SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

The purpose of this study was to assess the inflammatory nature of obesity and its effect on blood and bone marrow endothelial cell populations. Obese patients (BMI ≥30) had significantly higher concentrations of the inflammatory marker C-reactive protein (CRP) (P = 0.03) and lower concentrations of the anti-inflammatory cytokine interleukin-10 (IL-10) (P = 0.05). This cytokine profile is consistent with obesity being an inflammatory condition and is further supported by the significant correlation between total white blood cell count and BMI (r = 0.15; P = 0.035). High BMI was associated with significantly lower numbers of early endothelial cells (CD45/CD34+) in the bone marrow (r = −0.20; P = 0.0068). There was also a significant inverse correlation between BMI and a more mature endothelial cell phenotype (CD45/31+) in the blood (r = −0.17; P = 0.02). In addition, there was a significant correlation between BMI- and endothelial-related cells of hematopoietic origin (CD133+/VEGFR-2+) in the bone marrow (r = −0.26; P = 0.0007). Patients with higher plasma IL-10 and insulin-like growth factor (IGF) concentrations had higher numbers of endothelial phenotypes in the bone marrow suggesting a protective effect of these anti-inflammatory cytokines. In conclusion, this work confirms the inflammatory nature of obesity and is the first to report that obesity is associated with reduced endothelial cell numbers in the bone marrow of humans. These effects of obesity may be a potential mechanism for impaired tissue repair in obese patients.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Obesity, defined as BMI ≥30, has been described as an inflammatory process (1). Inflammation likely contributes to many of the obesity-related pathologies including cardiovascular disease and cancer. In the 19th and 20th centuries there was a major benefit from the development of the germ theory and resulting public health improvements including development of vaccines and antibiotics. This led to reduced deaths from infections and improvements in the human lifespan (2). The often catastrophic inflammatory responses associated with poor public health hygiene and infections, have given way to an equally important chronic inflammation associated with the characteristics of the modern lifestyle including obesity, and physical inactivity (3,4).

Two common systemic markers of acute and chronic inflammation are interleukin-6 (IL-6) and C-reactive protein (CRP). IL-6 is a potent stimulus for CRP production from the liver. Obesity has been proposed to represent a low-grade inflammatory process. Studies have reported elevated CRP, IL-6, and IL-8 in obese patients (5,6,7). Obesity as a risk factor for cancer, heart disease, and chronic diseases such as diabetes is relatively well established. It is likely that obesity-related inflammation plays a role in the development of cancer, heart disease, and diabetes (1,8,9).

IL-10 is the classic anti-inflammatory cytokine through its inhibition of nuclear factor-κB causing a subsequent decline in inflammatory cytokine production. The effect of obesity on IL-10 concentrations is not clear with conflicting reports (10). Insulin-like growth factor (IGF) is also considered an anti-inflammatory adipokine which is more closely tied to growth hormone levels which tend to decline in obesity. Generally, reports have not linked obesity with changes in IGF-1 concentrations (11).

There are data supporting an inflammatory component of cardiovascular disease due to endothelial cell damage and thrombosis, platelet activation, and instability of plaques. These processes lead to an increased risk of cardiac events and mortality (12). The role of obesity and accompanying inflammation is relatively well established in metabolic syndrome and diabetes, but only recently appreciated, is the effect of obesity and inflammation on the survival of bone marrow derived progenitor cells and related repair functions in cardiac disease (13,14,15,16).

Inflammation and the risk of cancer are associated and may involve angiogenesis. Many mediators of inflammation are also angiogenic. The link between angiogenesis and metastatic spread is well documented (17). Chronic inflammation associated with obesity may increase the risk of developing metastatic disease and may partially explain the poor prognosis associated with cancer in the obese. In addition to effects on prognosis, obesity is also a risk factor for the development of cancer (8,18). Generally, for tumors to become clinically detectable, a blood supply is required. Whether bone marrow and circulating progenitor cells contribute to tumor progression remains an area of controversy (19,20).

