CKD mice were fed identical diet but with 100 ppm iron (TD.210096, Envigo). Iron-deficient and iron-replete controls were age-matched mice on an adenine-free but composition-matched diet (TD.200065, TD.80396, Envigo). Severe ID was induced by phlebotomizing 4-week-old C57BL/6J mice (∼250 μL blood) and feeding them an iron-deficient diet (4 ppm iron; TD.80396, Envigo) for 4 weeks. They were then switched to a customized 0.2% adenine-rich diet that was iron deficient (4 ppm iron; TD.130826, Envigo) for an additional 8 weeks. Matching iron-replete CKD mice were fed an iron-replete diet (100 ppm iron; TD.200065, Envigo) starting at 4 weeks of age for 4 weeks before transitioning to a customized 0.2% adenine-rich diet that was iron replete (100 ppm iron; TD.210096, Envigo) for an additional 8 weeks. Concluding these experiments, mice were euthanized under 2.5% isoflurane, and samples were prepared as described in the supplemental Methods.

To assess the impact of ID on CKD in a genetic model, 5-week-old Alport mice were randomly assigned to either an iron-replete diet (58 ppm iron; TD.80394, Envigo) or an iron-deficient diet (4 ppm iron; TD.80396, Envigo) for 5 weeks. Concluding the experiments, mice were euthanized under 2.5% isoflurane. Constitutive Col4a3^–/– knockout mice are recognized as a hereditary model of human Alport syndrome and progressive CKD. When bred on a Sv129 background, Alport mice die of rapid kidney injury by 10 weeks of age.

Animal protocols were approved by the institutional animal care and use committee of the University of California, Los Angeles.

Additional details are provided in the supplemental Materials.

To explore the impact of ID on CKD-related complications, we generated adenine diet-induced nephropathy models with varying ID severities. To establish effects of moderate-ID CKD mice, C57BL/6J males were fed either an iron-replete (100 ppm Fe) or iron-deficient (4 ppm Fe) diet containing 0.2% adenine for 8 weeks. Mice fed a diet lacking adenine, but with equivalent iron content, served as non-CKD iron-replete or non-CKD moderate-ID controls (Figure 1A; supplemental Table 1). To establish severe-ID CKD mice, C57BL/6J males were phlebotomized (∼250 μL of blood) by submandibular puncture and given an equal volume of 0.9% NaCl by intraperitoneal injection at 4 weeks of age, subjected to a 4 ppm Fe diet for 4 weeks, and transitioned to a 4 ppm Fe + 0.2% adenine diet for 8 weeks. Age-matched, nonphlebotomized mice on a 100 ppm Fe diet, later switched to a 100 ppm Fe + 0.2% adenine diet for 8 weeks, served as iron-replete CKD counterparts (Figure 1B; supplemental Table 2). Non-CKD controls were excluded to focus on the effects of ID in CKD.

We evaluated tissue and serum iron concentrations in both models to determine ID severity. Comparing non-CKD mice, liver, spleen, and heart nonheme iron levels were nonsignificantly decreased, and kidney iron significantly reduced in the moderate-ID model (Figure 1C-F). Transcript levels of the iron importer transferrin receptor (Tfrc) were slightly increased in the heart, and protein (TFR1) was significantly elevated in the kidneys of mice with non-CKD moderate ID (Figure 1G-H), confirming tissue ID. On adenine diet, all organs (liver, heart, spleen, kidney) accumulated more iron compared with non-CKD mice with a negligible impact of a moderate-ID diet on organ iron levels compared with iron-replete mice (Figure 1C-F). No differences in heart Tfrc transcript and kidney TFR1 protein levels were detected between groups on adenine (Figure 1G-H). Thus, in this context of adenine-induced CKD, moderate ID did not lead to measurable tissue iron depletion, likely resulting from CKD-related iron restriction mechanisms.

