Blood Flow Regulates Metabolism in Hematopoietic Development

Abstract

Blood flow modifies oxygen availability and biomechanical forces within the vasculature of the embryo as the hematopoietic system develops. The aorta-gonad-mesonephros (AGM) envelops the largest artery in the body and is a critical site for the emergence of hematopoietic stem cells (HSCs). Herein, I discuss the role of hypoxia-inducible factors (HIFs) and force as determinants of metabolism and fate determination. To address the effects of blood flow on hematopoietic development, I employ mouse embryo models and biomimetic culture. Real-time cell metabolic analyses show that oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) are altered by flow in cultures of AGM cells. Additionally, mitochondrial permeability transition pore (mPTP) activity assays indicate that mPTP opening is regulated as precursors commit to a hematopoietic fate. Moreover, transcriptome data suggests that expression of Ppif, the gene that encodes the mPTP regulator Cyclophilin D, may be downregulated by fluid force. Together, these data implicate blood flow in regulation of metabolic activity in the AGM region, influencing bioenergetics and serving as a cue for HSC emergence. The broader significance of this research lies in its potential to accurately recapitulate the hematopoietic niche in vitro for study of blood disorders and improvement of cellular therapies.

1. Introduction

Oxygen deprivation is a natural cue in developing mammalian embryos [1]. Indeed, oxygen gradients provide critical spatial information necessary to coordinate development of the embryonic blood system, cardiovascular system, nervous system, and a multitude of other organs. Stem cells, particularly of the blood system, are known to occupy hypoxic niches, even into adulthood [2]. Yet, oxygen is the primary electron acceptor in countless biochemical reactions required for life, and abundant evidence indicates that insufficient oxygen delivery to embryonic tissues results in developmental defects and lethality. Thus, it is not surprising that stem cell niches are frequently found in close proximity to the vasculature [3].

In embryogenesis, hematopoietic stem cells (HSCs) and progenitors are specified from unique endothelial cells that line the arterial vasculature, termed hemogenic endothelium (HE) [4]. Within developing arterial blood vessels, cells change from an endothelial to hematopoietic identity in a process known as the endothelial-to-hematopoietic transition (EHT). HSCs capable of durable renewal of erythrocytes, platelets, and the majority of cells of the immune system arise from within an evolutionarily conserved “hotspot” of HSC fate specification - the aorta-gonad-mesonephros (AGM) - at approximately 10.5–11.5 days of murine (E11.5) and 5 weeks of human embryogenesis [5]. We and others have shown that blood flow is required for differentiation of HE capable of producing HSCs [6–10].

2. Flow-induced Cellular Signaling in Hematopoiesis

How blood flow promotes blood development is of great interest since this mechanistic information could aid in recapitulating the embryonic hematopoietic niche in culture. In fact, accurate biomimetic modeling of the hematopoietic niche remains elusive but, if successful, would redefine strategies available for hematologic disease modeling and therapeutic use of HSCs. In 2009, two reports were published demonstrating that blood flow in zebrafish [9] and biomechanical force caused by blood flow in mice [7] drive emergence of hematopoietic cells. Both studies showed that flow-induced nitric oxide generation contributes to blood development. In continuation of these studies, we examined the transcriptome for clues to the intracellular signaling that regulates hematopoiesis in response to blood flow [6]. We found that flow triggers calcium oscillations in the cytosol to elevate production of prostaglandin E₂, which thereby activates a signaling cascade through cAMP-PKA-CREB. Inhibition of this signaling pathway by blocking antibodies or chemical inhibitors blunted hematopoietic activity. We also reported that flow upregulates Notch, Wnt, and Hedgehog [6, 11–13], developmental pathways critical for hematopoiesis [14]. Interestingly, flow can be mimicked pharmacologically via nitric oxide donors [7, 9, 10] and a synthetic prostaglandin E₂ derivative [6, 8] to enhance hematopoietic activity from pluripotent stem cells and cells from the AGM. This information provides potential druggable targets to enhance hematopoiesis in vitro, yet we believe additional flow-responsive signaling is important for blood development.

Initiation of circulation within the embryo alters oxygen availability at the site of HSC emergence on the arterial wall. In fact, examination of gene expression from Ncx1-deficient embryos without a heartbeat in an independent study [15] and AGM cultures from our published work [6] suggests that flow alters signaling indicative of change in oxygen within the aorta. Using a platform that permits culture of embryonic cells with precise intensities of wall shear stress (WSS) modeled after blood flow in the mouse embryo, we find that hypoxia-inducible factor 1 (HIF-1) is a topmost enriched pathway in WSS compared to static (no flow) cultures of AGM. Likewise, HIF-1 is enriched in comparisons of several EHT cell types in wild type embryos and cardiac mutant littermates. Contributing to this enrichment are transcripts for HIF-1α and factors implicated in HIF regulation, including various PI3K subunits. How and whether this signaling is important for hematopoiesis remains largely unaddressed.

