Prevention of Bone Marrow Cell Apoptosis and Regulation of Hematopoiesis by Type I IFNs during Systemic Responses to Pneumocystis Lung Infection

David Taylor, Michelle Wilkison, Jovanka Voyich and Nicole Meissner
J Immunol May 15, 2011, 186 (10) 5956-5967; DOI: https://doi.org/10.4049/jimmunol.1003558

Abstract

We recently demonstrated that lack of type I IFN signaling (IFNAR knockout) in lymphocyte-deficient mice (IFrag−/−) results in bone marrow (BM) failure after Pneumocystis lung infection, whereas lymphocyte-deficient mice with intact IFNAR (RAG−/−) had normal hematopoiesis. In the current work, we performed studies to define further the mechanisms involved in the induction of BM failure in this system. BM chimera experiments revealed that IFNAR expression was required on BM-derived but not stroma-derived cells to prevent BM failure. Signals elicited after day 7 postinfection appeared critical in determining BM cell fate. We observed caspase-8– and caspase-9–mediated apoptotic cell death, beginning with neutrophils. Death of myeloid precursors was associated with secondary oxidative stress, and decreasing colony-forming activity in BM cell cultures. Treatment with N-acetylcysteine could slow the progression of, but not prevent, BM failure. Type I IFN signaling has previously been shown to expand the neutrophil life span and regulate the expression of some antiapoptotic factors. Quantitative RT-PCR demonstrated reduced mRNA abundance for the antiapoptotic factors BCL-2, IAP2, MCL-1, and others in BM cells from IFrag−/− compared with that in BM cells from RAG−/− mice at day 7. mRNA and protein for the proapoptotic cytokine TNF-α was increased, whereas mRNA for the growth factors G-CSF and GM-CSF was reduced. In vivo anti–TNF-α treatment improved precursor cell survival and activity in culture. Thus, we propose that lack of type I IFN signaling results in decreased resistance to inflammation-induced proapoptotic stressors and impaired replenishment by precursors after systemic responses to Pneumocystis lung infection. Our finding may have implications in understanding mechanisms underlying regenerative BM depression/failure during complex immune deficiencies such as AIDS.

Subjecting the host to infectious agents commonly results in rapid recruitment and accumulation of leukocytes at the site of insult to attain clearance and tissue homeostasis. In response to most bacterial as well as fungal infections, neutrophils constitute the first line of defense and are subsequently followed by cells of the acquired immune response. All leukocytes are derived from the bone marrow (BM), and circulating neutrophils in particular have, under steady-state conditions, a very short life span of ∼12 h (1). This can be extended after activation and recruitment to specific tissue sites to up to 2 d. However, neutrophils will subsequently undergo apoptosis, and the rate is controlled by the activation state and various antiapoptotic factors such as MCL-1 (2).

Hematopoiesis and inflammation are tightly linked (3). Upon infection at local sites, resident phagocytes release inflammatory mediators, such as TNF-α, IL-1, and IL-6 and thus elicit a systemic response (acute-phase response), resulting in increased production and release of neutrophils from the BM to sites of infection (4, 5). This process of supply and demand is tightly regulated. With regard to granulopoiesis, it is known that distinct transcription factors regulate basal (C/EBPα) and emergency (C/EBPβ) hematopoiesis (6). However, continuous increased demand of inflammatory cells during chronic inflammatory conditions can also be associated with abnormal hematopoiesis. Patients with chronic inflammation, such as rheumatic diseases, (7–9) or chronic infections often have additional blood abnormalities, including RBC anemia (10, 11), thrombocytopenia (12), or neutropenia either as lineage-specific presentations or pancytopenia (13). Neutropenia is generally associated with pancytopenia. In this regard, in virtually all patients with advanced-stage AIDS (stage IV), pancytopenia is the rule due to regenerative BM failure, and it appears that primarily on-demand hematopoiesis is suppressed (14). Mechanisms involved are certainly multifactorial; however, whether the underlying immune defect directly contributes to these BM dysfunctions has not been elucidated.

Clearly, immune-compromised patients with defects of lymphocyte functions and/or a defect of the type I IFN system, such as in AIDS (15–17), are at risk to develop opportunistic infections. In this regard, Pneumocystis infection of the lung is the most common opportunistic infection in AIDS causing severe pneumonia (Pneumocystis pneumonia); however, it also affects patients undergoing chemotherapy for cancer, anti–TNF-α therapy for some rheumatic diseases (18–20), or patients with inherited CD4 T cell defects (21). During Pneumocystis pneumonia, significant numbers of neutrophils are recruited from the BM to the lung, although they are unable to clear the pathogen (22, 23).

Type I IFNs are important regulators of the innate and acquired immune response and have also been therapeutically used for their antiviral and antiproliferative properties. However, when administered in high therapeutic dosages for hepatitis C or cancer treatment, type I IFNs can induce BM suppression (24). In contrast, we recently demonstrated that specifically in response to pulmonary infection with the fungal pathogen Pneumocystis, the lack of type I IFN signaling in lymphocyte-deficient mice (IFrag−/− mice) results in rapidly progressing BM failure (25). Lymphocyte-competent but IFNAR-deficient mice (IFNAR−/− mice) developed mild BM depression with induction of extramedullary hematopoiesis, whereas lymphocyte-deficient mice with intact IFNAR (RAG−/− mice) as well as wild-type mice demonstrate normal hematopoiesis after Pneumocystis lung infection. This occurred although the infection remained in the lung without evidence of dissemination. BM failure in IFrag−/− mice could be prevented via immune reconstitution with wild-type or IFNAR−/− splenocytes and specifically B cells (25).

In the current study, we further examined mechanisms involved in the induction of BM failure to discern how type I IFN signaling acted in protecting hematopoiesis in the absence of lymphocytes and under the stress of increased demand after Pneumocystis lung infection. By comparing responses in IFrag−/− versus RAG−/− mice, we demonstrate that IFNAR signaling is required on BM-derived but not stromal-derived cells to prevent BM failure after Pneumocystis lung infection. Failure is initiated by accelerated apoptosis of the neutrophil lineage, mainly via the extrinsic pathway of apoptosis, followed by the additional loss of myeloid precursors. BM failure is preceded by reduction of transcription or rapid degradation of mRNA for some key antiapoptotic and survival factors and increased mRNA abundance for proapoptotic TNF-α when assessed in total BM cells. This is also followed by evidence for oxidative stress in all myeloid and other precursors associated with caspase-9 activation.

These data show that a state of immunodeficiency involving the type I IFN system may negatively affect the regenerative capabilities of the BM in response to increased demand due to opportunistic lung infections with Pneumocystis by regulating the balance of proapoptotic and antiapoptotic mechanisms. This could have implications in understanding the causes underlying regenerative BM failure as it can occur in AIDS and possibly other BM-failure diseases associated with immune defects (14, 26).

Materials and Methods

Mice, treatment procedures, and BM chimera generation

C.B17 SCID mice, as a source for Pneumocystis murina organisms, were bred and maintained at the Montana State University Animal Resource Center. RAG1−/− mice (C57/BL6 background) were initially purchased from Jackson Laboratories (stock no. 002096) and then bred at Montana State University. IFrag−/− mice were generated by crossing IFNAR knockout mice (129SvEv background and maintained at Montana State University) with RAG1−/− mice (C57/BL6 background), as previously described (25), and have since been backcrossed two more times on a C57/BL6 background. Animals were kept in ventilator cages with sterilized food and water. Some mice received N-acetylcysteine (NAC) as an antioxidant treatment at 7.6 mM (1.25 g/l) in acidified drinking water starting 1 d prior to and continuing through the course of the experiment.

BM chimeric mice between IFrag−/− and RAG−/− were generated as previously described (19). Briefly, two groups of recipient IFrag−/− and Rag-1−/− mice were lethally irradiated with 9.5 Gy from a 156Cs source in a split dose, and 107 BM cells from IFRag−/− donor animals was i.v. injected into one group of RAG−/− recipient mice (IFrag−/− BM into RAG−/− mice) and into one group of IFrag−/− recipient mice (IFrag−/− BM into IFrag−/− mice), and 107 BM cells from RAG−/− mice was injected into IFrag−/− recipient mice (RAG−/− BM into IFrag−/−) and into RAG−/− recipient mice (RAG−/− BM into RAG−/− mice). Eight weeks after BM engraftment, the chimeric mice were infected with Pneumocystis species as described later.

