Alcohol Impairs the Myeloid Proliferative Response to Bacteremia in Mice via Inhibiting the Stem Cell Antigen-1 - Extracellular-Regulated Kinase Pathway¹

Abstract

Enhancement of stem cell antigen-1 (Sca-1) expression by myeloid precursors promotes the granulopoietic response to bacterial infection. However, the underlying mechanisms remain unclear. Extracellular-regulated kinase (ERK) pathway activation strongly enhances proliferation of hematopoietic progenitor cells. Here, we investigated the role of Sca-1 in promoting ERK-dependent myeloid lineage proliferation and the effects of alcohol on this process. Thirty minutes after intraperitoneal injection of alcohol, mice received intravenous challenge with 5 × 10⁷ E.coli for 8 or 24 h. A subset of mice received intravenous BrdU injection 20 h post challenge. Bacteremia increased Sca-1 expression, ERK activation, and proliferation of myeloid and granulopoietic precursors. Alcohol administration suppressed this response and impaired granulocyte production. Sca-1 expression positively correlated with ERK activation and cell cycling, but negatively correlated with myeloperoxidase content in granulopoietic precursors. Alcohol intoxication suppressed ERK activation in granulopoietic precursors and proliferation of these cells during bacteremia. Granulopoietic precursors in Sca-1^−/− mice failed to activate ERK signaling and could not increase CFU-GM activity following bacteremia. These data indicate that Sca-1 expression promotes ERK-dependent myeloid cell proliferation during bacteremia. Suppression of this response could represent an underlying mechanism for developing myelosuppression in alcohol abusing hosts with severe bacterial infection.

Introduction

Polymorphonuclear leukocytes (PMNs, referring specifically to neutrophilic granulocytes or neutrophils) represent the largest population of phagocytes in the circulation. They constantly patrol the bloodstream and can quickly migrate to tissue sites of infection to exert microbicidal activity. Through the course of bacterial infection, mediators generated from inflammatory tissues and substances derived from invading microbes, stimulate bone marrow granulopoietic activity by enhancing granulopoietic precursor cell proliferation, accelerating granulocyte maturation, and increasing the release of these phagocytes into the bloodstream (1, 2). Multiple humoral factors including granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-1 (IL-1), IL-6, and androgens have been reported to play different roles in the regulation of granulopoiesis and the granulopoietic response (3–7). Hematopoietic stem cells and progenitors at different stages of differentiation express Toll-like receptors (TLRs) and their co-receptors. TLR4, as well as co-receptors MD-2 and CD14, are required for the recognition of lipopolysaccharide (LPS) from Gram-negative bacteria (8). Recent studies from our group have revealed that in response to Escherichia coli infection or LPS stimulation, primitive hematopoietic precursors including hematopoietic stem cells upregulate their surface expression of stem cell antigen-1 (Sca-1, an 18 kDA glycophosphatidylinositol-anchored cell surface protein associated with cell cycle activation) (9–11). This enhancement of Sca-1 expression is associated with increases in hematopoietic precursor cell proliferation and their commitment towards granulocyte lineage development during bacterial infection (10–13).

It has long been recognized that excessive alcohol consumption injures bone marrow granulopoietic function (14–17). Such compromised innate immunity predisposes alcohol abusers to the development of severe bacterial infections (14, 18). Alcohol abusers with severe infections, particularly sepsis, frequently present with granulocytopenia which is an indicator of increased mortality (19). Examination of the bone marrow from alcoholic patients has shown vacuolated granulopoietic progenitors with a significantly reduced number of mature granulocytes (16, 20, 21). Yet, mechanisms underlying alcohol-induced impairment of marrow granulopoietic activity have remained unclear.

Activation of the extracellular-regulated kinase (ERK) pathway following ligand engagement of TLR4 has been known to provide a strong signal mediating the proliferative response of hematopoietic precursors (8, 22). Our current investigation focused on the role of enhanced Sca-1 expression in facilitating ERK signaling, as well as the associated myeloid and granulopoietic precursor cell proliferation in response to bacteremia. The results indicate that upregulation of Sca-1 expression promotes ERK-dependent myeloid cell proliferation following E. coli bacteremia. Alcohol intoxication suppresses this response which may represent an underlying mechanism for developing granulocytopenia in alcohol abusing hosts with severe bacterial infection.

