|Home | About | Journals | Submit | Contact Us | Français|
Tobacco use is associated with an increase in white blood cell count (WBC). This association has been attributed to bronchopulmonary inflammation and/or infection. It is not known if nicotine itself may play a role.
We determined if nicotine itself could effect WBC count, and determined if this was due to a direct effect on hematopoietic stem cells (HSC).
C57Bl6J mice received nicotine orally, and measurements of WBC count; bone marrow and spleen cellularity; and HSC count were made. To determine the functionality of HSCs, irradiated animals received bone marrow transplants from vehicle or nicotine treated mice.
Nicotine increased leukocytes in the peripheral blood, bone marrow and spleen. Peripheral red cell and platelet count were unaffected. Nicotine increased the frequency of HSC in the bone marrow. Isolated long-term HSCs from nicotine-treated mice transplanted into irradiated mice regenerated all hematopoeitic cell lineages, demonstrating functional competence of those HSCs. HSC expressed nicotinic acetylcholine receptors (nAChRs), as documented by FITC-conjugated alpha-bungarotoxin binding. Nicotine increased soluble Kit ligand, consistent with stem cell activation.
The data suggest a new mechanism for the increased WBC associated with tobacco use. The effect of nicotine to activate hematopoiesis may contribute to tobacco-related diseases.
Tobacco use is associated with an increased white blood cell count(1–3). This association may be of clinical significance as an increased white blood cell count is an independent risk factor for coronary heart disease and stroke.4–6,7 The increased WBC observed in smokers may not be entirely related to bronchopulmonary inflammation and/or infection.8 Nevertheless, cigarette smoke contains constituents such as phenols (e.g. 3-methylphenol) and carbonyls (e.g. acrolein) that are known irritants of the bronchopulmonary mucosa. Notably, cigarette smoke activates neutrophils and increases their generation of superoxide anion.9,10
Nicotine is a major active component of cigarette smoke and some adverse effects of smoking are attributable to this alkaloid.11 Although most of the toxicity of tobacco exposure is related to other components, nicotine is responsible for the neurobehavioral effects and addiction.12 Furthermore, nicotine stimulates the release of catecholamines, thereby increasing vascular tone, and inducing inotropic and chronotropic effects on the heart.13 These sympathomimetic effects have the untoward effect of increasing blood pressure and myocardial oxygen demand. Nicotine may also cause endothelial activation and vasodilator dysfunction, associated with the expression of adhesion molecules and chemokines that can facilitate vascular inflammation.14–16 Finally, nicotine can induce pathological angiogenesis.17,18 Whereas these effects might be anticipated to accelerate atherogenesis and precipitate major adverse cardiovascular events, short term use of nicotine patches has been shown to be safe in smokers with cardiovascular disease.19
More recently, non-neuronal nicotinic acetylcholine receptors (nAChRs) have been recognized to mediate some of the systemic effects of nicotine.14,17,20.21 The nAChRs are a class of ligand-gated ion channels that are variably permeable to sodium and calcium.22 Notably, circulating peripheral blood cells contain nAChRs.23,24 In the current study we tested the hypothesis that nicotine may directly affect WBC count.
Female C57/Bl6 mice were obtained from Jackson Laboratories and maintained by the Department of Laboratory Animals and Management at Stanford University. Mice were fed a normal chow diet (Purina) and given vehicle (0.4% saccharine in water) or nicotine solution (100 ug/ml nicotine in vehicle) to drink ad libitum. At 3, 6, 9, 14, 42 or 182 days after initiation of the study drug, groups of mice were sacrificed and the total cells from the bone marrow, spleen and periphery were isolated and prepared for assessment of cell number and composition. Unless otherwise stated, studies began with mice at 12 weeks of age.
Mice were anaesthetized by intraperitoneal injection of xylazine/ketamine (16 and 80 mg/kg body weight respectively) and blood was obtained via cardiac puncture and collected into standard tubes (coated with EDTA as an anticoagulant) for complete blood cell counts (CBC) and fluorescence activated cell sorting (FACS) or flow cytometric analysis as described below.
