- Email: [email protected]
Relationship of plasma apolipoprotein M with proprotein convertase subtilisin–kexin type 9 levels in non-diabetic subjects
Atherosclerosis 214 (2011) 492–494
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Letter to the Editor Relationship of plasma apolipoprotein M with proprotein convertase subtilisin–kexin type 9 levels in non-diabetic subjects
a r t i c l e Keywords: Apolipoprotein M Apolipoprotein B LDL cholesterol Obesity PCSK9
i n f o
a b s t r a c t Purpose: Apolipoprotein M (apoM) retards atherosclerosis development in murine models, and may be regulated by pathways involved in LDL metabolism. Proprotein convertase subtilisin–kexin type 9 (PCSK9) plays a key role in LDL receptor processing. We determined the extent to which plasma apoM is related to PCSK9 levels in subjects with varying degrees of obesity. Methods: We sought correlations between plasma apoM and PCSK9, measured using recently developed ELISAs, in 79 non-diabetic subjects. Results: ApoM and PCSK9 levels were both correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB (P < 0.05 to P < 0.001). ApoM correlated positively with PCSK9 in lean individuals (n = 37, r = 0.337, P = 0.041), but not in overweight subjects (n = 32, r = 0.125, P = 0.50) and in obese subjects (n = 10, r = −0.055, P = 0.88). Conclusions: The PCSK9 pathway may contribute to plasma apoM regulation in humans. The influence of PCSK9 on circulating apoM appears to be modified by adiposity. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Interest in the regulation of apolipoprotein (apo) M comes from the observation that apoM retards atherosclerosis development in murine models [1–3]. Its plasma level is related positively with HDL cholesterol and apoA-I [4,5]. Remarkably, apoM is also correlated positively with plasma total cholesterol and LDL cholesterol [4–8], although apoM is mainly associated with HDL with only a minor proportion being located in the LDL fraction [7]. ApoM deficiency decreases whereas apoM overexpression increases plasma cholesterol [3]. This effect is more pronounced in the setting of LDL receptor deficiency [9]. Conversely, in murine models of LDL receptor deficiency plasma apoM is increased, which in turn may impair LDL and VLDL removal from plasma [3]. Recently, plasma apoM was found to decrease after statin treatment in humans [10]. Taken together, these findings favour an intricate relation between apoM regulation and LDL cholesterol homeostasis. Proprotein convertase subtilisin–kexin type 9 (PCSK9) provides a key regulatory pathway for LDL receptor processing [11,12]. The mechanisms whereby PCSK9 promotes LDL receptor degradation involve intracellular targeting of internalized LDL receptors towards the lysosomal compartment. PCSK9 is a secreted protease and binds to the extracellular domain of the LDL receptor, where it acts as a chaperone targeting the LDL receptor towards intracellular degradation and preventing LDL receptor recycling to the cell surface. This implies that circulating PCSK9 is physiologically relevant [11,12]. Indeed, the LDL apoB fractional catabolic rate correlates inversely with plasma PCSK9 levels in humans [13]. Human studies have now shown that plasma apoB-containing lipoproteins levels are correlated positively with the PCSK9 concentration [13,14]. 0021-9150/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2010.10.041
Against this background we hypothesized that plasma apoM is positively related to the PCSK9 level. Here we determined the extent to which plasma apoM correlates with PCSK9 levels in nondiabetic subjects with varying degrees of obesity. 2. Materials and methods 2.1. Subjects The medical ethics committee of the University Medical Center Groningen, The Netherlands, approved the study. Participants (aged >18 years) were recruited by advertisement and provided written informed consent. Physical examination did not reveal abnormalities. Previously diagnosed diabetes mellitus, cardiovascular disease, renal insufficiency, hypertension, thyroid disorders, liver disease and pregnancy were exclusion criteria. Smokers and subjects who used >3 alcoholic drinks per day were also excluded. BMI was calculated as weight divided by height squared (kg/m2 ). Overweight and obesity were defined as BMI ≥ 25 kg/m2 and ≥30 kg/m2 , respectively. Homeostasis model assessment (HOMAir ) was used to assess insulin sensitivity, and was calculated with the formula: fasting plasma insulin × glucose/22.5. The subjects were studied after an overnight fast. 2.2. Laboratory analyses Venous blood samples were collected into EDTA-containing tubes (1.5 mg/mL). Plasma samples were stored at −80◦ C until analysis. Glucose was measured shortly after blood collection. Cholesterol and triglycerides were assayed by routine enzymatic
Letter to the Editor / Atherosclerosis 214 (2011) 492–494
2.3. Statistical analysis Data are given in mean ± SD or in median (interquartile range). Univariate relationships were calculated using Spearman’s rank correlation. Multivariable regression analysis was performed to the independent contribution of variables. Two-sided P-values <0.05 were considered significant.