Mouse studies manipulating genes encoding HIF protein subunits suggest that HIF-1 and HIF-2 are both critical for proper formation of the vasculature and for development of some blood cell lineages. The alpha subunits of HIF-1 and HIF-2 transcription factors are regulated by oxygen tension; whereas, the related HIF-1β protein, also known as aryl hydrocarbon receptor nuclear translocator (ARNT), forms heterodimers with HIF-1α and HIF-2α and is believed to be constitutively expressed [16]. Evidence from Arnt genetic knockout studies suggested that HIF signaling is required for proliferation of multilineage hematopoietic progenitors derived from embryos and embryonic stem cells [17]. Impaired hematopoiesis resulting from loss of HIF-α-Arnt signaling could be rescued by exogenous vascular endothethial growth factor (VEGF). Similarly, deletion of Hif1a, encoding the hypoxia-responsive alpha subunit of HIF-1, reduced myeloid multilineage and committed erythroid progenitors in HIF-1α-deficient embryos [18]. Unlike the prior study, defects in erythroid colonies and decreased levels of hemoglobin in erythroid cells could not be rescued by addition of cytokines like VEGF and erythropoietin. Instead, erythropoiesis could be improved by addition of a synthetic iron chelate that alleviates defective iron utilization, suggesting that HIF-1 is an important regulator of iron homeostasis which is required for erythropoiesis. These studies raise the possibility that HIF-1 regulates a panoply of intracellular processes in hematopoiesis, including metabolic pathways.

HIF-1 is widely recognized for strong association with metabolic adaptations affecting glucose metabolism via glycolysis, the pentose phosphate pathway (PPP), and oxidative phosphorylation (OXPHOS). Interestingly, evidence shows that HIF-1α protein is stabilized by hypoxia, but only when glucose is available [19]. Accumulation of HIF-1α leads to increase in glucose-6-phosphate (G6P), a key substrate utilized in the PPP for synthesis of NADPH, ribonucleotides, deoxyribonucleotides, and amino acid precursors [20]. These molecules are the building blocks for cycling cells, and evidence in cancer cells suggests that HIF-1α plays a predominantly pro-proliferative role. Yet, the effect of HIF-1 on cell cycling in healthy cells is more complex, especially in non-hypoxic environments [21].

3. Metabolic Adaptation in Blood Precursors

Blood flow increases the availability of oxygen to the arterial vasculature and can also facilitate increased utilization of oxygen by cells lining the vessel walls. In our studies, we find that consumption of oxygen in the AGM region is elevated by flow and that the force generated by blood flow triggers this adaptation (Fig. 1a). Conversely, extracellular acidification is reduced in the presence of WSS (Fig. 1b). We speculate that initiation of blood circulation in the embryo not only increases supply of oxygen to tissues where hematopoiesis begins but also that signaling driving intracellular oxygen consumption is stimulated by the frictional force generated by blood flow. A study analyzing single-cell transcriptomics showed that cells undergoing EHT in Ncx1 heartbeat mutant embryos fail to undergo a metabolic switch from glycolysis to OXPHOS that is found in wild-type cells. The researchers then mimicked the effects of hypoxia via stabilization of HIF-1α by inhibiting prolyl hydroxylases with dimethyloxalylglycine (DMOG) in embryo explant cultures. Upon chemical stabilization of HIF-1α, which presumably elevated glycolysis, intraembryonic hematopoiesis was decreased [15]. Authors speculate that severe hypoxia associated with a lack of circulation in the Ncx1 mutants prevents this metabolic switch toward elevated OXPHOS and suggest that the HE to pro-HSC transition is the stage of EHT that is sensitive to effects of circulation on the path to HSC commitment.

Fig 1.

Oxygen consumption and extracellular acidification under wall shear stress (WSS) or static (no flow) conditions in embryonic day 11.5 (E11.5) aorta-gonad-mesonephros (AGM) cultures. Three replicates are shown with bars indicating the standard error of the mean at each time point. A Oxygen consumption was determined in real-time cell metabolic assays using the Seahorse Mito Stress test assay following 36-hour treatment of cells with one of two intensities of WSS (2 or 5 dyne cm⁻²) or static conditions. OCR: Oxygen consumption rate. B Concurrent measurement of acidification of media serves as an indicator of glycolytic activity in the Seahorse MitoStress test assay. ECAR: Extracellular acidification rate.

