Evidence of a pivotal role for the distal part of the complement cascade in the diurnal release of hematopoietic stem cells into peripheral blood

Abstract

We found that diurnal activation of the three evolutionarily ancient proteolytic cascades in peripheral blood (PB), namely, the complement, coagulation, and fibrinolytic cascades, late at night or in the early morning hours, precedes the diurnal release of hematopoietic stem/progenitor cells (HSPCs) from bone marrow (BM) into PB in wild type mice. Moreover, activation of the distal part of the complement cascade (ComC), involving cleavage of the fifth component (C5), seems to play a crucial role in pharmacological mobilization of HSPCs. In order to shed more light on the role of diurnal rhythms in the egress of HSPCs, we studied diurnal changes in the number of circulating HSPCs in C5-deficient mice and did not observe diurnal changes in the number of these cells circulating in PB in C5^−/− animals. Based on this finding, we conclude that activation of the distal part of the ComC, C5 cleavage, and release of C5a and _desArgC5a, are required in executing the diurnal release of HSPCs from BM into PB. Moreover, the fact that C5^−/− mice still displayed normal activation of the coagulation and fibrinolitic cascades indicates that, of all the proteolytic cascades, the ComC is the dominant player regulating diurnal egress of HSPCs.

Introduction

A diurnal rhythm, according to its definition, is any biological process characterized by an endogenous oscillation with a period of about 24 hours (6), and an intrinsic circadian “clock” regulates all aspects of diurnal behavior and physiology (6). These 24-hour rhythms are seen in sleeping animals and are driven in mammals by diurnal fluctuations, e.g., in melatonin secretion by the pineal gland, core body temperature, and the plasma level of cortisol (6,11,14,21). In parallel, there is recent, somewhat underappreciated, evidence that three evolutionarily ancient serum proteolytic cascades, the complement cascade (ComC), the coagulation cascade (CoaC), and the fibrinolytic cascade (FibC), are involved in regulating certain aspects of circadian activities (19,24). Activation of these cross-reacting cascades is triggered in a circadian manner during deep sleep hypoxia in the late-night hours (10,19,24).

The number of circulating HSPCs in PB follows a circadian rhythm, with the peak occurring in the early morning hours and the nadir at night, and the timing of this peak has been attributed to enhanced tonus or fluctuation in tonus of the vegetative nervous system (10). In support of such a role for the vegetative nervous system, it has been shown that UDP-galactose:ceramide galactosyltransferase-deficient mice, which exhibit aberrant nerve conduction and do not release norepinephrine (NE) into the BM microenvironment, do not mobilize HSPCs (13). However, in contrast to mice, modification of the sympathetic output in humans, as seen in normal human HSPC volunteer donors receiving NE reuptake inhibitors (NRI) for depression or β2-blockers for hypertension, does not affect mobilization (4). Specifically, mobilization in these patients was neither enhanced by NRI administration nor suppressed by β2-blockers, as one would expect based on the murine results (20), which suggests differences between species or the involvement of other circadian mechanisms.

We recently reported that activation of the ComC, CoaC, and FibC is essential for release of hematopoietic stem progenitor cells (HSPCs) from bone marrow (BM) into peripheral blood (PB) during stress- or pharmacology-induced mobilization (5). Moreover, since all these proteolytic cascades show circadian activation at late night or in the early morning hours due to deep sleep hypoxia (19,24), we became interested in their role in diurnal release of HSPCs into PB. In particular, we became interested in activation of the distal part of the ComC, because the C5 cleavage fragments (C5a and _desArgC5a) seem to be important executors of the release of HSPCs from BM into the circulation (12,15,18). Therefore, we focused on the circadian oscillation in the number of circulating HSPCs in C5-deficient (C5^−/−) mice, which, unlike their wild type (WT) littermates, do not activate the distal part of the ComC.

