Skip to main content
NCBI home page
Blood logoLink to Blood
. 2025 Jun 18;146(8):984–997. doi: 10.1182/blood.2025029352

TNFα signaling restores steady-state hematopoiesis in a TNFαKO mouse model of anemia of inflammation∗

Amaliris Guerra 1,∗, Vania Lo Presti 1, Ding-Wen Chen 1,2, Ana Catarina Martins 3, Ariel Rivera 1, Nolan Hamilton 1, Pankja Sharma 1, Yelena Z Ginzburg 4, Carlo Castruccio Castracani 1, Carla Casu 1, Ritama Gupta 1, Raffaella Gozzelino 3, Edward M Behrens 5, Laura F Bennett 6, Robert F Paulson 6, Peter Kurre 1,2,7, Stefano Rivella 1,7,8,9,10,11
PMCID: PMC12783521  PMID: 40512402

Key Points

  • •

    Germ line deletion of Tnfα results in irreversible anemia in an anemia of inflammation mouse model.

  • •

    TNFα plays an important role in restoring steady-state erythropoiesis after inflammatory event.

Visual Abstract

graphic file with name BLOOD_BLD-2025-029352-ga1.jpg

Open in a new tab

Abstract

Anemia of inflammation (AI) is the second most common form of anemia and is prevalent in patients with chronic inflammatory states, such as infection, autoimmunity, and cancer. Interleukin 6 (IL-6) is well-known to induce the iron-sequestering hormone hepcidin, which results in iron-restricted anemia. The contributions of other proinflammatory cytokines, such as tumor necrosis factor-α (TNFα) and interferon gamma (IFNγ), are less understood in the pathophysiology of AI. This study investigated the role of TNFα in a mouse model of AI by administering heat-killed Brucella abortus (HKBA) to germ line TNFα knockout (KO) mice. We hypothesized that TNFα possessed an important role in restoring steady-state erythropoiesis after inflammatory insult. TNFαKO injected with HKBA displayed a chronic anemia, with elevated proinflammatory IL12p40 and IFNγ cytokines that did not resolve. However, IFNγKO and TNFαKO/FNγKO double knockout mice showed reduced inflammation and anemia following HKBA administration. Because IFNγKO displayed normal serum TNFα and IL12p40 levels, we hypothesized that the persistent anemia was IFNγ induced and TNFα was necessary for AI cessation. However, treatment with recombinant TNFα (rTNFα) accelerated death, while reducing IFNγ using an anti-IFNγ antibody (Ab) only briefly improved anemia. Only the combination of both the Ab and rTNFα together reversed the hyperinflammatory phenotype, restored erythropoiesis, and prevented death of TNFαKO + HKBA mice. Our data provide compelling evidence for an anti-inflammatory role of TNFα that is necessary for the restoration of erythropoiesis and mitigation of proinflammatory IFNγ action in a mouse model of AI.

Introduction

Anemia of inflammation (AI) is the second most prevalent anemia worldwide after iron deficiency anemia.1 Causes of AI include microbial pathogen infections and systemic autoimmune diseases.1 AI is the most common anemia among hospitalized and chronically ill patients, yet no targeted treatments are currently available.2 The biological mechanisms of AI remain speculative; however, a recent study of patients with COVID-19 showed that those with more pronounced anemic states exhibited higher acute systemic inflammatory responses.3 Strategies to mitigate AI typically focus on treating the underlying autoimmune, infectious, or malignant disease, using transfusion in severe and/or symptomatic cases, and combining iron therapy with erythropoiesis-stimulating agents.2 A deeper understanding of how other proinflammatory cytokines contribute to erythropoiesis in the etiology and pathophysiology of AI may open new therapeutic avenues.

Currently, tumor necrosis factor α (TNFα)-blocking agents are the gold standard for treating many autoimmune diseases. Clinical studies have shown that drugs targeting TNFα improve anemia as a secondary treatment outcome.4,5 TNFα is a potent proinflammatory cytokine with diverse functions across different immune cell types in response to various stimuli, including infection, injury, and inflammation.6 The impact of TNFα on erythropoiesis in AI remain poorly understood. In mouse studies, TNFα administration has been shown to induce anemia and hypoferremia.7,8 TNFα has also been shown to negatively regulate the interferon gamma (IFNγ)-dependent inflammatory response by restricting the duration of macrophage-derived interleukin 12 (IL-12) production in TNFα knockout (KO) mice.9 Monocyte and T helper cell 1 T-cell production of IFNγ is the primary activator of proinflammatory macrophages.10,11 Pathologically, excess IFNγ has been implicated in promoting myeloid differentiation, tissue damage, necrosis, inflammation, anemia, and bone marrow (BM) failure.11, 12, 13, 14, 15, 16, 17 IFNγ also induces erythrophagocytosis by macrophages and has been implicated in regulating hepcidin and other iron-related proteins.17 Additionally, both TNFα and IFNγ have been shown to inhibit early- and late-stage erythropoiesis in vitro and in vivo.7,18, 19, 20, 21, 22

This study aimed to explore the effects of TNFα and IFNγ on erythropoiesis during acute and chronic inflammation using a well-established heat-killed Brucella abortus (HKBA)-induced mouse model of AI. Our data showed that TNFαKO + HKBA mice develop macrocytic hypochromic anemia, leukocytosis, and die from BM failure. However, HKBA-injected mice lacking IFNγ or the double knockout (DKO) group exhibited reduced inflammation and normalized erythropoiesis. Interestingly, treatment with a combination of anti-IFNγ antibody (Ab) and recombinant TNFα reversed the hyperinflammatory phenotype, normalized erythropoiesis, and prevented premature death in TNFαKO + HKBA mice. Our data strongly suggest that TNFα plays a necessary anti-inflammatory role in AI by modulating IFNγ action.

Materials and methods

Mice housing

Wild-type (WT) mice (C57BL/6J – JAX:000664) and TNFαKO (B6;129S-Tnftm1Gkl/J – JAX:005540) were purchased from The Jackson Laboratory. IFNγKO mice were a generous gift from Edward M. Behrens. All mice were crossed onto a C57BL/6J, as shown in supplemental Figure 1 (available on the Blood website) and housed in a level C animal facility at Children’s Hospital of Philadelphia (CHOP). Both male and female age-matched (2-4 months) mice were used. The study followed the recommendations of the Institutional Animal Care and Use Committee (IACUC) regulation (protocol IACUC IAC 18-001173_AM03) at CHOP.