Switch in metabolism appears to be important for blood development based upon other studies in mice and humans. Indeed, one report shows that HE and other EHT populations have high basal respiration levels and utilize glutamine for the TCA cycle and mitochondrial respiration [22]. Respiration rate in EHT cells is elevated relative to HE cells, regardless of the presence or absence of glucose, supporting that OXPHOS increases during EHT and that glutamine is sufficient to fuel metabolism in EHT cell types. In fact, glutamine deprivation produced > 80% reduction in surface expression of the CD43 hematopoietic cell marker in differentiating HE cells, further supporting a potential role for mitochondrial metabolism in commitment to the hematopoietic fate. In another report by the same researchers, they observe that differentiation of HE toward primitive erythroid hematopoiesis is dependent on glycolysis in human induced pluripotent stem cell cultures [23]. In contrast, differentiation of HE toward definitive hematopoiesis is dependent upon OXPHOS-mediated cholesterol metabolism.

We recently showed that mitochondrial ultrastructure is important for blood development as well [24]. The mitochondrial permeability transition pore (mPTP) is a supramolecular channel that regulates exchange of solutes across the inner mitochondrial membrane. The inner mitochondrial membrane is highly impermeable but has been shown under specific conditions to lose barrier function, such as in hypoxia-induced ischemia, which can induce excess reactive oxygen species (ROS), or other conditions that drive a drop in pH. We have found that mPTP conductance shapes metabolic machinery as cells transit toward hematopoietic identity in the embryo. The pore demonstrates modest activity in early arterial endothelial (eAE) cells, followed by restriction of pore opening just prior to commitment to a hematopoietic fate (Fig. 2a). The transient closing of the mPTP appears to aid in stimulating biogenesis of cristae, or folds in the inner mitochondrial membrane [24]. Enforced closure of the mPTP by pharmacological or genetic means enhances hematopoiesis in vitro. HSCs also acquire an increased capacity to engraft, repopulate, and expand in the bone marrow of adult transplant recipient mice [24]. We also found that WSS applied to E10.5 AGM suppresses expression of one of the regulatory molecules of the mPTP, Cyclophilin D, encoded by the Ppif gene (Fig. 2b). By extension, one might predict that force could contribute to regulation of mPTP opening and could modulate some of the same biology impacted by the pore, including mitochondrial phenotypes.

Fig 2.

Mitochondrial permeability transition pore (mPTP) is regulated as cells commit to a hematopoietic fate and in response to force generated by blood flow. A Dye efflux from mitochondria via pore opening reveals high mPTP activity in E11.5 embryo pre-HSCs and low activity in mAE (One-way ANOVA, p<0.05). Low inverse mean fluorescence intensity (MFI) indicates more dye retention and less mPTP opening in mPTP MitoProbe assay. This suggests that pore conductance is suppressed as arterial endothelium matures and subsequently increased with acquisition of hematopoietic fate in hemogenic endothelium and pre-HSCs. eAE: early arterial endothelium; mAE: mature arterial endothelium; HE: hemogenic endothelium; Pro/pre-HSCI: pro-HSCs and pre-HSCs pooled; Pre-HSCII: pre-HSC II cells purported to precede definitive adult-type HSCs. B WSS of 5 dyne cm⁻² in E10.5 AGM cells produces a non-significant trend in suppression of Ppif expression with or without cyclic strain (stretching) and regardless of pulsatility of flow, suggesting that blood flow could participate in downregulation of mPTP activity vital for differentiation of eAE to mAE. Absolute values on the y-axis are from Illumina beadchip intensities after background correction and quantile normalization to housekeeping genes. One-way ANOVA, n.s. for 16-hour force type comparisons on left; T-test, n.s. for 36-hour static versus WSS comparison on right. Each condition represents three independent biological replicates. Bars indicate standard error of the mean.

Multiple aspects of blood flow, including increased oxygen availability and generation of WSS, contribute to a reprogramming of energy metabolism as cells commit to the hematopoietic fate (Fig. 3). An active area of research is centered on understanding how mechanotransduction regulates pathways that dictate how energetic and anabolic demands are met during EHT and whether signaling beyond the mPTP plays a role in that process. Future research should be aimed at understanding the effects of flow and the force it generates on metabolic adaptation and mitochondrial functions in blood development and disease.

Fig 3.