All mouse studies conformed to National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee at Montana State University.

Pneumocystis infection and enumeration

Experimental animals were intratracheally infected with 107 Pneumocystis nuclei in 100 μl lung homogenate from infected source mice diluted in PBS buffer, and Pneumocystis burden of infected animals was assessed microscopically by enumeration of trophozoite nuclei count in lung homogenates in 10–50 oil immersion fields as previously described. The limit of detection for this technique is log10 4.43 (27).

Collection and differentiation of BM cells

BM cells from femur and tibia was collected as previously described by flushing 2 ml PBS through the BM canal using a 26 1/2 g needle and brought into a single-cell suspension (28).

BM cells were diluted 1:10 in PBS. Cell numbers were enumerated, spun onto glass slides, and stained with Diff-Quick solution (Dade/Behring). Cell differentiation was performed based on morphology and staining pattern to distinguish myeloid (including myeloblast–myelocyte and metamyelocyte stage), erythroid cells, neutrophils, and eosinophils (28).

FACS analysis was also applied for BM cell subset differentiation using Abs for the cell surface molecules CD11b, Ly6G/6C (Gr-1), and Ter119, as described later, in combination with assessment for caspase-3, -8, and -9 activities at the single-cell level. Cell-sorting analysis confirmed that CD11b+/Gr-1− BM cells consisted of cells in the myeloblast–metamyelocyte stage of differentiation, whereas CD11b+/Gr-1+ cells consisted of both neutrophilic and eosinophilic granulocytes. Staining with the eosinophil marker CCR3 gave inconsistent results, and thus we were unable to distinguish between eosinophils and neutrophils by FACS analysis. The Ter119 marker identified cells of the erythroid lineage. Hematopoietic stem cells were identified by combining a lineage marker mix with Abs for Sca1, c-Kit, and CD150 (SLAM) followed by gating on cells negative for lineage marker expression but that were Sca1+, c-Kit+, and CD150+ (29).

Prior to FACS staining, RBC lysis of BM cell samples was performed using ACK lysis buffer. Cells were then suspended at 1 × 107/ml in FACS buffer (PBS/10% calf serum) containing Fc-block (mouse clone 24G2; Pharmingen) at a 1:800 dilution. Sets of 5 × 105 cells were stained with specific Abs: TER-119 (PerCP–Cy5-5, clone TER-119; BioLegend), anti-CD11b (Alexa Fluor 700, clone M1/70; BioLegend), anti–Ly-6G/6C (allophycocyanin–Cy7, clone RB6-8C5; Pharmingen). After cell surface staining, cells were assessed for the presence of total activated caspases in live cells using the CaspGLOW Red Caspase staining kit (Bio Vision), the presence of activated caspase-8 using CaspaseGLOW Red Caspase-8 staining kit (Bio Vision), the presence of activated caspase-9 using CaspGLOW Red Caspase-9 staining kit (Bio Vision), or the presence of activated caspase-3 using CaspGLOW Red Caspase-3 (Bio Vision). The manufacturer’s protocols were followed for these CaspGLOW products. Hematopoietic stem cells were identified as described above using a FITC-labeled lineage marker mix (Invitrogen), anti–c-Kit (allophycocyanin–Cy-7, clone 2B8; BioLegend), anti-CD150 (PerCP, clone TC15-12F12.2; BioLegend), and anti-Sca1 (allophycocyanin, clone D7; eBioscience). BCL-2 expression in BM cells was assessed using a PE-labeled mouse BCL-2–specific Ab and an isotype-specific irrelevant control Ab (PE Hamster anti-mouse Bcl-2 Set; BD Pharmingen), and an intracellular staining was performed according to the manufacturer’s protocol. For cell fixation and permeabilization, Cytofix–Cytoperm reagents from BD Pharmingen were used. Cell acquisition was performed using an LSRII flow cytometer (BD) and analyzed by FlowJo software.

Detection of oxidative stress in BM cells using dichlorodihydrofluorescein diacetate

Baseline reactive oxygen species (ROS) production in BM cells from IFrag−/− mice compared with that in RAG−/− mice was quantified during the course of Pneumocystis lung infection as an indicator of cell stress (30). Briefly, BM cell suspensions were diluted to 2 × 106/ml in MEM, and 5 × 105 cells were placed in flat-bottom, black, 96-well plates (Costar) in the presence of 8 μM cell-permeant dichlorodihydrofluorescein diacetate (H2DCF-DA= DCF), a fluorimetric detector of ROS. Each cell sample was measured in quadruplicate using a Fluoroskan Ascent FL fluorometer (Thermo Electron) read at 2-min intervals over the course of 120 min (excitation 485 nm and emission 530 nm). Oxidation of DCF (H2DCF-DA) resulted in increased fluorescence of the product relative to the amount of ROS production in the cells, which was displayed in arbitrary fluorescent units. Accumulated ROS production was compared between the experimental groups at 120 min.

Quantitative gene expression analysis by RT-PCR and protein analysis

Total RNA was harvested from total BM cells from three individual IFrag−/− and RAG−/− mice at days 0, 7, and 10 after Pneumocystis lung infection using an RNeasy Plus Mini Extraction Kit (Qiagen). RNA quality (RNA integrity number) was assessed with an Agilent BioAnalyzer following the Nano Kit Lab-On-A-Chip procedure (Agilent Technologies). RNA with an RNA integrity number of ≥7.6 was used to be analyzed using pathway-focused RT-PCR array systems for apoptosis (RT2 Profiler PCR Array Mouse Apoptosis PAMM-012C; SABiosciences) using an Applied Biosystems 7500 Fast Real-Time PCR machine. Ct values were gathered using 7500 Fast System SDS Software. Comparative data analysis was performed via the ΔΔCt method using the PCR Array Data analysis Web Portal (http://www.SABiosciences.com/pcrarraydataanalysis.php) to determine relative expression differences between the comparison groups. Changes of mRNA abundances by 2-fold and higher with a p value <0.05 were considered significantly different between the comparison groups as shown. Confirmation studies and relative quantification of additional genes was performed using QuantiFast SYBR Green RT-PCR Kit (no. 204154; Qiagen) with 100 ng RNA/reaction and following the manufacturer’s protocol. Specific primers were ordered from Qiagen: mouse β-Actin (no. QT01136772), G-CSF (no. QT00105140), GM-CSF (no. QT00251286), TNF-α (no. QT00104006). Data were analyzed with 7500 Fast System SDS Software. All assays were done in triplicate. TNF-α protein was detected in cell lysates of IFrag−/− and RAG−/− BM cells at day 9 postinfection. Cell lysates were generated in RIPA buffer, and 60 μg total protein was assayed in duplicate using a commercially available mouse TNF-α ELISA kit from R&D Systems according to the manufacturer’s protocol.

Bone histology

For detection of apoptosis of BM cells, in situ TUNEL staining was performed on decalcified and paraffin-embedded bone sections. Cleaned femur and tibia were placed in 10% buffered formalin solutions for 24 h. Bones were decalcified in 14% EDTA in DPBS, pH 7.4–7.6, at 4°C for approximately 1 wk, rinsed in tap water for 30 min, and placed into 70% ethanol until further processing. Paraffin-embedded bones were cut in 5-μm sections and H&E stained (Richard Allen Scientific). TUNEL staining was performed on adjacent slides using the Roche Apoptosis Kit (TUNEL assay) following the manufacturer’s protocol.

Colony-forming cell assay for mouse BM cells

Hematopoietic precursor cell activity in BM from IFrag−/− and RAG−/− mice was assessed by performing colony-forming cell (CFC) assays in methylcellulose media. For this, 105 BM cells per animal and group of each time point was plated in MethoCult GF M3534 media (StemCell Technologies), which has been formulated to support the optimal growth of granulocyte and macrophage precursor cells. Cells from each sample were plated in duplicate according to the manufacturer’s protocol in 35-mm sterile culture dishes (StemCell Technologies) placed in 100-mm Petri dishes in the presence of one 35-mm dish containing sterile water. Cultures were incubated for 7 d in a water-jacketed incubator maintained with 5% CO2. Colony recognition (granulocyte–macrophage [GM]-, granulocyte [G]-, and macrophage [M]-forming colonies) and enumeration was performed according to StemCell Technologies guidelines.