In severe-ID CKD mice, liver, spleen, heart, and kidney nonheme iron levels were significantly reduced compared with iron-replete CKD mice (Figure 1C-F). This indicates that baseline tissue iron levels were profoundly diminished despite CKD-related iron restriction mechanisms. In addition, we observed significant elevations in heart Tfrc transcript and kidney TFR1 protein levels in severe-ID CKD (Figure 1G-H). Interestingly, serum iron levels did not change between groups in either model (Figure 1I), likely due to reduced iron utilization for erythropoiesis after kidney injury.

Next, we assessed hematologic parameters in both models. In non-CKD controls, moderate ID caused expected reductions in mean corpuscular volume (MCV), hemoglobin, and hematocrit percentage (HCT%), increased zinc protoporphyrin (ZPP) levels, and a trend toward increased kidney erythropoietin (Epo) expression (Figure 2A-E). On adenine diet, CKD mice all had anemia. However, the difference between groups was blunted or even reversed. Between moderate-ID and iron-replete CKD mice, MCV, hemoglobin, HCT%, ZPP levels, and kidney Epo expression were similar (Figure 2A-E), indicating that moderate ID did not worsen erythropoietic parameters beyond the effect of CKD.

In severe-ID CKD mice, MCV, ZPP, and Epo expression levels were similar to iron-replete CKD mice as was the case for the moderate-ID group (Figure 2A,D-E), but hemoglobin and HCT% were surprisingly increased (Figure 2B-C). However, we cannot eliminate a batch effect in this experiment as complete blood count parameters in the severe-ID group were inadvertently measured on a separate day from the iron-replete group. Serum VEGF levels were significantly higher in severe-ID CKD, indicating greater systemic hypoxia in these mice, arguing against any functional improvement in the erythropoietic status of the mice (Figure 2F). Dehydration and hemoconcentration may have also contributed as total protein in serum of severe-ID CKD mice was on average 11% higher compared with iron-replete CKD. After the normalization of hemoglobin to total serum protein for each mouse,³⁸ severe-ID CKD and iron-replete CKD hemoglobin concentrations became similar and had decreased variability (Figure 2G). Nevertheless, the lack of dietary ID effects on erythropoiesis in either moderate- or severe-ID CKD suggests that anemia in this mouse model is primarily driven by CKD pathology, unlike tissue iron content in the liver, spleen, kidney, and heart, which is profoundly decreased by severe ID.

Inflammation is a hallmark of CKD, but it is not clear whether ID directly affects the inflammatory response in CKD. We evaluated hepatic and systemic inflammatory mediators in both models. In the absence of CKD, moderate ID had no effect on cytokine expression in the liver.

Relative to non-CKD controls, iron-replete CKD mice displayed nonsignificant increases in liver serum amyloid A1 (Saa1), tumor necrosis factor α (Tnfa), and Il6 (Figure 3A-B,D) transcript levels. Moderate ID in CKD mice further elevated liver Saa1 and Il6 levels to statistically significant levels compared with iron-replete CKD (Figure 3A,D). No differences were observed in liver Il1b messenger RNA (mRNA) expression (Figure 3C). In severe-ID CKD compared with iron-replete CKD, transcript levels of both liver Saa1 and Il6, including serum IL-6 protein levels, were significantly elevated (Figure 3A,D,F). Notably, these levels were higher in severe-ID relative to moderate-ID mice on adenine (Figure 3A,D,F). Spleen Il1b transcript levels were also significantly increased in severe-ID CKD (supplemental Figure 1B). We did not observe differences in liver Tnfa and Il1b or spleen Tnfa and Il6 transcript levels (Figure 3B-C; supplemental Figure 1A,C).