Various studies have reported that obesity affects endothelial progenitor cell numbers. A recent report by Tobler et al., described a negative correlation between BMI and the number of CD34+/CD133+/VEGFR-2+ cells in the blood suggesting that obesity has a negative effect on vascular repair (21). One of the difficulties interpreting the literature describing the effect of various progenitor cell phenotypes, is the accurate identification of these specific progenitor phenotypes. Several types of endothelial cell-related populations can be identified in the blood and bone marrow using flow cytometric phenotyping. The cells identified in the study by Tobler et al. (21), those with the CD133+/VEGFR-2+ phenotype, have been identified previously as endothelial progenitor cells but are more accurately categorized as endothelial-like cells of hematopoietic origin. CD45/CD34+ cells are a population of early endothelial cell progenitors while CD45/CD31+ phenotype identifies a more mature population (22,23). These populations exist both in the bone marrow and peripheral blood but, not surprisingly, these cells are less common and difficult to measure accurately in the blood which can limit the interpretation of blood data.

The purpose of this correlative study was to investigate the inflammatory nature of obesity and the effect of obesity on endothelial cell numbers in blood and bone marrow. The association between inflammatory cytokine concentrations in the plasma and endothelial cell phenotypes was also evaluated.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Human subjects

The National Institute of Aging supported this study of stem cells in humans with age and health status. Institutional review board approval was received to enroll and follow with informed consent, up to 240 persons undergoing total hip replacement. Exclusion criteria included a diagnosis of avascular necrosis, any abnormal bone marrow condition, a history of malignancy, or any previous chemotherapy or radiation therapy. Blood was drawn from each subject and samples of bone marrow were obtained from the trochanter.

Procedures

Samples. Patients had blood drawn into EDTA vacutainers in the preoperative area before surgery. Blood was centrifuged (1,000g) for 10 min and plasma collected, aliquoted, and stored at −80 °C for batched cytokine analysis using a sandwiched-type ELISA (R&D Systems, Minneapolis, MN). Bone marrow samples were collected from the upper femoral shaft of bone that would otherwise be discarded following a hip replacement, by the orthopedic surgeon at the time of surgery (24). A single-cell suspension was made by pulverizing the samples in a mortar and pestle using HBSS (Invitrogen, Carlsbad, CA) without Ca or Mg, containing 20% FBS (Hyclone, Logan, UT), 13.5 µg/ml DNAse (Sigma-Aldrich, St Louis, MO) and 10 U/ml sodium heparin (Elkins-Sinn, Cherry Hill, NJ).

Cytokine concentrations (IL-6, IL-10, VEGF, CRP, IL-8, IGF, SDF, adiponectin)

100–200 µl of plasma was analyzed for cytokine concentrations using sandwiched-type ELISAs according to the manufacturer's instructions (Quantikine; R&D Systems). The interassay and intra-assay coefficient of variations are <10% for all assays used.

Endothelial cell populations measurement

1 × 106 mononuclear cells from both bone marrow samples and peripheral blood were stained with fluorochrome-conjugated antibodies CD133PE, CD45FITC, CD34PE, CD31APC, and VEGFR2APC using standard phenotyping techniques (15 min, 4 °C). Analysis was performed using a FACSAria (Becton-Dickinson, San Jose, CA). Results were analyzed using FlowJo Software. Cells not expressing the common leukocyte marker CD45 are enriched for endothelial cells, when coupled with stem cell (CD34) and endothelial cell marker (CD31), relatively early (CD34+/CD45) and more mature endothelial populations (CD31+/CD45) can be identified. These cells of endothelial origin (CD45) are deemed true endothelial cells as they are capable of forming chimeric vessels when transplanted in mice (25). Another population of endothelial-related cells express CD45 (CD133+/VEGFR2+) but lack the functional capability of forming endothelial cell sprouts. However, these cells contribute to blood vessel formation by paracrine mechanisms and are identified as endothelial-like cells of hematopoietic origin (26,27).