The spleen was harvested, fatty tissue carefully removed and the spleen weighed. The spleen was then pulverized in the presence of HBSS/2% FBS(staining buffer - SB) and strained through a 0.34 um filter mesh into a 50 ml Falcon tube. Total final volume was about 15 ml. An aliquot of cells (about 106 cells) was processed for FACS analysis of B and T cells (CD8-FITC and CD4-PE and B220-Texas Red).
Bone marrow cells were removed from the tibial and femoral bones of both hindlimbs by aspiration using an 18 gauge needle connected to a syringe containing 1ml of staining media in HBSS/2% FBS. Subsequently the cells were resuspended in 10ml of HBSS/2% FBS, centrifuged (1000xg, 5 min, 4°C) and resuspended in 300 ul of ACK solution (a hypotonic solution consisting of 0.83% NH2Cl/0.1%KHCO3 in water) for 3–5min to lyse red blood cells. To stop the lysis, the cells were resuspended in 10 volumes of HBSS/2% FBS. The cells were then centrifuged (1000xg, 5min, 4°C), the supernatant removed and finally the cells were resuspended in 10 ml HBSS/2% FBS. The numbers of cells per ml were determined via standardized hemocytometric measurements and an aliquot of 1–2 million cells was reserved for flow cytometric analysis or sorting in the transplantation studies (see sections below). Cell suspensions (10 μl) were mixed with 190 ul Trypan blue. After 10 minutes of exposure, an aliquot of 15 μl was used for cell count using a standard hemocytometer and inverted light microscope.
We used hemoglobin content as a surrogate marker for bone marrow and splenic RBC count. We followed a modified protocol of the Drabkin assay.25 Briefly, isolated bone marrow and splenic cells were centrifuged to a pellet(1000xg, 5min, 4°C), the supernatant is drawn off and the cells resuspended in 500 ul of ACK solution for 5 minutes at 4°C. After 5 minutes an additional 500 ul of ACK solution is added to complete the lysis of rbcs. The cells are then centrifuged(1000xg, 5min, 4°C), and the supernatant removed. To assay for hemoglobin content, one ml of the supernatant is mixed with 4 ml of a potassium ferricyanide/sodium azide mix known as Drabkin’s reagent (Sigma, St. Louis, USA) and let stand in the dark for at least 15 minutes. The mixture is then assessed for absorbance at 540nm. Actual hemoglobin content is then determined by comparison to a standard curve of dilutions (0–0.72 mg/ml protein) to bovine hemoglobin (Sigma, St. Louis, USA).
All flow cytometric analyses and fluorescence-activated cell sorting (FACS) were performed at the Shared FACS Facility at Stanford University using methods described previously.26,27 To detect myeloid cells we used the fluorescently labeled antibodies GR1-FITC and MAC1-PE. To detect B cells, B-220-Texas Red was used.
Hematopoietic stem cells (HSCs) were identified by negative staining for FITC-conjugated antibodies staining against the following lineage antigens: CD3, CD4, CD5, CD8 (T-cells), B220 (B-cells), Grl (Granulocytes), Mac1 (macrophages) and Ter 119 (erythrocytes) and positive antibody staining with anti-Sca1-Texas Red and anti-cKit-APC. Fluorescent antibodies to PE-conjugated Flk2 were used to differentiate short-term (ST, Flk2-positive) from long-term (LT, Flk2-negative) HSCs.28 Myeloid progenitors were defined as c-kit+Lineage−Sca1− cells.
LT-HSC from 18-week old (control or nicotine-treated) mice (Ly5.1+ strain) were purified by FACS using the protocol described above.28 Purified LT-HSC were transferred via retro-orbital venous injection into irradiated host mice(Ly5.2+ strain). The host mice were not treated with nicotine. Transfer of 25 or 100 cells was accompanied by 3×105 Sca-1 depleted bone marrow cells from the host strain. After 3 months, the mice were sacrificed, and blood cells analyzed as described above, to determine if there was donor-derived reconstitution.
The snake venom peptide, alpha-bungarotoxin (abgt) is an irreversible antagonist with high affinity for the alpha7 and alpha 1 subunits of the nAChR, making it useful for identification of these receptor subunits.29–33 Total bone marrow cells or FACS sorted HSC were exposed to 10 ug/ml FITC-abgt (Sigma-Aldrich) for 2 hr at 20°C in HBSS/2% FBS. Control experiments to test for non-specific binding were performed by first adding 50-fold excess of unlabelled abgt for 30 min at 20°C followed by 1.5 hr addition of FITC-abgt. FITC positive cells were visualized either through fluorescence microscopy on a Nikon TE-2000U or through FACS analysis.