ID suppresses hepcidin production, but inflammation stimulates it.^39,40 As expected, Hamp transcript levels were greatly reduced (Figure 3E) and Hamp synthesis remained appropriately low relative to liver iron content in non-CKD moderate-ID compared with non-CKD iron-replete mice (Figure 3F). Interestingly, in the adenine model, Hamp transcript levels and ratios of Hamp synthesis relative to liver iron content were significantly and comparably elevated in both moderate-ID and iron-replete mice (Figure 3E-F), aligning with increased serum levels of the hepcidin-stimulator IL-6 (Figure 3G). Only severe ID partially suppressed Hamp expression despite a strong inflammatory stimulus (Figure 3E), as evident by a significant yet inappropriately high ratio of Hamp synthesis relative to liver iron content in severe-ID CKD mice (Figure 3F), highlighting the opposing influences of inflammation and severe ID on hepcidin transcription.

Because severe-ID CKD mice exhibit more pronounced systemic inflammation than moderate-ID CKD mice, we performed a multiplex bead immunoassay on sera from the severe-ID CKD model to gain further insight into the inflammatory milieu. Multiplex analysis of 32 cytokines and chemokines revealed significant elevations in serum tumor necrosis factor α (TNFα) and CXCL1, along with a nonsignificant increase in CCL4 in severe-ID CKD compared with iron-replete CKD (Figure 3H-J). Interestingly, serum CXCL2 was significantly reduced in severe-ID CKD (Figure 3K). Together, these data illustrate that the systemic inflammatory milieu in CKD is modulated by the severity of ID, which might contribute to and further aggravate CKD-associated pathologies.

Evidence suggests that dietary iron restriction can improve kidney function in animal models of CKD.^41,42 We evaluated whether moderate ID or severe ID affects kidney function in our CKD model. All CKD mice displayed significant kidney dysfunction as determined by elevated blood urea nitrogen (BUN) and serum creatinine levels (Figure 4A-B) vs non-CKD controls, with no difference between ID and iron-replete mice (in either the moderate-ID or severe-ID group).

CKD disrupts mineral homeostasis, with ID and CKD elevating FGF23 levels, a key regulator of serum phosphate.^43-45 In non-CKD controls, moderate ID increased FGF23 cleavage, resulting in elevated C-terminal FGF23 with normal intact FGF23 levels as expected (Figure 4C-E). CKD mice all had excessive levels of intact FGF23 and C-terminal FGF23. However, severe ID significantly increased FGF23 cleavage in CKD (Figure 4E). Moreover, serum phosphate was elevated in all CKD mice, unaffected by ID severity (Figure 4F).

Next, we measured kidney injury markers. All CKD groups had significantly elevated kidney neutrophil gelatinase-associated lipocalin (Ngal) and kidney injury marker 1 (Kim1) transcript levels, with no further effect from moderate or severe ID (Figure 4G-H). Supporting these findings, histologic analysis of kidney tissue sections revealed similar pathologic alterations in all CKD mice, including pronounced tubular atrophy and dilation, along with marked interstitial inflammatory cell infiltration (Figure 4I; supplemental Figure 1D). These data suggest that neither moderate ID nor severe ID affected the severity of kidney injury or dysfunction in adenine-induced nephropathy.

As our results demonstrate that ID augments inflammation in CKD (Figure 3), we explored whether ID modulates kidney fibrotic progression, as inflammation is a known contributor to kidney fibrosis.^5,46 Compared with non-CKD controls, both moderate-ID CKD and iron-replete CKD mice had significantly increased kidney fibronectin (Fn1), transforming growth factor β (Tgfb), α smooth muscle actin (Acta2), collagen type 1 α 1 chain (Col1a1), and collagen type 3 α 1 chain (Col3a1) transcript levels (Figure 5A-E). Interestingly, Col3a1 transcript levels were significantly increased in moderate-ID CKD compared with iron-replete CKD (Figure 5E). Severe-ID CKD also potentiated mRNA expression of kidney Acta2, Col1a1, and Col3a1 compared with iron-replete CKD counterparts (Figure 5C-E). Histopathologic analysis of kidney tissue sections revealed extensive collagen deposition in both the glomeruli and interstitium across CKD groups, but particularly potentiated in severe-ID CKD mice (Figure 5F; supplemental Figure 1E). Taken together, these results indicate that ID aggravates kidney fibrotic progression in adenine-induced nephropathy.