Materials and Methods

Animals

Male Balb/c mice (7–10 weeks old; Charles River Laboratories, Wilmington, MA) with a body weight of 21.87 ± 2.10 g were housed in a specific pathogen free facility with a 12 h light/dark cycle. Acute alcohol binge was administered by intraperitoneal (i.p.) injection of 20% ethanol (EtOH, 5g/kg) in saline. Blood alcohol levels achieved with this model are in the ranges of 106.3 to 132.8, 87.7 to 122.4, and 48.4 to 61.4 mM, respectively, at 90 min, 3 h, and 6 h post alcohol administration, as previously reported (11, 23). The alcohol dose used in our experiments was selected to model the kinetic changes in blood alcohol levels frequently observed in clinical patients with acute alcohol intoxication. Mice commonly awakened from the intoxication within 6 h post alcohol administration. Thirty minutes after i.p. injection with EtOH or saline, mice received intravenous (i.v.) challenge with 5 × 10⁷ Escherichia coli (E. coli; E11775 from the American Type Culture Collection, Manassas, VA; in 100μL saline/mouse) by penile vein injection under isoflurane anesthesia. Control mice received equal volume saline. Animals were sacrificed 8 h or 24 h after E.coli challenge. During the 24 h challenge, a cohort of animals received i.v. BrdU (1mg in 100uL of PBS/mouse; BD Biosciences, San Jose, CA) 4 h prior to sacrifice. Sca-1^−/− mice, considered congenic on the C57BL/6 background, were bred under specific pathogen free conditions in the Animal Care Facility of Louisiana State University Health Sciences Center (24)' (25). Studies on Sca-1^−/− animals were conducted using male, 7–10 week old mice with age and gender matched C57BL/6 controls (Charles River Laboratories, Wilmington, MA). Sca-1^−/− mice and the matched control C57BL/6 mice received i.v. challenge with 10⁸ E.coli or saline. Following sacrifice, heparinized blood was obtained by cardiac puncture. Total and differential white blood cell counts were performed using Wright-Geimsa stains. All experiments received prior approval from the Institutional Animal Care and Use Committee of Louisiana State University Health Sciences Center and were performed in adherence with National Institute of Health guidelines.

Flow cytometry

Phospho-specific flow cytometry enables the study of intracellular signaling events at the single cell level. To investigate intracellular signaling mechanisms, we developed a phospho-specific flow cytometry protocol for analyzing intracellular ERK signaling in distinct stages of myeloid and granulocyte lineage development. After 8 h E.coli challenge, bone marrow was collected and red blood cells in the bone marrow sample were lysed with Purescript RBC lysis solution (Qiagen, Valencia, CA) as previously described (10). Thereafter, nucleated bone marrow cells were fixed and permeabilized according to the previously reported protocol, with some modifications (26). Nucleated cells were washed in 1× PBS after RBC lysis. Cells were fixed immediately with 1.5% paraformaldehyde for 10 min. Fixed cells were washed twice with staining buffer (PBS + 0.5% BSA + 0.1% NaN₃). A mixed panel of biotinylated anti-mouse lineage markers [CD3ε (clone 145-2C11), CD45/B220 (clone RA3-6B2), CD11b/CD18 (Mac-1; clone M1/70), Gr1 (Ly6G/Ly6C, clone RB6-8C5), or TER119] or isotype control antibodies (clones A19–3, R35–95, A95–1) (BD Biosciences) were then incubated for 20 min at room temperature. PE-conjugated streptavidin (BD Biosciences; 10 ug/mL) and anti-Sca-1 (Ly6A/E, clone D7, BD Biosciences; 10 ug/mL) were then added to the incubation. The cell mixture was incubated another 20 min. Cells were washed once with staining buffer. Fixed cells were then permeabilized with 100% ice cold acetone for 10 min at 4°C. Acetone was chosen as a permeabilizing agent in order to preserve Sca-1 antigenicity (26). Following permeabilization, cells were washed twice with staining buffer. Fixed and permeabilized samples were incubated with fluorochrome conjugated anti-ckit (CD117, clone 2B8, BD Biosciences; 10 ug/mL), anti-Gr1 (granulocyte differentiation antigen-1 or Ly6G/Ly6C, clone RB6-8CF, BD Biosciences; 10 ug/mL), and anti-Phospho-p44/42 (E10; Thr202/Tyr204; Cell Signaling Technology, Danvers, MA; 10 ug/mL) 20 min at room temperature. Cells were then washed and re-suspended in staining buffer for FACS analysis. Phospho-ERK expression was quantified by integrated mean fluorescence intensity (iMFI, representing the percent positive cells multiplied by their mean channel fluorescence) (27). Preliminary results showed that fixation and permeabilization of murine bone marrow cells produced differential effects on cell surface staining (Supplemental Figure 1). Therefore, the sequence of fixation, permeabilization, and staining for each myeloid and granulocyte population was conducted to most reflect the immunostaining of live cells.

Myeloid progenitors were identified by the lineage (lin)⁻ckit⁺Sca-1⁻ (LKS⁻) surface phenotype (28). Granulocyte lineage cells were studied using the expression of Gr1. Immature granulopoietic precursors express low levels of Gr1 antigen (Gr1^lo cell; granulopoietic precursor). As these precursors terminally differentiate, Gr1 surface expression increases with mature granulocytes expressing the highest level of Gr1 (Gr1^hi cell; mature granulocyte). (29, 30)

Cell cycle activity was determined by measuring the DNA content of granulopoietic precursors following 24 h E.coli challenge using the DNA QC Particles kit (BD Biosciences). After surface staining, Gr1^loSca-1⁺ (Sca-1⁺granulopoietic precursors) and Gr1^loSca-1⁻ cells (Sca-1⁻ granulopoietic precursors) were sorted directly into DNA QC buffer. DNA content of sorted cells was then assessed by measuring the propidium iodide content of each cell type on a BD FACSAria. Cell cycle analysis was performed using ModFit LT software (Verity Software House, Topsham, ME, USA).