Anti–TNF-α treatment of IFrag−/− mice

Anti–TNF-α treatment of IFrag−/− mice was performed via biweekly i.p. injection of 250 μg hamster anti-mouse mAb anti–TNF-α Ab (clone TN3-19.12; Leinco Technology) as previously described (31).

Microscopy and statistical analysis

Microscopy was performed using a Zeiss Axiostar microscope (Zeiss, Jena, Germany).

Statistical analysis was performed using either a one-way ANOVA or two-way ANOVA followed by either a Tukey or a Bonferroni post test.

Results

Hematopoietic cells of IFrag−/− mice undergo apoptosis in response to Pneumocystis lung infection

As previously reported, IFrag−/− but not RAG−/− mice develop rapidly progressing bone BM failure in response to Pneumocystis lung infection. Furthermore, BM failure in IFrag−/− mice can be prevented by immune reconstitution with splenocytes from wild-type or IFNAR−/− mice, and the rescuing activity is specifically provided by B cells (25). To determine whether apoptosis plays an essential role in Pneumocystis-induced BM failure, we assessed morphological features of BM cells from Pneumocystis-infected IFrag−/− mice with BM failure compared with those of BM cells from immune-reconstituted and infected IFrag−/− mice without BM failure. Morphological analysis revealed that BM cells from infected and unreconstituted IFrag−/− mice demonstrated nuclei fragmentation and cell membrane blebbing compared with BM cells from infected but immune-reconstituted IFrag−/− mice without BM failure (Fig. 1A). Such changes are consistent with apoptosis. To confirm increased apoptosis in the BM of Pneumocystis-infected IFrag−/− mice, bone sections from the two comparison groups were H&E stained to assess BM cellularity in situ (Fig. 1B), and adjacent sections were TUNEL stained to detect DNA fragmentation (Fig. 1C). BM cellularity was clearly reduced in Pneumocystis-infected IFrag−/− mice compared with that in immune-reconstituted and infected littermates (Fig. 1B). Moreover, despite reduced BM cellularity, there was an ∼5-fold greater signal for TUNEL-positive cells (green dots/field of view in Fig. 1C) in unreconstituted and Pneumocystis-infected IFrag−/− mice compared with that in immune-reconstituted and infected IFrag−/− mice (Fig. 1C).

FIGURE 1.
FIGURE 1.

IFrag−/− BM cells undergo apoptosis after Pneumocystis lung infection. BM cells were harvested at day 16 after Pneumocystis infection from lymphocyte-reconstituted and unreconstituted mice, and cytospins were performed for morphological evaluation and differentiation using a Diff-Quik (Siemens) stain. A, Cytospins from unreconstituted IFrag−/− mice demonstrated cell loss and morphological evidence of apoptosis in remaining cells such as nuclear fragmentation (see arrowhead). B, H&E-stained specimens from formalin-fixed and paraffin-embedded bone sections through the epiphysis area of femurs from both comparison groups show decreased cellularity in BM from unreconstituted IFrag−/− mice compared with that in BM from immune-reconstituted IFrag−/− mice. C, TUNEL staining on adjacent tissue sections using FITC-labeled nucleotides confirms increased apoptosis in BM from unreconstituted IFrag−/− mice compared with that in BM from reconstituted IFrag−/− mice. Images were taken using an Axio Imager with digital imaging system using ×200 (A) or ×100 (B, C) magnification.

Type I IFN signaling is relevant on BM-derived cells and not on radio-resistant stromal cells to prevent BM failure in lymphocyte-deficient mice.

To elucidate further the mechanism involved in the induction of BM failure in our model, and how type I IFN signaling acts to prevent it, we focused our analysis on the comparison of the BM responses between IFrag−/− and RAG−/− mice during the course of Pneumocystis lung infection. To determine whether IFNAR signaling is essential on BM-derived cells or on radio-resistant stromal cells, BM chimeric mice were generated by lethally irradiating two sets of IFrag−/− mice as well as RAG−/− mice and then reconstituting one group of IFrag−/− mice with BM from RAG−/− donor mice (RAG−/− BM into IFrag−/−) and one group as a control with BM from IFrag−/− littermates (IFrag−/− BM into IFrag−/−). Conversely, each group of irradiated RAG−/− mice was reconstituted either with BM from IFrag−/− donor mice (IFrag−/− BM into RAG−/−) or with BM from RAG−/− littermates (RAG−/−BM into RAG−/−). After BM engraftment, all groups were Pneumocystis-infected, and the BM response was assessed at day 16 postinfection and compared with the response of IFrag−/− and RAG−/− mice with previously unmanipulated BM and with that of uninfected control groups. RAG−/− mice engrafted with IFrag−/− BM experienced BM failure in response to Pneumocystis lung infection, whereas IFrag−/− mice that were engrafted with RAG−/− BM demonstrated normal hematopoiesis in response to the infection (Fig. 2A). All control groups behaved as expected: irradiated IFrag−/− mice engrafted with IFrag−/− BM and nonchimeric IFrag−/− mice developed BM failure, and RAG−/− mice engrafted with RAG−/− BM as well as nonchimeric RAG−/− mice maintained hematopoiesis after Pneumocystis lung infection. Thus, type I IFN signaling is required on BM-derived cells to maintain hematopoiesis during the systemic stress response to Pneumocystis lung infection. Notably, in BM chimeric RAG−/− mice that had received IFrag−/− BM (IFrag−/−into RAG−/− mice), BM failure appeared to have progressed more rapidly compared with that in control groups in which both stromal and BM cells lacked IFNAR. This more pronounced loss of BM cells did not appear to be triggered by systemic signals released due to higher Pneumocystis lung burdens, as no significant difference between the infected groups could be detected (Fig. 2B). Nevertheless, BM differentiations demonstrated that regardless of whether IFNAR was lacking on only BM cells or on both BM and stromal cells, loss of band neutrophils was equally severe when assessed by relative and absolute numbers (Fig. 2C, 2D). However, in mice in which both stromal and BM was lacking IFNAR, BM eosinophils and their precursors appeared to be more resistant to induction of cell death and lingered longer in the marrow (Fig. 2E, 2F).

FIGURE 2.
FIGURE 2.

Type I IFN signaling on BM-derived cells prevents BM failure after Pneumocystis lung infection. BM chimeric mice were generated between IFrag−/− and RAG−/− mice. After complete engraftment at 6 wk posttransplantation, mice were Pneumocystis-infected and the BM response evaluated at day 16 postinfection. A, BM cell counts flushed from one femur and tibia per mouse with six mice per group. Data demonstrate that RAG−/− mice that received IFrag−/− BM developed BM failure after Pneumocystis lung infection whereas IFrag−/− mice receiving RAG−/− BM did not (white bars). All other comparison groups served as controls to the experiment. B, Pneumocystis lung burden of all experimental groups was evaluated microscopically in lung homogenates. The detection limit of the method is 4.43 log10. C–F, BM cell differentiation was performed microscopically using Diff-Quik–stained BM cytospins. C and D, Relative and absolute neutrophil numbers in BM of all comparison groups. E and F, Relative and absolute eosinophil numbers. Differences were compared with uninfected control groups. *p < 0.05, **p < 0.01, ***p < 0.001.

BM failure in IFrag−/− mice is associated with increased oxidative cell stress and global caspase activity.

Previous analysis had determined that BM failure in IFrag−/− mice becomes obvious between days 7 and 10 and is commonly severe by day 16 postinfection. Neutrophil counts in the BM are sensitive indicators of this progression. Fig. 3A shows comparative cytospins from BM samples of IFrag−/− and RAG−/− mice at days 0, 7, 10, and 16 postinfection. The data demonstrate a progressive loss of neutrophils during the course of infection and morphological changes consistent with apoptosis in younger myelocytes and metamyelocytes from IFrag−/− but not from RAG−/− mice. Significant differences between the comparison groups were evident by day 10 postinfection (Fig. 3B). FACS analysis also verified the complete loss of neutrophils by disappearance of the CD11b+Gr-1hi–expressing granulocyte population at day 16 postinfection (Fig. 3C, compare 1.58% in IFrag−/− versus 61.5% in RAG−/− mice).

FIGURE 3.
FIGURE 3.

