Schistosome infection impacts hematopoiesis

Abstract

Helminth infections are common in animals. However, the impact of a helminth infection on the function of hematopoietic stem cells (HSCs) and other hematopoietic cells has not been comprehensively defined. Here we describe the hematopoietic response to infection of mice with Schistosoma mansoni, a parasitic flatworm which causes schistosomiasis. We analyzed the frequency or number of hematopoietic cell types in the bone marrow, spleen, liver, thymus, and blood, and observed multiple hematopoietic changes caused by infection. Schistosome infection impaired bone marrow HSC function after serial transplantation. Functional HSCs were present in the infected liver. Infection blocked bone marrow erythropoiesis and augmented spleen erythropoiesis, observations consistent with the anemia and splenomegaly prevalent in schistosomiasis patients. This work defines the hematopoietic response to schistosomiasis, a debilitating disease afflicting more than 200 million people, and identifies impairments in HSC function and erythropoiesis.

Introduction

Schistosomiasis is a parasitic disease caused by infection with Schistosoma spp. flatworms. It afflicts more than 200 million people in Africa, the Middle East, South-East Asia, and South America (1, 2). Vaccines are not available. Treatment can clear adult parasites but is ineffective against immature parasites, does not prevent reinfection, nor does it reverse multi-organ immunopathology (3). As a result, the chronic symptoms of schistosomiasis contribute substantially to the global disability burden, creating a cycle of poverty and infection (4–7). Schistosomes are shed as larvae from freshwater snails, infect humans by penetrating the skin, and can live in the circulation for decades (8), continuously laying eggs which lodge in liver, bladder, and other organs. Egg antigens trigger a Th2 response which dominates the chronic phase of the disease and is the main cause of pathology (1, 3, 9, 10). Several schistosomiasis symptoms including anemia, splenomegaly, and inflammation suggest hematopoietic involvement. However, the effect of schistosomiasis on HSCs and restricted progenitors is mostly uncharacterized.

Until recently most humans were likely to be parasitized (11). The prevalence of infection with schistosomes or other helminths was as high as 50% in some populations before the modern era (12, 13), or in modern untreated populations living in endemic areas (14). Primates and other animals are parasitized in the wild (15) and in some areas as many as 90% of primates have a history of schistosomiasis (16). Thus, HSCs and the hematopoietic system have likely evolved under near-constant pressure from schistosomes and other helminths. Recent studies have examined the impact on HSC function of various infections (17), including Mycobacterium avium, (18, 19), E. coli (20), Streptococcus (21), Plasmodium (22), Ehrlichia muris (23), Leishmania (24), C. albicans (25), Salmonella (26, 27) and several viruses (28–30). Most of these infections induce a strong proinflammatory Th1 response. The impact of Th2 response-dominated helminth infections on hematopoiesis has been much less characterized (31–33) and the effects of schistosomes or other helminths on HSCs and hematopoiesis have not been comprehensively analyzed. To address this, we investigated the impact of S. mansoni infection on HSCs and hematopoiesis in mice.

Materials and Methods

Mice.

Mice were on a C57BL/Ka background. Both male and female mice were used in all studies. Young adult mice were infected at the ages of 10–21 weeks and were either analyzed or used as donors for transplantation at 7 weeks after schistosome infection. C57BL/Ka-Thy-1.1 (CD45.2) and C57BL/Ka-Thy-1.2 (CD45.1) mice were used for transplantation experiments. Mice were housed in the Animal Resource Center of UT Southwestern and all procedures were approved by the UT Southwestern Institutional Animal Care and Use committee.

Schistosome infection.

Each mouse was infected with around 200 Schistosoma mansoni (NMRI strain) cercariae released from infected Biomphalaria glabrata snails (Schistosome Resource Center) by percutaneous tail exposure (34). Mice were restrained, and their tail was dipped in water containing cercariae for 30 minutes.

Cell isolation and hematopoietic analysis.