Our study confirms diurnal activation of the ComC, CoaC, and FibC in WT animals at late night or in the early morning hours preceding the release of HSPCs from BM into PB. The fact that we did not observe diurnal changes in the number of circulating cells in PB in C5^−/− mice confirms the pivotal role of the ComC in executing diurnal release of HSPCs from BM into PB and that activation of only the CoaC and FibC, which are normal in C5^−/− mice, is not sufficient to maintain diurnal release of HSPCs into PB.

Materials and Methods

Animals

Experiments were performed on two strains of mice: C57BL/6 (WT) and complement cascade protein 5 (C5^−/−)-deficient male mice (Jackson Laboratory, Bar Harbor, ME, USA). Mice were accustomed to alternating periods of 12 hours light and 12 hours darkness. Light was turned on at 6 AM (ZT0), and the number of circulating white blood cells (WBC), Sca-1⁺c-Kit⁺Lin⁺ (SKL cells), Sca-1⁺Lin⁻CD45⁺ HSCs, clonogenic CFU-GM progenitors, non-hematopoietic Sca-1⁺Lin⁻CD45⁻ very small embryonic-like stems cells (VSELs), CD45⁻CD31⁻CD44⁺CD51⁺ mesenchymal stromal cells (MSCs) and CD45⁻CD51⁻CD44⁻CD31⁺ endothelial progenitors (EPCs) were measured at 7 AM (ZT1), 11 AM (ZT5), 7 PM (ZT13), and 3 AM (ZT21). At the same time points, we evaluated activation of the ComC (by C5a ELISA), the CoaC (by thrombin/antithrombin ELISA), and the FibC (by plasmin/antiplasmin complex ELISA). Animal studies were approved by the Animal Care and Use Committee of the University of Louisville (Louisville, KY, USA).

Peripheral blood parameters

To obtain leukocyte counts, blood samples were collected from the retro-orbital plexus of mice into microvette EDTA-coated tubes (Sarstedt Inc., Newton, NC, USA) and run within 3 h of collection on a HemaVet 950 hematology analyzer (Drew Scientific Inc., Oxford, CT, USA; http://www.drew-scientific.com). Additional plasma was collected for ELISAs. For CFU-GM, SKL, HSCs, VSELs, MSCs and EPCs analysis, blood was collected from the vena cava.

Enumeration of the number of colony-forming unit-granulocyte/macrophages (CFU-GMs) mobilized into PB

After PB red blood cell lysis (BD Pharm Lyse Buffer, San Jose, CA, USA), nucleated cells were washed and counted, and 1 × 10⁶ cells were resuspended in 30% RPMI-1640 culture medium (Corning Co, Corning, NY, USA) with 70% human methylcellulose base medium (R&D Inc, Minneapolis, MN, USA) supplemented with 25 ng/ml recombinant murine GM-CSF and 10 ng/ml recombinant murine IL-3 (PeproTech, Rocky Hill, NJ, USA). After 5 days of culture, the numbers of CFU-GM colonies were scored using an inverted microscope (Olympus, Center Valley, PA, USA). For evaluation of circulating CFU-GM the following formulas were used: (number of white blood cells (WBCs) × number of CFU-GM colonies)/number of WBCs plated = number of CFU-GM per μl of PB.

FACS analysis of circulating cells

After PB red blood cell lysis, the following monoclonal antibodies were employed to stain Sca-1⁺c-Kit⁺Lin⁺ (SKL cells), Sca-1⁺Lin⁻CD45⁺ HSCs, Sca-1⁺Lin⁻CD45⁻ VSELs, CD45⁻CD31⁻CD44⁺CD51⁺ MSCs and CD45⁻CD51⁻CD44⁻CD31⁺ EPCs: lineage markers - anti-mouse CD45R/B220–PE (clone RA3–6B2), anti-mouse TCRβ–PE (clone H57–597), anti-mouse γδ T Cell–PE (clone GL3), anti-mouse CD11b–PE (clone M1/70), anti-mouse Ter119–PE (clone TER-119) and anti-mouse Gr-1–PE (clone RB6–8C5); anti-mouse Ly-6A/E-Biotin (clone E13–161.7) + Streptavidin-PE-Cy5; anti-mouse CD45-APC-Cy7 (clone 30-F11); anti-mouse CD117-FITC (clone 2B8); anti-mouse CD44-APC-Cy7 (clone IM7); anti-mouse CD45-V450 (clone 30-F11); anti-mouse CD31-APC (clone 390); anti-mouse CD51-Biotin (clone RMV-7) + Streptavidin-FITC, as described. All monoclonal antibodies were added at saturating concentrations (0.5–2.0μg/100μl), and the cells were then incubated for 30 min on ice, washed twice, resuspended in RPMI-1640 + 2% fetal bovine serum, and analyzed with an LSR II flow cytometer (BD, San Diego, CA, USA). For evaluation of circulating SKL cells, HSCs, VSELs, MSCs and EPCs the following formulas were used: (number of WBCs × number of mobilized cells)/number of gated WBCs = number of mobilized cells per μl of PB.