Day 7 postinfection marks a turning point, and neutrophil loss is a sensitive indicator of BM failure-progression in IFrag−/− mice. Comparative cytospin analysis of BM flushes from IFrag−/− versus RAG−/− mice is depicted at days 0, 7, 10, and 16 postinfection demonstrating progressive loss of first band neutrophils, and then younger myeloid precursors in IFrag−/− but not RAG−/− mice. Loss of neutrophils was assessed microscopically on Diff-Quik–stained BM cytospins. Pictures were taken using an Axio Imager with digital imaging system (original magnification ×100) (A) and the percentage of BM-neutrophil determined by differential cell count using Diff-Quik–stained cytospin slides (B). FACS analysis allowed for the identification of granulocytes (mainly neutrophils) via staining for Gr-1 versus CD11b. Additional BM cell FACS analysis confirms loss of CD11b+Gr-1hi–expressing granulocytes in BM from IFrag−/− but not RAG−/− mice at day 16 postinfection (C). Statistical analysis was performed using a two-way ANOVA for B. Values shown are means (±SEM), n = 5. **p < 0.01, ***p < 0.001.

ROS production has been shown to induce apoptotic cell death in neutrophils (32, 33) and is causally linked to the development of progressive BM failure in models for Fanconi anemia (34) and ataxia telangiectasia (35). Thus, baseline ROS production in BM cells from IFrag−/− compared with RAG−/− mice was assessed in response to Pneumocystis lung infection. We established that ROS accumulation in H2DCF-DA–loaded BM cells from uninfected IFrag−/− and RAG−/− mice was essentially equivalent. However, at day 10 after Pneumocystis infection, ROS levels were significantly higher in cells from IFrag−/− mice compared with those in cells from RAG−/− mice (data not shown). To test further whether BM failure was indeed causally linked to increased ROS production, a kinetic study was performed to investigate the effects of neutralizing ROS production via antioxidant treatment in IFrag−/− mice. In this experiment (Fig. 4A), we compared BM responses of infected IFrag−/− mice either untreated (open circle) or orally treated with NAC (open triangle) and those of infected RAG−/− mice (open square). We found that ROS production in BM cells from IFrag−/− mice progressively increased over the course of infection, whereas ROS production remained low in BM cells from RAG−/− mice. Furthermore, NAC treatment reduced ROS production significantly in BM cells of infected IFrag−/− mice (Fig. 4A). Accumulation of ROS in BM cells from untreated IFrag−/− mice over the course of infection was associated with a progressive decrease in cell numbers. However, despite significant reduction of ROS levels, NAC treatment could only postpone but not prevent the induction of BM failure in IFrag−/− mice. In this group, BM cell numbers remained normal until day 10 postinfection but then also rapidly decreased by day 16 postinfection (Fig. 4B).

FIGURE 4.
FIGURE 4.

BM failure in IFrag−/− mice is associated with increasing ROS production and global caspase activation. ROS production in BM cells from Pneumocystis-infected IFrag−/− and RAG−/− mice was assessed at days 0, 7, 10, and 16 postinfection using H2DCF-DA dye loading for fluorescent conversion upon oxidation relative to the presence of cellular ROS. This was also compared with NAC-treated IFrag−/− mice, and the effects on BM failure progression were assessed. ROS production was measured as relative units in BM cells from IFrag−/− (open circles) to RAG−/− (open squares) and NAC-treated IFrag−/− mice (open triangles) (A). BM cell counts flushed from two hind legs were performed over the course of infection (B). Activated global caspase activity was also assessed on single-cell level in Pneumocystis-infected IFrag−/− versus RAG−/− mice using FACS analysis over the course of infection. Depicted are representative histogram plots in IFrag−/− versus RAG−/− BM over the course of infection showing the percentage of cells positive for activated caspases (C), and data from all treatment groups are summarized (D). The percentage of neutrophils in BM differentiation of the comparison groups over the course of infection is demonstrated (E). Statistical analysis was performed using a two-way ANOVA for B. Values shown are means (±SEM), n = 5. *p < 0.05, **p < 0.01, ***p < 0.001.

Activated caspases initiate and facilitate apoptosis. The percentage of cells positive for global caspase activity was determined by FACS analysis in all comparison groups throughout the course of infection. Comparative histogram plots show staining for global caspase activity (mainly executioner caspases 1, 3, 4, 7, 10) at the single-cell level for BM cells from IFrag−/− and RAG−/− mice over the course of infection (Fig. 4C) and are summarized for all comparison groups (Fig. 4D): At day 0, minimal activity for caspases was detected (1.5% of BM cells from both IFrag−/− and RAG−/− mice). Over the course of Pneumocystis lung infection, this increased to maximum 5% in RAG−/− BM cells at day 16 postinfection. However, in untreated IFrag−/− mice, the number of caspase+ cells rapidly increased to 11% by day 7, 20% by day 10, and 30% by day 16 postinfection. In NAC-treated IFrag−/− mice, this progression was dampened with 5–8% cells being caspase+ by day 7 and at day 10. These profiles were consistent with prolonged normal cell counts throughout day 10, but were then followed by a rapid increase of caspase+ cells to 30% by day 16. Increasing percentages of caspase+ BM cells inversely correlated with relative (and absolute) neutrophil numbers (Fig. 4E). Multicolor FACS analysis demonstrated that particularly CD11b+Gr-1hi–expressing cells (identifying granulocytes) were the first to become caspase+ (data not shown) consistent with accelerated apoptosis in this cell subset.

Mechanisms initiating both the intrinsic and extrinsic pathways of apoptosis are involved in the induction of BM failure in IFrag−/− mice.

Executioner caspase activation is induced via either the intrinsic or extrinsic pathway of apoptosis (36, 37). Understanding the pathway responsible for apoptosis induction in this model of Pneumocystis-induced BM failure could provide key insight into how type I IFNs act to prevent BM failure. Thus, a kinetic study (days 0, 7, 10, and 16 postinfection) was performed to determine whether either the initiator caspase-8 (extrinsic pathway) or caspase-9 (intrinsic pathway) was predominately activated. Furthermore, using FACS analysis we assessed whether this also led to activation of the primary executioner caspase-3 in BM cells from IFrag−/− compared with those from RAG−/− mice. Representative FACS plots of BM cells from IFrag−/− versus RAG−/− mice over the course of infection display all cells stained for CD11b, Gr-1 (blue dots), and either activated caspase-8, -9, or -3 (yellow dots) (Fig. 5A). This approach demonstrated in this experiment that the percentage of cells positive for the initiators caspase-8 and caspase-9 was still identical in both comparison groups at day 7 postinfection. However, by day 10 postinfection, the percentage of cells positive for both initiators caspase-8 and caspase-9 and executioner caspase-3 was significantly higher in IFrag−/− mice than that in RAG−/− mice and progressed this way throughout day 16 postinfection (Fig. 5B). This was consistent with cell loss (data not shown; see Fig. 4). When caspase activity was assessed on a cell-subset level, we found that under steady-state conditions, caspase activities were negligible in the CD11b+Gr-1hi subset (granulocytes) between the groups. However, at day 10 postinfection, there was a significant difference in the percentage of cells positive for caspase-8 but not caspase-9 between the comparison groups in the CD11b+Gr-1hi subset (granulocytes) due to an increase of caspase-8+ granulocytes in IFrag−/− BM. This implicated caspase-8 as an early initiator of increased neutrophil apoptosis (Fig. 5C). In contrast, CD11b+Gr-1−/(low) myeloid precursors in both comparison groups already demonstrated significant caspase activity under steady-state conditions (day 0, ∼10% of cells are positive for initiator and executioner caspases). In response to the infection, this activity persisted unchanged in IFrag−/− mice out to day 10 and then increased for both initiator caspases to ∼25%. In contrast, in RAG−/− mice, caspase activity actually decreased from baseline throughout day 10 and returned to baseline levels at day 16 postinfection (Fig. 5D).

FIGURE 5.
FIGURE 5.