Bone marrow cells were obtained by flushing femurs and tibia with 25G needle, or crushing femurs, tibias, vertebrae, and pelvic bones with a mortar and pestle in staining medium consisting of Ca²⁺/Mg²⁺-free Hank’s balanced salt solution (HBSS; Gibco), supplemented with 2% heat-inactivated bovine serum (Gibco). Spleens and thymuses were mechanically dissociated by trituration in staining medium. Livers were enzymatically digested for 30 minutes at 37°C in 1.5 ml RPMI-1640 (Sigma), containing 250 μg/ml liberase (Roche) and 100 μg/ml DNase I (Roche). Cell suspensions were filtered through a 40 μm strainer. Cell number was assessed with a Vi-CELL cell viability analyzer (Beckman Coulter). Blood was collected by cardiac puncture using a 25G needle and mixed in a tube containing 5 μl 0.5M EDTA. Complete blood cell counts were determined using a hemavet HV950 (Drew Scientific). For hematopoietic analysis, 40 μl blood was lysed in 1 ml of ammonium chloride buffer (155mM NH₄Cl; 10 mM KHCO₃; 0.1 mM EDTA) for 10 minutes at 4°C. Cells were incubated with fluorescently conjugated antibodies for 90 minutes on ice when using CD34 antibody or for 30 minutes at 4°C. Cells were washed with staining media and resuspended in staining media containing 1 μg/ml DAPI or 1 μg/ml propidium iodide for live/dead discrimination. Cell populations were defined with the following markers: CD150⁺CD48⁻Lineage⁻Sca-1⁺Kit⁺ hematopoietic stem cells (HSCs), CD150⁻CD48⁻Lineage⁻Sca-1⁺Kit⁺ multipotent progenitor cells (MPPs) (35), CD150⁻CD48⁺Lineage⁻Sca-1⁺Kit⁺ hematopoietic progenitor cells (HPC-1), CD150⁺CD48⁺Lineage⁻Sca-1⁺Kit⁺ hematopoietic progenitor cells (HPC-2) (36), CD34⁺CD16/32⁻Lineage⁻Sca-1⁻Kit⁺ common myeloid progenitors (CMPs), CD34⁺CD16/32⁺Lineage⁻Sca-1⁻Kit⁺ granulocyte–monocyte progenitors (GMPs), CD34⁻CD16/32⁻Lineage⁻Sca-1⁻Kit⁺ megakaryocyte–erythroid progenitors (MEPs) (37), Mac-1⁺ myeloid cells, Mac1⁺CD115⁺Ly6C⁺Ly6G⁻ inflammatory monocytes, Mac1⁺CD115⁻Ly6C^mid/highSiglecF⁺ eosinophils, Mac1⁺CD115⁻Ly6C^midLy6G⁺ neutrophils, CD11c⁺ dendritic cells (DCs), CD11c⁺Mac1⁺Ly6C⁺ monocytic DCs (moDCs), CD11c⁺Mac1⁻ Mac1⁻DCs, Mac1⁻B220⁺ B cells, Mac1⁻CD3⁺ T cells, Mac1⁻B220⁻CD3⁻CD71^midTer119⁻ immature erythroid progenitors, Mac1⁻B220⁻CD3⁻CD71⁺Ter119⁻ erythroid progenitors, and Mac1⁻B220⁻CD3⁻CD71⁺Ter119⁺ erythroid progenitors. The Lineage cocktail for HSCs and progenitors consisted of CD2, CD3, CD5, CD8, Ter119, B220, and Gr-1 antibodies. T cell progenitor populations in the thymus were defined with the following markers, after excluding Mac-1⁺, B220⁺, and Ter119⁺ cells: CD4⁺CD8⁺ double-positive (DP), CD3⁺CD4⁺CD8⁻ (CD4⁺ single positive, CD4⁺SP), CD3⁺CD4⁻CD8⁺ (CD8⁺SP), CD4⁻CD8⁻ double-negative (DN), CD4⁻CD8⁻CD44⁺CD25⁻ (DN1), CD4⁻CD8⁻CD44⁺CD25⁺ (DN2), CD4⁻CD8⁻CD44⁻CD25⁺ (DN3), CD4⁻CD8⁻CD44⁻CD25⁻ (DN4), and CD3⁻CD4⁻CD8⁺ immature single positive (ISP). Mature T cell populations were defined by the following markers, after excluding Mac-1⁺, B220⁺, and Ter119⁺ cells: CD4⁺ (CD4⁺ cells), CD4⁺CD44⁻CD62L⁺ (naïve CD4⁺ cells), CD4⁺CD44⁺CD62L⁺ (CD4⁺ central memory cells), CD4⁺CD44⁺CD62L⁻CD69⁺ (CD4⁺ resident memory cells), CD4⁺CD44⁺CD62L⁻CD69⁻CD103⁻ (CD4⁺ effector memory cells), CD8⁺ (CD8⁺ cells), CD8⁺CD44⁻CD62L⁺ (naïve CD8⁺ cells), CD8⁺CD44⁺CD62L⁺ (CD8⁺ central memory cells), CD8⁺CD44⁺CD62L⁻CD69⁺ (CD8⁺ resident memory cells), and CD4⁺CD44⁺CD62L⁻CD69⁻CD103⁻ (CD8⁺ effector memory cells). For all flow cytometry experiments, combinations of antibodies were used against the following cell-surface markers, conjugated to Violet Fluor 450, Brilliant Violet 421, Brilliant Violet 510, FITC, PE, PE594, PerCP-Cy5.5, PE-Cy7, APC, eFluor 660, Alexa Fluor 700, RedFluor 710, APC/Fire 750, or APC-Cy7: CD2 (RM2–5, Tonbo Biosciences/BioLegend), CD3 (17A2, Tonbo Biosciences/BioLegend), CD4 (GK1.5, Tonbo Biosciences/BioLegend), CD5 (53–7.3, BioLegend), CD8a (53–6.7, Tonbo Biosciences/BioLegend), CD11b (Mac1) (M1/70, Tonbo Biosciences/BioLegend), CD11c (N488, BioLegend), CD16/32 (93, BioLegend), CD25 (PC61, BioLegend), CD34 (RAM34, Invitrogen), CD44 (M7, Tonbo Biosciences), CD45.1 (A20, Tonbo Biosciences), CD45.2 (104, Tonbo Biosciences/BioLegend), CD45R (B220) (RA.3–6B2, Tonbo Biosciences/BioLegend), CD48 (HM48–1, BioLegend), CD62L (MEL-14, Tonbo Biosciences), CD69 (H1.2F3, Tonbo Biosciences), CD71 (R17217, BioLegend), CD103 (2E7, BioLegend), CD115 (CSF-1R) (AFS98, BioLegend), CD117 (c-Kit) (2B8, BioLegend), CD150 (TC15–12F12.2, BioLegend), Ly-6A/E (Sca-1) (D7, BioLegend), Ly-6C (HK1.4, BioLegend), Ly-6G (Gr-1) (RB6–8C5, Tonbo Biosciences; 1A8, BioLegend), SiglecF (E50–2440, BD Biosciences) or Ter119 (TER119, Tonbo Biosciences/BioLegend). Analysis and sorting were performed using the FACSAria flow cytometer (BD Biosciences) or a FACSCanto (BD Biosciences). Data were analyzed using FlowJo (Flowjo LLC) or FACSDiva (BD Biosciences).

