Myelopoiesis is temporally dynamic and is regulated by lifestyle to modify multiple sclerosis

Abstract

Monocytes and neutrophils from the myeloid lineage contribute to multiple sclerosis (MS), but the dynamics of myelopoiesis during MS are unclear. Here we uncover a disease stage-specific relationship between lifestyle, myelopoiesis and neuroinflammation. In mice with relapsing-remitting experimental autoimmune encephalomyelitis (RR-EAE), myelopoiesis in the femur, vertebrae and spleen is elevated prior to disease onset and during remission, preceding the peaks of clinical disability and neuroinflammation. In progressive EAE (P-EAE), vertebral myelopoiesis rises steadily throughout disease, while femur and splenic myelopoiesis is elevated early before waning later during disease height. In parallel, sleep disruption or hyperlipidemia and cardiometabolic syndrome augment M-CSF generation and multi-organ myelopoiesis to worsen P-EAE clinical symptoms, neuroinflammation, and spinal cord demyelination, with M-CSF blockade abrogating these symptoms. Lastly, results from a previous trial show that Mediterranean diet restrains myelopoietic activity and myeloid lineage progenitor skewing and improves clinical symptomology of MS. Together, our data suggest that myelopoiesis in MS is dynamic and dependent on disease stage and location, and that lifestyle factors modulate disease by influencing M-CSF-mediated myelopoiesis.

Fig. 1. Myelopoiesis and blood monocytosis precede and predict clinical disability and neuroinflammation in relapsing-remitting EAE.

a Schematic design of experiment. Relapsing-remitting EAE was induced in female SJL/L mice, 8 weeks old, by administering proteolipid protein (PLP_139-151) mixed with complete Freund’s adjuvant (CFA) by sub cutaneous injection on day 0. Naïve SJL/L mice were used as controls. Mean clinical disease scores over the course of 35 days shows multiphasic episodes of disability with an initial peak, followed by remission, and subsequent relapse. Tissue was collected during peak disease and remission for analysis. (n = 7) (b) Quantification of CD45⁺ leukocyte populations in the spinal cord (n = 6 naïve, n = 8 peak, n = 9 remission). c Quantification of Ly6C^hi monocytes and neutrophils in the blood (n = 6 naïve, n = 7 peak, n = 8 remission). d Flow cytometry analysis of bone marrow hematopoietic progenitors. Quantification of CD45⁺ cells, LSKs, and MPP3 leukocyte progenitors in the femur bone marrow (n = 3 naïve, n = 5 peak, n = 5 remission, except MPP3s and LSKs where n = 6 naive) (e) Measurement of growth factors in femur bone marrow fluid by ELISA (n  = 6 naïve, n = 9 peak, n = 10 remission) (f) Quantification of LSKs and MPP3 leukocyte progenitors in vertebral bone marrow (n = 5/time point). g Measurement of splenic weight (n = 6 naïve, n = 9 peak, n = 10 remission) and enumeration of LSKs and MPP3 leukocyte progenitors in the spleen (n = 3 naïve, n = 5 peak, n = 5 remission). h Mean clinical disease scores of RR-EAE mice over the course of 35 days with corresponding blood Ly6C^hi monocytes and neutrophils quantification at days 0, 8, 14, 21, and 35 dpi (n = 7). i Correlation of the number of circulating Ly6C^hi monocytes and neutrophils at pre-onset (day 8) with the subsequent peak clinical score (n = 7). Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. a–h are a combination of 2, 3 individual experiments, i is from a representative experiment. b–g one-way ANOVA, I Pearson’s correlation. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. EAE experimental autoimmune encephalomyelitis, PLP proteolipid protein, CFA complete Freud’s adjuvant, LSK Lin^-Sca1⁺cKit^+, MPP3 myeloid-biased multipotent progenitor 3, GM-CSF granulocyte-macrophage colony-stimulating factor, IL-3 interleukin-3, M-CSF macrophage colony stimulating factor, dpi days post immunization.

Fig. 2. Myelopoiesis is temporally dynamic in medullary and extramedullary organs in progressive EAE.