Mice treatment with heat-killed Brucella abortus

HKBA preparation and intraperitoneal injection of HKBA were done as previously described.23

Hematological parameters

Blood samples were collected by retro-orbital bleeding. Blood samples were analyzed at CHOP by the Institutional Clinical and Translational Science Award Research Center. Serum samples were analyzed as previously described.24

Biotinylated red blood cell injection

In vivo, erythrocyte biotinylation for lifespan measurements in WT and TNFαKO mice were done as previously described.25,26

Fluorescence-activated cell sorting analyses of erythroid, lymphoid, myeloid, and HSC populations

Erythroid populations of the BM and spleen were assessed by flow cytometry using erythroid markers (CD71, CD44, and Ter119) as previously described.27 The LSK fraction was identified as Lin− cKIT+ ScaI+, and the LT-hematopoeitic stem cells (HSC) was defined as Lin− cKIT+ ScaI+ Flt3− CD34− or Lin− cKIT+ ScaI+ CD150+ CD48− based on the availability of reagents in the laboratory at the time of the experiment. Cellularity was determined by counting cells using AccuCheck Counting Beads (Thermo catalog number PCB100). Lymphoid populations were selected using CD45, CD3, and B220. T cell naïve (CD62Lhigh CD44low/neg), memory (CD62Lhigh CD44high), and effector (CD62Llow/neg CD44high) cells were assessed using CD4, CD8, CD62L, TCRβ, CD44, and anti-TNFα and anti-IFNγ to measure their intracellular TNFα and IFNγ levels. Cells were fixed using BioLegend's True-Nuclear Transcription Factor Buffer. Myeloid populations were identified using CD115, F4/80, Gr1 (Ly-6G and Ly-6C), CD206, and MCHII to discriminate proinflammatory macrophage phenotypes. Protocols were performed as previously described.28 All samples were recorded using FACSCanto II flow cytometer (BD, Franklin Lakes, NJ). Data were analyzed with FlowJo 10.2 (Tree Star, Inc, Ashland, OR).

Transplanting into CD45.1 mice

To generate mixed chimeras, CD45.1 recipient mice received 10E+05 LT-HSC as calculated by fluorescence-activated cell sorting (FACS) on red blood cell (RBC) lysed BM cells and determining the frequency of LT-HSC (Lin−ScaI+CD150+CD48−) per sample. CD45.1 mice were lethally irradiated with 2 administrations of 500 cGy separated by 4 hours, for a total of 1000 cGy, using an X-Rad320 Irradiator (Precision, Madison, CT).29 RBC lysed BM containing 10E+05 LT-HSC was intravenously injected into lethally irradiated recipient mice.

In vitro assessment of TNFα and IFNγ production in CD4+TCRβ+ and CD8+TCRβ+

BM and spleen cell suspensions from WT and TNFαKO mice were plated in complete RPMI media and treated with PMA:ionomycin and brefeldin A. After a 24-hour incubation at 37°C and 5% CO2, cells were stained for CD45+TCRβ+CD4+ or CD45+TCRβ+CD8+ T cells, fixed, and then permeabilized and stained for intracellular TNFα or IFNγ.

In vivo IFNγ depletion and rTNFα administration

TNFαKO + HKBA mice were treated 6 to 8 weeks after HKBA treatment based on their hemoglobin (Hb) values. Animals with Hb below 6 g/dL were sacrificed as required by our IACUC protocol. Mice with Hb between 6 and 8 g/dL were treated either 200 mg anti-IFNγ Ab to deplete IFNγ and/or 0.25 mg of recombinant TNFα (rTNFα) via retro-orbital injection every other day for 3 doses. The treatment course was adapted from a previous study.30

Statistical analysis

All statistics were calculated using GraphPad Prism (v10.4). Statistics data are presented as mean and standard deviation (SD). Statistical tests are as indicated in figure legends.

Results

TNFαKO + HKBA mice exhibited irrecoverable chronic anemia

We assessed complete blood cell counts (CBCs) over 6 to 8 weeks (Figure 1A-C). TNFαKO mice injected with HKBA displayed severe anemia, characterized by reduced RBC, Hb, and a concomitant increase in reticulocytes (RETs) starting 2 weeks after HKBA administration. While WT + HKBA mice began to recover by 4 weeks, TNFαKO + HKBA mice showed persistent reductions in RBC counts and Hb concentrations, along with elevated RET counts, from 2 weeks until the end of the study at 6 to 8 weeks. Additionally, white blood cell (WBC) counts were significantly elevated in TNFαKO + HKBA mice compared with WT + HKBA (Figure 1D). These results suggest immune dysregulation due to the absence of TNFα.

Figure 1.

Figure 1.

Open in a new tab

TNFKO mice treated with HKBA display a chronic anemia phenotype. TNFαKO + HKBA displayed sustained reductions in (A) RBC, (B) Hb, and (C) RET and elevated (D) WBC over the 8-week time course. Chronic anemia in TNFαKO + HKBA was not due to hypoferremia. (E) Serum iron levels, (F) hepcidin levels, (G) erythroferrone (ERFE), and (H) FGL1 show that TNFαKO + HKBA mice were not iron restricted compared with WT + HKBA mice. (I) TNFαKO + HKBA displayed lower levels of IL6 4 to 6 hours after HKBA injection compared with WT + HKBA. (J) Serum EPO levels in TNFαKO + HKBA mice were similar to WT + HKBA controls until the 6-week time point, where serum EPO was elevated in TNFαKO + HKBA compared with WT + HKBA. IL12p40 (K) and IFNγ (L) were abnormally elevated 4 weeks after HKBA injection in TNFαKO + HKBA mice. (M) TNFαKO + HKBA animals had the lowest probability of survival compared with WT + HKBA. Statistics data are presented as mean and standard deviation (SD). Data were analyzed by 2-way analysis of variance (ANOVA). Both male and females 2 to 3 months old were used. For survival curve, statistics were analyzed using the log-rank (Mantel-Cox) test. Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001. Number of animals used is missing (uniformization of n = or N =).

Because hypoferremia limits erythropoiesis, we assessed whether iron restriction was responsible for the unresolved anemia. Serum analysis of iron markers showed that serum iron (Figure 1E) was only elevated in TNFαKO + HKBA mice compared with WT + HKBA at the last collection time point. Serum hepcidin levels were reduced in TNFαKO + HKBA mice compared with WT + HKBA mice (Figure 1F). Furthermore, there were no differences between serum erythroferrone (Figure 1G) or liver fibrinogen-like-protein 1 (Fgl1) mRNA levels relative to hypoxanthine-guanine phosphoribosyltransferase (HPRT) (Figure 1H). This indicates that limited iron availability was not the cause for depressed erythropoiesis. We also evaluated serum IL-6 and erythropoietin (EPO), as both can directly or indirectly modulate hepcidin regulation. Our findings show that IL-6 was moderately elevated at the 4- to 6-hour cytokine storm time point in TNFαKO + HKBA mice but was significantly lower than in WT + HKBA controls (Figure 1I). This indicates that hepcidin expression in this AI model may be more strongly influenced by the erythropoietic state than by inflammation. We measured serum EPO levels, noting that between 4 to 6 hours and 1 week, EPO levels in TNFαKO + HKBA mice were comparable to those in WT + HKBA mice. However, by 6 weeks, EPO levels were significantly higher in the serum of TNFαKO + HKBA mice (Figure 1J). These results exclude hepcidin-induced iron restriction or low EPO levels as underlying causes of persistent anemia in TNFαKO + HKBA mice.