To determine myeloperoxidase (MPO) expression, bone marrow samples were prepared as described above. After RBC lysis, nucleated marrow cells were suspended in 1× PBS. Antibodies against surface antigens Sca-1 and Gr1 were added into the cell suspension. After 20 min incubation at room temperature, cells were washed with 1× PBS and permeabilized with 1× FACS Permeabilization Solution 2 (BD Biosciences) for 10 min. Following permeabilization, samples were washed with staining buffer, and anti-MPO (Abcam, Cambridge, MA; 10 ug/mL) was added and incubated for 30 min. Cells were then washed with staining buffer, and PE-conjugated anti-mouse IgG1 secondary antibody (BD Biosciences; 10 ug/mL) was added and incubated for 30 min. After incubation, samples were washed with staining buffer and fixed with 1% paraformaldehyde in 1× PBS in readiness for FACS. MPO content was quantified by iMFI.

Western Blot analysis

Bone marrow was collected following an 8 h challenge with 5 × 10⁷ E.coli in the presence and absence of acute alcohol administration. Protein was extracted with a lysis buffer (10mM Tris-HCl, 1% Triton X-100, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 2mM sodium orthovanadate [Na3VO4], 1mM phenylmethyulsulfonyl fluoride, 50mM sodium fluoride, 5 ug/mL aprotinin, 5 ug/mL pepstatin, and 5 ug/mL leupeptin, pH 7.6). Protein concentration was determined by BCA protein assay kit (Thermo Fisher Scientific, Rockford, IL). Twenty micrograms of protein sample was resolved using 8% SDS-PAGE ready gel (Bio-Rad Laboratories, Hercules, CA) and transferred to a PVDF membrane (Bio-Rad Laboratories). Phosphorylation of ERK1/2 was detected by incubating the PVDF membrane with anti-Phosphop44/42 antibody (phospho-p44/42 MAPK; Thr202/Tyr204; Cell Signaling Technology) and then horse radish peroxidase-conjugated goat anti-rabbit IgG (1:1000 with blocking buffer) secondary antibody. Determination of bound antibody was conducted using the ECL Western Blotting Detection System (GE Healthcare, Piscataway, NJ) and imaged using the KODAK Gel Logic 2200 Imaging System (Carestream Molecular Imaging, Rochester, NY) with KODAK Molecular Imaging Software.

CFU-GM assay

Following 24 h E.coli challenge, sorted lin⁻ckit⁺ (enriched myeloid progenitors with a small fraction of hematopoietic stem cells; myeloid precursors) and granulopoietic precursors from wild type and Sca-1^−/− animals were cultured in Methocult 3534 media (StemCell Technologies, Vancouver, BC, Canada). One milliliter of medium, containing either 100 enriched myeloid precursors or 2000 granulopoietic precursors, was plated on to a 35-mm Nunclon dish (Nunc, Rochester, NY). Each sample was cultured in triplicate for 7 days at 37°C in an atmosphere of 5% CO₂. Colonies containing ≥ 50 cells were enumerated.

Statistical Analysis

Data are presented as mean ± SEM. The sample size is indicated in each figure legend. Statistical analysis was conducted using unpaired Student's t-test (for comparison between two groups) and one-way ANOVA, followed by Student-Newman-Keuls test. Asterisk and bars with different letters in each panel are statistically different (p≤0.05). Regression analysis was performed with the Microsoft Excel Analysis ToolPak (Microsoft, Redmond, WA).

Results

Granulopoietic Response to Bacteremia

We first sought to determine how the bone marrow supported increased granulocyte turnover during bacteremia. Twenty four hours post E.coli challenge, the number of mature granulocytes was significantly reduced in the bone marrow with a concomitant increase in the immature granulopoietic precursor population (Figure 1A). These alterations in the marrow granulocyte population likely represent mobilization of the bone marrow reserve of mature granulocytes into the circulation and subsequent expansion of marrow granulopoietic precursors in response to bacterial infection (31, 32). Alcohol administration further decreased the number of mature granulocytes but did not alter the increase in marrow granulopoietic precursor fraction during bacteremia (Figure 1A). E.coli challenge significantly increased the proliferation and expansion of Sca-1⁺granulopoietic precursors (Figure 1B and C). Although, the number of Gr1^hiSca-1⁺BrdU⁺ cells (BrdU⁺Sca-1⁺mature granulocytes) similarly increased, the total number of Gr1^hiSca-1⁺ cells (Sca-1⁺mature granulocytes) was reduced during bacteremia (Figure 1B and C). Alcohol treatment suppressed the expansion and proliferation of marrow Sca-1⁺granulopoietic precursors. The total number and proliferation rate of marrow Sca-1⁺mature granulocytes were also reduced by alcohol during bacteremia.

Figure 1.

Granulopoietic response to bacteremia. (A) Changes in the marrow granulopoietic precursor and mature granulocyte cell populations during 24 h E.coli challenge. (B) Number of marrow Sca-1⁺granulopoietic precursor cells and Sca-1⁺mature granulocytes during bacteremia. (C) Proliferative response of marrow Sca-1⁺granulopoietic precursor cells and Sca-1⁺mature granulocytes. (D) Total granulocyte compartment (Gr1^lo+Gr1^hi), and the mature granulocyte/granulopoietic precursor represents the percentage of mature granulocytes to granulopoietic precursors contributing to granulocyte lineage proliferation. This ratio can be an indicator of the rate of granulopoietic precursor differentiation into a mature granulocyte. N=5. Bars with different letters are statistically different (p≤0.05). BMCs: bone marrow cells.