Bone marrow and liver reconstitution assays.

Recipient mice (CD45.1) were irradiated using an XRAD 320 X-ray irradiator (Precision X-Ray) with two doses of 540 rad (1080 rad in total) delivered at least 3 hours apart. Bone marrow cells were injected into the retro-orbital venous sinus of anesthetized recipients. Seven weeks prior to transplant, donor mice were either infected with 200 cercariae or left uninfected (controls). For competitive transplants, 5 × 10⁵ CD45⁺-selected bone marrow cells from infected or from uninfected donor (CD45.2) mice and 5 × 10⁵ competitor (CD45.1;CD45.2) cells were mixed and transplanted by injection into the retro-orbital venous sinus of anesthetized recipients. Recipient mice were maintained on antibiotic water (Baytril 0.08 mg/ml) for 1 week pre-transplantation, and for 4 weeks after transplantation. Blood was obtained from the tail veins of recipient mice every four weeks for at least 16 weeks after transplantation. Red blood cells were lysed in ammonium chloride lysis buffer. The remaining cells were stained with antibodies against CD45.2, CD45.1, C11b (Mac1), CD115, Ly6G, Ly6C, CD45R (B220), and CD3 and analyzed by flow cytometry. For the secondary bone marrow transplants, 1 × 10⁷ bone marrow cells from primary recipients were harvested 5 months after primary transplant and transplanted into lethally irradiated CD45.1 secondary recipients. For the competitive liver reconstitution assays, 2 × 10⁶ CD45⁺-selected liver cells from infected or from uninfected donor (CD45.2) mice and 4 × 10⁵ bone marrow competitor (CD45.1;CD45.2) cells from uninfected mice were mixed and transplanted into lethally irradiated CD45.1 recipient mice. Bone marrow cells for analysis of transplant recipient mice or for secondary transplantations was obtained by crushing femurs, tibias, vertebrae, and pelvic bones.

Histological evaluation, immunohistochemistry and immunofluorescence.

Livers and spleens from control and Schistosoma mansoni-infected mice were collected and fixed in 10% formalin for at least 24 hours. After fixation, tissues were submitted to the UTSW Histo Pathology Core for processing, paraffin embedding, sectioning and hematoxylin & eosin staining. H&E sections of livers and spleens from uninfected control and infected mice were evaluated for pathology, hematopoiesis, erythropoiesis, and inflammation by a pathologist (W.C). Immunohistochemistry was performed on formalin-fixed, paraffin-embedded livers and spleens. Sections were gradually rehydrated and heat-induced Tris-EDTA buffer (10mM Tris base, 1 mM EDTA, 0.1% Tween, pH 9.0) was used for antigen retrieval. Sections were washed in PBS with 0.1% Tween and blocked in PBS with 10% BSA containing 0.1% Tween for 1h. Sections were stained overnight with the following antibodies: myeloperoxidase (MPO; EPR20257, Abcam; 1/500), or CD61 (SJ19–09, Thermo Fisher; 1/100). Secondary antibody was IgG-HRP (Cell Signaling, 7074; 1/1000) and the signal was visualized by using DAB peroxidase substrate kit according to the manufacturer’s instructions (Vector Laboratories, SK-4100). Sections were counterstained with Mayer’s hematoxylin (Sigma, MHS32) and mounted with Vectashield Express mounting medium (H-5700). Images were acquired on an Olympus BX53 microscope. For immunofluorescence, dissected livers were fixed overnight in 4% paraformaldehyde (Fisher Scientific, AAJ19943K2) on a rotator at 4°C. Livers were then washed with PBS, rotated overnight in 20% sucrose at 4°C before being frozen in OCT compound on dry ice. Livers were sectioned into 8 μm slices using a cryostat (Leica CM3050S) and dried at 57 °C for 1 hour. Antigens were unmasked in 10 mM citric acid, 0.05% Tween20 buffer (pH 6) using a pressure cooker at 7.5 psi for 17 min and then cooled for 30 min. Sections were washed in PBS with 0.1% Tween and blocked with 10% BSA, 0.1% Tween in PBS for 1 hour at room temperature. Sections were stained overnight in a moist chamber at 4°C with a combination of the following primary antibodies: c-Kit (R&D Biosciences, AF1356; 1/150), desmin (Epredia, RB9014P0; 1/200), CD31 (Abcam, ab28364; 1/50), and MPO (Abcam, ab208670; 1/100). Sections were stained for 1 hour in the dark with DAPI and a combination of the following secondary antibodies: rabbit Alexa Fluor 568 (Invitrogen, A10042; 1/200), goat Alexa Fluor 647 (Invitrogen, A32849; 1/200), rabbit Alexa Fluor 488 (Invitrogen, A32790; 1/200). Slides were mounted with Aqua-Poly/Mount (Polysciences, 18606–20). Images were acquired on a Zeiss LSM780 inverted confocal microscope using the tiling function and tiled images were then stitched together using Zen software (Zeiss) to create a composite image of the tissue.

