The authors have declared that no competing interests exist.
Conceived and designed the experiments: AGQ FG. Performed the experiments: MA LV LL JC. Analyzed the data: AGQ FG. Wrote the paper: AGQ JC MA.
Assessment of serum concentration of lipopolysaccharide (LPS)-binding protein (LBP) has been suggested as a useful biomarker to indicate activation of innate immune responses to microbial products. We investigated LBP concentrations and associations with demographics, lifestyle factors, and common metabolic abnormalities in adults. We also examined if LBP concentrations were associated with common polymorphisms in genes coding for LBP (rs2232618), CD14 (rs2569190), and TLR4 (rs4986790), the molecules responsible for the innate immune response to LPS, or serum levels of soluble CD14 (sCD14) and proinflammatory cytokines.
Serum LBP was measured with a commercial immunoassay in a random sample of the adult population (n = 420, 45% males, age 18–92 years) from a single municipality.
Serum LBP concentrations increased with age (P<0.001) and were higher in individuals who were overweight or obese than in normal-weight individuals (P<0.001). Similarly, LBP concentrations were higher in individuals with metabolic syndrome than in individuals without it (P<0.001). Among metabolic syndrome components, LBP concentrations were independently associated with abdominal obesity (P = 0.002) and low concentrations of HDL-cholesterol (P<0.001). Serum LBP concentrations tended to be independently associated with smoking (P = 0.05), but not with alcohol consumption. Likewise, there was not significant association between LBP concentrations and gene polymorphisms. Concentrations of LBP significantly correlated with serum levels of proinflammatory cytokines (IL-6 and IL-8), sCD14, and with liver enzymes.
Serum LBP concentrations increased with age. Overweight, obesity, and having metabolic syndrome (particularly, low HDL cholesterol levels) were associated with higher LBP concentrations. These findings are consistent with microbial exposure playing a role in these inflammatory, metabolic abnormalities.
A dysregulated response to bacteria or their products such as lipopolysaccharide (LPS, endotoxin) underlies many common inflammatory diseases
Assessment of LBP concentrations in serum or plasma has been suggested as a useful marker during systemic infectious complications
The present study aimed to investigate LBP concentrations in a general adult population and the potential relationships with demographic factors, lifestyle factors, and common metabolic abnormalities. In addition, we investigated whether LBP levels were associated with common SNPs in molecules at the interface of the innate immune response to LPS (CD14, TLR4, and LBP itself), serum concentrations of sCD14, and proinflammatory cytokine levels.
This cross-sectional study was included in a survey of the general population in the municipality of A-Estrada (Spain), as detailed elsewhere
The source study (FIS1306/99) was reviewed and approved by the Institutional Review Board of the Complejo Hospitalario Universitario from Santiago de Compostela (Spain). The present study (PGIDIT06PXIB918313) was reviewed and approved by the Clinical Research Ethics Committee from Galicia (Spain). Written informed consent was obtained from each participant in the study, which conformed to the current Helsinki Declaration.
All individuals underwent a physician-administered questionnaire. Alcohol consumption was evaluated as the number of standard drinking units (glasses of wine [∼10 g], bottles of beer [∼10 g], and spirits [∼10 g]) regularly consumed per week. Individuals were classified as abstainers/occasional drinkers (<10 g/week), light-moderate drinkers (10–200 g/week) or heavy drinkers (≥210 g/week). Consumers of at least one cigarette per day were considered smokers.
Body mass index (BMI) was calculated as the weight (in kilograms) divided by the square of height (in meters). Individuals were classified according to BMI as normal-weight (<25 kg/m2), overweight (25–30 kg/m2), or obese (>30 kg/m2). According to the Adult Treatment Panel III
Serum LBP was measured by chemiluminescent enzyme immunoassay (Immulite, Siemens Medical Solutions, Gwynedd, UK). The upper limit of the calibration range is 200 µg/mL and the analytical sensitivity is 0.2 µg/mL. The intra- and interassay coefficients of variation for the assay were <11%. According to the manufacturer, a study performed on 160 apparently healthy volunteers yielded a mean of 5.3 µg/mL, a 95th percentile of 8.4 µg/mL and an absolute range of 2.0 to 15.2 µg/mL.
Serum levels of liver enzymes were measured in an Olympus AU-400 analyzer (Olympus, Tokyo, Japan). The serum levels of gamma-glutamyl transferase (GGT) in this population have been reported elsewhere
DNA was extracted from peripheral blood leukocytes using standard protocols. The SNPs in the LBP, CD14 and TLR4 genes were genotyped using TaqMan™ validated assays (Applied Biosystems, Foster City, CA). The polymorphisms studied included LBP +1306T/C (rs2232618, p.Phe436Leu), CD14 -159C/T (rs2569190), and TLR4+896A/G (rs4986790, p.Asp299Gly). The LBP rs2232618 is a functional SNP that has been associated with host susceptibility to sepsis and multiple organ dysfunction in patients with major trauma
The Mann-Whitney U-test and the Jonckheere-Terpstra trend test were used to compare LBP concentrations among groups. The Spearman’s rank test was used to assess correlations. Linear regression was used for multivariate analyses. For that purpose, LBP concentrations (dependent variable) were log10-transformed to normalize the distribution. P-values lower than 0.05 were considered statistically significant.