Caspase-8 and caspase-9 activation appear relevant to the induction of BM failure in IFrag−/− mice. Activity for the initiator caspases 8 and 9 versus the primary executioner caspase-3 was assessed on a single-cell level in combination with cell surface marker staining using FACS analysis in BM cells from IFrag−/− versus RAG−/− mice in a time course after infection. A, Representative FACS plots are shown. Blue dots represent all cells stained, and yellow dots represent cells additionally positive for the specific caspase activity. The left panel show plots for IFrag−/− BM cells, and the right panel shows cells from RAG−/− BM cells. B–D, The analysis of all mice per group and time points are summarized. The percentage of all BM cells positive for caspase-8, -9, or -3 is depicted in B. The percentage of all granulocytes/neutrophils positive for specific caspase activities (gated on CD11b+GR-1hi) is depicted in C. The percentage of all other myeloid cells/myeloid precursors (gated on CD11b+Gr-1low/−) positive for the specific caspase activities is depicted in D. Statistical analysis was performed using a two-way ANOVA. Values shown are means (±SEM), n = 5. *p < 0.05, **p < 0.01, ***p < 0.001.

To gain further insight into the mechanisms underlying the initiation of both pathways of apoptosis in BM cells from IFrag−/− mice, pathway-focused PCR arrays related to apoptotic and antiapoptotic mechanisms were performed (results of the complete study can be found in the Gene Expression Omnibus database under accession number GSE27835 under the following link: http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE27835). For this experiment, three experimental groups were compared: IFrag−/−, RAG−/−, and IFrag−/− mice receiving NAC treatment. Three independent BM cell samples per group at days 0 and 7 postinfection were analyzed. In this experiment, gene expression of all assessed genes was not significantly different at day 0 between the comparison groups (data not shown). However, at day 7 postinfection, mRNA abundance of important antiapoptotic genes known to interfere with caspase activation, such as Bcl-2, Birc2 (IAP), Pim-2, and Mcl-1, were significantly reduced in infected IFrag−/− mice compared with that in infected RAG−/− mice and preceded the induction of BM failure in these mice (Fig. 6A). Furthermore, quantitative PCR analysis for some (apoptosis-inducing) cytokines and growth factors not only confirmed data of the PCR array but also demonstrated reduced mRNA abundance for the two important myeloid growth factors GM-CSF and G-CSF as well as an increased mRNA abundance for the proapoptotic factor TNF-α in BM of IFrag−/− compared with that in BM of RAG−/− mice at day 7 postinfection (Fig. 6B). Notably, mRNA abundance for the same genes assessed in BM cells from NAC-treated IFrag−/− mice at day 7 postinfection demonstrated Bcl-2 and Pim-2 mRNA abundance already significantly reduced, but not for Mcl-1 and Birc2 (Fig. 6C). In addition, mRNA abundance for GM-CSF was also already diminished, and TNF-α increased compared with levels in BM cells from RAG−/− mice. However, G-CSF levels were not different from those in RAG−/− BM cells yet (Fig. 6D). Thus, BM cells of NAC-treated IFrag−/− mice demonstrated an intermediate phenotype regarding mRNA abundance for some key cell survival and proapoptotic regulators consistent with the delayed progression of BM failure found in this group. Some of the factors differentially regulated on mRNA level between the two comparison groups were also assessed on protein level to confirm the findings. Indeed, increased TNF-α protein levels were also detected in BM lysates from IFrag−/− but not RAG−/− mice in response to Pneumocystis lung infection when assessed by ELISA (Fig. 6E). In addition, intracellular staining for BCL-2 expression in BM cells demonstrated a reduced percentage of BCL-2–expressing myeloid progenitor cells in IFrag−/− mice compared with that in RAG−/− mice and a particular decrease of cells with high expression for BCL-2 (Fig. 6F). This is consistent with increasing caspase-9 activity in these subsets over time and with decreased mRNA abundance for the latter as assessed in total BM cells.

FIGURE 6.
FIGURE 6.

Reduced mRNA for key antiapoptotic genes and myeloid growth factors in IFrag−/− BM cells precedes BM failure. Relative quantification of mRNA abundance of IFrag−/− versus RAG−/− BM cells at day 7 postinfection for some antiapoptotic genes, growth factors, and cytokines was performed. A, Fold difference of mRNA abundance for some key antiapoptotic regulators such as Bcl-2, IAP-2, Mcl-1, and Pim-2. For this, mRNA from three animals per group was analyzed in triplicate. B, Fold difference of mRNA abundance for the growth factors GM-CSF and G-CSF as well as the apoptosis-inducing cytokine TNF-α. For this, mRNA of six mice per group was analyzed in triplicate. C and D, The corresponding data for mRNA abundance differences comparing NAC-treated IFrag−/− mice with RAG−/− mice at day 7 postinfection. E, TNF-α protein concentrations in BM cell lysates from IFrag−/− compared with RAG−/− mice at day 10 postinfection (n = 4). F, The percentage of myeloid progenitor cells (gated on CD11b+Gr-1−/low cells and set as 100%) expressing high levels of BCL-2 protein in IFrag−/− versus RAG−/− BM at day 10 after Pneumocystis infection as analyzed by intracellular FACS staining (n = 4). Values shown are means (±SEM). *p < 0.05, **p < 0.01, ***p < 0.001.

Hematopoietic precursor cell activity decreases over the course of Pneumocystis lung infection in BM of IFrag−/− mice, and TNF-α activity is a contributing factor

Caspase activity was also increased in myeloid precursor cells of IFrag−/− mice. Thus, the question arose whether loss of neutrophils in our system was predominately due to lack of replenishment rather than accelerated apoptosis of band neutrophils themselves. Therefore, CFC assays from BM cells of both IFrag−/− and RAG−/− mice were performed in methylcellulose-based media formulated to support optimal growth of granulocyte and macrophage progenitors. BM cells from both comparison groups were harvested and equal cell numbers seeded into culture at days 0, 7, 10, and 16 postinfection. Colony formation was assessed microscopically 7 d postseeding. Fig. 7A shows the total BM cell numbers in the comparison groups during the course of infection, which recapitulates the previous findings. Colony numbers established with BM cells from the comparison groups at each time point are shown in Fig. 7B. The data demonstrate a significant reduction in colony numbers from seeded BM cells of IFrag−/− mice after day 7 postinfection compared with that of seeded BM cells from RAG−/− mice, indicating a reduction in viable BM precursor cells from IFrag−/− mice. Microscopic differentiation of established colonies revealed that the relative distribution of GM-, G-, and M-forming colonies was not different in the comparison groups up to day 7 (Fig. 7C–E). However, after day 7, particularly the relative (and absolute) number of G-forming colonies significantly increased in BM cultures from RAG−/− mice but decreased in those from IFrag−/− mice (Fig. 7D). To assess whether increased TNF-α activity found in BM lysates from IFrag−/− mice negatively affected myeloid precursor viability, two groups of IFrag−/− mice were Pneumocystis-infected. One of the groups received anti–TNF-α treatment via i.p. injection (250 μg twice weekly) whereas the other infected group did not. BM cells were harvested at day 16 postinfection for counts, differentiation, as well as assessment of precursor cell activity in CFC assays. Fig. 8A demonstrates that total BM cell numbers were low and not significantly different between the infected IFrag−/− groups. However, cell numbers were consistently higher in the TNF-α–treated group compared with those in the untreated but infected IFrag−/− group. Microscopic BM differentiation revealed that neutrophils were also rapidly lost in the anti–TNF-α–treated group (data not shown). However, myeloid precursor cells remained significantly higher in anti–TNF-α–treated IFrag−/− mice (Fig. 8B) and also appeared microscopically more viable (data not shown). Colony assays performed with BM cells from the comparison groups confirmed this observation. Fig. 8C shows significantly higher CFCs in BM cell cultures from anti–TNF-α–treated IFrag−/− mice compared with those in BM cultures from untreated but infected IFrag−/− mice indicating a negative effect of TNF-α on BM precursor cell viability in our system.

FIGURE 7.
FIGURE 7.

Myeloid precursor cell activity decreases in BM from IFrag−/− mice in response to Pneumocystis lung infection. Hematopoietic precursor cell activity in BM from IFrag−/− and RAG−/− mice was assessed by performing CFC assays in methylcellulose media at days 0, 7, 10, and 16 postinfection. BM cells (105) per animal and group of each time point were plated in MethoCultR GF M3534 media (StemCell Technologies), which supports the optimal growth of granulocyte and macrophage precursor cells. A, BM cell numbers flushed from two hind legs of each mouse per group over the course of infection. B, Colony numbers counted in each group and time point. C–E, The percentage of respective colonies identified in each group and time point. Colonies were identified as GM (C), G (D), and M (E). To show representative distributions, colony differentiation data were not plotted if total colony counts were below 10. Values shown are means (±SEM), n = 4 per group and time point. *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 8.
FIGURE 8.