Plasma NOx was measured using the Griess reaction via a colorimetric assay (Cayman Chemicals). Plasma cotinine, a stable metabolite of nicotine, was measured by ELISA (OraSure Technologies). Plasma MMP-9 and soluble Kit ligand (SKitL) were measured by Quantikine ELISA (R&D Systems).
Comparisons between control and nicotine-treated mice were analyzed either by Student’s t-test (unpaired, two-sided) or 1-way ANOVA for experiments with more than two groups. One-way ANOVAs with post-hoc tests for linear trends were performed using GraphPad Prism version 4.00 for Windows. All data expressed as mean + SEM. Except where otherwise mentioned, the sample size, N, represents number of animals studied for both control and nicotine-treated mice. P-values of less than 0.05 were considered to be statistically significant.
The administration of nicotine ad libitum in the drinking water resulted in plasma cotinine levels (40–80 ng/ml) that are similar to those in light to moderate smokers.34 Plasma cotinine was undetectable in vehicle-treated animals. The administration of nicotine had no effect on red blood cell count (Figure 1a) mean corpuscular volume or mean corpuscular hematocrit count. Platelet count was also unaffected. By contrast, nicotine had striking effects on white blood cell count. After a transient decline at 3 days, there was a sustained increased in WBC count, peaking at 6 weeks (N=10 in each group, p<0.01, Figure 1a) was observed in nicotine treated mice. However, by 182 days, the WBC count was not different between nicotine-treated and control animals.
In a second set of animals, the previous observations were confirmed. Absolute numbers of circulating wbcs for control vs. nicotine-treatment at 6 weeks were 4.6+0.3 K/ul vs and 7.2+0.7 K/ul (N=10, p<0.01, control vs. nicotine) respectively. The increase in cell count occurred across all leukocyte classes, and accordingly there was no alteration in the relative percentages of neutrophils, lymphocytes, and monocytes. Once again, we observed that nicotine did not change circulating RBC counts (9.7+0.6 vs 8.9+1.0 (M/ul) for control vs. nicotine treated mice respectively, N=10, NS).
To address the mechanisms of the increased WBC, we studied the cellularity and cell composition of the bone marrow and spleen in this set of animals. For these analyses, red blood cells were removed by lysis to focus on the leukocyte populations. Notably, nicotine caused changes in bone marrow and spleen cellularity that mirrored the temporal course of its effects on peripheral white blood cell counts (Figure 1b). Total cell counts in the bone marrow and spleen were transiently reduced at 3 days followed by an increase (Figure 1b). By six weeks of nicotine exposure, total cell counts in the bone marrow and spleen were significantly elevated (N=15 in each group; p<0.005 for bone marrow and p<0.004 for spleen; Figure 1b). Similarly, nicotine induced directionally and temporally similar changes in spleen weight. Nicotine administration caused a transient fall in spleen: body weight ratio at 3 days (0.33+/−0.03 vs. 0.39+/−0.01; Nicotine vs. Control; p< 0.001; N=8 in each group), which normalized at two weeks, (0.36+/−0.02 vs. 0.38+/−0.02; Nicotine vs. Control; p=NS; N=8 in each group) progressing to a significant elevation in the ratio at 6 weeks (0.43+/−0.02 vs. 0.38+/−0.03; Nicotine vs. Control; p<0.001; N= 14 in each group).
In a separate series of studies, we confirmed the effects of nicotine to increase circulating wbc levels, and to increase bone marrow cellularity. At 6 weeks the increased white blood cell count was associated with increased bone marrow cellularity (N=15 in each group; p<0.001 versus control; Figure 1c). To determine if the increase in bone marrow or spleen cellularity was secondary to a specific leukocyte subpopulation, flow cytometric analysis was performed. The administration of nicotine did not induce a change in the frequency of neutrophils, B lymphocytes, T lymphocytes or monocytes in the bone marrow. Similar observations were made in the spleen, except for a slightly increased frequency of CD8 T cells (15% relative increase over control).