HKBA is a potent T helper cell 1 adjuvant that leads to strong IL-12 and IFNγ responses. By 4 weeks, serum IL12p40 (a proinflammatory subunit of IL-12) and IFNγ levels in WT + HKBA mice return to baseline, but not in TNFαKO + HKBA mice (Figure 1K-L). We analyzed the lifespan and clearance of transfused biotinylated RBCs, observing the most significant reductions in RBCs in TNFαKO + HKBA mice (supplemental Figure 2A-B). These results support the notion that sustained IFNγ elevations in TNFαKO + HKBA mice contributed to the observed anemia phenotype by increasing clearance of RBCs. Additionally, IFNγ-driven BM failure due to chronic inflammation and stem cell exhaustion has been reported in other studies.17 Given that TNFαKO + HKBA mice also exhibited a decline in platelet counts (supplemental Figure 3), we investigated whether the high levels of serum IFNγ in these mice resulted in premature death (Figure 1M) due to IFNγ-dependent BM failure. BMs from WT and TNFαKO mice treated with HKBA were assessed at 8 weeks for cellularity, Lin−ScaI+cKIT+ (LSK) cells, and long-term hematopoietic stem cells (LT-HSC) (Lin−cKIT+ScaI+CD34−Flt3−) (supplemental Figure 4A-D). We observed that, although TNFαKO + HKBA mice had reduced BM cellularity (supplemental Figure 4B), both LSK and LT-HSC absolute counts were significantly elevated in TNFαKO + HKBA mice compared with WT + HKBA mice (supplemental Figure 4C-D). To explore if the cause of death in TNFαKO + HKBA mice was due to an HSC defect, we transplanted BM from WT + HKBA and TNFαKO + HKBA mice (8 weeks post-HKBA) into irradiated CD45.1 mice (supplemental Figure 5A-B). Our findings show that CD45.1 mice transplanted with BM from TNFαKO + HKBA mice had a lower probability of survival (supplemental Figure 5A) and lower percentages of engraftment (supplemental Figure 5B). These data suggest that elevated IFNγ may result in BM failure in TNFαKO + HKBA mice.

Absence of IFNγ ameliorated anemia, normalized WBC, and corrected erythropoiesis

Because elevated IFNγ was a key feature in TNFαKO + HKBA mice, we tested if concurrent loss of TNFα and IFNγ would improve AI. Both IFNγKO + HKBA and DKO + HKBA mice displayed milder anemic phenotypes relative to TNFαKO + HKBA mice (Figure 2A-C), with RBC, Hb, and hematocrit similar to WT + HKBA mice (supplemental Tables 1 and 2). The probability of survival of IFNγKO + HKBA and DKO + HKBA was comparable to WT + HKBA controls (supplemental Figure 6). Additionally, the mean corpuscular volume, mean corpuscular Hb concentration, and RET counts were altered only in TNFαKO + HKBA mice for most time points analyzed (Figure 2D-F). Altogether, these observations indicate these mice are unable to correct anemia, generate normal RBC, and restore steady-state erythropoiesis without Tnfα after IFNγ exposure. In the WBC compartment, only in TNFαKO + HKBA monocytes and lymphocytes remained elevated by the end of the study compared with the other mice (Figure 2G-I; supplemental Table 2). Our results suggested that the absence of IFNγ dampened both innate and adaptive immune responses, while the absence of Tnfα caused an exacerbation in the immune compartment when IFNγ was present. Representative blood smears at 8 weeks illustrate these findings (Figure 2J). To gain insights regarding the abscence of Tnfα and/or Ifnγ on erythropoiesis, BM and spleens were assessed by FACS (supplemental Figure 7A). There were no differences in the percentage of the erythroid cell fraction in the BM as measured by Ter119+ at T0 between untreated mice (Figure 3A). However, at 2 weeks, erythropoiesis was dramatically reduced in the BMs of WT + HKBA and TNFαKO + HKBA, but not in IFNγKO + HKBA or DKO + HKBA compared with their baseline values. Erythroid percentages remained low only in TNFαKO + HKBA BMs, while all other groups fully recovered (supplemental Figure 7B). Interestingly, IFNγKO + HKBA and DKO + HKBA had statistically significant elevations in Ter119+ cell percentages at 8 weeks compared with baseline values (Figure 3A), but no statistically significant differences in the absolute number of (nonerythroid + erythroid) cells (Figure 3B-C). This finding may be a consequence of reduced anemic phenotype and lacking Ifnγ, which is a key player in erythrocyte recycling. Only TNFαKO + HKBA mice showed reduced BM cellularity and erythropoiesis. In the spleen, the erythroid fraction was increased at the 2-week time point in WT + HKBA and TNFαKO + HKBA compared with baseline values, whereas IFNγKO + HKBA and DKO + HKBA showed no major differences from baseline values until 6 to 8 weeks (Figure 3D-F). Spleen-to-body-weight ratios in IFNγKO + HKBA and DKO + HKBA showed no splenomegaly at any time point (Figure 3G). However, TNFαKO + HKBA showed increased splenomegaly at 2 and after 6 weeks (Figure 3G-H). The increase in spleen size at 2 weeks was consistent with WT + HKBA mice when stress erythropoiesis was induced after the initial inflammatory response, leading to an accumulation of new erythrocytes that increased spleen mass. In the WT + HKBA mice, the resolution of inflammation occurred by 4 weeks, and a reduction in spleen size was observed at this time (Figure 3G). However, TNFαKO + HKBA mice showed elevated RET but low RBC and Hb levels at 4 weeks. TNFα has been implicated in preserving the stemness and self-renewal capacity of stress erythroid progenitors in the spleen.31,32 Our data indicate an inefficient stress erythropoiesis response in TNFαKO mice, which may be due to an important role in splenic erythroid progenitor expansion.33

Figure 2.

Figure 2.

Open in a new tab

HKBA-treated TNFαKO, and IFNγKO show opposite phenotypes in AI. WT (gray), TNFαKO (magenta), IFNγKO (green), and DKO (blue) were subjected to the HKBA. WT (n = 9-25), TNFαKO (n = 18-58), IFNγKO (n = 12-30), and DKO (n = 3-30) mice were followed for 6 to 8 weeks after HKBA injection. (A-C) IFNγKO + HKBA and DKO + HKBA mice showed a milder anemic phenotype by (A) RBC, (B) Hb, and (C) hematocrit (HCT) compared with WT + HKBA and TNFαKO + HKBA mice. (D) TNFαKO + HKBA showed a macrocytic phenotype that was not observed in WT + HKBA, IFNγKO + HKBA, or DKO + HKBA mice. (E) Mean corpuscular Hb concentrations (MCHCs) showed reduced Hb per cell in TNFαKO + HKBA only 2 weeks after HKBA administration. (F) As early as 1 week, the DKO + HKBA mice exhibited the most accelerated reticulocyte recovery, followed by the IFNγKO + HKBA and WT + HKBA, while the TNFαKO + HKBA group exhibited elevated reticulocytes beyond the 4-week time point. (G) The WBC counts comparison revealed that both IFNγKO + HKBA and DKO + HKBA mice exhibited normalized WBC counts with levels similar to the WT + HKBA group, in contrast to TNFαKO + HKBA mice that experienced an increase and sustained WBC counts. (H) Lymphocyte counts (I) and monocyte counts remained elevated in TNFαKO + HKBA. (J) Analysis of blood smears at 8 weeks shows a reduced number of RBCs with a macrocytic phenotype and increased leukocytes in TNFαKO + HKBA compared with all other groups, which show recovery. Statistics data are presented as mean and standard deviation (SD). Both male and female 2 to 3 months old were used. Data were analyzed by the 2-way analysis of variance (ANOVA). Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

Figure 3.

Figure 3.