Blood flow regulates energy metabolism via cooperative signalling downstream of mechanotransduction and oxygen sensing in cells undergoing transition to the hematopoietic fate. As cells differentiate under the effects of blood flow, WSS and increased oxygen availability support a switch from glycolysis to OXPHOS that is essential for the acquisition of hematopoietic identity. Transient closing of the mPTP is important for the mitochondrial maturation necessary for hematopoietic stem cells, but a role for force in regulating that closure remains an active area of research.

4. Experimental Methods

4.1. Mouse Embryos and Culture of AGM

All animal experiments were approved under the Animal Welfare Committee protocol AWC-23-0111 by the UTHealth Houston Animal Welfare Committee. Copulation plugs were identified following pairing of C57BL/6 male and female pairs, defining day 0.5 of embryonic gestation. Embryos were collected on embryonic day E10.5 or E11.5 for microdissection of the AGM region, dissociation of cells, and culture under static or WSS conditions. Embryos of both sexes were analyzed in parallel and/or combined into single samples.

Embryonic tissue was dissociated to single-cell suspension with Accutase (STEMCELL Technologies) at 37°C with periodic agitation. Cells were resuspended in M5300 Myelocult medium (STEMCELL Technologies) enriched with 10 ng/ml murine SCF (Peprotech), 24 μM HEPES (Invitrogen), non-Essential Amino Acids (Gibco), Sodium Pyruvate (Gibco), and Penicillin-Streptomycin (Gibco). Cells were strained through a 70-micron filter and seeded into μ-Slide VI 0.4 channel slides (ibidi GmbH) at 400,000 cells cm⁻². Following cell attachment for 6 hours, cells were cultured under static conditions or in the presence of WSS (5 dyne cm⁻²). WSS was applied using a previously described microfluidic flow system for either 16 or 36 hours [25]. Cells were retrieved by dissociation with Accutase, or RNA was prepared by direct lysis on ibidi slides with Buffer RLT Plus (RNeasy Plus Micro, Qiagen) and processed for analysis of Ppif expression in Illumina BeadChip assays according to our prior report [6]. Briefly, transcriptome data was generated on Illumina MouseWG-6 v2.0 R2 expression beadchips, followed by background correction and quantile normalization to housekeeping genes with GenomeStudio (Illumina) Normalized values from the beadchips are reported as absolute values for Ppif expression.

4.2. Measurement of Metabolism by Real-Time Cell Metabolic Assay

The Seahorse XFe96 Analyzer was used to measure oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) as surrogates of oxidative phosphorylation and glycolysis, respectively, using 40,000 cells per well dissociated from the AGM exposed to static or WSS conditions. Chemical compounds injected were components of the Seahorse XFp Cell Mito Stress Test (Agilent), and included oligomycin, carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP), rotenone, and antimycin A.

4.3. Quantification of mPTP Opening

Following dissociation of E11.5 AGM cells to single-cell suspension, dead cells were labeled by incubation for 15 min in PBS with Zombie NIR viability dye (1:1000, Biolegend). Cells were subsequently incubated on ice for 30 min with antibodies against EHT surface markers EPCR/CD201 (EPCR-APC, 1:100, clone eBio1560, eBiosciences), cKit (cKit-PerCP-Cy5.5, 1:100, clone 2B8, Biolegend) VE-Cadherin (VEC-BV421, 1:100, clone 7B4, BD Biosciences), CD44 (CD44-PE, 1:100, clone IM7, Biolegend) and CD45 (CD45-BV510, 1:100, clone 30-F11, Biolegend). The opening of the mPTP was quantified by the MitoProbe Transition Pore Assay Kit (ThermoFisher). MitoProbe reagents were then applied according to manufacturer instructions and included labeling of cells at 37°C with 0.01 μM calcein AM, 0.4 mM CoCl₂, and 0.5 μM ionomycin reagents in calcium-containing Hank’s Buffered Saline Solution. Cells were analyzed immediately following loading of calcein dye by flow cytometry. Fluorescence of all events acquired was averaged using FlowJo version 10.10.0 software and inverse values were determined by subtraction from maximum fluorescence.

4.4. Statistical Analysis

Numerical data were plotted using GraphPad Prism 10 and analyzed by SigmaPlot 12.5 for statistical significance. Differences were determined by one-way ANOVA for multiple comparisons or the unpaired t-test for two groups. Significance was defined as p < 0.05 and n.s. indicates non-significant difference. Pooled data are reported as ± standard error of the mean (SEM) and individual biological replicates depicted as individual data points, when feasible.