In saline-treated control mice, proliferative granulocyte lineage committed cells (BrdU⁺Gr1⁺; Table 1) accounted for approximately 18% of all cellular proliferation in the bone marrow. The proliferative activity of marrow B cells (CD19⁺), monocytes (F4–80⁺), or erythroid precursors (TER119⁺) was similar to that of marrow granulocytes (Gr1⁺). Multipotent precursors (lin⁻) and T cells (CD3⁺) contributed a lesser percent (Table 1). Alcohol administration alone decreased the ratio of proliferating mature granulocyte to granulopoietic precursor cells in the bone marrow. Both granulopoietic precursors and mature granulocytes had incorporated BrdU during the last 4 h of E.coli challenge. Since mature granulocytes are post mitotic and arise from granulopoietic precursor maturation, Gr1^hiBrdU⁺ cells (BrdU⁺mature granulocytes) may represent newly produced neutrophils from granulopoietic precursors since the time of BrdU injection (29, 30, 33). Therefore within this BrdU⁺ population, the mature granulocyte: granulopoietic precursor may serve as an indicator for the activity of granulocyte maturation in the bone marrow. After 24 h E.coli challenge, enhanced marrow Sca-1 expression corresponded with an increase in marrow granulocyte production (Table 1). Greater granulocyte production resulted from an increase in the proliferation of marrow hematopoietic (lin⁻) and granulopoietic precursors during bacteremia. Expanded production of granulocytes during bacteremia also corresponded with decreased B cell and monocyte proliferation (Table 1). Production of marrow T cells and erythroid precursors remained relatively unchanged. Alcohol administration suppressed marrow granulocyte production and reduced the BrdU⁺mature granulocyte: granulopoietic precursor ratio in the bone marrow during bacteremia (Figure 1D).

Table 1.

Table 1A Constitution of Sca-1⁺BrdU⁺ cells in the bone marrow following 24 h E.coli challenge. Percentage Sca-1⁺ is shown per BrdU⁺ cell population. Table 1B. Constitution of BrdU⁺ cells in the bone marrow.

Table 1a: Constitution of Sca-1+ BrdU+ cells in the bone marrow following 24 h E.coli challenge – % Sca-1+ (SEM) of each population
Cell Type	Saline	EtOH	E. coli	EtOH + E. coli
Multipotent Precursors	< 0.1 (0.0)	< 0.1 (0.0)	56.8 (1.9)	11.3 (2.2)
Granulocyte Precursors	2.6 (0.5)	1.6 (0.4)	19.1 (1.1)	3.6 (0.8)
Mature Granulocytes	11.4 (1.1)	7.3 (1.9)	31.1 (2.0)	8.1 (0.6)
Erythroid Precursors	0.6 (0.1)	0.6 (0.2)	1.0 (0.1)	0.5 (0.0)
Monocytes	0.8 (0.2)	1.1 (0.1)	16.8 (1.6)	3.9 (0.6)
T Cells	3.4 (0.4)	2.8 (0.6)	53.7 (3.5)	15.9 (1.5)
B Cells	0.8 (0.1)	1.3 (0.2)	18.1 (1.7)	4.6 (0.6)

Table 1b: Constitution of BrdU+ cells in the bone marrow following 24 h E.coli challenge - % Total BMCs (SEM)
Cell Type	Saline	EtOH	E. coli	EtOH + E. coli
Multipotent Precursors	2.7 (0.2)	4.5 (0.7)	5.3 (0.4)	3.6 (0.3)
Granulocyte Precursors	6.5 (0.5)	12.3 (2.5)	20.2 (0.8)	22.0 (1.3)
Mature Granulocytes	11.6 (1.0)	10.9 (1.6)	10.8 (0.4)	4.0 (0.3)
Erythroid Precursors	27.5 (2.1)	38.5 (4.9)	27.1 (2.2)	43.2 (1.9)
Monocytes	26.2 (1.8)	19.8 (2.0)	19.9 (1.2)	16.7 (0.3)
T Cells	0.5 (0.1)	0.6 (0.1)	0.8 (0.0)	0.7 (0.0)
B Cells	24.9 (2.0)	13.4 (4.1)	15.9 (1.2)	9.9 (0.6)

Proportion of Multipotent Precursors (lin⁻), Granulocyte Precursor (Gr1^lo), Mature Granulocyte (Gr1^hi), Erythroid Precursor (TER-119⁺), Monocyte (F4-80⁺), T cell (CD3⁺), and B cell (CD3⁺) populations of the proliferating cells as a percentage of total bone marrow. N=5. Mean (SEM). BMCs: Bone marrow cells.