Open in a new tab

Simultaneous germ line deletion of IFNγ corrects erythropoiesis in DKO + HKBA mice. Erythropoiesis was assessed in BM and spleens (A-F) by FACS using the CD71, CD44, and Ter119 markers. WT (gray), TNFαKO (magenta), IFNγKO (green), and DKO (blue) were subjected to the HKBA, and mice were collected at various time points for 8 weeks after HKBA injection. (A) Time course analysis of the percentage of the total frequency of erythroid cells (Ter119+) present in the BM shows that WT + HKBA and TNFαKO + HKBA have suppressed erythropoiesis at 2 weeks, whereas IFNγKO + HKBA and DKO + HKBA do not. By the end of the time course, WT + HKBA erythroid percentage of total frequency returned to baseline, whereas IFNγKO + HKBA and DKO + HKBA were mildly elevated compared with baselines. Depressed erythropoietic output continued in TNFαKO + HKBA mice at 6 to 8 weeks. BM total cell counts (B) and total number of erythroid cells (C) were suppressed only in TNFαKO + HKBA mice at 6 to 8 weeks, but no statistical differences were observed in IFNγKO + HKBA or DKO + HKBA compared with WT + HKBA controls. (D) Splenic extramedullary erythropoiesis showed WT + HKBA and TNFαKO + HKBA had increased percentages of total frequency at 2 weeks. Only TNFαKO + HKBA remained elevated at 4 weeks and returned to baseline by 6 to 8 weeks. Spleen (E-F) total cell counts (E) and Ter119+ cells (F) display elevated erythroid counts in TNFαKO + HKBA mice at the 8-week time point. (G) Spleen-to-body weight ratios showed persistent splenomegaly in the TNFαKO + HKBA group that deviated from normal WT + HKBA controls only at 8 weeks. (H) Representative image of the spleen size comparison between groups at 8 weeks after HKBA. WT (n = 4-6), TNFαKO (n = 3-6), IFNγKO (n = 3-6), and DKO (n = 3-6) for each time point. Both male and female 2 to 3 months old were used. Data were analyzed by the 1- or 2-way analysis of variance (ANOVA). Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

TNFαKO + HKBA mice displayed a dysregulated immune phenotype

The inefficient erythropoietic stress response, but large spleen size, high WBC count, high lymphocyte count, and elevated serum IFNγ levels led us to assess the WBC compartment in HKBA-treated mice. Assessment of T cells (CD45+CD3+) and B cells (CD45+B220+) (supplemental Figure 8A) at T0 showed no differences in the percentages of lymphocytes in the BM (Figure 4A) or spleens (Figure 4C). Total T-cell counts in TNFαKO + HKBA mice showed that T-cell numbers were significantly increased compared with the other groups at 8 weeks (Figure 4C). There were no differences in T-cell percentages in any other group at any time point. Similar results were observed in TNFαKO + HKBA mouse spleens, where T cells were elevated by both percentage and cell counts (Figure 4D). Additionally, immunohistochemistry using an anti-CD3 antibody showed disorganization of the white pulp in the spleens of TNFαKO + HKBA (supplemental Figure 9A) and evidence of T-cell infiltration, as suggested by the elevated anti-CD3 reactivity in the liver compared with the other genotypes (supplemental Figure 9B). Next, we further characterized the CD3+ subpopulations by CD4+ (supplemental Figure 10A-B) and CD8+ (supplemental Figure 11A-B) subtypes. We found that both BM and splenic CD4+ memory cells and effector cells (supplemental Figure 10C-D), as well as CD8+ memory cells and effector cells (supplemental Figure 11C-D), were significantly elevated in TNFαKO + HKBA compared with controls. CD8+ naïve T cells in the BM were also elevated compared with controls (supplemental Figure 11C). To gain insights regarding elevated serum IFNγ in TNFαKO + HKBA, we activated T cells in vitro and measured intracellular TNFα and IFNγ mean fluorescence intensity by FACS (supplemental Figure 12A-B) in CD4+ and CD8+ T cells. Our results show that CD4+ T cells produce similar IFNγ levels in T cells from BM samples; however, in spleens, IFNγ was reduced in TNFαKO CD4+ T cells compared with WT controls (supplemental Figure 12A). Analysis of CD8+ T cells from the BM and spleen showed a different result, where CD8+ T cells displayed higher IFNγ mean fluorescence intensity levels in both BM and spleen samples (supplemental Figure 12B). This suggests that CD8+ T cells may play a significant role in producing the IFNγ serum levels observed in TNFαKO + HKBA. Our data point to a potential intrinsic regulatory role for TNFα on IFNγ production at the cellular level. Taken together, our data support other studies showing a crucial suppressive role of TNFα in T-cell regulation.30,34,35

Figure 4.

Figure 4.

Open in a new tab

Absence of TNFα results in an aberrant lymphocyte profile that is corrected in DKO + HKBA mice. WT (gray), TNFαKO (magenta), IFNγKO (green), and DKO (blue) were injected with HKBA, and mice were assessed at various time points for 8 weeks after HKBA injection. (A) FACS analysis of BM and spleen (B) over the 8 weeks shows that only the TNFαKO + HKBA had elevations in T cells (CD45+CD3+), which never return to baseline levels. (C) Absolute counts of T-cell population in the BM showed that the significant expansion of T cells persistent at the 8-week time point in TNFαKO + HKBA but not the other groups. (D) Absolute counts of T-cell populations in the spleen showed that the significant increase in T-cell population was sustained until the last time point (8 weeks) in TNFαKO + HKBA but not the other groups. (E) The BM percentage of B cells (CD45+B220+) showed that this cell lineage was significantly decreased in the TNFαKO + HKBA and WT + HKBA group at 2 weeks, but only WT + HKBA recovered to baseline levels by 8 weeks. No difference was found in the IFNγKO and DKO at any time compared with baseline levels. (F) In the spleen, B-cell percentages were reduced at 2 weeks in the TNFαKO + HKBA and WT + HKBA groups, but only WT + HKBA recovered to a steady-state level. No difference was found in the IFNγKO and DKO at any time compared with baseline levels. (G) Absolute counts of BM B cells showed that this B-cell population was significantly reduced only in the TNFαKO + HKBA compared with the other groups. (H) No significant differences were found in absolute B-cell counts in the spleen. WT (n = 4-6), TNFαKO (n = 3-6), IFNγKO (n = 3-6), and DKO (n = 3-6) for each time point. Both male and female 2 to 3 months old were used. Data were analyzed by the 1- or 2-way analysis of variance (ANOVA). Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

Assessment of the B-cell fraction in the BM showed that the percentage and total number of B cells were unchanged in IFNγKO + HKBA and DKO + HKBA mice at any point but were reduced at 2 weeks in WT + HKBA and TNFαKO + HKBA mice (Figure 4E). In the spleen, we observed the same trend in B-cell percentages as in the BM for all groups analyzed (Figure 4F). However, BM B-cell populations in TNFαKO + HKBA mice remained suppressed after 2 weeks until 8 weeks (Figure 4G). In the spleen, the absolute number of B cells were not different from other groups (Figure 4H). Additionally, we assessed the percentage and abundance of BM and splenic macrophages (CD45+Gr1–CD115intF4/80+) (Figure 5A-B) and used a variety of markers discriminating between proinflammatory and nonproinflammatory cell types (supplemental Figures 13 and 14). In TNFαKO + HKBA mice, the proportion of macrophages in the BM elevated after 4 weeks, whereas other groups return to steady stage (Figure 5A). However, their absolute counts were approximately the same as WT + HKBA (Figure 5C). In the spleen, TNFαKO + HKBA showed significantly more macrophages at 8 weeks (Figure 5D), which may be explained by the relatively large spleen size of TNFαKO + HKBA compared with the other groups. This result demonstrates the critical role of TNFα in restoring the B-cell and macrophage pools during the resolution of inflammation. In summary, while IFNγKO + HKBA and DKO + HKBA models appeared to have a mild immune response, TNFαKO + HKBA mice displayed a severely dysregulated immune system with increased IFNγ-producing T cells, a diminished B-cell profile, and elevated macrophages.