Neutralization of TNF-α in IFrag−/− mice during Pneumocystis lung infection positively affects precursor cell viability. Anti–TNF-α treatment of IFrag−/− mice was performed via biweekly i.p. injection of 250 μg hamster anti-mouse mAb anti–TNF-α Ab (clone TN3-19.12; Leinco Technology) and the effects on BM cell numbers, differentiation, and precursor cell activity assessed at day 16 postinfection in comparison with untreated but infected IFrag−/− mice as well as uninfected IFrag−/− mice. A, Total BM cell numbers from two hind legs. B, Total number of myeloid precursor cells in BM. C, Total colony numbers counted from equal number of BM cells plated in CFC assay. Values shown are means (±SEM), n = 4 per group. Statistical analyses between the three comparison groups were generated using a one-way ANOVA and a Tukey post hoc test. **p < 0.01, ***p < 0.001 (compared with infected but untreated IFrag−/− mice).

Discussion

Systemic responses to local infections are necessary to allow recruitment of inflammatory cells from the BM and signal increased demand to the system (5). We recently demonstrated that type I IFN signaling is instrumental for the maintenance of on-demand hematopoiesis after pulmonary Pneumocystis infection (25). In this study, we further examined mechanisms underlying the development of BM failure in the absence of type I IFN signaling after Pneumocystis lung infection.

In the presented studies, type I IFN signaling was found to be relevant on BM-derived cells and not on stromal-derived cells to maintain hematopoiesis after the systemic response to Pneumocystis lung infection by modulating neutrophil and precursor cell apoptosis. Neutrophils are high-turnover cells that are continuously replenished from BM precursors and form the first line of defense after acute-phase response signals during inflammation. Neutrophils are also recruited to the lung in response to Pneumocystis infection (22, 23). Their short life span can be extended via suppression of their intrinsic apoptotic program with growth factor stimulation (32). Type I IFNs are proinflammatory cytokines that, when administered in therapeutic dosages, can be proapoptotic and result in BM suppression (38, 39). In contrast, recent research also demonstrated that type I IFNs extend neutrophil life span similarly to G-CSF in culture (40) and during certain inflammatory responses in vivo, which can also contribute to prolonged inflammation (41). The regulation of antiapoptotic factors has been shown to be critically involved in this process (42).

Mechanisms underlying BM failure in response to Pneumocystis infection in our model are certainly multifactorial. We demonstrated that lack of type I IFN signaling in lymphocyte-deficient mice (IFrag−/−) resulted in increased apoptosis of hematopoietic cells. BM-residing band neutrophils were the first cells affected in IFrag−/− mice. In addition to possibly accelerated apoptotic turnover, CFC assays of precursor cells indicated that band neutrophils may also not be sufficiently replenished by their progenitor cells as they also undergo apoptosis in vivo and successively lost their ability to form colonies under optimal culture conditions in vitro.

Events near day 7 postinfection appeared vital in determining the fate of BM cells. Transcriptional analysis using quantitative RT-PCR revealed that apoptosis in the BM of IFrag−/− mice was preceded by significantly reduced mRNA abundance for relevant antiapoptotic regulators such as IAP2 (Birc2), MCL-1, BCL-2, and others at day 7 postinfection. Indeed protein expression levels for BCL-2 were significantly lower in myeloid precursor cells from infected IFrag−/− compared with those in myeloid precursor cells from RAG−/− mice in response to Pneumocystis lung infection and were associated with increasing percentage of caspase+ cells in this subset. BCL-2 is only expressed in hematopoietic cells up to the promyelocyte stage but not in band neutrophils (43, 44). Decreasing expression of BCL-2 in these subsets is consistent with increasing global caspase activity in these precursors and their declining viability as evident by diminishing activity in CFC assays. However, various other antiapoptotic molecules have been identified to be important in directly prolonging neutrophil survival. One of these factors is IAP2 (Birc2), which is downregulated in IFrag−/− mice after Pneumocystis lung infection. IAP2 expression is upregulated in response to type I and type II IFNs via STAT3-mediated pathways, resulting in prolonged human neutrophil survival in vitro (42). This is consistent with our unpublished finding that RAG/STAT1−/− mice do not develop BM failure in response to Pneumocystis lung infection (N. Meissner, unpublished observation). STAT1 is an important signaling molecule for IFNAR, and we had anticipated BM failure to develop in these mice.

Although downregulation of IAP2 in BM of IFrag−/− mice could be a direct link to the apoptosis of band neutrophils in our system, MCL-1 is also known to play a major role in controlling the rate at which neutrophils undergo apoptosis (2, 45). Cellular levels of MCL-1 are regulated by the rate of transcription as well as the rate of protein turnover, which is accelerated by TNF-α (46). Both TNF-α mRNA abundance as well as protein levels were found increased in BM from IFrag−/− but not RAG−/− in response to Pneumocystis lung infection. Yet, treatment of infected IFrag−/− mice with neutralizing anti–TNF-α Ab could not significantly prevent loss of neutrophils from the BM. However, it accomplished prolonged survival of BM myeloid precursor cells, which also maintained their ability to form colonies in CFC assays compared with BM cells from untreated but infected IFrag−/− mice. These findings suggest that increased TNF-α levels in IFrag−/− BM do not directly affect the life span of band neutrophils but that of their precursor cells. However, the continued loss of neutrophils in anti–TNF-α–treated IFrag−/− mice must have resulted either from additional lack of sufficient growth factors to allow for rapid replenishment from viable precursor cells or from the presence of other apoptosis-inducing factors specifically targeting band neutrophils, or a combined effect.

Although loss of neutrophils was a sensitive indicator of BM failure in our model, precursors of the myeloid and erythroid lineage also showed morphological changes consistent with apoptosis and subsequently disappeared from the BM of infected IFrag−/− mice (25). Single-cell analysis revealed that neutrophil death in IFrag−/− mice was associated with a slightly more pronounced activation of initiator caspase-8 over caspase-9 followed by increased executioner caspase-3 activity. Apoptosis of all other lineages was associated with an equal activation of both initiator caspases. Indeed, differences of caspase activity in the myeloid precursor subset of each comparison group were initially due to a decrease in baseline caspase activity for initiator caspases 8 and 9 as well as the executioner caspase-3 in cells from RAG−/− mice beginning at day 7 and were significant at day 10 after Pneumocystis infection. Only after day 10 was there an actual increase in the baseline percentage of precursor cells positive for caspases from IFrag−/− mice compared with RAG−/− mice. These findings could point to a potential difference in the induction of myeloid growth factor activity between the two comparison groups to meet increased hematopoietic demand in response to Pneumocystis lung infection. Caspase-9 is activated by internal stressors such as growth factor deficiencies (37), and its activity was increased during the progression of BM failure particularly in early precursors of IFrag−/− mice. This was associated with reduced mRNA abundance for key myeloid growth factors such as GM-CSF and G-CSF was well as reduced mRNA abundance of the antiapoptotic factor PIM-2 in BM cells from IFrag−/− mice at day 7 postinfection. PIM-2 expression has been demonstrated to confer long-term resistance to a variety of apoptotic signals in hematopoietic cells. However, growth factor withdrawal results in rapid decline of PIM-2 transcripts and cell apoptosis (47).

Activation of the intrinsic pathway of apoptosis is associated with generation of ROS due to cell stress (48). Baseline ROS production indeed increased in BM cells from IFrag−/− but not RAG−/− mice by day 10 postinfection and continued to increase over time as BM failure progressed. At day 10, many band neutrophils had already disappeared from IFrag−/− BM. Thus, the source for ROS was likely the remaining precursor cells, and their loss and dysfunction resulted in lack of band neutrophil replenishment.

Models for inherited BM-failure diseases, such as Fanconi anemia, have demonstrated an increased sensitivity to inflammation-induced (TNF-α mediated) ROS production in hematopoietic precursor cells (34, 49) and protection from BM failure via antioxidant treatment. Antioxidant treatment of IFrag−/− mice during Pneumocystis lung infection also significantly suppressed ROS production in BM cells. However, BM failure could only be delayed but not prevented. These data indicate that increased ROS production was likely an additional stressor or a result of other underlying mechanisms of cell stress but not the primary cause of BM failure in our model.