Plasma concentration of C5a

The concentration of C5a was measured by employing the commercially available, highly sensitive enzyme-linked immunosorbent assay (ELISA) kit K-ASSAY (Kamiya Biomedical Company, Seattle, WA, USA), according to the manufacturer’s protocol. For analysis, PB was obtained by retro-orbital plexus bleeding into cold microvette EDTA-coated tubes (Sarstedt Inc.). Subsequently, blood was centrifuged at 2000 x g for 20 min at 4 °C to obtain plasma.

Activation of the coagulation cascade and the fibrinolysis cascade

Thrombin/antithrombin (TAT) and plasmin/antiplasmin (PAP) complexes were measured by employing ELISA assays, according to the manufacturer’s protocols (USCN Life Science, Wuhan, China). For this measurement, we used plasma collected from the retro-orbital plexus to avoid hemolysis.

Statistical analysis

Arithmetic means and s.d. were calculated using Excel. Statistical significance was defined as P<0.05. Data were analyzed using Student’s t-test for unpaired samples (Excel).

Results

Diurnal activation of the ComC, CoaC, and FibC in C5-deficient and WT mice

Based on the observation that the ComC is activated in a circadian rhythm-dependent manner in mice (5,19,24), we became interested in activation of all three proteolytic cascades in C5^−/− and WT littermates (Figure 1). As expected, we found that the ComC became gradually activated, as measured by the plasma level of the C5 cleavage fragment (C5a) in WT animals from late night through the morning hours (ZT21–ZT5, Figure 1A). For obvious reasons, C5a was not detected in C5^−/− mice (data not shown).

Figure 1. Diurnal activation of the CoaC, FibC, and ComC in WT and C5^−/− mice.

Panel A. Diurnal activation of the ComC in WT mice measured by ELISA to detect C5a concentration in plasma, with the peak occurring between ZT1 (7 AM) and ZT5 (11 AM). Data are pooled from two independent experiments (n=6 mice each). *P<0.05. Panel B. Diurnal activation of the CoaC in WT and C5^−/− mice measured by ELISA to detect TAT concentration in plasma, with the peak at ZT21 (3 PM). Data are pooled from two independent experiments (n=6 mice each). *P<0.05, **P<0.01. Panel C. Diurnal activation of the FibC in WT and C5^−/−mice measured by ELISA to detect PAP concentration in plasma, with the peak between ZT21 (3 PM) and ZT1 (7 AM). Data are pooled from two independent experiments (n=6 mice each). *P<0.05, **P<0.01. Please note that all parameters studied in this work were measured at set ZT points only.

By contrast, activation of the CoaC and FibC, as measured by the plasma level of TAT and PAP, respectively, occurred at late night (ZT21) and somewhat preceded the peak of activation of the ComC in WT mice (Figure 1B, C). This finding tends to support the involvement of crosstalk between the CoaC and FibC in ComC activation (5). Figure 1B, C also shows a similar pattern for the CoaC and FibC in WT as in C5^−/− mice.