Quantitative RT-PCR.

Livers were homogenized in 1 ml TRIzol LS reagent (Life Technologies, 10296010) using the FastPrep-24^™ 5G bead beater (MP Biomedicals). Total RNA was isolated by chloroform extraction, precipitated with isopropanol, washed with 75% ethanol and RNA pellets were dissolved in RNase/DNase-free water. Reverse transcription of RNA was performed using the iScript cDNA synthesis kit (BioRad, 1708891) according to the manufacturer’s instructions and cDNA was diluted 10-fold prior to use. 20 ng cDNA was used for quantitative RT-PCR with iTaq Universal SYBR Green Supermix (BioRad, 1725122) and a CFX384 Real-Time System (BioRad). Data were analyzed using the BioRad CFX Maestro software. GAPDH expression levels were used for normalization. The following primers were used: Gapdh: 5’-GGAGAGTGTTTCCTCGTCCC-3’, 5’-ACTGTGCCGTTGAATTTGCC-3’; Gm-csf (Csf2): 5’-GATATTCGAGCAGGGTCTACG-3’, 5’-AGGCTGTCTATGAAATCCGC-3’; M-csf (Csf1): 5’- ACCCAGGATGAGGACAGAC-3’, 5’-AGGAAGATGGTAGGAGAGGG-3’; Scf (Kitl): 5’-TCAAGAGGTGTAATTGTGGACG-3’, 5’- GGGTAGCAAGAACAGGTAAGG-3’.

Statistical analysis.

Most figure panels show the pooled results from mice we analyzed from multiple independent experiments. Mice were allocated to experiments randomly. For most experiments the operator was not blinded to the treatment. Uninfected littermate controls, or uninfected controls from litters of the same parental strains born a few days apart were used for experiments. Prior to analyzing the statistical significance of differences among treatments, we tested whether data were normally distributed and whether variance was similar among treatments. To test for normal distribution, we performed the Shapiro-Wilk test when 3 ≤ n < 20 or the D’Agostino & Pearson test when n ≥ 20. To test if variability significantly differed among treatments, we performed F-tests. If the data did not significantly (p < 0.01 for at least one treatment) deviate from normality, we used a parametric test, otherwise data were log-transformed and tested for a significant deviation from normality. If the log-transformed data passed normality, a parametric test was used on the transformed data. If both the untransformed and log-transformed data did not pass the normality test, a non-parametric test was used on the untransformed data. To assess the statistical significance of a difference between two treatments, we used a t-test for data that was normally distributed and had equal variability, or a t-test with Welch’s correction for data that was normally distributed and had unequal variability, or a Mann-Whitney test for data that was not normally distributed. To assess the statistical significance of differences between treatments when multiple measurements were taken across time, we used a repeated measures mixed-effects model for data for which some values were missing. To determine the statistical significance between treatments for the presence of multi-lineage reconstitution, we used a Fisher’s exact test.

Results

The effects of schistosome infection on hematopoiesis.

To understand the impact of schistosomiasis on HSCs and hematopoietic progenitors, we infected mice with the human pathogen Schistosoma mansoni by tail exposure in water containing cercariae. In this model, eggs deposited in the liver trigger granuloma formation and schistosomiasis pathology starting from ~ 5 weeks post-infection (10, 38, 39). We analyzed the hematopoietic system in the bone marrow and spleen of mice 7 weeks post-infection (Fig. 1A). As compared to uninfected mice, bone marrow cellularity did not significantly change after infection (Fig. 1B). The frequencies of CD150⁺CD48⁻Lineage⁻Sca-1⁺Kit⁺ HSCs and CD150⁻CD48⁻Lineage⁻Sca-1⁺Kit⁺ multipotent progenitors (MPPs) were unchanged (Fig. 1C–D). Infection increased the frequency of CD150⁻CD48⁺Lineage⁻Sca-1⁺Kit⁺ hematopoietic progenitor cells (HPC-1) (Fig. 1E) and decreased the frequency of CD34⁺CD16/32⁻Lineage⁻Sca-1⁻Kit⁺ common myeloid progenitors (CMPs) and CD34⁻CD16/32⁻Lineage⁻Sca-1⁻Kit⁺ megakaryocyte-erythroid progenitors (MEPs) (Fig. 1F, 1G). The observed decline in CMP and MEP frequency was similar to a previous study examining the effects of schistosome infection in Apoe-deficient mice on a high-fat diet (40). The frequencies of CD150⁺CD48⁺Lineage⁻Sca-1⁺Kit⁺ (HPC-2) and CD34⁺CD16/32⁺Lineage⁻Sca-1⁻Kit⁺ granulocyte-monocyte progenitors (GMPs) did not change (Fig. 1H, 1I). The spleen size and cellularity significantly increased after infection (Supplemental Fig. 1A, 1B), however the frequency of HSCs and most progenitor cell types in the spleen did not change (Supplemental Fig. 1C–P). This suggests that in contrast to other infectious or inflammatory challenges, the spleen is not a reservoir for multilineage hematopoiesis in schistosomiasis despite its increased size.

Figure 1. The effects of schistosome infection on the frequency of hematopoietic cells in the bone marrow and blood cells.