Signals activating both the intrinsic and extrinsic pathways of apoptosis appear involved in the induction of BM failure in our model. Caspase-8 activation is elicited via death receptor-mediated signals provided by cytokines such as FASL, TNF-α, TRAIL, TWEAK, and others via their specific receptors (50–52). Analysis of death receptor expression levels on BM cells from IFrag−/− versus RAG−/− mice by FACS revealed no significant differences until day 16 postinfection (data not shown). Whereas TNF-α mRNA abundance and protein concentrations in BM cells from IFrag−/− mice were elevated, other mediators, including TRAIL, were not increased. However, we found reduced mRNA abundance for the TRAIL-decoy receptor osteoprotegerin (data not shown), which, if also reduced on protein level, could allow for increased TRAIL activity in IFrag−/− BM due to reduced decoy activity.

Based on these data, it appears that complex mechanisms of cell regulation are disrupted during the systemic response to Pneumocystis lung infection in the absence of type I IFN signaling. TNF-α and TRAIL are important apoptosis-inducing cytokines relevant to the regulation of hematopoiesis (51, 52). We hypothesize that their potential increased activity, in combination with reduced expression of key growth factors for the myeloid lineage (GM-CSF and G-CSF), may have important implications in the pathogenesis of BM failure in our system by affecting cellular turnover. What signals in the lung elicited in response to Pneumocystis lung infection due to the absence of IFNAR signaling result in such a profound deregulation of BM homeostasis is currently not clear and under investigation.

Overall, our system demonstrates a key regulatory role of type I IFN signaling in balancing proapoptotic and antiapoptotic signals in the BM after the systemic inflammatory response initiated during pulmonary infection with the opportunistic fungal pathogen Pneumocystis. Our previous results had demonstrated that Pneumocystis lung burden is only very low at the time BM failure is progressing (25), and a pulmonary affliction may thus not be a primary presenting symptom. Collectively, our findings could have implications for understanding how regenerative BM failure might be triggered in the context of immune-deficiency syndromes that also involve the type I IFN system such as AIDS.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Dr. David Pascual and Dr. Mark Quinn for critical reading of the manuscript.

Footnotes

  • This work was supported by the National Institutes of Health (RO1HL090488) and Center of Biomedical Research Excellence (2P20RR020185-06).

  • The array data presented in this article have been submitted to the Gene Expression Omnibus database under accession number GSE27835.

  • Abbreviations used in this article:

    BM
    bone marrow
    CFC
    colony-forming cell
    G
    granulocyte
    GM
    granulocyte–macrophage
    H2DCF-DA
    dichlorodihydrofluorescein diacetate
    M
    macrophage
    ROS
    reactive oxygen species.

  • Received October 26, 2010.
  • Accepted February 25, 2011.