Figure 5.

Figure 5.

Open in a new tab

Absence of TNFα and IFNγ in DKO mice resulted in a normalized macrophage profile. WT (gray), TNFαKO (magenta), IFNγKO (green), and DKO (blue) were subjected to HKBA, and mice were collected at various time points for 8 weeks after HKBA injection. (A) FACS analysis of BM and spleen (B) over the 8 weeks shows that only the TNFαKO + HKBA had an elevated percentage of macrophages (CD45+Gr1–CD115intF4/80+) 2 weeks after AI induction by HKBA, with no resolution only in the BM. (C) Absolute counts of macrophages were not statistically different in TNFαKO + HKBA compared with controls. (D) Macrophage absolute cell counts were elevated in the spleen of TNFαKO + HKBA. WT (n = 4-6), TNFαKO (n = 3-6), IFNγKO (n = 3-6), and DKO (n = 3-6) for each time point. Both male and female 2 to 3 months old were used. Data were analyzed by the 1- or 2-way analysis of variance (ANOVA). Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

Loss of IFNγ restores IL12p40 levels

TNFα and IL12 are 2 major mediators of the early inflammatory response, playing critical roles in macrophage and/or dendritic cell activation and bridging innate and adaptive phases of immunity.36 We assessed IFNγ and IL12p40 cytokine levels at acute (4-6 hours) and resolution time points (4 weeks) after HKBA administration. Cytokine measurements of IL12p40 during the first acute response were suppressed in TNFαKO + HKBA and DKO + HKBA; however, they remained elevated in only TNFαKO + HKBA by 4 weeks (Figure 6A). Serum IFNγ levels were similar at 4 to 6 hours between WT + HKBA and TNFαKO + HKBA but remained elevated at 4 weeks in TNFαKO + HKBA (Figure 6B). Serum TNFα levels in IFNγKO + HKBA were comparable to WT + HKBA levels at 4 to 6 hours but returned to baseline by 4 weeks (Figure 6C). Additionally, IFNγKO + HKBA mice displayed elevated IL-6 levels at 4 to 6 hours similar to WT + HKBA mice, whereas DKO + HKBA mice had lowered IL-6 levels similar to TNFαKO + HKBA (Figure 6D). Serum levels of iron were normal in IFNγKO + HKBA and DKO + HKBA compared with WT + HKBA by 6 to 8 weeks (Figure 6E). There were no differences in serum hepcidin, EPO, erythroferrone, or FGL1 in IFNγKO + HKBA and DKO + HKBA compared with WT + HKBA (Figure 6F-I). These data are in agreement with other studies that show that TNFα is necessary for resolving IL-12 and IFNγ-dependent responses.37

Figure 6.

Figure 6.

Open in a new tab

IFNγKO + HKBA mice showed normal serum TNFα levels. (A) Serum analysis of IL12-p40, at T0, 4 to 6 hours and 4weeks, showed IL12-p40 levels were reduced in TNFαKO + HKBA and DKO + HKBA groups at 4 to 6 hours after HKBA, compared with WT + HKBA and IFNγKO + HKBA, but remained elevated only in TNFαKO + HKBA at 4 weeks. (B) IFNγ serum levels at 4 to 6 hours after HKBA injection were not different from TNFαKO + HKBA compared with WT + HKBA; however, the level was elevated at 4 weeks. (C) TNFα serum levels at 4 to 6 hours after HKBA injection were not statistically different between WT + HKBA or IFNγKO + HKBA, and returned to baseline by 4 weeks. (D) Serum analysis of IL-6, at T0, 4 to 6 hours, and 4 weeks, showed IL-6 levels were reduced in TNFαKO + HKBA and DKO + HKBA groups at 4 to 6 hours after HKBA, compared with WT + HKBA and IFNγKO + HKBA, but were undetectable in all groups by 4 weeks. (E) Serum iron levels in IFNγKO + HKBA and DKO + HKBA mice were similar to WT + HKBA at baseline and at 6 to 8 weeks. (F) Serum hepcidin and (G) EPO levels in IFNγKO + HKBA and DKO + HKBA mice were not different from WT + HKBA. (H) Erythroferrone (ERFE) levels were significantly reduced in IFNγKO + HKBA and DKO + HKBA mice compared with WT + HKBA and TNFαKO + HKBA at 2 weeks. (I) No differences were found in liver FGL1 mRNA qPCR measurements relative to HPRT at T0 or 6 to 8 weeks among the groups. WT (n = 4-6), TNFαKO (n = 3-6), IFNγKO (n = 3-6), and DKO (n = 3-6) for each time point. Both male and female 2 to 3 months old were used. Data were analyzed by the 1- or 2-way analysis of variance (ANOVA). Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

TNFαKO + HKBA phenotype is rescued by anti-IFNγ Ab and rTNFα treatment

Several studies have shown TNFα exhibits anti-inflammatory properties.30,37 Because TNFαKO + HKBA displayed sustained elevated levels of IFNγ, and lack of Ifnγ in the DKO did not result in the persistent anemia phenotype, we hypothesized that either administration of TNFα or reduction in IFNγ was necessary for resolution of anemia. Initial treatments with rTNFα at the 6- to 8-week time point resulted in mortality of TNFαKO + HKBA within 1 to 2 days (Figure 7A). Next, we attempted to resolve AI in TNFαKO + HKBA by reducing circulating IFNγ by administration of an anti-IFNγ Ab. TNFαKO + HKBA mice were injected between 6 and 8 weeks with an anti-IFNγ Ab every other day, at doses previously reported.38 Results showed that anti-IFNγ Ab treated TNFαKO + HKBA mice showed improvements 1 week after anti-IFNγ Ab (supplemental Figure 15). However, these mice reverted to their anemic state by 2 weeks with reduced RBC, Hb, and increased RET (supplemental Figure 15). These findings suggest that although anti-IFNγ Ab treatment initially improved CBC parameters in TNFαKO + HKBA mice, it was insufficient to correct AI. Because IFNγKO + HKBA mice showed elevated TNFα cytokine levels in serum comparable to WT + HKBA, during the first acute inflammatory phase, and reducing IFNγ alone via an anti-IFNγ Ab was insufficient to rescue AI (as shown in supplemental Figure 15), we hypothesized that TNFα was necessary for recovery. Treatment of TNFαKO + HKBA at 8 weeks with anti-IFNγ Ab and rTNFα showed restored RBC, Hb, and WBC levels to baseline within the first week after anti-IFNγ Ab and rTNFα (9 weeks after HKBA injection) (Figure 7B-D). This result was sustained until the analysis of the mice at 20 weeks (Figure 7A-I; supplemental Figure 16). Analysis of the erythropoietic compartment of the BM by FACS showed restoration of erythropoiesis (supplemental Figure 16A). The quantification of the absolute number of Ter119+ cells (supplemental Figure 16B) and by erythroblast subtypes (supplemental Figure 16C) in the BM was no different from that of WT + HKBA controls. A similar result was observed in the spleen, where the absolute number of Ter119+ cells (supplemental Figure 16D) and erythroblast subtypes (supplemental Figure 16E) were not statistically different from WT + HKBA controls. Assessment of BM cellularity (Figure 7E), LSK fraction (Figure 7F), and LT-HSC pool (Lin–, cKIT+, ScaI+, CD150+CD48–) (Figure 7G) cell numbers showed that TNFαKO + HKBA mice treated with anti-IFNγ Ab and rTNFα were not statistically different from WT + HKBA animals at 20 weeks. The spleen-to-body weight ratio was also normalized in TNFαKO + HKBA mice treated with anti-IFNγ Ab and rTNFα (Figure 7H). Last, blood smears of TNFαKO + HKBA mice treated with anti-IFNγ Ab and rTNFα showed improvements in cell morphology and hypochromia and were indistinguishable from WT + HKBA at 20 weeks or other controls (Figure 2J; Figure 7I).