The distribution of serum LBP concentrations in this population is shown in
Horizontal lines represent median values, boxes represent the interquartile range, whiskers represent the range, and dots represent extreme values (higher than the 75th percentile plus 1.5 times the interquartile range). The P-values (top of the figure) reflect comparisons to the group of 18–30 year-old (Mann-Whitney test). NS, not significant (P>0.05); *, P<0.005.
In univariate analyses, overweight and obese individuals had higher LBP concentrations than normal-weight individuals. Similarly, individuals with metabolic syndrome had higher LBP concentrations than those without metabolic syndrome (
Horizontal lines represent median values, boxes represent the interquartile range, whiskers represent the range, and dots represent extreme values (higher than the 75th percentile plus 1.5 times the interquartile range). P-values we obtained with the Jonkheere-Terpstra test for trend.
Correlation coefficient was obtained with the Spearman’s rank test.
No. | LBP (µg/mL) | sCD14 (µg/mL) | |
|
|||
Female (reference) | 231 | 6.97 (5.71–8.55) | 3.38 (2.89–3.80) |
Male | 189 | 7.39 (5.73–9.15) | 3.30 (2.95–3.80) |
|
|||
18–50 years (reference) | 191 | 6.46 (5.24–8.24) | 3.17 (2.77–3.62) |
>50 years | 229 | 7.58 (6.29–9.37) |
3.49 (3.08–3.94) |
|
|||
Abstainers (reference) | 195 | 7.07 (5.66–8.78) | 3.23 (2.83–3.81) |
Light-moderate drinkers | 141 | 7.29 (5.54–8.82) | 3.43 (2.94–3.80) |
Heavy drinkers | 84 | 7.06 (6.02–8.86) | 3.33 (3.03–3.77) |
|
|||
Non-smokers (reference) | 329 | 7.11 (5.85–8.74) | 3.37 (2.90–3.80) |
Smokers | 91 | 7.18 (5.42–9.15) | 3.23 (2.98–3.70) |
|
|||
Normal weight (reference) | 112 | 5.90 (5.09–7.67) | 3.34 (2.89–3.80) |
Overweight | 180 | 7.29 (5.96–8.78) |
3.35 (2.91–3.84) |
Obese | 127 | 7.75 (6.35–9.47) |
3.34 (2.89–3.80) |
|
|||
Absent (reference) | 315 | 6.82 (5.48–8.40) | 3.28 (2.89–3.76) |
Present | 105 | 8.02 (6.63–9.82) |
3.48 (3.09–3.89) |
Data are medians and interquartile ranges (within parenthesis).
P<0.05;
P<0.001 (comparison with the reference category, Mann-Whitney test).
Body mass index was unavailable for one individual.
Covariates | Coefficient (slope) | Standard error | P-value |
0.002 | 0.0003 | <0.001 | |
|
|||
Female | (reference) | ||
Male | 0.024 | 0.016 | 0.12 |
|
|||
Abstainers | (reference) | ||
Light-moderate drinkers | −0.022 | 0.016 | 0.16 |
Heavy drinkers | −0.017 | 0.021 | 0.40 |
|
|||
Non-smokers | (reference) | ||
Smokers | 0.036 | 0.018 | 0.05 |
0.006 | 0.001 | <0.001 | |
|
0.552 | 0.043 |
Linear regression analysis. LBP concentrations (dependent variable) was log10-transformed in order to normalize their distribution. All listed covariates entered the equation. Age was introduced in years and BMI was introduced in kg/m2; the remaining variables were introduced as “1 = present or yes” and “0 = absent or not”. Complete data were available for 419 individuals. The model explained 13.6% of the variability of serum LBP concentrations (R square, 0.136).