Next, we evaluated the role of ID in kidney calcification by analyzing local inflammatory cytokine expression, which triggers calcification pathways.⁴⁷ Kidney Tnfa and Il6 transcript levels were similarly elevated in moderate-ID and iron-replete mice on adenine vs controls (supplemental Figure 2A-B). Measuring expression of calcification genes, kidney runt-related transcription factor 2 (Runx2), sex determining region box 9 (Sox9), and osteopontin (Opn) transcript levels were significantly and similarly elevated in moderate-ID and iron-replete mice on adenine (supplemental Figure 2C-E).

In severe-ID CKD mice, kidney Tnfa, Il6, Runx2, Sox9, or Opn levels had no significant differences from iron-replete CKD mice (supplemental Figure 2). Thus, ID neither potentiates kidney tissue inflammation nor contributes to the pathogenesis of kidney calcification in adenine-induced CKD.

Vascular calcification stiffens arteries, exacerbating hypertension, heart failure, and mortality, particularly in CKD.^10,48 We investigated the role of ID in cardiac calcification by assessing inflammation. Cardiac Tnfa mRNA demonstrated a nonsignificant increase in both moderate-ID and iron-replete CKD mice vs controls (supplemental Figure 3A). We observed no changes in cardiac Il1b or Il6 transcript levels (supplemental Figure 3B-C). Measuring expression of calcification genes, cardiac Runx2, Sox9, and Opn transcript levels were significantly and uniformly elevated in both moderate-ID and iron-replete mice on adenine (supplemental Figure 3D-F). These data indicate that although cardiac inflammation was mild, signaling pathways involved in cardiovascular calcification were activated in adenine-induced CKD, but unaffected by moderate ID.

In severe-ID CKD mice, cardiac Tnfa, Il1b, Il6, Runx2, Sox9, and Opn levels demonstrated no significant differences from iron-replete CKD mice (supplemental Figure 3). These data confirm that ID does not affect cardiac inflammation or calcification in adenine-induced CKD.

Given that systemic inflammation and anemia are risk factors for heart disease, we explored whether ID affects cardiac injury and/or fibrosis. Cardiac hypertrophy, indicated by heart-to-body weight ratio, was present in moderate-ID and iron-replete CKD mice vs controls (Figure 6A). Furthermore, mRNA expression of cardiac adult alpha-myosin heavy chain (Myh6) was significantly and similarly reduced, whereas fetal beta-myosin heavy chain (Myh7) was significantly and similarly increased in both moderate-ID and iron-replete CKD mice (Figure 6B-C). The transition from adult to fetal myosin heavy-chain isoforms signifies the reactivation of fetal gene programs related to cardiac hypertrophy.^49-52 In addition, cardiac fibrosis was observed in both moderate-ID and iron-replete CKD mice, as evidenced by marked and uniformly elevated cardiac Tgfb and Col1a1 transcript levels, corroborated by histopathologic analysis of heart tissue sections revealing consistent collagen deposition within the myocardium (Figure 6D-E,G; supplemental Figure 1F). No changes were observed in cardiac Acta2 mRNA expression (Figure 6F).

Evaluating pathologic cardiac remodeling in severe-ID CKD mice, heart-to-body weight ratio and mRNA expression of cardiac Myh6 and Myh7 reflected similar cardiac hypertrophy between severe-ID CKD and iron-replete CKD mice (Figure 6A-C). However, severe ID amplified the expression of Tgfb, Col1a1, and Acta2 (Figure 6D-F), resulting in a modest, albeit nonsignificant, increase in myocardial collagen deposition (Figure 6G), indicating a slight worsening of myocardial fibrosis. Collectively, these data demonstrate that the severity of ID does aggravate the expression of cardiac fibrosis markers, but without substantially affecting CKD-associated cardiomegaly.