To determine if the increase in WBC and in cellularity of the bone marrow and spleen was associated with an amplification of HSCs, we performed flow cytometric analyses for HSCs. In the bone marrow, nicotine administration for six weeks increased HSC frequency 2.5 fold (Figure 2a). With G-CSF treatment, increase in HSC number is seen in the bone marrow within 2–3 days after exposure.35 Nicotine, by way of contrast, did not significantly increase HSC pools by 3–14 days of exposure (data not shown). Although the Drabkin’s assay detected higher levels of hemoglobin in the spleen of nicotine-treated animals at 6 weeks, we detected no significant increase in HSC in the spleen or blood at any time point. We next determined if there was an increase in the relative populations of long-term (LT-HSC) vs. short-term HSCs (ST-HSC). Both LT-HSC and ST-HSC subpopulations were increased in the nicotine treated animals at 6 weeks (Figure 2c and d).
We examined the relationship between total BM cell number and HSC number at 6 weeks of treatment with vehicle or nicotine. There was a positive correlation between the frequency of HSC and the total BM cell number (Figure 3a–b). A similar relationship between HSC frequency and spleen/body weight ratio was observed (data not shown). Flow cytometric analysis however yielded no significant change in HSCs in the spleen for nicotine-treated mice (HSCs in the spleen were infrequent in both groups, 10–20 HSCs/million cells). In preliminary studies, we found evidence for a nicotine-induced increase in myeloid progenitors. Specifically, after 6 weeks of treatment, nicotine-exposed mice (n=5) manifested a 2-fold increase in the frequency (p<0.005), and a 4-fold increase in the number (p< 0.001), of myeloid progenitor cells compared to vehicle-treated animals (n=5).
It is possible that nicotine would increase HSC number but not HSC function. Alternatively, a nicotine-induced increase in HSC proliferation could alter replicative capacity. To assess the function of HSCs, we isolated LT-HSCs (by FACS) from animals exposed to nicotine or vehicle for six weeks. The LT-HSCs were transplanted into lethally irradiated mice. Three months after transplantation of LT-HSCs, peripheral blood was obtained from the recipient mice and analyzed by FACS for B-cells (B220), T-cells (CD3), and macrophages (Mac1). Recipient mice displayed multilineage reconstitution of all major mononuclear cell types. Thus no significant differences in the ability of LT-HSCs from the nicotine treated animals to reconstitute the bone marrow (Figure 3c) were observed.
It is possible that nicotine acts directly on HSCs through the nAChR. RT-PCR revealed detectable levels of alpha7-nAChR subunit expression in the murine bone marrow (data not shown) but not in the spleen. To determine the level and distribution of nAChR protein expression on HSC, we incubated whole bone marrow mononuclear cells (4a) or purified HSC (4b) with the high-affinity alpha7-nAChR ligand, alpha bungarotoxin (abgt), conjugated to FITC and used FACS to detect for HSCs expressing nAChRs. The FACS data showed that a subpopulation of total bone marrow mononuclear cells expressed a receptor for agbt (4a). This binding was specific, as a 50-fold excess of unlabelled abgt abrogated the increase in in fluorescence intensity observed with FITC-abgt (4a). Similarly, HSC (~40%) manifested specific FITC-abgt staining (Figure 4b).
Subsequently, we used FACS to isolate Sca1+ cells from bone marrow, and then stained this HSC-enriched fraction with FITC-labeled alpha-bungarotoxin.30,33 As shown in Figure 4c, many Sca1+ cells stained very faintly for alpha-bungarotoxin binding, with about 30% of cells staining strongly.
Plasma levels of nitric oxide and sKit ligand have been used as markers of stem cell activation.36–42 At six weeks of nicotine administration, we observed increases in plasma levels of NOx (Figure 5a). Plasma sKitL levels were also elevated by nicotine (Figure 5b), within the range of reported levels for sKitL in the circulation (20–200 pg/ml).43 Plasma levels of MMP-9, were not elevated (Figure 5c) by nicotine.