Enhanced marrow Sca-1 expression and ERK activation during bacteremia

To study the correlation between upregulation of Sca-1 expression and activation of the ERK pathway during bacteremia, we customized a phospho-specific flow cytometry protocol for studying specific myeloid and granulopoietic cell subpopulations. ERK activation is important for G₁/S phase transition during proliferation of hematopoietic precursors (22). Western blot analysis demonstrated that ERK signaling was strongly activated in the bone marrow 8 h post-E.coli challenge (Figure 2A). Alcohol administration inhibited this response. Phospho-specific flow cytometry showed that enhanced Sca-1 expression by myeloid and granulopoietic precursors corresponded with increased ERK activation in the bone marrow 8 h post E.coli challenge (Figure 2B). Alcohol suppressed both of these responses. Further, bacteremia induced a significant increase in the number of marrow Sca-1⁺pERK⁺ cells and the contribution of Sca-1⁺ cells to total marrow cells with activated ERK signaling (Figure 2C). Alcohol suppressed the increase in number of Sca-1⁺pERK⁺ cells in the bone marrow but did not affect the percentage of Sca-1-expressing cells in ERK-activated marrow cells. These results suggest that alcohol inhibited marrow ERK activation by suppressing the expansion of marrow Sca-1⁺ cells during bacteremia. Moreover, we found that ERK signaling was upregulated in myeloid progenitors and granulopoietic precursors during bacteremia. Mature granulocytes failed to increase ERK signaling, which may correspond to their post-mitotic, differentiated state. Alcohol suppressed ERK activation in marrow myeloid progenitors and granulopoietic precursors during bacteremia, as well as baseline ERK signaling in mature granulocytes (Figure 3A and B). ERK activation in the granulopoietic precursor population was greater in cells expressing Sca-1. Following 8 h E.coli challenge, enhanced Sca-1 expression by Sca-1⁺granulopoietic precursor cells was associated with a further increase in ERK activation (Figure 3C). Alcohol suppressed the increased ERK signaling in Sca-1⁺granulopoietic precursor cells during bacteremia (Figure 3C). Enhanced Sca-1 expression by granulopoietic precursors also correlated (R = 0.6037) with ERK pathway activation in these cells during E.coli challenge (Figure 3D). Bacteremia did not induce an increase in ERK signaling in Sca-1⁻granulopoietic precursor cells. Suppression of Sca-1 expression in granulopoietic precursors by alcohol correlated with reduced ERK activation during E.coli challenge (Figure 3D).

Figure 2.

Enhanced marrow Sca-1 expression and ERK activation during bacteremia. (A) Representative Western blot of phospho-ERK expression during 8 h E.coli challenge. Quantification of N=4. (B) Representative dot plots of enhanced Sca-1 and pERK expression in the bone marrow during 8 h bacteremia. (C) Increase in the number of marrow Sca-1⁺pERK⁺ cells and percentage of Sca-1⁺ in pERK⁺ cells during challenge. Data are combined from two independent experiments. N=5–8. Bars with asterisks or different letters are statistically significant (p≤0.05). BMCs: bone marrow cells.

Figure 3.

Enhanced ERK activation by myeloid and granulopoietic precursors during bacteremia. ERK activation by marrow (A) myeloid progenitor and (B) granulopoietic precursor cells during 8 h E.coli challenge. (C) Sca-1⁺granulopoietic precursor cells have greater ERK activity than Sca-1⁻granulopoietic precursor cells. (D) Sca-1 expression by granulopoietic precursor cells directly correlates with ERK activation in this cell type. Data are combined from two independent experiments. N=5–8. Bars with different letters are statistically significant (p≤0.05). BMCs: bone marrow cells.

Enhanced Sca-1 Expression and Granulocyte Cell Cycle Activity

Since Sca-1 expression correlated with ERK activation in myeloid and granulopoietic precursors, we then examined whether enhanced Sca-1 expression was also associated with an increase in cell cycle activity. In control animals, marrow Sca-1⁺granulopoietic precursor cells had greater cell cycle activity than Sca-1⁻granulopietic precursor cells (Figure 4A and B). Following 24 h E.coli challenge, the percentage of cycling Sca-1⁺granulopoietic precursor cells slightly decreased (Figure 4B). However, Sca-1⁻granulopietic precursor cell cycle activity declined markedly, and fewer proliferating Sca-1⁻granulopietic precursor cells were resident in the bone marrow during bacteremia (Figure 4B and C). The number of proliferating marrow Sca-1⁺granulopoietic precursor cells was significantly fewer than Sca-1⁻granulopietic precursor cells in the absence of infection (Figure 4C). This is reflective of the far greater marrow Sca-1⁻granulopietic precursor cell population in controls. With an increase in the marrow Sca-1⁺granulopoietic precursor cell population during bacteremia (Figure 1B), there were significantly more proliferating Sca-1⁺granulopoietic precursor cells than Sca-1⁻granulopietic precursor cells in the bone marrow (Figure 4C). Alcohol suppressed the proliferation of Sca-1⁺granulopoietic precursors during bacteremia (Figure 4B). Suppression of Sca-1⁺granulopoietic precursor cell cycle activity by alcohol was associated with a decrease in the number of proliferating Sca-1⁺granulopoietic precursors in the bone marrow during bacteremia as determined by BrdU incorporation.

Figure 4.

Enhanced Sca-1 expression and cell cycle activity of granulopoietic precursors during bacteremia. (A) Representative ModFit analysis of marrow Sca-1⁺granulopoietic precursor and Sca-1⁻granulopoietic precursor cells after 24 h E.coli challenge. Increased propidium iodide incorporation indicates greater DNA content per cell, as seen during the S/G₂ phases of the cell cycle. Changes in (B) percent cell cycle activity and (C) number of cycling marrow Sca-1⁺granulopoietic precursor and Sca-1⁻granulopoietic precursor cells during bacteremia. Data are combined from two independent experiments. N=9–10. Bars with different letters are statistically significant (p≤0.05). BMCs: bone marrow cells.