(A) Schematic overview of analysis of the hematopoietic and blood system 7 weeks after Schistosoma mansoni infection. (B-N) The frequency of HSCs, progenitors and mature cell types in the bone marrow of S. mansoni infected mice or uninfected controls (n = 7–13 mice per treatment). (O-R) Blood cell counts of white blood cells, neutrophils, lymphocytes, and monocytes of infected or uninfected mice (n = 9–13 mice per treatment). (S-W) The frequency of immune cell types in the blood of infected or uninfected mice (n = 4–11 mice per treatment). For all graphs both male and female mice were used. All graphs show mean ± s.d. In all figures, *p < 0.05, **p < 0.01, ***p < 0.001. Statistical significance was assessed with a t-test with Welch’s correction (B-C, E, G-H, J-K, O-Q, S and U-W), a t-test (F, I, L-M, R and T), or a Mann-Whitney test (D and N).

In the myeloid lineage, infection preferentially increased the frequency of bone marrow monocytes and eosinophils but not neutrophils or dendritic cells (Fig. 1J–M). Infection significantly decreased bone marrow B cell frequency (Fig. 1N) in agreement with a recent report (41). The infected spleen had an increased frequency and number of most myeloid lineage cell types (Supplemental Fig. 1Q–AB). Infected mice had a higher white blood cell count than uninfected mice, and increased numbers or frequencies of neutrophils, monocytes, eosinophils, and dendritic cells in the blood (Fig. 1O–W). The development of major T cell progenitor cell types in the thymus was not significantly impacted by infection (Supplemental Fig. 2). This contrasts with the severe impact of many other infections on thymus cellularity (42). To determine the effects of infection on T cells, we analyzed T-cell subpopulations in the blood, bone marrow, spleen, and liver. In the marrow, the frequency of total CD4⁺ and effector memory CD4⁺ T cells increased and the frequency of CD8⁺ T cells did not change (Supplemental Fig. 3A–F). Infection did not cause significant changes in frequencies of T cell subsets in the spleen (Supplemental Fig. 3G–L). The frequency of naïve CD4⁺ T cells declined in the blood (Supplemental Fig. 3M–R). The liver had an increased frequency of total CD4⁺ or CD8⁺ T cells, particularly of a resident memory immunophenotype (Supplemental Fig. 3S–V). Therefore, schistosome infection caused several changes in the frequency of hematopoietic and immune cells in the marrow, spleen, blood, and liver.

Infection impairs bone marrow HSC function after serial transplantation.

To test if infection changes HSC activity, donor bone marrow cells from infected or uninfected mice were mixed with competitor bone marrow cells from uninfected mice and transplanted into lethally irradiated recipients (Fig. 2A). This classical competition assay determines changes in long-term HSC activity irrespectively of changes in phenotypic surface markers, which can change after infection (17). There was no significant difference in donor cell reconstitution capacity between infected and uninfected mice (Fig. 2B–E, Supplemental Fig. 4A, 4B). At 16 weeks after transplant, there was no significant difference in reconstituted lineage composition between blood cells from infected as compared to uninfected mice (Supplemental Fig. 4C–G) suggesting infection did not cause long-term cell-intrinsic changes in HSC differentiation potential. Despite the maintenance of hematopoietic reconstitution in the peripheral blood, the bone marrow of transplant recipients had proportionately fewer infected donor-derived HSCs and some restricted progenitors as compared to uninfected donor-derived HSCs or progenitors (Fig. 2F, 2G). This suggested an impairment in HSC self-renewal after transplantation. To test HSC self-renewal after the primary transplant and the ability of HSCs to respond to repeated challenge, bone marrow cells from primary recipient mice were transplanted into lethally irradiated secondary transplant recipients 5 months after primary transplant (Fig. 2H). The blood reconstitution capacity of bone marrow cells from infected mice significantly decreased as compared to uninfected mice (Fig. 2I–L, Supplemental Fig. 4H, 4I). The proportion of infected donor-derived myeloid progenitors and total hematopoietic cells in the marrow of secondary transplant recipients decreased (Fig. 2M, 2N). Therefore, schistosome infection impaired HSC function after serial transplant.

Figure 2. Schistosome infection impairs bone marrow HSC function.

(A) Schematic overview of experiments to assess bone marrow HSC function after infection. 5 × 10⁵ CD45.2⁺ donor cells from bone marrow of mice infected with S. mansoni for seven weeks, or from uninfected mice, were mixed with 5 × 10⁵ CD45.1⁺CD45.2⁺ competitor bone marrow cells from uninfected mice and transplanted to each lethally irradiated CD45.1⁺ recipient mouse (n = 5 donor and 21–25 recipient mice per treatment). (B-E) Donor cell reconstitution of CD45⁺, myeloid, B, and T cells in the blood at the indicated time points. (F) The percentage of donor-derived hematopoietic stem and progenitor cells in the bone marrow at 5 months after transplantation (n = 21–25 mice per treatment). (G) The percentage of donor-derived myeloid, B and T cells in the bone marrow at 5 months after transplantation (n = 15 mice per treatment). (H) Schematic overview of the secondary transplantation. 1 × 10⁷ bone marrow cells from primary recipient mice were harvested 5 months after primary transplant and transplanted into each lethally irradiated secondary recipient mouse (n = 3 donor mice and 14–15 recipient mice per treatment). (I-L) Donor cell reconstitution of CD45⁺, myeloid, B and T cells in the blood at the indicated time points after secondary transplantation. (M) The percentage of donor-derived hematopoietic stem and progenitor cells in the bone marrow after secondary transplantation (n = 13–14 mice per treatment). (N) Representative flow plots of donor-derived Lineage⁻ cells (left), Lin⁻Sca-1⁺Kit⁺ (LSK) cells (middle), and Lin⁻Sca-1⁻Kit⁺ (LK) cells (right). Arrows indicate the respective parent population. All graphs show mean ± s.d. Statistical significance was assessed with a repeated measures mixed model (B-E, and I-L), a t-test (F-G, and M, HSC and HPC-2 cells), a t-test with Welch’s correction (M, HPC-1; CMP; GMP; MEP; and CD45⁺ cells) or a Mann-Whitney test (F, HPC-1 cells; and M, MPP cells).