The salient findings of this investigation are: 1) Systemic administration of nicotine increases the WBC; 2) this effect is associated with an increase in nucleated cells in the bone marrow and spleen, as well as 3) increased numbers of short and long-term hematopoietic stem cells, and 4) plasma markers of HSC mobilization. The HSCs from nicotine-treated animals are capable of engraftment and reconstitution of both myeloid and lymphoid cell lineages, indicating that the increase in HSC numbers does not alter replicative capacity. These data suggest the existence of novel pathway for hematopoiesis mediated by nicotinic acetylcholine receptors (nAChRs). This hypothesis is supported by the demonstration of nAChRs on a subpopulation of HSCs.
Our data are consistent with a previous report by Koval and colleagues.44 They showed that in mice deficient for the α7 nAChR there was a reduction in myeloid and erythroid lineages. This latter finding is nearly the mirror image of our finding that nicotine-induced activation of nAChR in the bone marrow increases the proliferation of HSCs and increases the cellularity of the bone marrow. Notably, in α7 deficient mice, alpha-bungarotoxin binding disappeared from the bone marrow, confirming the specificity of the alpha-bungarotoxin for α7 nAChRs in the bone marrow cells. These data are consistent with our documentation of α7 nAChRs on HSCs.
Similarly, Pandit and colleagues provided evidence that nicotine could stimulate hematopoiesis.45 Specifically, in BALB/c mice, a subcutaneous depot of nicotine, released over 21 days, increased hematopoiesis in the spleen (although had little effect on the bone marrow). In the spleen there was a striking increase in total colony formation as well as eosinophil-, granulocyte-macrophage-, and B-lymphocyte-specific progenitors. The major difference between our study, and that of Pandit et al.,45 was the site of increased hematopoiesis induced by nicotine(spleen in Pandit et al, and bone marrow in our investigation). Perhaps this difference was related to the different strains of mice using in their study (BALB/c) versus ours (C57/Bl6).
We did find that nicotine increased spleen to body weight ratio, as well as hemoglobin content, phenomenon which could be due to extramedullary hyperplasia in the spleen and/or increased cellular sequestration. However we did not observe an increase in splenic HSCs with nicotine. In the absence of any increase in splenic hematopoietic stem cells, presumably the increased content of RBCs and WBCs in the spleen is due to sequestration in our model.
The nAChR is a ligand-gated calcium channel formed by a pentameric complex of nAChR subunits. The endogenous ligand is acetylcholine, which is a known neurotransmitter activating neuronal and neuromuscular nAChRs. We and others have described non-neuronal nAChRs that mediate other functions requiring intercellular communication including angiogenesis and inflammation.17,20,23,24 Nicotine hijacks this pathway to induce pathological neovascularization such as that observed in tumors, atherosclerotic plaque, or choroidal neovascularization.17,20 The current data provide a novel mechanism for another tobacco-related pathology, ie. an increased WBC. Smokers are known to have elevated white blood cell counts, which may be in part due to chronic irritation of the bronchopulmonary mucosa. In addition, smokers are at greater risk of bronchopulmonary infection. We have found evidence that nicotine may play a role in the increased WBC associated with tobacco.
In the first few days of nicotine exposure, we observed a transient reduction in WBC count in the periphery and bone marrow. We speculate that this early reduction in WBC counts in the peripheral blood and bone marrow may be due to sequestration of circulating leukocytes by the vascular endothelium and perhaps their infiltration into parenchymal tissues. Nicotine is known to enhance the adherence of hematopoietic stem/progenitor cells (HSPC) to the endothelium, an effect blocked by mecamylamine.46 Furthermore, nicotine induces the expression of endothelial adhesion molecules and chemokines, and increases endothelial transmigration.47 These actions of nicotine may have the effect of increasing WBC retention in peripheral tissues.
Smoking increases plasma fibrinogen levels and viscosity, as well as platelet count and hematocrit.48 The erythrocytosis is believed to be secondary to an increase in carboxyhemoglobin, resulting in hypoxia-induced increases in erythropoietin. In our study, platelet and red blood cell counts in the circulation were not significantly affected. The lack of a comparable increase in peripheral rbc and platelet count suggests that precursor cells for these lineages are insensitive to nACR stimulation, or their levels are more tightly (homeostatically) regulated; or they are sequestered in the spleen. In preliminary studies, we have observed evidence for the latter mechanism, ie. an increase in spleen and bone marrow hemoglobin content (a surrogate for erythrocyte mass). The effect of nicotine on erythropoiesis and thrombopoiesis requires further study.