References

  1. ↵
    1. Dale D. C.,
    2. L. Boxer,
    3. W. C. Liles
    . 2008. The phagocytes: neutrophils and monocytes. Blood 112: 935–945.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Kennedy A. D.,
    2. F. R. DeLeo
    . 2009. Neutrophil apoptosis and the resolution of infection. Immunol. Res. 43: 25–61.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Kincade P. W.
    2006. Supplying the demand for granulocytes. Nat. Immunol. 7: 701–702.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Heinrich P. C.,
    2. I. Behrmann,
    3. S. Haan,
    4. H. M. Hermanns,
    5. G. Müller-Newen,
    6. F. Schaper
    . 2003. Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochem. J. 374: 1–20.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Heinrich P. C.,
    2. J. V. Castell,
    3. T. Andus
    . 1990. Interleukin-6 and the acute phase response. Biochem. J. 265: 621–636.
    OpenUrlFREE Full Text
  6. ↵
    1. Hirai H.,
    2. P. Zhang,
    3. T. Dayaram,
    4. C. J. Hetherington,
    5. S. Mizuno,
    6. J. Imanishi,
    7. K. Akashi,
    8. D. G. Tenen
    . 2006. C/EBPbeta is required for ‘emergency’ granulopoiesis. Nat. Immunol. 7: 732–739.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Utsinger P. D.,
    2. W. J. Yount
    . 1976. Lymphopenia in Sjögren’s syndrome with rheumatoid arthritis: relationship to lymphocytotoxic antibodies, cryoglobulinemia, and impaired mitogen responsiveness. J. Rheumatol. 3: 355–366.
    OpenUrlPubMed
    1. Schaller J. G.
    1975. Immunodeficiency and autoimmunity. Birth Defects Orig. Artic. Ser. 11: 173–184.
    OpenUrlPubMed
  9. ↵
    1. Wagner U.,
    2. S. Kaltenhäuser,
    3. M. Pierer,
    4. B. Wilke,
    5. S. Arnold,
    6. H. Häntzschel
    . 2002. B lymphocytopenia in rheumatoid arthritis is associated with the DRB1 shared epitope and increased acute phase response. Arthritis Res. 4: R1.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Weiss G.,
    2. L. T. Goodnough
    . 2005. Anemia of chronic disease. N. Engl. J. Med. 352: 1011–1023.
    OpenUrlCrossRefPubMed
  11. ↵
    1. LaMonte A. C.,
    2. M. E. Paul,
    3. J. S. Read,
    4. M. M. Frederick,
    5. D. D. Erdman,
    6. L. L. Han,
    7. L. J. Anderson
    Women and Infants Transmission Study. 2004. Persistent parvovirus B19 infection without the development of chronic anemia in HIV-infected and -uninfected children: the Women and Infants Transmission Study. J. Infect. Dis. 189: 847–851.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Evatt B. L.
    2005. HIV Infection and thrombocytopenia. Curr. Hematol. Rep. 4: 149–153.
    OpenUrlPubMed
  13. ↵
    1. Levine A. M.,
    2. D. T. Scadden,
    3. J. A. Zaia,
    4. A. Krishnan
    . 2001. Hematologic Aspects of HIV/AIDS. Hematology Am Soc Hematol Educ Program 2001: 463–478.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Moses A.,
    2. J. Nelson,
    3. G. C. Bagby Jr..
    1998. The influence of human immunodeficiency virus-1 on hematopoiesis. Blood 91: 1479–1495.
    OpenUrlFREE Full Text
  15. ↵
    1. Siegal F.
    2003. Interferon-producing plasmacytoid dendritic cells and the pathogenesis of AIDS. Res. Initiat. Treat. Action 8: 10–13.
    OpenUrlPubMed
    1. Soumelis V.,
    2. I. Scott,
    3. F. Gheyas,
    4. D. Bouhour,
    5. G. Cozon,
    6. L. Cotte,
    7. L. Huang,
    8. J. A. Levy,
    9. Y. J. Liu
    . 2001. Depletion of circulating natural type 1 interferon-producing cells in HIV-infected AIDS patients. Blood 98: 906–912.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Levy J. A.,
    2. I. Scott,
    3. C. Mackewicz
    . 2003. Protection from HIV/AIDS: the importance of innate immunity. Clin. Immunol. 108: 167–174.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Kaur N.,
    2. T. C. Mahl
    . 2007. Pneumocystis jiroveci (carinii) pneumonia after infliximab therapy: a review of 84 cases. Dig. Dis. Sci. 52: 1481–1484.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Kaur N.,
    2. T. C. Mahl
    . 2004. Pneumocystis carinii pneumonia with oral candidiasis after infliximab therapy for Crohn’s disease. Dig. Dis. Sci. 49: 1458–1460.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Komano Y.,
    2. M. Harigai,
    3. R. Koike,
    4. H. Sugiyama,
    5. J. Ogawa,
    6. K. Saito,
    7. N. Sekiguchi,
    8. M. Inoo,
    9. I. Onishi,
    10. H. Ohashi,
    11. et al
    . 2009. Pneumocystis jiroveci pneumonia in patients with rheumatoid arthritis treated with infliximab: a retrospective review and case-control study of 21 patients. Arthritis Rheum. 61: 305–312.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Sepkowitz K. A.
    2002. Opportunistic infections in patients with and patients without acquired immunodeficiency syndrome. Clin. Infect. Dis. 34: 1098–1107.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Thomas C. F., Jr..,
    2. A. H. Limper
    . 2004. Pneumocystis pneumonia. N. Engl. J. Med. 350: 2487–2498.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Swain S. D.,
    2. T. W. Wright,
    3. P. M. Degel,
    4. F. Gigliotti,
    5. A. G. Harmsen
    . 2004. Neither neutrophils nor reactive oxygen species contribute to tissue damage during Pneumocystis pneumonia in mice. Infect. Immun. 72: 5722–5732.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Kowdley K. V.
    2005. Hematologic side effects of interferon and ribavirin therapy. J. Clin. Gastroenterol. 39(1, Suppl)S3–S8.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Meissner N.,
    2. M. Rutkowski,
    3. A. L. Harmsen,
    4. S. Han,
    5. A. G. Harmsen
    . 2007. Type I interferon signaling and B cells maintain hemopoiesis during Pneumocystis infection of the lung. J. Immunol. 178: 6604–6615.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Isgrò A.,
    2. A. Aiuti,
    3. W. Leti,
    4. C. Gramiccioni,
    5. A. Esposito,
    6. I. Mezzaroma,
    7. F. Aiuti
    . 2005. Immunodysregulation of HIV disease at bone marrow level. Autoimmun. Rev. 4: 486–490.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Harmsen A. G.,
    2. M. Stankiewicz
    . 1990. Requirement for CD4+ cells in resistance to Pneumocystis carinii pneumonia in mice. J. Exp. Med. 172: 937–945.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Bolliger A. P.
    2004. Cytologic evaluation of bone marrow in rats: indications, methods, and normal morphology. Vet. Clin. Pathol. 33: 58–67.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Wagers A. J.
    2005. Stem cell grand SLAM. Cell 121: 967–970.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Krejsa C. M.,
    2. G. L. Schieven
    . 2000. Detection of oxidative stress in lymphocytes using dichlorodihydrofluorescein diacetate. Methods Mol. Biol. 99: 35–47.
    OpenUrlPubMed
  31. ↵
    1. Williams R. O.,
    2. M. Feldmann,
    3. R. N. Maini
    . 1992. Anti-tumor necrosis factor ameliorates joint disease in murine collagen-induced arthritis. Proc. Natl. Acad. Sci. USA 89: 9784–9788.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Kobayashi S. D.,
    2. J. M. Voyich,
    3. A. R. Whitney,
    4. F. R. DeLeo
    . 2005. Spontaneous neutrophil apoptosis and regulation of cell survival by granulocyte macrophage-colony stimulating factor. J. Leukoc. Biol. 78: 1408–1418.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Scheel-Toellner D.,
    2. K. Wang,
    3. R. Craddock,
    4. P. R. Webb,
    5. H. M. McGettrick,
    6. L. K. Assi,
    7. N. Parkes,
    8. L. E. Clough,
    9. E. Gulbins,
    10. M. Salmon,
    11. J. M. Lord
    . 2004. Reactive oxygen species limit neutrophil life span by activating death receptor signaling. Blood 104: 2557–2564.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Du W.,
    2. Z. Adam,
    3. R. Rani,
    4. X. Zhang,
    5. Q. Pang
    . 2008. Oxidative stress in Fanconi anemia hematopoiesis and disease progression. Antioxid. Redox Signal. 10: 1909–1921.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Ito K.,
    2. A. Hirao,
    3. F. Arai,
    4. S. Matsuoka,
    5. K. Takubo,
    6. I. Hamaguchi,
    7. K. Nomiyama,
    8. K. Hosokawa,
    9. K. Sakurada,
    10. N. Nakagata,
    11. et al
    . 2004. Regulation of oxidative stress by ATM is required for self-renewal of haematopoietic stem cells. Nature 431: 997–1002.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Yan N.,
    2. Y. Shi
    . 2005. Mechanisms of apoptosis through structural biology. Annu. Rev. Cell Dev. Biol. 21: 35–56.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Leist M.,
    2. M. Jäättelä
    . 2001. Four deaths and a funeral: from caspases to alternative mechanisms. Nat. Rev. Mol. Cell Biol. 2: 589–598.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Platanias L. C.
    2005. Mechanisms of type-I- and type-II-interferon-mediated signalling. Nat. Rev. Immunol. 5: 375–386.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Verma A.,
    2. D. K. Deb,
    3. A. Sassano,
    4. S. Uddin,
    5. J. Varga,
    6. A. Wickrema,
    7. L. C. Platanias
    . 2002. Activation of the p38 mitogen-activated protein kinase mediates the suppressive effects of type I interferons and transforming growth factor-beta on normal hematopoiesis. J. Biol. Chem. 277: 7726–7735.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Wang K.,
    2. D. Scheel-Toellner,
    3. S. H. Wong,
    4. R. Craddock,
    5. J. Caamano,
    6. A. N. Akbar,
    7. M. Salmon,
    8. J. M. Lord
    . 2003. Inhibition of neutrophil apoptosis by type 1 IFN depends on cross-talk between phosphoinositol 3-kinase, protein kinase C-delta, and NF-kappa B signaling pathways. J. Immunol. 171: 1035–1041.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Xin L.,
    2. D. A. Vargas-Inchaustegui,
    3. S. S. Raimer,
    4. B. C. Kelly,
    5. J. Hu,
    6. L. Zhu,
    7. J. Sun,
    8. L. Soong
    . 2010. Type I IFN receptor regulates neutrophil functions and innate immunity to Leishmania parasites. J. Immunol. 184: 7047–7056.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Sakamoto E.,
    2. F. Hato,
    3. T. Kato,
    4. C. Sakamoto,
    5. M. Akahori,
    6. M. Hino,
    7. S. Kitagawa
    . 2005. Type I and type II interferons delay human neutrophil apoptosis via activation of STAT3 and up-regulation of cellular inhibitor of apoptosis 2. J. Leukoc. Biol. 78: 301–309.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Hockenbery D. M.,
    2. M. Zutter,
    3. W. Hickey,
    4. M. Nahm,
    5. S. J. Korsmeyer
    . 1991. BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl. Acad. Sci. USA 88: 6961–6965.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Lagasse E.,
    2. I. L. Weissman
    . 1994. bcl-2 inhibits apoptosis of neutrophils but not their engulfment by macrophages. J. Exp. Med. 179: 1047–1052.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    1. Moulding D. A.,
    2. C. Akgul,
    3. M. Derouet,
    4. M. R. White,
    5. S. W. Edwards
    . 2001. BCL-2 family expression in human neutrophils during delayed and accelerated apoptosis. J. Leukoc. Biol. 70: 783–792.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Cross A.,
    2. R. J. Moots,
    3. S. W. Edwards
    . 2008. The dual effects of TNFalpha on neutrophil apoptosis are mediated via differential effects on expression of Mcl-1 and Bfl-1. Blood 111: 878–884.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Fox C. J.,
    2. P. S. Hammerman,
    3. R. M. Cinalli,
    4. S. R. Master,
    5. L. A. Chodosh,
    6. C. B. Thompson
    . 2003. The serine/threonine kinase Pim-2 is a transcriptionally regulated apoptotic inhibitor. Genes Dev. 17: 1841–1854.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Simon H. U.,
    2. A. Haj-Yehia,
    3. F. Levi-Schaffer
    . 2000. Role of reactive oxygen species (ROS) in apoptosis induction. Apoptosis 5: 415–418.
    OpenUrlCrossRefPubMed
  49. ↵
    1. Zhang X.,
    2. D. P. Sejas,
    3. Y. Qiu,
    4. D. A. Williams,
    5. Q. Pang
    . 2007. Inflammatory ROS promote and cooperate with the Fanconi anemia mutation for hematopoietic senescence. J. Cell Sci. 120: 1572–1583.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    1. Daniel P. T.,
    2. T. Wieder,
    3. I. Sturm,
    4. K. Schulze-Osthoff
    . 2001. The kiss of death: promises and failures of death receptors and ligands in cancer therapy. Leukemia 15: 1022–1032.
    OpenUrlCrossRefPubMed
  51. ↵
    1. Secchiero P.,
    2. G. Zauli
    . 2008. Tumor-necrosis-factor-related apoptosis-inducing ligand and the regulation of hematopoiesis. Curr. Opin. Hematol. 15: 42–48.
    OpenUrlCrossRefPubMed
  52. ↵
    1. Gupta S.
    2002. A decision between life and death during TNF-alpha-induced signaling. J. Clin. Immunol. 22: 185–194.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section