Hematopoietic activity in the liver of infected mice.

The liver of infected mice harbors schistosome eggs which trigger granuloma formation. To test if schistosomiasis elicits multilineage hematopoiesis in the liver, we analyzed the frequency of phenotypic HSCs and other hematopoietic progenitors 7 weeks after infection with S. mansoni. The liver of infected mice had an increased frequency of phenotypic HSCs, HPC-1, HPC-2, CMPs, GMPs and MEPs (Fig. 3A–F). Among mature cells, the frequency of inflammatory monocytes, eosinophils, dendritic cells, and monocytic dendritic cells also increased after infection, the frequency of neutrophils was unchanged, and the frequency of B cells decreased (Fig. 3G–L). The infected liver was enlarged (Fig. 3M). Its architecture was distorted by chronic inflammation in the portal triads and increased hematopoiesis. Lodged S. mansoni eggs surrounded by mixed histiocytes/granulocytes formed granulomas (Fig. 3N). Liver sinuses harbored hematopoietic clusters with erythroid and megakaryocyte cell production (Fig. 3N). Immunohistochemistry staining for MPO and CD61 confirmed a dramatic increase of granulocytes throughout the liver, including in the granulomas and sinuses, and a mild increase in megakaryocytes (Fig. 3O, 3P). Therefore, inflammation and hematopoiesis increase in infected liver. To test the localization of hematopoietic cells in infected livers, we assessed immunofluorescence staining for the hematopoietic stem and progenitor marker c-kit in combination with other markers. In uninfected liver, almost all c-kit⁺ cells were also positive for the endothelial cell marker CD31 and localized around blood vessels, suggesting they were endothelial cells. By contrast in infected liver some c-kit⁺ cells were negative for CD31, localized within the hepatic tissue, and had the round morphology of hematopoietic cells, consistent with a hematopoietic progenitor identity (Fig. 4A). In infected liver, CD31⁻c-kit⁺ cells localized near cells which were positive for desmin, a marker of hepatic stellate cells, consistent with the ubiquitous presence of desmin⁺ cells (Fig. 4B). Infected livers harbored clusters of MPO-positive cells, including in the periphery of granulomas and in the liver parenchyma, some of which co-stained with c-kit suggesting granulopoietic foci (Fig 4C–E). Infection increased the expression in the liver of hematopoietic stem and progenitor cell growth factors including SCF, M-CSF, and GM-CSF (Fig. 4F–H), suggesting the presence of niches to support active hematopoiesis in the infected liver.

Figure 3. Schistosome infection promotes liver hematopoiesis.

(A-L) Frequency of hematopoietic stem, progenitor, and mature immune cells in the liver of schistosome infected or uninfected control mice (n = 4–12 mice per treatment). (M) Liver weight after schistosome infection. (N) Representative images of H&E staining of livers from uninfected control (left) and infected mice (middle and right). Middle panel shows inflammation composed of mixed histiocytes and granulocytes associated with heavy egg lodging (red arrow) in infected mice. Right panel shows a microabscess (red arrow) and increased hematopoiesis including erythropoiesis (green arrow) and megakaryopoiesis (black arrow) in infected mice. (O) Representative images of immunohistochemical staining for MPO⁺ granulocytes in livers from uninfected (left) or infected mice (middle and right). (P) Representative images of immunohistochemical staining for CD61⁺ megakaryocytes (black arrows) in livers from uninfected (left) or infected mice (right). All graphs show mean ± s.d. Statistical significance was assessed with a Mann-Whitney test (A), a t-test with Welch’s correction (B-C, E, H, K, and M), and a t-test (D, F-G, I-J, and L).

Figure 4. Localization of c-Kit⁺ cells in the liver of schistosome-infected mice.

(A-B) Representative images of immunofluorescence staining for (A) CD31⁺ endothelial and c-kit⁺CD31⁻ hematopoietic progenitor cells, and (B) desmin⁺ hepatic stellate cells and c-kit⁺ hematopoietic progenitor cells. White arrows indicate c-kit⁺ hematopoietic progenitor cells in (A-B). (C-E) Representative images of immunofluorescence staining for MPO⁺ granulocytes and c-kit⁺ hematopoietic progenitor cells (magnified image in C, white arrows in D, E) in livers from schistosome infected mice, including in (D) granulomas and the (E) liver parenchyma. For (A, B, D, and E) a green pseudocolor was used for Alexa Fluor 568. (F-H) Quantitative RT-PCR analysis of HSC growth factors in livers from uninfected control and infected mice (n = 6–8 mice per treatment). All graphs show mean ± s.d. Statistical significance was assessed with a t-test (F, H) or a t-test with Welch’s correction (G).