Covariates | Coefficient (slope) | Standard error | P-value |
|
|||
Absent | (reference) | ||
Present | 0.051 | 0.016 | 0.002 |
|
|||
Absent | (reference) | ||
Present | 0.005 | 0.017 | 0.78 |
|
|||
Absent | (reference) | ||
Present | 0.083 | 0.016 | <0.001 |
|
|||
Absent | (reference) | ||
Present | −0.008 | 0.018 | 0.65 |
|
|||
Absent | (reference) | ||
Present | 0.003 | 0.019 | 0.87 |
|
0.694 | 0.021 |
Linear regression. LBP concentrations (dependent variable) was log10-transformed in order to normalize their distribution. Coefficients are adjusted for all listed variables, as well as for age and sex. Age was introduced in years; the remaining variables were introduced as “1 = present or yes” and “0 = absent or not”. Metabolic syndrome components were defined by the Adult Treatment Panel III criteria
Male sex tended to be independently associated with higher serum LBP concentrations (
The LBP +1306T/C SNP (rs2232618) allele frequencies were 0.91 and 0.09, respectively. Carriers of the C allele tended to have higher serum LBP concentrations, although the difference was not statistically significant (P = 0.1,
No. | LBP (µg/mL) | sCD14 (µg/mL) | |
|
|||
TT (reference) | 196 | 7.03 (5.50–8.80) | 3.27 (2.79–3.69) |
CT or CC | 41 | 7.39 (5.98–9.99) | 3.34 (2.90–3.97) |
|
|||
CC (reference) | 63 | 7.34 (5.64–9.09) | 3.19 (2.72–3.55) |
CT or TT | 190 | 7.03 (5.41–8.81) | 3.38 (2.90–3.79) |
|
|||
AA (reference) | 232 | 7.09 (5.56–8.84) | 3.30 (2.89–3.74) |
AG or GG | 21 | 7.14 (5.17–10.7) | 3.15 (2.75–3.65) |
Data are medians and interquartile ranges (within parenthesis).
DNA samples were available for 237–253 out of 420 individuals.
P<0.05 (comparison with the reference category, Mann-Whitney test).
Serum sCD14 concentrations were found to be significantly correlated with those of LBP (correlation coefficient 0.149, P = 0.002). Similar to LBP, sCD14 concentrations increased with age. Contrary to LBP, sCD14 concentrations were not significantly associated either body mass index or metabolic syndrome (
There was a statistically significant correlation between serum LBP concentrations and concentrations of proinflammatory cytokines, particularly IL-6 and IL-8 (
IL-6 | IL-8 | TNF-α | AST | ALT | GGT | |
|
0.125 | 0.107 | 0.038 | 0.007 | 0.126 | 0.213 |
|
0.01 | 0.02 | 0.43 | 0.88 | 0.01 | <0.001 |
IL-6, interleukin-6; IL-8, interleukin-8; TNF-α, tumor necrosis factor alpha; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, gamma-glutamyl transferase.
Lipopolysaccharide-binding protein (LBP) might be a biomarker to indicate activation of innate immune responses in response to microbial compounds, particularly LPS
The finding that LBP levels increased with age in adults has not been reported previously. On the contrary, a decrease of LBP with age was reported in a selected sample of individuals being studied for cardiovascular risk
An increased BMI was strongly associated with increased serum levels of LBP in the general adult population studied. Alterations in gut microbiota and gut permeability to LPS are characteristics of obesity and related metabolic disorders
The association of LBP concentrations with low HDL-cholesterol, found in the present study, was independent of BMI and additional confounders. This finding confirms previous observations in selected samples
Alcohol consumption was not associated with LBP levels in this study. Previous studies have shown that alcohol intake enhances intestinal permeability
Serum LBP concentrations were associated with serum levels of proinflammatory cytokines (IL-6 and IL-8). This finding is consistent with the role of LBP as an acute phase response protein
The human LBP gene is located on chromosome 20
Strength of this study is its population-based design including randomly selected individuals with a broad age range (18 to 92 years). Admittedly, the study has also limitations. First, the epidemiological, cross-sectional nature of this study does not allow for a causal inference and does not provide a mechanistic insight into the findings. Second, LBP concentrations were measured in samples that had been stored frozen for years. To the best of our knowledge, there is no available information about stability or instability of LBP in frozen samples over time. It should be noted that the observed measurements were similar to those reported in different populations using fresh samples, including those reported by the manufacturer in apparently healthy volunteers. Moreover, it should be noted that the study was not aimed to investigate absolute LBP concentrations but the association of LBP concentrations with demographic, lifestyle, and metabolic variables. Along this line, the potential misclassification bias would be non-differential, i.e., the unreliability in LBP measurements must be expected to be equal among obese and non-obese individuals. Therefore, it can only result in bias toward the null (type 2 error) and not a type 1 error, i.e. it could have obscured an effect but it could not cause a false effect to be seen. Third, blood levels of LPS were not measured in the present study because samples were neither obtained nor stored under nonpyrogenic, sterile conditions. It should be noted that LPS is not the only microbial ligand for LPS
The serum/plasma LBP commercial assay is intended as a tool for management of patients with infectious complications, because increased LBP concentrations may be indicative of exposure to bacterial compounds (particularly, LPS) and may be prognostic of disease progression