Figure 7.

Figure 7.

Open in a new tab

Combined administration of anti-IFNy Ab and rTNFα at 8 weeks after HKBA reverses the TNFαKO + HKBA phenotype. Survival curve (A) of mice treated with anti-IFNy Ab, rTNFα, or a combination of both show that TNFαKO + HKBA + rTNFα mice had the lowest probability of survival compared with all other mice. TNFαKO + HKBA + anti-IFNy Ab were similar to TNFαKO + HKBA. Only TNFαKO + HKBA + rTNFα + anti-IFNy Ab displayed a probability of survival similar to WT + HKBA. (B) RBC, (C) Hb, and (D) WBC levels improve 1 week after combined or double combination treatment and return to baseline 20 weeks after HKBA treatment (12 weeks after combination of anti-IFNy Ab and rTNFα treatment). (E-H) BM analysis at 20-weeks’ time point of WT + HKBA (gray) and TNFαKO + HKBA + rTNFα + anti-IFNyAb (magenta) mice. (E) Total cell counts of BM cells from the left femur, (F) total number of LSK cells (Lin–cKIT+ScaI+), (G) total number of HSC cells (Lin–cKIT+ScaI+CD150+CD48–), and (H) spleen-by-body weight ratios, showed the normalization of BM and spleen in TNFαKO + HKBA treated animals. (I) Analysis of blood smears at 20 weeks after HKBA (12 weeks after combination anti-IFNy Ab and rTNFα treatment) showed corrections in RBC and leukocyte proportions and profiles. Statistics data are presented as mean and standard deviation (SD) (n = 3-6). Data were analyzed by the 1-way analysis of variance (ANOVA) or the Student t-test. For survival curve, statistics were analyzed using the log-rank (Mantel-Cox) test. Statistical significance was considered as follows: ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

Discussion

The early stages of the inflammatory response in AI are reported to be IL-6 dependent, leading to the elevation of the iron-regulatory hormone hepcidin and iron-restricted erythropoiesis.1,39,40 In 2014, Gardenghi et al characterized the role of IL-6 and hepcidin in erythropoiesis using a well-established AI model induced by a single intraperitoneal injection of HKBA.23 The study raised questions regarding the role of other cytokines in the etiology of AI. This study investigated the effects of TNFα and IFNγ loss using the HKBA mouse model of AI.

Under immunologically unperturbed conditions, TNFαKO and IFNγKO mice were not different from WT mice. However, the importance of these cytokines became apparent under pathogen-induced inflammation. TNFαKO resulted in macrocytic, hypochromic anemia characterized by reticulocytosis low RBC numbers, and abnormally elevated serum IFNγ levels. Chronic anemia in TNFαKO + HKBA mice was not due to iron restriction. In contrast, DKO mice, which lacked TNFα and IFNγ, exhibited blunted inflammation and AI markers. Consequently, we focused on the immune profile to gain insights into the role of these cytokines in AI.

TNFα has been reported to contain anti-inflammatory properties by directly limiting T-cell–mediated responses or indirectly regulating the expression of IL-12.41, 42, 43, 44 We observed a sustained increase in T cells in TNFαKO + HKBA mice in both the BM and spleen, particularly in CD4+ and CD8+ memory and effector cells.

Macrophages were also elevated by absolute counts in the spleen. It is plausible that high IL12p40 levels in TNFαKO + HKBA mice result from increased proinflammatory macrophages (supplemental Figures 13 and 14). Nonproinflammatory macrophages were also elevated but were ineffective in counterbalancing IL12p40 in the absence of TNFα and/or high IFNγ levels. In vitro assessment of CD4+ and CD8+ T-cell subtypes from BM of WT and TNFαKO mice using a generic activator that bypasses the TCR complex, showed that CD8+ T cells lacking Tnfα produced more IFNγ than WT controls (supplemental Figure 12). This finding suggests that Tnfα may play an intrinsic role in modulating IFNγ synthesis in T cells. To further support an anti-inflammatory role by TNFα, levels of IFNγ and IL12p40 were continuously elevated in TNFαKO + HKBA until concurrent treatment with anti-IFNγ Ab and rTNFα was able to reverse the lethal anemia phenotype and restore CBC profiles to baseline. Based on these observations, lack of Tnfα was identified as disrupting a safeguard mechanism that modulated IL12p40-IFNγ action in these mice.

TNFα has been demonstrated to provide essential prosurvival signals to HSCs during the resolution of inflammation.30 Increased IFNγ levels during chronic inflammatory stress negatively affect HSC homeostasis by skewing HSC toward differentiation, impeding self-renewal, and promoting HSC exhaustion.45 In our study, we show TNFα is essential resolving AI. The absence of Tnfα resulted in elevated IFNγ, expansion of the HSC pool, and an inability to activate TNFα-mediated anti-inflammatory protective mechanisms, resulting in BM failure.

Currently, treatments that block TNFα are central to managing various autoimmune diseases. However, some patients do not respond to TNFα blocking treatment, and others show exacerbation of a preexisting autoimmune disease or the onset of new autoimmune conditions following anti-TNFα treatment.46 Considering our findings, it is difficult not to question whether the exacerbation of autoimmune diseases or the onset of new ones in patients being treated with anti-TNFα therapies may result from reduced TNFα levels during a new inflammatory trigger. This work highlights the need for deeper investigations into the anti-inflammatory roles of TNFα and associated pathways to develop alternative anti-TNFα therapies and/or combination approaches to treat autoimmune disorders. Further nuanced analysis of the proinflammatory and anti-inflammatory actions of cytokines during AI would aid both in the translational development of agents to mechanistically target AI in numerous conditions, as well as provide new options for the management of patients who are not candidates for anti-TNFα therapy.

Taken together, our work supports an essential role for TNFα in resolving the inflammatory response, characterizes a new model to investigate the role of TNFα in the etiology of AI, and provides potential clues for exacerbations and, yet unexplained adverse effects in some patients treated with TNFα inhibitors.

Conflict-of-interest disclosure: S.R. is a member of the scientific advisory board for Ionis Pharmaceuticals and Meira GTx; has acted as a consultant for Disc Medicine, Protagonist, and La Jolla Pharmaceutical Company; and is a coinventor for patents US8058061 B2 C12N 20111115 and US7541179 B2 C12N 20090602. The consulting work and intellectual property of S.R. did not affect this research's design, conduct, or reporting. The remaining authors declare no competing financial interests.