Seventy-nine non-diabetic Caucasian subjects participated (Table 1). BMI ranged from 19.4 to 40.4 kg/m2 . Forty-two subjects (53%) were overweight (n = 32) or obese (n = 10). Plasma insulin and HOMAir were higher (P < 0.001), but glucose was similar (P = 0.81) in subjects with BMI ≥ 25 kg/m2 vs. lean subjects (not shown). In the whole group, plasma apoM ranged from 0.53 to 1.95 mol/L, and PCSK9 ranged 43 from to 339 g/L. ApoM was lower in subjects with BMI ≥ 25 kg/m2 (1.04 ± 0.26 mol/L) compared to lean individuals (1.14 ± 0.22 mol/L, P = 0.046). In contrast, PCSK9 levels were not decreased in overweight or obese subjects (BMI ≥ 25 kg/m2 : 162 ± 62 g/L vs. BMI < 25 kg/m2 : 146 ± 52 g/L, P = 0.22). Plasma apoM was not different between men (1.06 ± 0.18 mol/L) and women (1.13 ± 0.31 mol/L, P = 0.26), but PCSK9 levels were lower in the men (141 ± 54 g/L) than in the women (172 ± 59 g/L, P = 0.023). Table 1 Clinical characteristics, plasma glucose, insulin, homeostasis model assessment (HOMAir ), plasma lipids, apolipoproteins (apos), proprotein convertase subtilisin–kexin type 9 (PCSK9) levels, and correlation coefficients with apoM and PCSK9 in 79 non-diabetic subjects. Variable
Correlation coefficient of variable with apoM
PCSK9
55 ± 10 44/35 131 ± 19
0.028
−0.120
−0.115
0.022
82 ± 11
−0.133
−0.079
25.9 ± 3.9 5.6 ± 0.7 6.7 (4.3–8.4) 1.57 (1.03–2.20)
−0.298** 0.174 0.014 0.042
−0.021 −0.141 0.245* 0.219*
5.72 ± 1.00 4.25 ± 1.06 3.54 ± 0.88 1.46 ± 0.40 1.27 (0.88–1.91) 0.96 ± 0.24 1.42 ± 0.22 1.09 ± 0.25 155 ± 58
0.365*** 0.303** 0.273** 0.132 0.210* 0.253* 0.216*
0.357*** 0.398*** 0.292** −0.130 0.337** 0.278** 0.034 0.149
Data in mean (SD) or in median (interquartile range). BMI, body mass index. Spearman’s correlation coefficients are shown. * P ≤ 0.05. ** P ≤ 0.01. *** P ≤ 0.001.
lean overweight obese
1.5
1.0
0.5
0.0
3. Results
Age (years) Sex (m/f) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) BMI (kg/m2 ) Glucose (mmol/L) Insulin (mU/L) HOMAir (mU × mmol/(l2 × 22.5)) Total cholesterol (mmol/L) Non-HDL cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) Triglycerides (mmol/L) ApoB (g/L) ApoA-I (g/L) ApoM (mol/L) PCSK9 (g/L)
2.0
ApoM (µmol/L)
methods. HDL cholesterol was measured with a homogeneous enzymatic colorimetric test. Non-HDL cholesterol was calculated as the difference between total cholesterol and HDL cholesterol. LDL cholesterol was calculated with the Friedewald formula. ApoAI and apoB were measured by immunoturbidimetry. ApoM was assayed using a sandwich ELISA based on two highly specific monoclonal antibodies as described [4]. Plasma PCSK9 was measured using a sandwich ELISA exactly as described [13]. The inter-assay coefficients of variation were 4.9% and 5.6%, respectively
493
0
100
200
300
pcsk9 (µg/L) Fig. 1. Scatterplot showing relationship of plasma apolipoprotein M (apoM) with proprotein convertase subtilisin–kexin type 9 (PCSK9) concentrations in 37 lean subjects, 32 overweight subjects and 10 obese subjects. Spearman’s correlation coefficients: lean subjects r = 0.337, P = 0.041; overweight subjects r = 0.125, P = 0.50; obese subjects r = −0.055, P = 0.88.