BMI calculation

BMI was calculated using patient weight in kilograms and height in meters obtained at the preoperative clinical appointment. Weight was divided by height squared and reported as kilogram per meter squared (kg/m2). Obesity was defined as BMI ≥30 kg/m2 (1).

Data and statistical analysis

Cytokine concentrations were compared in obese patients (BMI ≥30 kg/m2) and nonobese patients and using Mann-Whitney Rank sum test. Simple linear regression was used to analyze the trends in endothelial and endothelial-like cell subsets and plasma cytokine concentrations. Statistically significant differences were defined as P ≤ 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Patients

A total of 233 patients were consented into this study. Table 1 describes patient characteristics with patients categorized as obese (BMI ≥30) and nonobese. While there was a wide range in ages in both obese and nonobese patients, obese patients were significantly younger than those who were defined as nonobese.

Table 1.  Characteristics of obese and nonobese patients
inline image

Blood and bone marrow subsets and obesity

Data was consistent with a decline in populations of endothelial cells and endothelial-like cells in the blood and bone marrow in obese patients. Higher BMI was associated with significantly lower number of differentiated endothelial cells in the blood (CD45/CD31+) (Figure 1a) and lower numbers of early endothelial cell progenitors (CD45−/CD34+) in the bone marrow (Figure 1b). Obesity was also associated with a lower number of endothelial-like hematopoietic progenitor cells in the bone marrow (CD133+/VEGFR-2+) (Figure 1c).

image

Figure 1. Relationship between BMI and endothelial cells (EC) and endothelial-like cells of hematopoietic origin (ECH) in blood and bone marrow. (a) EC (CD45/CD31+), measured as a percentage of blood mononuclear cells, significantly decline with increasing BMI. (b) Early EC (CD45/CD34+), measured as a percentage of bone marrow mononuclear cells, significantly decline with increasing BMI and (c) ECH (CD133+/VEGFR-2+), measured as a percentage of bone marrow mononuclear cells, significantly decline with increasing BMI.

Download figure to PowerPoint

Obesity and plasma cytokines

CRP was significantly higher, and IL-10 significantly lower, in obese compared to nonobese patients (Figure 2a,b), whereas IL-6 and IGF were not significantly different between obese and nonobese (Figure 2c,d). VEGF, IL-8, and SDF-1 levels were not significantly related to obesity (data not shown). Adiponectin concentrations strongly trended lower in patients who were defined as obese (BMI ≥30) (P = 0.06) (Figure 2e).

image

Figure 2. Comparison of cytokine plasma concentrations in nonobese (BMI <30) vs. obese (BMI ≥30). (a) C-reactive protein plasma concentrations (mcg/ml) significantly increase in obese compared to nonobese patients. (b) Interleukin-10 (IL-10; pg/ml) plasma concentrations significantly decrease in obese compared to nonobese patients. (c) IL-6 (pg/ml) not different between obese and nonobese patients. (d) Insulin growth factor-1 (IGF; ng/ml) not different between obese and nonobese patients. (e) Adiponectin (ng/ml) trends lower in obese compared to nonobese patients.

Download figure to PowerPoint

Blood and bone marrow phenotypes and plasma cytokine concentrations

To discern possible mechanisms by which obesity may influence stem cell numbers, the associations between circulating cytokine levels and endothelial cell subsets was evaluated. Plasma cytokine concentrations variously correlated with blood and bone marrow endothelial cell subset numbers. The anti-inflammatory cytokines, IL-10 (Figure 3a,b) and IGF (Figures 4a,b) directly correlated with bone marrow endothelial cell phenotypes but not endothelial-like cells of hematopoietic origin (Figures 3c and 4c). The classic inflammatory markers CRP and IL-6 did not correlate significantly with any of the blood and bone marrow endothelial cell phenotypes. In the blood, endothelial-like cells of hematopoietic origin (CD133+/VEGFR-2+) correlated inversely with IL-10 (Figure 5c) while early endothelial cells (CD45/CD34+) (Figure 5b) and more differentiated endothelial cell populations (CD45/CD31+) (Figure 5a) directly correlated with IGF concentrations.