a Schematic design of experiment. Progressive EAE was induced in female C57BL/6J mice, 9 weeks old, by administering myelin oligodendrocyte glycoprotein (MOG_35-55) mixed with complete Freud’s adjuvant (CFA) by subcutaneous injection of day 0 and intraperitoneal injection of PTX on days 0 and 1. Naïve C57BL/6 mice were used as controls. Mean clinical disease scores over the course of 21 days shows monophasic disease. Tissue was collected during the progressive and chronic phases of the model. (n = 16) (b) Quantification of CD45⁺ leukocyte populations in the spinal cord (n = 5 Naïve, n = 12 progressive, n = 14 chronic). c Quantification of Ly6C^hi monocytes and neutrophils in the blood (n = 13 Naïve, n = 14 progressive, n = 10 chronic). d Flow cytometry analysis of bone marrow hematopoietic progenitors. Quantification of LSKs and MPP3 leukocyte progenitors in the femur bone marrow by ELISA (n = 15 Naïve, n = 14 progressive, n = 10 chronic). e Measurement of M-CSF in femur bone marrow (n = 14 Naïve, n = 8 progressive, n = 5 chronic). f Quantification of LSKs and MPP3 leukocyte progenitors in vertebral bone marrow (n = 16 Naïve, n = 14 progressive, n = 9 chronic). g Measurement of splenic weight (n = 11 Naïve, n = 14 progressive, n = 11 chronic) and enumeration of LSKs and MPP3 leukocyte progenitors in the spleen (n = 6 Naïve, n = 8 progressive, n = 6 chronic). One-way ANOVAs. h Circulating blood leukocytes in non-MS population controls (n = 473,905) and MS patients (n = 1918) from the UK Biobank. Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are a combination of 2, 3 individual experiments. b-g one-way ANOVA, h two sided Mann-Whitney U test. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. EAE experimental autoimmune encephalomyelitis, MOG myelin oligodendrocyte glycoprotein, CFA complete Freud’s adjuvant, PTX pertussis toxin, LSK Lin^-Sca1⁺cKit⁺, MPP3 myeloid-biased multipotent progenitor 3, M-CSF macrophage colony stimulating factor, ELISA enzyme-linked immunosorbent assay, MS multiple sclerosis.

Fig. 3. Sleep disruption elevates myelopoiesis and worsens clinical severity, neuroinflammation, and demyelination in P-EAE.

a Enumeration of circulating total CD45⁺ leukocytes and mature myeloid cells in WT and Hcrt^−/− mice (n = 7 WT, n = 13 Hcrt^−/−). b Flow cytometry analysis of bone marrow hematopoietic progenitors. Quantification of LSKs and MPP3 hematopoietic progenitors in the femur bone marrow (n = 4 WT, n = 9 Hcrt^−/−). c Quantification of LSKs and MPP3s in vertebral bone marrow (n = 5 WT, n = 9 Hcrt^−/−). d Enumeration of LSKs and MPP3 leukocyte progenitors in the spleen (n = 5 WT, n = 10 Hcrt^−/−). e Progressive EAE (MOG_35-55/CFA) was induced in WT and Hcrt^−/− mice. Mean clinical disease scores of diseased mice over the course of 21 days (n = 13/group). f Quantification of CD45⁺ cell subtypes in the spinal cord of diseased mice at peak EAE (n = 8 WT, n = 5 Hcrt^−/−). g Representative images of fluoromyelin-stained, diseased spinal cords and quantification of demyelination of white matter at peak EAE (n = 6/group). h Quantification of LSKs and MPP3s in the bone marrow, 21 days dpi (n = 5 WT, n = 6 Hcrt^−/−). i Measurement of M-CSF in femur bone marrow, 8 days dpi, by ELISA (n = 4/group). j Quantification of LSK and MPP3 hematopoietic progenitors in the vertebral bone marrow, 21 days dpi (n = 5/group). k Quantification of LSKs and MPP3s in the spleen, 21 days dpi (n = 5/group). L Circulating blood leukocytes in individuals with MS who reported sleeping 2–5 h (n = 70) or 8+ h (n = 438) per night, from the UK Biobank. Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. a–e are a combination of 2, 3 individual experiments, f–k are from a representative experiment. Two sided Mann Whitney U tests, except (e) two-way ANOVA. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. EAE experimental autoimmune encephalomyelitis, WT wild-type, Hcrt^−/− hypocretin knockout, LSK Lin^-Sca1⁺cKit⁺, MPP3 myeloid-biased multipotent progenitor 3, M-CSF macrophage colony stimulating factor, ELISA enzyme-linked immunosorbent assay, dpi days post-immunization, MS multiple sclerosis.

Fig. 4. Hyperlipidemia and cardiometabolic syndrome exacerbates myelopoiesis, increases immune cell infiltration of the spinal cord, and worsens clinical scoring in P-EAE.