Acknowledgments

This work was supported by National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health (NIH) HLBI T32 Hematology Research Training Program grant T32 HL07439; National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), NIH, Post-Doctoral Diversity Supplement grant NIH-R01DK090554-09S1; Bridge to Faculty (CHOP K99-like award) grant 726070063-16; NIDDK, NIH, Early Career Investigator Diversity Supplement grant NIH-3R01DK133475-02S1 (A.G.); National Institute of Allergy and Infectious Diseases (NIAID), NIH Training grant 5T32AI074551 (L.F.B.); NHLBI, NIH grant HL146528; NIDDK, NIH grant DK138865; United States Department of Agriculture, National Institute of Food and Agriculture Hatch Project PEN04960, accession number 7006577 (R.F.P.); NIDDK, NIH grants DK107670 and DK095112 (Y.Z.G.); NIAID, NIH grant R01AI121250 (E.M.B.); and the Iron and Red Blood Cells Research Fund grant 5260550000.

Authorship

Contribution: S.R., A.G., V.L.P., P.K., R.F.P., R. Gozzelino, and A.C.M. designed research and completed data interpretation; V.L.P., C.C., C.C.C., A.R., N.H., P.S., R. Gupta, and L.F.B. assisted with data collection and analysis; D.-W.C. assisted with bone marrow transplantation, experimental design, and assessment of bone marrow failure; E.M.B. provided the interferon gamma mouse model; Y.Z.G. assisted with manuscript revisions; and A.G. wrote the manuscript.

Footnotes

∗

A.G. and V.L.P. are joint first authors.

The data that support the findings of this study are available from the corresponding author, Amaliris Guerra (amalirisguerra@gmail.com), on reasonable request.

The online version of this article contains a data supplement.

There is a Blood Commentary on this article in this issue.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Supplementary Material

Supplemental Figures
BLOOD_BLD-2025-029352-mmc1.pdf (2.5MB, pdf)
Supplemental Tables
BLOOD_BLD-2025-029352-mmc2.pdf (49.5KB, pdf)