Alcohol-induced suppression of Sca-1⁺granulopoietic precursor cell proliferation also coincided with a slight increase in the number of proliferative marrow Sca-1⁻granulopietic precursor cells during bacteremia (Figure 4C). However, the cell cycle activity of this population did not change with infection. These data may partially explain the reduced ratio of mature granulocyte: granulopoietic precursor of proliferative cells in the bone marrow with alcohol administration during bacteremia (Figure 1A). Our results suggest that the Sca-1⁺granulopoietic precursor population became the predominant cell population contributing to granulocyte production in the bone marrow during the granulopoietic response to bacteremia, and the expansion of this population was inhibited by alcohol.

Sca-1 Expression and Maturation of Granulopoietic Precursors

We next measured myeloperoxidase (MPO) content as a marker of granulocyte maturation to determine whether Sca-1 expression also correlated with increased differentiation of granulopoietic precursors. Proliferation of granulopoietic precursors terminates at the myelocyte stage, as these cells begin to produce the antimicrobial components characteristic of mature granulocytes (34). Production of MPO, the principle enzyme of cytoplasmic, peroxidase-positive granules, increases from the myeloblast to myelocyte stages (35, 36). As expected in control animals, immature granulopoietic precursor cells represented a lesser percentage than mature granulocytes of the total MPO⁺ population in the bone marrow. Sca-1⁺ cells composed relatively little of either granulocyte population (Figure 5A). Following 24 h E.coli challenge, the number of MPO⁺Sca-1⁻ cells was reduced and coincided with a slight increase in the number of marrow MPO⁺Sca-1⁺ cells. This increase however, did not achieve statistical significance (Figure 5B). Focusing on granulopoietic precursor development, we found that Sca-1⁺granulopoietic precursor cells had greater MPO content than Sca-1⁻granulopoietic precursor cells in control animals (data not shown). During bacteremia, the MPO content of Sca-1⁺granulopoietic precursors was reduced, while the MPO content of Sca-1⁻granulopioetic precursors remained unchanged. Alcohol exacerbated the decrease in MPO content of Sca-1⁺granulopoietic precursor cells but did not affect the MPO content of Sca-1⁻granulopioetic precursor cells during bacteremia (Figure 5C). Additionally, we found that the reduction in MPO content of Sca-1⁺granulopoietic precursors inversely correlated (R = 0.6178) with the increase in number of marrow Sca-1⁺granulopoietic precursor cells during E.coli challenge (Figure 5D). With alcohol administration prior to challenge, MPO content and Sca-1⁺granulopoietic precursor cell number remained inversely correlated (R = 0.6987) but were further reduced during bacteremia (Figure 5D). The relationship between Sca-1 expression and MPO content was specific to granulopoietic precursors (Supplemental Figure 2). There was no correlation between Sca-1 and MPO in mature granulocytes.

Figure 5.

Sca-1 expression and maturation of granulopoietic precursors. (A) Composition of MPO⁺ cells in control mice. (B) Changes in the MPO⁺ cell population during 24 h E.coli challenge. (C) Sca-1 expression and MPO content in granulopoietic precursor cells during bacteremia. Increase in Sca-1⁺granulopoietic precursor cells during 24 hr E.coli challenge inversely correlates with MPO content in this cell population (D). Data are combined from two independent experiments. N=5–10. Bars with different letters are statistically significant (p≤0.05). BMCs: bone marrow cells.

CFU-GM Production by Sca-1^−/− Myeloid and Granulopoietic Precursors during Bacteremia

To further determine the importance of Sca-1-associated proliferation in increasing granulocyte production during bacteremia CFU-GM assays were performed using myeloid and granulopoietic precursors from wild type C57BL/6 and Sca-1^−/− mice. As expected in the absence of infection, CFU-GM production by myeloid precursors (30%; 30/100 lin⁻ckit⁺ cells) was higher than downstream granulopoietic precursor cells (0.5%; 10/2000 Gr1^lo cells) in wild type controls. The CFU-GM activity of marrow Sca-1^−/− myeloid precursors was significantly reduced compared to that of wild type cells. We observed that Sca-1^−/− granulopoietic precursor cells partially compensated for this deficiency with greater CFU-GM production than wild type granulopoietic precursor cells in the absence of infection (Figure 6A and B). These results agree with earlier findings by Ito et al (25).

Figure 6.

CFU-GM activity of myeloid and granulopoietic precursors from Sca-1^−/− mice during bacteremia. Number of CFU-GM produced and the fold induction with infection by sorted marrow (A) myeloid precursor and (B) granulopoietic precursor cells from wild type and Sca-1^−/− mice following 24 h E.coli challenge. ERK activation in (C) bone marrow and (D) granulopoietic precursor cells of wild type and Sca-1^−/− mice during bacteremia. N=5–8. Bars with asterisks or different letters are statistically significant (p≤0.05). BMCs: bone marrow cells.