a Measurement of total and free cholesterol in the plasma of WT, WT + HFD, Apoe^−/−, and Apoe^−/−+HFD mice (n = 8 WT, n = 10 WT + HFD, n = 5 Apoe^−/−, n = 7 Apoe^−/−+HFD). b Quantification of circulating Ly6C^hi monocytes and neutrophils (n = 20 WT, n = 10 WT + HFD, n = 8 Apoe^−/−, n = 16 Apoe^−/−+HFD). c Quantification of LSK and MPP3 hematopoietic progenitors in the femur bone marrow of WT and Apoe^−/−+HFD mice (n = 7 WT, n = 8 Apoe^−/−+HFD). d Measurement of M-CSF in femur bone marrow fluid by ELISA (n = 14 WT, n = 8 Apoe^−/−+HFD, student’s unpaired t-test). e Quantification of LSKs and MPP3s in the vertebral bone marrow (n = 8/group). f Quantification of LSKs and MPP3s in the spleen (n = 4 WT, n = 5 Apoe^−/−+HFD). g Progressive EAE (MOG_35-55/CFA) was induced in WT, WT + HFD, Apoe^−/−, and Apoe^−/−+HFD. Mean clinical disease scores of diseased mice over the course of 21 days (n = 15 WT, n = 4 WT + HFD, n = 7 Apoe^−/−, n = 15 Apoe^−/−+HFD). h Representative images of fluoromyelin-stained diseased spinal cords and quantification of demyelination of white matter in WT and Apoe^−/−+HFD mice at peak EAE (n = 6 WT, n = 5 Apoe^−/−+HFD). i Quantification of CD45⁺ cells in the spinal cord at peak EAE (n = 8 WT, n = 5 Apoe^−/−+HFD). j Quantification of LSKs and MPP3 hematopoietic progenitors in the femur bone marrow, 21 days dpi (n = 6 WT, n = 7 Apoe^−/−+HFD). k Measurement of growth factors in femur bone marrow fluid, 21 days dpi, by ELISA (n = 4 WT, n = 6 Apoe^−/−+HFD). l Correlation of blood monocytes and neutrophils, prior to EAE induction, with subsequent clinical disease score (n = 26). m Circulating blood leukocytes in individuals with MS and a BMI of 18–25 (n = 329), 25–30 (n = 329), or 30+ (n = 222), from the UK Biobank. Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. a-g are a combination of 2–3 experiments, h–k are from a representative experiment. a, b one-way ANOVA, c–f, h–k, m two sided Mann Whitney U tests, g two sided ANOVA, l Pearson’s Correlation. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. EAE experimental autoimmune encephalomyelitis WT wild-type, HFD high-fat diet, Apoe^−/− Apolipoprotein E knockout, LSK Lin^-Sca1⁺cKit⁺, MPP3 myeloid-biased multipotent progenitor 3, GM-CSF granulocyte-macrophage colony stimulating factor, IL-3 interleukin-3, M-CSF macrophage colony stimulating factor, ELISA enzyme-linked immunosorbent assay, AUC area under the curve, dpi days post-immunization, MS multiple sclerosis, BMI body mass index.

Fig. 5. M-CSF blockade blunts sleep disruption- and cardiometabolic syndrome-induced myelopoiesis and disease in P-EAE mice.

a Experimental schematic of M-CSF blockade with an ɑ-M-CSF antibody in Hcrt^−/− or Apoe^−/− HFD mice beginning at induction of P-EAE. Mice were injected every other day until symptom onset then daily thereafter. b Ly6C^hi monocyte in the blood at pre-onset (day 8) in Hcrt^−/− mice with IgG or ɑ-M-CSF (n = 3 IgG, n = 7 ɑ-M-CSF) (c) Myelopoiesis parameters at day 21 in Hcrt^−/− mice with IgG or ɑ-M-CSF (n = 9 IgG, n = 7 ɑ-M-CSF). d Clinical scoring in Hcrt^−/− mice with IgG or ɑ-M-CSF (n = 10 IgG, n = 7 ɑ-M-CSF). e Ly6C^hi monocyte in the blood at pre-onset (day 8) in Apoe^−/− HFD mice with IgG or ɑ-M-CSF (n = 3 IgG, n = 6 ɑ-M-CSF). f Myelopoiesis parameters at day 21 in Apoe^−/− HFD mice with IgG or ɑ-M-CSF (n = 7 IgG, n = 8 ɑ-M-CSF). g Clinical scoring in Apoe^−/− HFD mice with IgG or ɑ-M-CSF (n = 8 IgG, n = 7 ɑ-M-CSF). Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are from a representative experiment. b, e two sided Mann Whitney U tests, d, g two-way ANOVA. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. EAE experimental autoimmune encephalomyelitis, HFD high-fat diet, Apoe^−/− Apolipoprotein E knockout, LSK Lin^-Sca1⁺cKit⁺, MPP3 myeloid-biased multipotent progenitor.

Fig. 6. Mediterranean diet restricts hematopoietic activity in individuals with MS.

a Schematic of human diet intervention study. b Change in NFI-MS, MSIS, and EDSS by mediterranean diet (n = 13 habitual diet, n = 16 mediterranean diet). c Enumeration of circulating PBMCs in participants after 6 months on their habitual diet or the Mediterranean diet (n = 14 habitual diet, n = 17 mediterranean diet). d Flow cytometry analysis of lineage^-CD34⁺ HSPCs. Quantification of HSPCs in MS patients after 6 months on their habitual diet or the Mediterranean diet (n = 13 habitual diet, n = 14 mediterranean diet). e Quantification of circulating CD115⁺ HSPCs (n = 14 habitual diet, n = 14 mediterranean diet). f Quantification of circulating CD123⁺ HSPCs n = 14 habitual diet, n = 15 mediterranean diet). Data presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two sided Mann Whitney U tests. Created in BioRender. McAlpine, C. (2025) https://BioRender.com/kvggmpp. MS multiple sclerosis, LDL low density lipoprotein, BMI body mass index, NFI-MS neurological fatigue index-multiple sclerosis, MSIS multiple sclerosis impact scale, EDSS expanded disability status scale, PBMCs peripheral blood mononuclear cells, HSPCs hematopoietic stem and progenitor cells.