References

  • 1.Ganz T. Anemia of inflammation. N Engl J Med. 2019;381(12):1148–1157. doi: 10.1056/NEJMra1804281. [DOI] [PubMed] [Google Scholar]
  • 2.Weiss G, Ganz T, Goodnough LT. Anemia of inflammation. Blood. 2019;133(1):40–50. doi: 10.1182/blood-2018-06-856500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Crooks CJ, West J, Morling JR, et al. Anaemia of acute inflammation: a higher acute systemic inflammatory response is associated with a larger decrease in blood haemoglobin levels in patients with COVID-19 infection. Clin Med. 2023;23(3):201–205. doi: 10.7861/clinmed.2022-0436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kalliolias GD, Ivashkiv LB. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat Rev Rheumatol. 2016;12(1):49–62. doi: 10.1038/nrrheum.2015.169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Papadaki HA, Kritikos HD, Valatas V, Boumpas DT, Eliopoulos GD. Anemia of chronic disease in rheumatoid arthritis is associated with increased apoptosis of bone marrow erythroid cells: improvement following anti-tumor necrosis factor-alpha antibody therapy. Blood. 2002;100(2):474–482. doi: 10.1182/blood-2002-01-0136. [DOI] [PubMed] [Google Scholar]
  • 6.Holbrook J, Lara-Reyna S, Jarosz-Griffiths H, McDermott MF. Tumour necrosis factor signalling in health and disease. F1000Res. 2019;8 doi: 10.12688/f1000research.17023.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Roodman GD, Johnson RA, Clibon U. Tumor necrosis factor alpha and the anemia of chronic disease: effects of chronic exposure to TNF on erythropoiesis in vivo. Adv Exp Med Biol. 1989;271:185–196. doi: 10.1007/978-1-4613-0623-8_19. [DOI] [PubMed] [Google Scholar]
  • 8.Alvarez-Hernández X, Licéaga J, McKay IC, Brock JH. Induction of hypoferremia and modulation of macrophage iron metabolism by tumor necrosis factor. Lab Invest. 1989;61(3):319–322. [PubMed] [Google Scholar]
  • 9.Hodge-Dufour J, Marino MW, Horton MR, et al. Inhibition of interferon gamma induced interleukin 12 production: a potential mechanism for the anti-inflammatory activities of tumor necrosis factor. Proc Natl Acad Sci U S A. 1998;95(23):13806–13811. doi: 10.1073/pnas.95.23.13806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.De Benedetti F, Prencipe G, Bracaglia C, Marasco E, Grom AA. Targeting interferon-γ in hyperinflammation: opportunities and challenges. Nat Rev Rheumatol. 2021;17(11):678–691. doi: 10.1038/s41584-021-00694-z. [DOI] [PubMed] [Google Scholar]
  • 11.Kak G, Raza M, Tiwari BK. Interferon-gamma (IFN-γ): exploring its implications in infectious diseases. Biomol Concepts. 2018;9(1):64–79. doi: 10.1515/bmc-2018-0007. [DOI] [PubMed] [Google Scholar]
  • 12.Lee SK, Silva DG, Martin JL, et al. Interferon-γ excess leads to pathogenic accumulation of follicular helper T cells and germinal centers. Immunity. 2012;37(5):880–892. doi: 10.1016/j.immuni.2012.10.010. [DOI] [PubMed] [Google Scholar]
  • 13.Vadhan-Raj S, Al-Katib A, Bhalla R, et al. Phase I trial of recombinant interferon gamma in cancer patients. J Clin Oncol. 1986;4(2):137–146. doi: 10.1200/JCO.1986.4.2.137. [DOI] [PubMed] [Google Scholar]
  • 14.de Bruin AM, Voermans C, Nolte MA. Impact of interferon-g on hematopoiesis. Blood. 2014;124(16):2479–2486. doi: 10.1182/blood-2014-04-568451. [DOI] [PubMed] [Google Scholar]
  • 15.Raefsky EL, Platanias LC, Zoumbos NC, Young NS. Studies of interferon as a regulator of hematopoietic cell proliferation. J Immunol. 1985;135(4):2507–2512. [PubMed] [Google Scholar]
  • 16.Mullarky IK, Szaba FM, Kummer LW, et al. Gamma interferon suppresses erythropoiesis via interleukin-15. Infect Immun. 2007;75(5):2630–2633. doi: 10.1128/IAI.01836-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Libregts SF, Gutiérrez L, de Bruin AM, et al. Chronic IFN-γ production in mice induces anemia by reducing erythrocyte life span and inhibiting erythropoiesis through an IRF-1/PU.1 axis. Blood. 2011;118(9):2578–2588. doi: 10.1182/blood-2010-10-315218. [DOI] [PubMed] [Google Scholar]
  • 18.Johnson CS, Pourbohloul SC, Furmanski P. Negative regulators of in vivo erythropoiesis: interaction of IL-1 alpha and TNF-alpha and the lack of a strict requirement for T or NK cells for their activity. Exp Hematol. 1991;19(2):101–105. [PubMed] [Google Scholar]
  • 19.Rusten LS, Jacobsen SE. Tumor necrosis factor (TNF)-alpha directly inhibits human erythropoiesis in vitro: role of p55 and p75 TNF receptors. Blood. 1995;85(4):989–996. [PubMed] [Google Scholar]
  • 20.Johnson CS, Cook CA, Furmanski P. In vivo suppression of erythropoiesis by tumor necrosis factor-alpha (TNF-alpha): reversal with exogenous erythropoietin (EPO) Exp Hematol. 1990;18(2):109–113. [PubMed] [Google Scholar]
  • 21.Morceau F, Dicato M, Diederich M. Pro-inflammatory cytokine-mediated anemia: regarding molecular mechanisms of erythropoiesis. Mediators Inflamm. 2009;2009 doi: 10.1155/2009/405016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Felli N, Pedini F, Zeuner A, et al. Multiple members of the TNF superfamily contribute to IFN-gamma-mediated inhibition of erythropoiesis. J Immunol. 2005;175(3):1464–1472. doi: 10.4049/jimmunol.175.3.1464. [DOI] [PubMed] [Google Scholar]
  • 23.Gardenghi S, Renaud TM, Meloni A, et al. Distinct roles for hepcidin and interleukin-6 in the recovery from anemia in mice injected with heat-killed Brucella abortus. Blood. 2014;123(8):1137–1145. doi: 10.1182/blood-2013-08-521625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Guerra A, Oikonomidou PR, Sinha S, et al. Lack of Gdf11 does not improve anemia or prevent the activity of RAP-536 in a mouse model of β-thalassemia. Blood. 2019;134(6):568–572. doi: 10.1182/blood.2019001057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Parrow NL, Li Y, Feola M, et al. Lobe specificity of iron binding to transferrin modulates murine erythropoiesis and iron homeostasis. Blood. 2019;134(17):1373–1384. doi: 10.1182/blood.2018893099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Dholakia U, Bandyopadhyay S, Hod EA, Prestia KA. Determination of RBC survival in C57BL/6 and C57BL/6-Tg(UBC–GFP) mice. Comp Med. 2015;65(3):196–201. [PMC free article] [PubMed] [Google Scholar]
  • 27.Chen K, Liu J, Heck S, Chasis JA, An X, Mohandas N. Resolving the distinct stages in erythroid differentiation based on dynamic changes in membrane protein expression during erythropoiesis. Proc Natl Acad Sci U S A. 2009;106(41):17413–17418. doi: 10.1073/pnas.0909296106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Purton LE. Adult murine hematopoietic stem cells and progenitors: an update on their identities, functions, and assays. Exp Hematol. 2022;116:1–14. doi: 10.1016/j.exphem.2022.10.005. [DOI] [PubMed] [Google Scholar]
  • 29.Chappell ME, Breda L, Tricoli L, et al. Use of HSC-targeted LNP to generate a mouse model of lethal α-thalassemia and treatment via lentiviral gene therapy. Blood. 2024;144(15):1633–1645. doi: 10.1182/blood.2023023349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Yamashita M, Passegué E. TNF-α coordinates hematopoietic stem cell survival and myeloid regeneration. Cell Stem Cell. 2019;25(3):357–372.e7. doi: 10.1016/j.stem.2019.05.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Ruan B, Paulson RF. Metabolic regulation of stress erythropoiesis, outstanding questions, and possible paradigms. Front Physiol. 2022;13 doi: 10.3389/fphys.2022.1063294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Paulson RF, Hariharan S, Little JA. Stress erythropoiesis: definitions and models for its study. Exp Hematol. 2020;89:43–54.e2. doi: 10.1016/j.exphem.2020.07.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Bennett LF, Liao C, Quickel MD, et al. Inflammation induces stress erythropoiesis through heme-dependent activation of SPI-C. Sci Signal. 2019;12(598) doi: 10.1126/scisignal.aap7336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Salomon BL, Leclerc M, Tosello J, Ronin E, Piaggio E, Cohen JL. Tumor necrosis factor α and regulatory T cells in oncoimmunology. Front Immunol. 2018;9 doi: 10.3389/fimmu.2018.00444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zheng J, Umikawa M, Zhang S, et al. Ex vivo expanded hematopoietic stem cells overcome the MHC barrier in allogeneic transplantation. Cell Stem Cell. 2011;9(2):119–130. doi: 10.1016/j.stem.2011.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Ma X. TNF-α and IL-12:a balancing act in macrophage functioning. Microbes Infect. 2001;3(2):121–129. doi: 10.1016/s1286-4579(00)01359-9. [DOI] [PubMed] [Google Scholar]
  • 37.Masli S, Turpie B. Anti-inflammatory effects of tumour necrosis factor (TNF)-α are mediated via TNF-R2 (p75) in tolerogenic transforming growth factor-β-treated antigen-presenting cells. Immunology. 2009;127(1):62–72. doi: 10.1111/j.1365-2567.2008.02933.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Alam Z, Devalaraja S, Li M, et al. Counter regulation of spic by NF-κB and STAT signaling controls inflammation and iron metabolism in macrophages. Cell Rep. 2020;31(13) doi: 10.1016/j.celrep.2020.107825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Guerra A, Parhiz H, Rivella S. Novel potential therapeutics to modify iron metabolism and red cell synthesis in diseases associated with defective erythropoiesis. Haematologica. 2023;108(10):2582–2593. doi: 10.3324/haematol.2023.283057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A, Ganz T. Hepcidin, a putative mediator of anemia of inflammation, is a type II acute-phase protein. Blood. 2003;101(7):2461–2463. doi: 10.1182/blood-2002-10-3235. [DOI] [PubMed] [Google Scholar]
  • 41.Zakharova M, Ziegler HK. Paradoxical anti-inflammatory actions of TNF-α: inhibition of IL-12 and IL-23 via TNF receptor 1 in macrophages and dendritic cells. J Immunol. 2005;175(8):5024–5033. doi: 10.4049/jimmunol.175.8.5024. [DOI] [PubMed] [Google Scholar]
  • 42.Speiser DE, Sebzda E, Ohteki T, et al. Tumor necrosis factor receptor p55 mediates deletion of peripheral cytotoxic T lymphocytes in vivo. Eur J Immunol. 1996;26(12):3055–3060. doi: 10.1002/eji.1830261235. [DOI] [PubMed] [Google Scholar]
  • 43.Kanaly ST, Nashleanas M, Hondowicz B, Scott P. TNF receptor p55 is required for elimination of inflammatory cells following control of intracellular pathogens. J Immunol. 1999;163(7):3883–3889. [PubMed] [Google Scholar]
  • 44.Siegmund D, Wajant H. TNF and TNF receptors as therapeutic targets for rheumatic diseases and beyond. Nat Rev Rheumatol. 2023;19(9):576–591. doi: 10.1038/s41584-023-01002-7. [DOI] [PubMed] [Google Scholar]
  • 45.Morales-Mantilla DE, King KY. The role of interferon-gamma in hematopoietic stem cell development, homeostasis, and disease. Curr Stem Cell Rep. 2018;4(3):264–271. doi: 10.1007/s40778-018-0139-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Salomon BL. Insights into the biology and therapeutic implications of TNF and regulatory T cells. Nat Rev Rheumatol. 2021;17(8):487–504. doi: 10.1038/s41584-021-00639-6. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figures
BLOOD_BLD-2025-029352-mmc1.pdf (2.5MB, pdf)
Supplemental Tables
BLOOD_BLD-2025-029352-mmc2.pdf (49.5KB, pdf)

Articles from Blood are provided here courtesy of The American Society of Hematology

RESOURCES