Although the clinical literature supports a linkage between nicotine and the increased WBC associated with tobacco use,1,48 there are limited and conflicting mechanistic data. The existence of nACh receptors on leukocytes23,49 has been demonstrated. These receptors may mediate the effect of nicotine (in vitro exposure for 2 weeks) to suppress clonal expansion of hematopoeitic and T-cell lineages of murine bone marrow isolates.50–52 Furthermore, longer term exposure of nicotine (10−7M and above) in vitro significantly down-regulates several nicotinic acetylcholinergic (nAChR) receptor subunits in leukemic T-cell lines.50 By contrast, in vivo studies found that chronic exposure to nicotine increased the surface density of select subunits of nAChR in peripheral mononuclear cells.23,24 The discordance in these observations may be related in part to differences in the duration and dose of nicotine exposure. Indeed, in another model system (endothelial cell proliferation and angiogenesis) nicotine has biphasic effects that are dose- and time-dependent.20,53 Indeed, long-term administration of nicotine downregulates vascular alpha 7 nAChRs and impairs angiogenesis.54
In this regard, chronic exposure to nicotine may adversely affect hematopoiesis. Koval and colleagues(44) found that after long-term exposure to oral nicotine, there was a reduction in the murine bone marrow of myeloblasts, promyelocytes and erythroblasts. This apparent discordance with our data is explained by the fact that they exposed mice to nicotine in the drinking water at a concentration of 200ug/ml for 10 months, which was twice as much nicotine for almost twice as long as the studies carried out by us. They suggested that their results were due to receptor desensitization of α7-containing nicotinic receptors which play a major role in regulating hematopoiesis. Of note, in our study, the effect of nicotine to increase the WBC count seemed to diminish between days 42 to 182. So in our study, even at a lower dose of oral nicotine, we may have begun to induce some nAChR desensitization.
The response of the hematopoietic system to nicotine may be different during development. Serobyan et al studied the effects of nicotine on the murine fetal bone marrow.55 They provided evidence that nicotine impaired colonization of fetal bone marrow by HSPC immigrating from the fetal liver, thereby reducing the number of mature lymphocytes, and the strength of the immune response in neonates. Differences between our work and theirs are likely due the models (adult versus fetal mice) and/or the mode of nicotine administration (intravenous versus oral). Notably, nicotine accumulates in fetal tissues, “likely due to immature enzymatic activity in the placenta and fetal liver”, to quote Serobyan and co-workers.55
In any event, our studies indicate that nicotine may have direct effects on hematopoietic stem cell proliferation and activation. Stem cells have both differentiation and self-renewal capacities. Short-term HSCs (which are flk2+, KLS cells) have a limited capacity for self-renewal, whereas long-term HSCs (flk2−, KLS cells) have an indefinite capacity for self-renewal. Both LT-HSC and ST-HSC may differentiate into any of the leukocyte lineages. We found that nicotine increased the number of both subtypes of HSCs as well as numbers of each lymphoid and myeloid cell lines in a sustained manner. The functionality of the long-term HSCs in the nicotine-treated animals was confirmed by documenting their ability to establish multilineage reconstitution of leukocyte lineages in lethally irradiated mice.
Indeed, the reconstitution levels upon transplantation of nicotine-treated HSC tended to be greater than those from vehicle-treated animals. This observation is consistent with our data showing that nicotine treatment increases HSC (Figure 2) and WBC (Figure 1) numbers and may indicate a proliferative effect of nicotine on HSC, ie. there may be an enhancement of HSC engraftment and/or regeneration of the hematopoietic system with stimulation of the HSC nAChR.
Although the effects of nicotine on the bone marrow are similar to those of granulocyte colony stimulating factor (GCSF), there are important differences. The effects of GCSF on mobilization and proliferation of HSCs, with attendant leukocytosis, occur within 2–3 days35 and immature HSCs can appear in the circulation. Like GCSF, nicotine increases plasma levels of nitric oxide and sKit ligand, biomarkers of stem cell mobilization,36–42,56 Nitric oxide is believed to nitrosylate and thereby activate matrix metalloproteinases, which then cleave the tethered ligand cKitL.57 This produces sKitL and releases HSCs from their niche, potentially inducing their proliferation and migration. We did not observe an increase in plasma levels of MMP-9, but we did not test for an increase in expression or activation of MMP-9 locally in the bone marrow.