During the granulopoietic response to bacteremia, the CFU-GM activity of wild type myeloid precursors and granulopoietic precursor cells increased 1.5 fold and 3 fold, respectively (Figure 6A and B). During this response, ERK signaling was also activated in total bone marrow and granulopoietic precursor cells (Figure 6C and D). In Sca-1^−/− mice, bacteremia did not stimulate a statistical increase in CFU-GM production by marrow myeloid precursor or granulopoietic precursor cells (Figure 6A and B). Impaired CFU-GM production by Sca-1^−/− granulopoietic precursors during bacteremia was also associated with a failure to activate ERK signaling in total bone marrow and marrow granulopoietic precursor cells (Figure 6C and D). These results suggest that granulopoietic precursors maintain a large reserve capacity to enhance granulocyte production during bacterial infection that is facilitated by upregulation of Sca-1 expression.

Discussion

Hematopoietic stem and progenitor cells respond to systemic infection by expansion, differentiation, and mobilization (37). Our previous studies have shown that enhanced surface expression of Sca-1 by marrow primitive hematopoietic precursors, including hematopoietic stem cells, plays an important role in the granulopoietic response to bacteremia. In response to Gram-negative or Gram-positive bacterial infection, enhanced Sca-1 expression by marrow hematopoietic precursors corresponds with enhanced proliferation and preferential commitment towards myeloid lineage development (10, 11, 38). Sca-1 expression can also be upregulated in granulopoietic precursors, leading to increased mitosis and subsequent granulocyte production (Supplemental Figure 3) (23). Results presented here suggest that upregulated Sca-1 expression may promote granulocyte lineage expansion through enhancing ERK pathway activation.

Excessive alcohol consumption dysregulates multiple structural, cellular, and humoral aspects of the immune system, including granulopoiesis. We have previously reported that alcohol suppresses enhanced Sca-1 expression by primitive hematopoietic precursors and marrow granulocyte lineage expansion during bacterial infection (11, 23). Our current study extends this line of investigation by demonstrating that alcohol suppresses the proliferation of granulopoietic precursors by impairing their upregulation of Sca-1 expression and subsequent activation of the ERK signaling pathway during the granulopoietic response.

Granulocyte production is dynamically regulated in the bone marrow during bacterial infection. Our current data show that the marrow granulopoietic precursor population expanded following E.coli challenge, coinciding with reduction of the marrow mature granulocyte reserve. Following 24 h bacteremia, activated marrow granulopoiesis was associated with an increase in proliferation of myeloid and granulopoietic precursors and development into mature granulocytes. A prominent feature of this process was the significant upregulation of Sca-1 expression by cells in both myeloid and granulopoietic precursor populations, which correlated with their robust increase in proliferation and granulocyte production.

ERK activation plays an important role in hematopoietic precursor proliferation and myeloid lineage commitment (22, 39). Therefore, we studied the role of Sca-1 in promoting ERK signaling during the granulopoietic response to bacteremia. Our data demonstrate that enhanced Sca-1 expression closely correlated with ERK activation in the bone marrow during E.coli challenge. Bacteremia induced a nearly 3-fold increase in the contribution of Sca-1⁺ cells to total marrow cells with ERK activation. Increased ERK signaling occurred in myeloid and granulopoietic precursors but not mature granulocytes during bacteremia. Moreover, enhanced Sca-1 expression by granulopoietic precursors directly correlated with increased ERK pathway activation. These data support previous reports that ERK signaling is important for granulopoietic precursor expansion rather than differentiation (22). We further assessed whether greater ERK activation by Sca-1⁺granulopoietic precursors also corresponded with greater cell cycle activity. Our data demonstrate that Sca-1⁺granulopoietic precursor cells were more actively involved in cell cycling than Sca-1⁻granulopoietic precursor cells in the absence of infection. Moreover, during the granulopoietic response to bacteremia, the Sca-1⁺granulopoietic precursor cell population became the predominant precursor cell type contributing to marrow granulocyte production.

Multiple host- and pathogen-derived factors are involved in stimulating granulopoiesis during a systemic bacterial infection. Although inflammatory cytokines, including G-CSF, GMCSF, IL-1, and IL-6 are well recognized regulators of granulopoiesis, increasing evidence suggests that pathogens can directly modify marrow hematopoiesis through TLR stimulation of hematopoietic precursors (5, 6, 8, 40, 41). The ERK pathway is a signaling cascade activated by ligand engagement of both TLR and cytokine receptors, which are critically involved in the granulopoietic response (2). Activation of ERK signaling enhances marrow granulopoiesis by promoting the proliferation over differentiation of granulopoietic precursors during infection. Activated ERK negatively regulates the master myeloid lineage transcription factor CEBP-α (CCAAT/enhancer-binding protein alpha). Upon stimulation, CEBP-α suppresses further granulopoietic precursor proliferation in favor of terminal granulocyte differentiation (42, 43). Since enhanced Sca-1 expression promoted ERK activation in granulopoietic precursors, we then studied whether Sca-1 expression was also related to their terminal differentiation during bacteremia. As predicted, expansion of the marrow Sca-1⁺granulopoietic precursor cell pool inversely correlated with the MPO content of this population. As granulopoietic precursors develop into more mature myelocytes, their proliferation ceases and MPO content increases. Enhanced Sca-1 expression by granulopoietic precursors may therefore promote proliferation over subsequent differentiation of granulopoietic precursors by supporting increased ERK activation during bacteremia. Alcohol intoxication impaired the upregulation of Sca-1 expression and ERK activation by granulopoietic precursors in response to bacteremia. Concurrently, alcohol suppressed Sca-1⁺granulopoietic precursor cell cycle activity, as well as BrdU incorporation into Sca-1⁺granulopoietic precursors during bacteremia. Alcohol administration also exacerbated the decrease in MPO content of Sca-1⁺granulopoietic precursor cells following systemic infection with E. coli. These data indicate that marrow granulopoietic precursors are one of the major cell types targeted by alcohol in alcohol-induced impairment of the granulopoietic response to bacteremia.