To test if the increased frequency of phenotypic HSCs corresponded to increased HSC function, we transplanted 2 × 10⁶ CD45⁺ cells from the liver of infected or uninfected donor mice in competition with 4 × 10⁵ bone marrow cells from uninfected mice into lethally irradiated recipient mice (Fig. 5A). The peripheral blood of recipient mice contained significantly more donor-derived cells from infected livers than from uninfected livers (Fig. 5B–E). Nine out of ten mice receiving donor cells from infected liver showed donor-derived multilineage reconstitution as compared to only one out of fourteen mice receiving donor cells from uninfected liver (Fig. 5F). The bone marrow of transplant recipients had a significantly higher proportion of HSCs and restricted progenitors derived from infected donor livers as compared to uninfected donor livers (Fig. 5G, 5H). Thus, functional HSCs were present in the liver of infected mice.

Figure 5. HSC activity in the liver.

(A) Schematic overview of experiments to assess HSC function in the infected liver. 2 × 10⁶ CD45⁺ donor cells from livers of infected or uninfected control mice were mixed with 4 × 10⁵ CD45⁺ competitor bone marrow cells from uninfected mice and transplanted to each lethally irradiated recipient mouse (n = 3 donor and 10–14 recipient mice per treatment). (B-E) Donor cell reconstitution of CD45⁺, myeloid, B and T cells in the blood at the indicated time points after transplantation. (F) The proportion of recipient mice which showed multilineage reconstitution after competitive transplantation of donor CD45⁺ cells from the liver of infected mice or uninfected controls. Multilineage reconstitution was defined as > 2% donor cell chimerism in peripheral blood myeloid, B and T cells at 16 weeks after transplantation. (G) The percentage of donor-derived hematopoietic stem and progenitor cells in the bone marrow at 16 weeks after transplantation (n = 10–14 mice per treatment). (H) Representative flow plots of chimerism analysis of LSK cells. All graphs show mean ± s.d. Statistical significance was assessed with a repeated measures mixed model (B-E), a Fisher’s exact test (F), a Mann-Whitney test (G, HSC; MPP; HPC-1; and GMP), a t-test with Welch’s correction (G, HPC-2; CMP; and CD45⁺), and a t-test (G, MEP).

Schistosome infection blocks marrow erythropoiesis and increases spleen erythropoiesis.

Schistosome-infected mice were anemic and thrombocytopenic (Fig. 6A–D). They also showed increased red blood cell distribution width, a marker of anemia and inflammation which in humans correlates with increased mortality (43) (Fig. 6E). Chronic anemia is one of the most prevalent and disabling symptoms of schistosomiasis (44). Several causes have been proposed including anemia of inflammation, blood loss, erythrocyte spleen sequestration, or autoimmunity (44). Schistosome-infected mice showed a sharp decrease in CD71⁺Ter119⁺ erythroid progenitors and an accumulation of CD71^midTer119⁻ immature progenitors in the marrow (Fig. 6F–I) suggesting a block in marrow erythropoiesis. In contrast the spleen of infected mice had an increased frequency of CD71⁺Ter119⁻ and CD71⁺Ter119⁺ erythroid progenitors (Fig. 6J–M). Therefore, erythropoiesis in schistosomiasis infection shifted from the bone marrow to the spleen. Splenic architecture was disrupted by reactive follicular hyperplasia with germinal centers in the white pulp and erythropoiesis and megakaryopoiesis in the red pulp (Fig. 6N, 6O). Staining for CD61 and MPO showed megakaryopoiesis and the presence of small numbers of granulocytes in spleens of infected mice (Fig. 6P, 6Q). S. mansoni egg lodging in spleens was rare, and granuloma formation was observed in only one out of five assessed spleens from infected mice. Consequently, inflammation associated with histiocytes and granulocytes was relatively mild in the spleen as compared to the liver (Fig. 6N). Our results are consistent with the idea that splenomegaly after S. mansoni infection is largely driven by extramedullary erythropoiesis rather than inflammation.

Figure 6. Schistosome infection blocks bone marrow erythropoiesis and increases spleen erythropoiesis.

(A-E) Analysis of red blood cells, hemoglobin, hematocrit, platelets, and red blood cell distribution width of S. mansoni infected or uninfected mice. Analysis was performed 7 weeks after infection. (F-H) The frequency of CD71^midTer119⁻ immature progenitors, CD71⁺Ter119⁻ and CD71⁺Ter119⁺ erythroid progenitors in the bone marrow of infected mice or uninfected control mice (n = 7–14 mice per treatment). (I) Representative flow plots of progenitors quantified in (F-H). (J-L) The frequency of CD71^midTer119⁻ immature progenitors, CD71⁺Ter119⁻ and CD71⁺Ter119⁺ erythroid progenitors in the spleen of infected mice or uninfected control mice (n = 5–11 mice per treatment). (M) Representative flow plots of progenitors quantified in (J-L). (N-O) Images of H&E staining of spleens from uninfected or infected mice showing follicular hyperplasia with germinal centers in the white pulp (black arrow in N), significant increased hematopoiesis including erythropoiesis in the red pulp (green arrow in N), rare inflammation composed of mixed histiocytes and granulocytes associated with Schistosoma egg lodging (red arrow in N), and (O) significant increased hematopoiesis including erythropoiesis in the red pulp within the splenic cords/sinuses. (P-Q) Representative images of immunohistochemical staining for CD61⁺ megakaryocytes (black arrows in P), and MPO⁺ granulocytes (red arrow in Q) in the spleen of infected mice or uninfected controls. (R) Graphical summary. All data show mean ± s.d. Statistical significance was assessed with a t-test (A-C, E, G, and J-K) and a t-test with Welch’s correction (D, F, H, and L).