Evaluating these findings in a genetic CKD model, we used Alport (Col4a3^–/–) mice, which develop progressive CKD with systemic inflammation, mild ID, anemia,^53,54 and kidney and cardiovascular dysfunction with concurrent fibrosis.^37,54 Testing whether ID worsens these complications, we generated moderate ID in Col4a3^–/– mice by subjecting males to an iron-deficient (4 ppm Fe) diet for 5 weeks (supplemental Figure 4A). In comparison to Col4a3^–/– mice fed an iron-replete (58 ppm Fe) diet, moderate-ID Col4a3^–/– mice displayed increased kidney TFR1 protein levels and reduced kidney iron, serum iron, MCV, hemoglobin, and HCT% (supplemental Figures 4D,F-G and 5A-C), consistent with expected effects of moderate systemic ID.

We next explored the impact of moderate ID on CKD-associated complications in the Alport model. We detected no substantial alterations in systemic inflammation (supplemental Figure 6) but increased kidney Tnfa (supplemental Figure 9A). There was no effect of moderate ID on kidney injury or dysfunction (supplemental Figure 7), kidney fibrosis (supplemental Figure 8), kidney inflammation and calcification (supplemental Figure 9), cardiovascular inflammation and calcification (supplemental Figure 10), or pathologic cardiac remodeling, except for increased Acta2 expression in the heart of iron-deficient animals (supplemental Figure 11). Taken together, these data indicate that moderate ID does not strongly modulate CKD-associated complications in Alport mice, despite impairing erythropoiesis in this model.

More than 180 years ago, anemia was first connected to CKD.⁵⁵ Today, numerous studies confirm the widespread prevalence of ID and anemia in patients with CKD, and these conditions have been associated with CKD progression, organ fibrosis, vascular calcification, and cardiomyopathy.^26-28,34 Impairment of hemoglobin synthesis is a relatively late consequence of ID, with anemia only manifesting in a third of iron-deficient individuals.⁵⁶ Furthermore, ID can have adverse effects on cell and tissue function independently of the effects of anemia. Thus, understanding the pathologic consequences of ID in CKD can help optimize iron supplementation strategies to improve outcomes by both correcting tissue ID and improving anemia. This study directly assesses the impact of different degrees of ID on common CKD-associated complications using mouse models. Using both dietary-induced and genetic mouse models of CKD, we induced moderate or severe ID and evaluated a broad array of potential effects. We demonstrate that although ID does not modulate vascular calcification, kidney injury, kidney dysfunction, or pathologic cardiac remodeling, ID does potentiate systemic inflammation and kidney and cardiac fibrosis. These effects appear independent of anemia in the adenine model, as erythropoietic parameters were not worsened by either moderate or severe ID. It seems that in the setting of CKD, where erythropoiesis is suppressed from the effects of renal toxins and systemic inflammation, ID, even if severe, has much less effect on the erythron (Figure 2) compared with nonerythroid tissues (Figure 1).

In the CKD group with moderate ID, where mice were fed a low-iron diet, neither tissue iron levels nor erythroid parameters differed from those of the iron-replete CKD group receiving 100 ppm iron. Although initially puzzling, this can likely be attributed to hepcidin-mediated iron restriction resulting from functional ID in adenine-induced CKD. As a result, even mice receiving 100 ppm iron were unable to effectively use dietary iron, rendering both groups similarly iron deficient/restricted. It is therefore not surprising that we observed small or no differences in CKD complications between the 2 groups given that they differed little in their iron or erythropoietic status. The second ID model we used (severe ID), which relied on lowering iron stores through phlebotomy before the onset of CKD, allowed us to evaluate the effect of true tissue ID on CKD pathology. Interestingly, compared with their non-CKD counterparts, moderate-ID CKD mice had significantly higher iron accumulation observed in the liver, spleen, heart, and kidney. This is likely the consequence of iron being redistributed from the erythroid/RBC compartment to other tissues, both because of decreased utilization of iron by nascent erythropoiesis and because of accelerated destruction of mature RBCs.