Diurnal changes in the number of circulating HSPCs in WT mice and the lack of a similar pattern in C5^−/− animals

To address the role of the activation of the distal part of the ComC and C5a in the diurnal release of HSPCs, we analyzed circadian changes in WBCs, Sca-1⁺Lin⁻CD45⁺ HSCs, and Sca-1⁺c-Kit⁺Lin⁻ (SKL) cell numbers as well as in the numbers of circulating clonogenic CFU-GMs in WT (Figure 2 panel A) and C5^−/− mice (Figure 2 panel B). We found that WT mice displayed a circadian pattern of cell numbers circulating in PB, with the peak occurring in the late morning hours (Figure 2 panel A), which correlated with the highest level of circulating C5a (Figure 1A). In contrast to WT mice, C5-deficient mice did not show any significant circadian changes in the number of circulating HSPCs (Figure 2 panel B).

Figure 2. Diurnal changes in the number of circulating WBCs, HSCs, SKL cells, and CFU-GMs in WT mice and C5^−/− mice.

Panel A - The highest number of cells circulating in PB was observed in WT mice at ZT5 (11 AM). Data are pooled from two independent experiments (n=6 mice each). *P<0.05. Panel B - Lack of diurnal changes in the number of circulating WBCs, HSCs, SKL cells, and CFU-GMs in C5^−/− mice. Data are pooled from two independent experiments (n=6 mice each). *P<0.05. Please note that all parameters studied in this work were measured at set ZT points only.

Lack of diurnal changes in the circulation of non-hematopoietic stem cells

Finally, we became interested in whether circulation of other types of stem cells, such as mesenchymal stem cells (MSCs), endothelial progenitors (EPCs), and very small embryonic-like stem cells (VSELs), are affected by a diurnal rhythm (Figure 3). In contrast to the significant changes in the number of circulating HSPCs observed in WT animals, we did not observe any changes in the number of circulating MSCs, EPCs, and VSELs in these mice (Figure 3A). Also, as expected, no changes were observed in C5^−/− mice (Figure 3B).

Figure 3. Lack of diurnal changes in the number of circulating MSCs, EPCs, and VSELs in WT and C5^−/− mice.

Panel A. Data are pooled from two independent experiments in WT animals (n=6 mice each). Panel B. Data are pooled from two independent experiments in C5^−/− animals (n=6 mice each).

Discussion

The salient observation of this study is that activation of the distal part of the ComC is an important executor of the diurnal release of stem cells from BM into PB, as C5-deficient mice do not show significant circadian changes in the number of circulating HSPCs in PB.

It has been proposed that circadian changes in biological systems are regulated by diurnal oscillations in the levels of certain hormones (2,6,7,9,11,14,21). It is well known, for example, that an intrinsic circadian rhythm plays an important role in regulating the biology of stem cells (6,11,14,21). Specifically, a circadian rhythm affects their proliferation and thereby regulates the steady-state number of HSPCs circulating in PB, with the peak in the early morning hours and the nadir at night (6,10,11,14,21). In our study we confirmed diurnal changes in the level of circulating HSPCs in WT animals. However, we did not observe significant changes in the level of other types of circulating stem cells, including MSCs, EPCs, and VSELs, in our animals.

Since the level of catecholamines increases in PB in the morning hours (21), it has been proposed that circadian changes in their levels play an important role in determining circadian changes in the number of circulating HSPCs. However, while catecholamines are involved in sympathoadrenomedullary regulation of cardiovascular, respiratory, and metabolic functions, extensive study in humans has demonstrated that the morning increase in the level of catecholamines is mainly in response to changes in activity and posture rather than by an endogenous circadian surge of plasma catecholamines (21). This finding may explain some discrepancies in the literature between the apparent role of catecholamines in the release of HSPCs in humans (20) versus mice (10). In support of a different role of catecholamines in human, sympathetic output in normal human HSPC volunteer donors receiving noradrenaline (NE) reuptake inhibitors (NRI) for depression or β2-blockers for hypertension induces mobilization in a similar manner as in normal controls (4) and was neither enhanced by NRI administration nor suppressed by β2-blockers (4). The potential role of catecholamines in regulating circadian release of HSPCs from BM into PB has been additionally challenged by a recent report (20), where an infusion of beta receptor agonists into normal volunteers increased the number of circulating leucocytes but did not affect the number of circulating HSPCs (20).