image

Figure 3. Relationship between interleukin-10 (IL-10; pg/ml) plasma concentrations and endothelial cell (EC) and endothelial-like cell of hematopoietic origin (ECH) in the bone marrow. (a) Early EC (CD45/CD34+), measured as a percentage of bone marrow mononuclear cells, significantly increase with increasing IL-10 plasma concentrations. (b) EC (CD45/CD31+), measured as a percentage of bone marrow mononuclear cells, significantly increase with increasing IL-10 plasma concentrations. (c) ECH (CD133+/VEGFR-2+), measured as a percentage of bone marrow mononuclear cells, not correlated with IL-10 plasma concentrations.

Download figure to PowerPoint

image

Figure 4. Relationship between insulin growth factor-1 plasma concentrations (IGF; ng/ml) and bone marrow endothelial cell (EC) and endothelial-like cells of hematopoietic origin (ECH). (a) Early EC (CD45/CD34+), measured as a percentage of bone marrow mononuclear cells, significantly increase with increasing IGF plasma concentrations. (b) EC (CD45/CD31+), measured as a percentage of bone marrow mononuclear cells, significantly increase with increasing IGF plasma concentrations. (c) ECH (CD133+/VEGFR-2+), measured as a percentage of bone marrow mononuclear cells, not correlated with IGF plasma concentrations.

Download figure to PowerPoint

image

Figure 5. Relationship between insulin growth factor-1 (IGF; ng/ml) and interleukin-10 (IL-10; pg/ml) plasma concentrations and blood endothelial cell (EC) and endothelial-like cells of hematopoietic origin (ECH) subsets. (a) EC (CD45/CD31+), measured as a percentage of blood mononuclear cells, significantly increase with increasing IGF plasma concentrations. (b) Early EC (CD45/CD34+), measured as a percentage of blood mononuclear cells, significantly increase with increasing IGF plasma concentrations. (c) ECH (CD133+/VEGFR-2+), measured as a percentage of blood mononuclear cells, significantly decrease with increasing IL-10 plasma concentrations.

Download figure to PowerPoint

Total WBC count, inflammatory cytokines, and BMI

The white blood corpuscles (WBC) count for patients directly correlated with both CRP and IL-6 (Figure 6a,b). IGF-1 and IL-10 did not significantly correlate with WBC count (Figure 6d,e). There was also a direct correlation between BMI and WBC count suggesting that excess body weight modifies this clinical measure of infection and inflammation (Figure 6c).

image

Figure 6. Relationship between plasma cytokine concentrations and BMI vs. total white blood cell (WBC) count. (a) WBC significantly increases with increasing C-reactive protein (CRP; mcg/ml) plasma concentrations. (b) WBC significantly increases with increasing interleukin-6 (IL-6; pg/ml) plasma concentrations. (c) WBC significantly increases with BMI. (d) WBC not correlated with insulin growth factor-1 (IGF; ng/ml) plasma concentrations. (e) WBC not correlated with interleukin-10 (IL-10; pg/ml) plasma concentrations.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Because of the increasing numbers of obese individuals in the population, it is critical to investigate the mechanism of obesity-related declines in health. This study presents new findings that endothelial cell and endothelial-related cells of hematopoietic origin numbers in the blood and bone marrow are adversely affected by obesity. Elevated CRP and lower IL-10 plasma concentrations in the obese patients in this study were consistent with the inflammation associated with visceral obesity. Plasma concentrations of IL-10 and IGF directly correlated with bone marrow CD45/CD34+ and CD45/CD31+ but not CD133+/VEFGR2 cell numbers. The effects of IL-10 and IGF on bone marrow endothelial cell subsets suggests a protective effect of these cytokines on the bone marrow microenvironment. The decline in bone marrow endothelial cell subsets and endothelial-like cells of hematopoietic origin in persons with high BMI is consistent with the deleterious effects of inflammation associated with obesity. It indicates a profound negative effect of obesity on endothelial progenitor cells and potentially vascular repair.