Clinical studies have reported that persistent inflammation, even at low to moderate levels, exacerbates CKD-related comorbidities.^57-59 However, the effects of ID on chronic inflammation remain poorly characterized. In our study, severe ID intensified the systemic inflammatory milieu, as severe-ID CKD mice displayed increased liver and splenic inflammatory markers and elevated plasma cytokines compared with iron-replete CKD mice (Figure 3; supplemental Figure 1). Even the moderate-ID CKD group had more liver inflammation, as reflected by higher Saa1 (Figure 3). Furthermore, we detected increased circulating TNFα levels in severe-ID CKD mice (Figure 3). Given that TNFα is known to exacerbate anemia,^60-62 its upregulation in our adenine model may potentiate the effects of ID, creating a self-perpetuating cycle that amplifies systemic inflammation. This heightened inflammatory milieu could further drive the activation and recruitment of inflammatory cells, such as neutrophils and macrophages, to affected tissues. Such mechanisms could aggravate CKD-associated pathologies. Supporting this postulate, a recent report demonstrates that iron dextran administration in CKD blunts the production of proinflammatory cytokines.²⁶ Although our Alport (Col4a3^–/–) mice fed an iron-replete diet displayed increased inflammation relative to adenine-fed mice, moderate ID in Alport mice did not amplify systemic inflammation (supplemental Figures 4-6). These findings suggest that although ID can augment the inflammatory milieu in CKD, it may not contribute further when the inflammatory response is already highly activated.

Inflammation and ID are also frequently associated with fibrotic development in CKD, although a recent study suggests that fibrosis may arise independently of inflammation.⁶³ In our analysis of the adenine model, severe ID, resulting in profound iron depletion across multiple tissues, exacerbated kidney fibrotic gene expression and interstitial fibrosis, despite similar expression of markers of kidney injury, impaired kidney function, and kidney tissue inflammation when compared with iron-replete CKD mice (Figures 4 and 5; supplemental Figure 2). A plausible mechanism explaining our data might involve intracellular ID in kidney macrophages, where it is reported to drive a profibrotic phenotype in CKD.²⁶ Although moderate ID did not aggravate this profibrotic response, it also failed to further intensify tissue ID relative to iron-replete CKD. However, in the context of severe ID, CKD mice also displayed increased systemic levels of the fibrosis-associated cytokines CXCL1 and CCL4, compared with iron-replete CKD mice (Figure 3). It is also possible that severe ID may contribute to worse CKD by precipitating a more severe acute kidney injury after the initial exposure to adenine.

Given that CXCL1 promotes fibrosis by recruiting immune cells to injured tissues and CCL4 contributes to oxidative stress, fibroblast activation, and immune cell recruitment,^64-67 coupled with iron's critical role in immune cell function,⁶⁸ we hypothesize that ID induces distinct metabolic reprogramming of kidney immune cells relative to other tissues. Consequently, this metabolic rewiring augments the sensitivity of kidney immune cells to inflammatory stimuli, thereby amplifying their profibrotic activity. Similarly, in the heart, severe ID potentiated expression of multiple fibrotic markers in adenine-CKD but moderate ID elevated Acta2 expression in Alport mice (Figure 6; supplemental Figures 10 and 11). This suggests potential similarities in metabolic rewiring between the kidney and the heart in response to different levels of ID. Nonetheless, additional studies are required to test this hypothesis.