Discussion

The impact of schistosomiasis on hematopoietic stem cells.

Our findings provide a framework to understand how S. mansoni infection affects hematopoiesis (Fig. 6R). The frequency of bone marrow HSCs is not reduced after infection, and HSCs from infected mice can reconstitute primary recipients but are compromised in their ability to reconstitute secondary recipients. The molecular mechanisms that regulate HSC function in response to schistosome infection remain unknown. Schistosomiasis elevates the levels of several cytokines. Some type 2 cytokines, including IL-4 and IL-25 promote the development of restricted hematopoietic progenitors without impacting HSCs (45, 46), and consistent with this a type 2 immune response did not impact HSC frequency in mice infected with Trichuris muris (33). Other schistosomiasis-associated cytokines can have either protective or detrimental effects on HSCs. For example, eosinophilia caused by IL-5 overexpression negatively regulates HSC frequency and function through CCL-6 production (47). Conversely, IL-10 protects HSCs in inflammatory contexts and promotes HSC function during regeneration (48). Thus, the effects on HSC function we observe may reflect a balance of positive and negative effects from multiple cytokines. Populations in areas endemic for schistosomiasis face several hematopoietic challenges, including a high burden of anemia and of other infectious diseases such as malaria. Our findings that HSC function in serial transplantation is reduced after schistosome infection suggests that schistosomiasis impairs the ability of the hematopoietic system to regenerate after repeated challenge.

Erythropoiesis is blocked in the marrow but not the spleen.

It is thought that inflammation is the most common cause of anemia in schistosomiasis (49). Our results suggest that anemia in schistosomiasis is partly caused by a sharp block in bone marrow erythropoiesis. This block is accompanied by a striking increase in spleen erythropoiesis. Inflammation is known to reduce marrow erythropoiesis (50) and consistent with this schistosomiasis increased inflammatory monocytes in the bone marrow and blood. The mechanisms for the arrest in erythropoiesis are not known but may involve cytokines known to be elevated in schistosomiasis, such as IL-10 which can cause anemia (51). It is interesting that schistosomiasis arrested erythropoiesis in the marrow but not in the spleen. The inflammatory response in the spleen after infection is mild, and the spleen of infected mice does not harbor proportionately more HSCs or myeloid progenitors than the spleen of uninfected mice. Therefore, splenomegaly in schistosomiasis is not a consequence of general extramedullary hematopoiesis or inflammation but specifically of splenic erythropoiesis. Our results suggest that an important function of the spleen is to make erythrocytes in schistosome infections.

Hematopoiesis in the liver.

We show that HSCs and restricted progenitors are present in the liver after infection, as assayed by immunophenotypic, immunofluorescence and transplantation experiments. CD45⁺ cells from liver of infected mice competed against bone marrow cells from uninfected mice at a 5:1 donor:competitor ratio produced an average reconstitution of 20%, or 1:4 donor:competitor cells after transplant. This suggests that functional HSC frequency is 20-fold lower in infected liver than bone marrow. Given that the bone marrow typically contains more CD45⁺ cells than the liver, systemic output of blood cells from liver HSCs in infection is likely minor as compared to bone marrow hematopoiesis. However, hematopoiesis in the infected liver, including in the periphery of granulomas, may be important for the local immune response. The production of monocyte-derived macrophages in the infected liver has been suggested to be protective (52–54), and, consistent with our results, proliferating myeloid lineage cells are found in the periphery of granulomas (55). The fact that HSC or immature progenitor frequency is not elevated in spleen of infected mice suggests some specificity to liver hematopoiesis as opposed to general activation of extramedullary hematopoiesis. Are HSCs and hematopoietic progenitors in the infected liver supported by specialized niches? Our results show that in the infected liver, c-kit⁺ hematopoietic progenitors localize close to hepatic stellate cells and the expression of HSC and progenitor growth factors increases. Hepatic stellate cells serve as the major HSC niche cell type in fetal liver hematopoiesis by secreting the c-kit ligand SCF (56). They can also express known HSC niche factors in adult mice, including SCF after schistosome infection (57) or CXCL12 in other contexts (58). Thus, we hypothesize that during schistosome infection HSCs migrate to the liver and are supported by reactivation of a dormant liver hematopoietic niche.

Key Points.

• Schistosome infection impairs bone marrow HSC function after serial transplantation

• Infection blocks marrow erythropoiesis and increases spleen erythropoiesis

• Hematopoietic activity in the liver of infected mice

Abbreviations used in this article:

BM

bone marrow

CMP

common myeloid progenitor

DN

double negative thymocyte

DP

double positive thymocyte

GMP

granulocyte–monocyte progenitor

HPC-1

hematopoietic progenitor cell subset 1

HPC-2

hematopoietic progenitor cell subset 2

HSC

hematopoietic stem cell

LK

Lineage⁻Sca1⁻Kit⁺

LSK

Lineage⁻Sca1⁺Kit⁺

MEP

megakaryocyte–erythroid progenitor

MPP

multipotent progenitor cell

MPO

myeloperoxidase

SCF

stem cell factor