Nicotine recruits circulating endothelial precursors to sites of tumor neovascularization in murine models.58 This is consistent with previous observations that nicotine can induce the release of vascular endothelial growth factor as well as fibroblast growth factor from endothelial cells, and that systemic administration of nicotine can increase circulating levels of these growth factors.59, 60 Furthermore, it is likely that bone marrow derived endothelial progenitor cells share a common precursor with hematopoeitic stem cells.56,61 In explanation of our data and that of Natori and colleagues,58 there may be a common precursor for endothelial cells and leukocytes that is responsive to stimulation by endogenous acetylcholine or nicotine. Indeed, we found PCR, FACS and immunohistochemical evidence for nACh receptors on a subpopulation of HSCs. The presence of alpha-bungarotoxin binding sites on 20–40% of the HSCs suggests an important role for endogenous cholinergic regulation of stem cell activation and proliferation.
Exogenous nicotine may hijack this pathway to stimulate HSC activation, proliferation, and differentiation. If so, the question arises as to whether chronic stimulation by nicotine (as with smokers) could induce a “replicative exhaustion”. Indeed, in chronic smokers, circulating endothelial progenitor cells are reduced, and tobacco cessation improves the numbers of circulating EPCs.62 However, we did not observe a decline in the numbers of long-term HSCs, nor their functionality after six weeks of nicotine exposure. More chronic exposure to nicotine could ultimately induce a “replicative exhaustion”. Indeed, it could be argued that there is subtle evidence within our data for an early nicotine-induced decline in replicative capacity. Figure 3 shows that, when both nicotine-treated and control animals are considered, there is a positive correlation between the frequency and number of HSCs, and the cellularity of the bone marrow. This is not an unexpected finding. However, the slope of this correlation tends to be shallower when considering only the nicotine-treated animals. This observation could reflect a decline in the replicative capacity in the HSC from nicotine-treated mice. Another possibility is that the nicotine-induced increase in HSC may saturate the limited number of HSC niches in the bone marrow thereby rendering a more shallow correlation between HSC frequency and total bone marrow number. The possibility that stem cell niches in the bone marrow may become saturated has experimental support.63,64
In summary, we find that chronic exposure to nicotine induces an increase in WBC and in cellularity of the bone marrow and spleen. All white blood cell lineages seem equally affected. These effects are associated with an increase in hematopoietic stem cells in the bone marrow. A subpopulation of the HSCs expresses the nACh receptor. Nicotine may exert its effects directly through this receptor, and/or by activating mediators of HSC mobilization. These findings suggest a novel mechanism for the increased WBC associated with tobacco, and suggest a role for cholinergic regulation of hematopoiesis.
This study was supported by grants from the National Institutes of Health (RO1 HL-75774, R01 CA098303, R21 HL085743, RC2HL103400, 1U01HL100397), the Tobacco Related Disease Research Program of the University of California (18XT-0098), Philip Morris USA Inc, and by the Cancer Research Institute (for ECF). Dr. Cooke is an inventor on Stanford University patents related to therapeutic modulation of angiogenesis by agonists or inhibitors of the nACh receptors. These patents have been licensed to Comentis Inc., in which company Dr. Cooke has equity.
Edwin Chang: Collection and/or assembly of data; Data analysis and interpretation, Manuscript writingC. Camilla Forsberg: Collection and/or assembly of data; Data analysis and interpretation, Manuscript writing
Jenny Wu: Collection and/or assembly of data; Data analysis and interpretation
Bingyin Wang: Collection and/or assembly of data
Susan S. Prohaska: Collection and/or assembly of data; Data analysis and interpretation
Rich Allsopp: Collection and/or assembly of data; Data analysis and interpretation
Irving L. Weissman: Financial support, final approval of manuscript
John P. Cooke: Financial support, conception and design, administrative support, manuscript writing, final approval of manuscript