Beyond its role in fibrosis, ID may also contribute to cardiovascular complications in CKD, particularly by modulating vascular calcification. Although vascular calcification is a major driver of cardiovascular mortality in CKD,^48,69,70 the impact of ID on cardiac injury, especially through its effects on kidney and cardiac calcification, remains largely unexplored. Although few studies directly suggest that ID influences vascular calcification,^31-33 our findings indicate that the pathogenesis of heart failure occurs independently of ID-induced changes in kidney and cardiac calcification (supplemental Figures 2-3 and 9-10). Furthermore, we observed no significant ID-attributable differences in heart injury across both CKD models (Figure 6; supplemental Figure 11). However, we did not directly evaluate cardiomyocyte intracellular ID or mitochondrial dysfunction, both of which are known to be associated with impaired cardiac function.⁷¹ 

This study has several strengths. Using both an adenine-induced nephropathy model and a genetic Col4a3^–/– (Alport) model, we provide a comprehensive assessment of models of CKD pathology. In addition, we established different gradations of ID, enabling a nuanced understanding of how different degrees of ID affect CKD. This approach better reflects heterogeneity observed in human patients with CKD. However, this study also has notable limitations. First, only male mice were used, as they exhibit a consistent CKD phenotype. Although this choice is common in the field, it limits generalizability of our findings to females. Second, the adenine model is prone to inducing a severe, often lethal CKD phenotype. Given that ID did not exacerbate kidney injury despite presence of extensive fibrotic lesions after 8 weeks of adenine, a time-course evaluation could help clarify whether ID accelerates kidney injury at earlier stages compared with iron-replete conditions. Third, rapid progression of kidney injury in Alport mice leads to mortality by 10 weeks of age; thus, a 5-week 4-ppm Fe diet may not fully capture long-term effects of moderate ID in this model. Last, although we assessed cardiac fibrosis, we did not evaluate functional cardiac parameters, leaving this aspect of CKD-related cardiac effects unexplored.

In conclusion, our study establishes models of moderate and severe ID in adenine nephropathy and moderate ID in Alport mice. We found that ID amplifies systemic inflammation and worsens both kidney and cardiac fibrosis, particularly in adenine-induced CKD. Notably, ID did not affect other CKD-related complications, including kidney injury and dysfunction, vascular calcification, or cardiomyopathy, in either model. Our data suggest that iron supplementation may be beneficial to combat inflammation-induced kidney damage and offer a more comprehensive understanding of its therapeutic potential. Future studies exploring targeted iron supplementation strategies to mitigate kidney and cardiac fibrosis in CKD complicated with ID are warranted to translate these findings into clinical practice.

The authors greatly appreciate the support of the University of California Los Angeles Integrated Molecular Technologies Core and Translational Pathology Core Laboratory for their expertise in conducting the serum 32-plex cytokine/chemokine immunoassay and tissue processing/staining for histopathologic analysis.

The authors acknowledge financial support from the National Institutes of Health: K01-DK-127004 (V.S.), IK2-CX-002195 (S.S.), R01-DK-126680 (T.G.), R01-DK-136691 (E.N.), and T32-HL-072752 (B.C.).

Contribution: M.Z. and A.L. performed the experiments and analyzed the data; G.J., J.D.O., and V.S. assisted with the experiments; S.S., T.G., E.N., and M.R.H. helped conceive and plan the project and assisted with data interpretation and manuscript editing; B.C. conceived the project, designed and performed the experiments, analyzed the data, prepared the figures, and wrote the manuscript; and all authors discussed the results, read and contributed edits to the manuscript, and approved the final version.

Conflict-of-interest disclosure: T.G. is a shareholder and a scientific advisor of Intrinsic LifeSciences and Silarus Therapeutics, and a consultant for Silence Therapeutics, Intrinsic LifeSciences, Ionis Pharmaceuticals, Disc Medicine, City Therapeutics, Chugai, Dexcel, and Bristol Myers Squibb. E.N. is a shareholder and a scientific advisor of Intrinsic LifeSciences and Silarus Therapeutics, and a consultant for Disc Medicine, Ionis Pharmaceuticals, Protagonist, Vifor, GlaxoSmithKline, Chiesi, and Novo Nordisk. The remaining authors declare no competing financial interests.

Correspondence: Brian Czaya, Center for Iron Disorders, Department of Medicine, David Geffen School of Medicine at UCLA, 43-229 CHS, 10833 Le Conte Ave, Los Angeles, CA 90095; email: [email protected].