To confirm the role of Sca-1 in achieving maximal granulopoietic precursor proliferation during bacterial infection, we challenged Sca-1^−/− mice with intravenous E.coli. Previous studies with Sca-1^−/− mice have identified deficiencies in hematopoietic and mesenchymal stem cell self-renewal and myeloid progenitor cell production (25, 44). Our studies have also found Sca-1^−/− mice to have a suppressed marrow granulopoietic response and reduced production of circulating granulocytes during bacteremia (23). Because maximal granulocyte production is dependent on lineage-committed precursors, we determined the approximate stage of granulocyte development that Sca-1 was most important for enhancing granulocyte production during bacteremia (45). Our observations indicate that Sca-1 expression is necessary for granulocyte production by myeloid precursors in the absence of infection. Sca-1^−/− granulopoietic precursors partially compensate for this deficiency with greater proliferation than wild type cells, in support of earlier observations (25). Following 24 h E.coli bacteremia, granulocyte production increased in wild type, but not Sca-1^−/− myeloid precursor cells. In wild type animals, bacteremia induced a greater increase in granulocyte production by downstream granulopoietic precursor cells. In response to 24 h E.coli challenge, granulocyte production by wild type granulopoietic precursors increased 3-fold, doubling the increase observed in wild type myeloid precursor cells. In contrast, Sca-1^−/− granulopoietic precursor cells were unable to further enhance granulocyte production during bacteremia. This impaired capacity for Sca-1^−/− granulopoietic precursors to contribute to granulocyte lineage expansion during bacteremia was associated with failure to activate ERK signaling. After 8 h E.coli challenge, the ERK pathway was activated in granulopoietic precursor cells from wild type mice but was absent in Sca-1^−/− mice.

Growing evidence suggests that Sca-1 expression is regulated by both pro-inflammatory and anti-inflammatory stimuli. Components of a pro-inflammatory milieu including IFN-α/γ, IL-6, TNF-α, and LPS can upregulate Sca-1, while anti-inflammatory factors including TGF-β have been shown to downregulate Sca-1 expression (10, 46). Therefore, Sca-1 expression and its role in hematopoietic precursor proliferation are dynamically regulated during the generation and resolution of the host immune response to bacteremia. It is likely that LPS-TLR4 signaling plays a direct role in upregulating Sca-1 expression at the initial stage of E. coli infection in our model system. As the host response to infection proceeds, other tissue-generated inflammatory mediators may subsequently join in the regulation of Sca-1 expression by hematopoietic cells. Further investigation on the dynamic regulation of the Sca-1 response will provide additional information for understanding this important host defense process.

Suppression of Sca-1 induction and ERK signaling in myeloid precursors by alcohol may result from impaired pathogen recognition (11, 47). Szabo and colleagues have shown that alcohol can impair LPS-mediated activation of phagocytes by inducing the redistribution of TLR4 complex components within membrane lipid rafts, impeding receptor clustering, and suppressing the transduction of intracellular signaling cascades (48). Because Sca-1 is a glycophosphatidylinositol-anchored protein that localizes to and may facilitate interactions within lipid rafts, the inhibition of Sca-1 expression by alcohol might be an underlying mechanism for the suppression of pro-proliferative signaling during the granulopoietic response to bacteremia (9, 49).

In our current investigation, we incorporated phospho-specific flow cytometry to understand the cell signaling events related to enhanced Sca-1 expression in myeloid and granulopoietic precursors. Phospho-specific flow cytometry is a significantly more powerful tool for studying cellular signaling events than conventional Western blot, because simultaneous analysis of cell surface marker expression and signaling can be resolved at a single cell level. This technique however, involves several technical hurdles, including the negative effects of fixation and permeabilization (fix/perm) on cell surface staining (26, 50). In this study, cell surface staining was not performed prior to fixation to avoid the possibility that monoclonal antibodies would stimulate intracellular signaling events. Our protocol optimized fixation/permeabilization procedures, specifically for studying intracellular signaling at different stages of myeloid lineage development during the host immune response to infection.

In summary, our current study reveals that enhanced Sca-1 expression facilitates activation of the ERK pathway and supports the proliferation of myeloid and granulopoietic precursors during bacteremia. Enhanced Sca-1 expression favors further proliferation over terminal differentiation of these precursors. While both myeloid and granulopoietic precursors contribute to granulocyte lineage expansion, granulopoietic precursors have a larger reserve capacity to support increased granulocyte production during bacteremia that may be facilitated by increased Sca-1 expression. Suppression of enhanced Sca-1 expression by alcohol correlates with impaired ERK pathway activation and proliferation of granulopoietic precursors, resulting in impaired marrow granulocyte lineage expansion. These results highlight a critical role of Sca-1 in the granulopoietic response and a target for alcohol-induced myelosuppression in the alcohol abusing host.