Inflammation has both protective functions, which activate the innate and adaptive immune system, promote clotting, and effecting repair as well as having tissue damaging properties. Inflammation can be characterized as either acute or chronic. The duration of acute inflammation is short and characterized by redness and edema, with neutrophil emigration. While chronic inflammation can be preceded by an acute inflammatory process, it is often preceded by a subacute insult. The pathophysiology of chronic inflammation is very different from acute inflammation and characterized by lymphocyte and macrophage responses which lead to initial neovascularization and ultimately fibrosis (28).

Obesity has been characterized as a chronic inflammatory condition associated with elevated IL-6, CRP, and IL-8 plasma concentrations (29). In general, inflammation related to obesity has been characterized as “low-grade” inflammation (30). Visceral fat readily secretes inflammatory cytokines such as IL-6 and may be more associated with abdominal obesity (31,32). This increase in IL-6 concentrations from visceral fat is greatest in the interstitial fluid of fat deposits and not the plasma (31,32). In obesity, IL-6 may act as a paracrine cytokine, because IL-6 related to visceral fat drains into the portal system and the liver is able to clear a significant portion of the IL-6 delivered to the liver (32). Increased IL-6 in the case of obesity may act locally on the liver leading to CRP elevation systemically, without a major elevation in plasma IL-6 concentrations (32). The profile described above was seen in this study with higher CRP but not IL-6 concentrations of in the plasma of obese patients.

In this present work IL-10 was significantly lower in the plasma of patients categorized as obese. IL-10 can be characterized as an anti-inflammatory cytokine, and a decline in plasma concentrations IL-10 in obese patients further contributes to the inflammatory environment associated with obesity.

The significant decline in endothelial cells in the bone marrow of obese patients in this study may have serious health consequences as these cells are integral to tissue repair processes. The inflammatory profile suggested by elevated CRP and lower IL-10 plasma concentrations in obese patients may damage the bone marrow compartment leading to lower numbers of progenitor cells observed in this study. Experiments in vitro have demonstrated that CRP lowers the number of endothelial progenitor cells in culture and this link between inflammation and a decline in progenitor cell numbers is a proposed explanation for increased cardiovascular risk (33,34). In addition to this effect on progenitor cell numbers, a recent report showed obesity impaired endothelial cell function including reduced adhesion, migration, and homing capacity. This functional impairment results in reduced angiogenesis. These effects may be a result of modified angiogenic cytokine secretion (34).

The inflammatory cytokines, CRP, and IL-6, were significantly correlated with WBC count and has previously been reported as a predictor of cardiac morbidity and mortality (35,36,37). The increase in WBC count in obese patients is an interesting finding and may allow this commonly measured clinical test to be used in concert with inflammatory markers to assess risk in obese patients. Future work on obesity should include total WBC as a covariate along with other more established measures of chronic inflammation such as CRP.

In conclusion, given the increasing numbers of obese individuals in the population, it is critical to investigate the mechanisms of obesity-related declines in health. We report a significant decline in endothelial cell numbers in the bone marrow of obese patients, which may have serious health consequences, since these cells are integral to tissue repair processes. These negative effects on bone marrow endothelial cell populations may at least in part, be related to obesity-related chronic inflammation. This may lead to the beneficial use of drugs, like statins, which have anti-inflammatory activities as well as promoting weight loss, although weight loss has been difficult to achieve for the majority of obese patients.

ACKNOWLEDGEMENT

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

We thank Judy Lane, Connie Feschuk, and Dana Swartz for the excellent clinical and technical support and the UNMC and Creighton University Flow Cytometry Core Facilities for progenitor cell phenotyping. These studies were supported in part by the National Institute of Aging (AG024912) and this support is gratefully acknowledged.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References