Therefore, it is possible that other more potent regulators of diurnal rhythms could be involved in this process, and we became interested in circadian activation of the three evolutionarily ancient proteolytic cascades, the ComC, CoaC, and FibC. Of note, all three cascades show crosstalk, as several proteolytic enzymes generated during their activation may enzymatically activate proteins in other cascades (1). For example, as has been convincingly demonstrated, thrombin shows C5 convertase activity (1). We have also demonstrated that crosstalk between the CoaC and the ComC due to the C5 convertase activity of thrombin contributes to mobilization of HSPCs (5).

As mentioned above, all these cascades become activated in PB during deep sleep-related hypoxia that triggers lactic acidosis and free radical release (19,24). The circadian activation of these cascades is supported by the observation that hemolytic crises observed in paroxysmal nocturnal hemoglobinuria (PNH) patients occur at night and the early morning hours (by circadian activation of the ComC) (16) as well as by the fact that the highest incidence of stroke is in the early morning hours (by circadian activation of the CoaC) (24). Activation of the ComC, as we have recently reported, also plays a crucial role in PNH patients in defective BM-retention and egress of HSPCs into PB (17).

The presence of circadian rhythm activation of the ComC and CoaC has already been demonstrated in mice (3,24), and here we confirmed this phenomenon (Figure 1A, B). In addition, we show for the first time that FibC is also activated in these animals in a diurnal manner (Figure 1C)—most likely as a response to activation of the CoaC and ComC.

In the past we have demonstrated that activation of the distal part of the ComC is crucial to executing egress of HSPCs from BM into PB, as mice deficient in C5 are poor mobilizers (12). It is known that the C5 cleavage fragment anaphylatoxins C5a and _desArgC5a are the most potent activators of granulocytes and monocytes, which upon activation in the BM microenvironment secrete several proteolytic and lipolytic enzymes involved in attenuation of BM-retention mechanisms of HSPCs in BM niches (8,16). Furthermore, in the chemotactic response to an increase in C5a and _desArgC5a in BM sinusoids, granulocytes are the first cells to leave the BM and pave the way for HSPCs to egress through the endothelial barrier in the BM sinusoids (12,18). Here, we show that C5^−/− mice do not show diurnal changes in the number of HSPCs circulating in PB, which lends additional support to a role for C5 cleavage fragments in the egress from BM into PB. Furthermore, since C5-deficient mice are still able to activate the CoaC and FibC, our results indicate that the ComC seems to play the most important role, and lack of activation of the distal part of the ComC cannot be compensated by activation of the other proteolytic cascades. Thus, even if, as postulated, the CoaC and FibC (22,23) play a role in mobilization of HSPCs, they need activation of the ComC to execute this process (Figure 4).

Figure 4. Diurnal changes in number of circulating in peripheral blood HSPCs as a function of diurnal activation of ComC, CoaC and FibC.

It is well known that deep sleep hypoxia (ZT21) activates all these ancient proteolytic cascades due to lactic acidosis and release of free radicals. Furthermore, ComC ma by additionally activated by products of activated CoaC and FibC such as for example thrombin that possesses C5 convertase activity and plasmin, respectively. This leads to increase of ComC activation in early morning hours (ZT1 – ZT5) and diurnal egress of HSPCs from BM into PB.

In this study we did not measure diurnal changes in the level of hormones, and we are aware that further studies are needed to see whether activation of the ComC affects hormonal mediators, such as corticosteroids (11) or melatonin (2), that are involved in circadian changes in the proliferation of HSPCs.

In conclusion, all ancient proteolytic cascades show diurnal changes in WT mice, and studies in C5-deficient mice have demonstrated that, despite diurnal changes in activation of the CoaC and FibC, diurnal activation of the distal part of the ComC plays a crucial role in diurnal egress of HSPCs from BM and cannot be replaced by activation of the other proteolytic cascades.