Plasma apoM was correlated inversely with BMI, but was unrelated to insulin and HOMAir . PCSK9 was correlated positively with insulin and HOMAir but was not correlated with BMI (Table 1). ApoM as well as PCSK9 levels were correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB. Additionally, apoM was related positively with apoA-I and PCSK9 correlated positively with triglycerides. Despite relationships of both apoM and PCSK9 with apoB-containing lipoproteins, the correlation of apoM with PCSK9 was not significant in the whole group (P = 0.19; Table 1). Since the presence of overweight could confound the putative relationship between PCSK9 and apoM, we determined relationships of apoM with PCSK9 in lean, overweight and obese subjects separately. As shown in Fig. 1, apoM did correlate positively with PCSK9 in lean subjects (r = 0.337, P = 0.041). However, apoM was not correlated with PCSK9 in overweight subjects (n = 32, r = 0.125, P = 0.50), in obese subjects (n = 10, r = −0.055, P = 0.88) and in overweight + obese subjects combined (n = 42, r = 0.099, P = 0.53). Multivariable regression analysis showed that in lean subjects apoM remained correlated positively with PCSK9 (ˇ = 0.413, P = 0.021) after adjustment for BMI and apoB. In contrast in overweight + obese subjects combined, apoM was unrelated to PCSK9 (ˇ = 0.011, P = 0.94) after controlling for BMI and apoB. 4. Discussion Here we demonstrate that plasma apoM is correlated positively with the PCSK9 level in lean subjects without diabetes mellitus. No such relationship was observed in overweight and obese individuals. Our findings support the hypothesis that the PCSK9 pathway may be involved in apoM regulation in humans, and suggest that the possible role of the PCSK9 pathway on apoM homeostasis may be modified by adiposity. Both plasma apoM and PCSK9 concentrations were correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB, as expected [4–8,13,14]. PCSK9 facilitates LDL receptor degradation [11,12]. In lean individuals, the positive relationship of apoM with PCSK9 levels was independent of BMI and apoB-containing lipoproteins. This result, therefore, agrees with the
494
Letter to the Editor / Atherosclerosis 214 (2011) 492–494
concept that the clearance of circulating apoM is affected by the LDL receptor-pathway in humans [3,9]. Relevant differences were found for the association of apoM and PCSK9 with obesity and insulin sensitivity. ApoM was correlated inversely with BMI, in accord with other data [5,6,8]. The relationship of apoM with insulin and insulin sensitivity was not significant. Plasma PCSK9 was not significantly related to BMI, but was correlated positively with plasma insulin and HOMAir . Positive relationships of plasma PCSK9 with insulin and HOMAir were reported recently using a different ELISA [14]. Plasma PCSK9 levels varied 100-fold in that study [14], exceeding the range of values that was currently observed. That report also showed a positive relationship of PCSK9 with BMI, but this correlation was weaker than that with insulin and HOMAir [13]. In vitro studies have suggested that insulin decreases apoM [15], but may increase PCSK9 expression [16]. These apparently contrasting effects of insulin could in part be responsible for the differences in associations of apoM compared to PCSK9 with insulin and insulin sensitivity. Although the precise mechanisms responsible for a possible modifying effect of adiposity on the contribution of PCSK9 to apoM regulation are unknown and our findings need to be replicated in a larger cohort, metabolic factors related to insulin-regulated pathways could be involved. In conclusion, the present study suggests that PCSK9, which plays a key role in LDL receptor degradation, may affect apoM regulation in lean individuals.
[7] Christoffersen C, Nielsen LB, Axler O, et al. Isolation and characterization of human apolipoprotein M-containing lipoproteins. J Lipid Res 2006;47:1833–44. [8] Ooi EM, Watts GF, Chan DC, et al. Association of apolipoprotein M with high-density lipoprotein kinetics in overweight-obese men. Atherosclerosis 2010;210:326–30. [9] Christoffersen C, Pedersen TX, Gordts PL, et al. Opposing effects of apolipoprotein M on catabolism of apolipoprotein B-containing lipoproteins and atherosclerosis. Circ Res 2010;106:1624–34. [10] Kappelle PJWH, Ahnström J, Dikkeschei LD, et al. Plasma apolipoprotein M responses to statin and fibrate administration in type 2 diabetes mellitus. Atherosclerosis. doi:10.1016/j.atherosclerosis.2010.07.048. [11] Horton JD, Cohen JC, Hobbs HH. PCSK9: a convertase that coordinates LDL catabolism. J Lipid Res 2009;50(Suppl.):S172–7. [12] Lambert G, Charlton F, Rye KA, Piper DE. Molecular basis of PCSK9 function. Atherosclerosis 2009;203:1–7. [13] Chan DC, Lambert G, Barrett PH, et al. Plasma proprotein convertase subtilisin/kexin type 9: a marker of LDL apolipoprotein B-100 catabolism? Clin Chem 2009;55:2049–52. [14] Lakoski SG, Lagace TA, Cohen JC, Horton JD, Hobbs HH. Genetic and metabolic determinants of plasma PCSK9 levels. J Clin Endocrinol Metab 2009;94:2537–43. [15] Wolfrum C, Howell JJ, Ndungo E, Stoffel M. Foxa2 activity increases plasma high density lipoprotein levels by regulating apolipoprotein M. J Biol Chem 2008;283:16940–9. [16] Costet P, Cariou B, Lambert G, et al. Hepatic PCSK9 expression is regulated by nutritional status via insulin and sterol regulatory element-binding protein 1c. J Biol Chem 2006;281:6211–8.
Paul J.W.H. Kappelle 1 Department of Endocrinology, University Medical Center Groningen and University of Groningen, Groningen, The Netherlands
Acknowledgements
Gilles Lambert a,b,1 The Heart Research Institute, Sydney, Australia b Université de Nantes, Faculté de Médecine, Nantes, France a
This study is in part supported by grant 2001.00.012 from the Dutch Diabetes Research Foundation (to RPFD), the Danish National Research Council (217-07-0352), the Novo Nordisk Foundation and the Rigshospitalet, Copenhagen (to LBN), the Swedish Research Council (#07143), the Swedish Heart-Lung Foundation, the A. Påhlsson Foundation and the Wallenberg Foundation (to BD) and a grant in aid G-08S-3700 from the Heart Research Foundation of Australia (to GL). We thank Ms. Francine Petrides for excellent technical assistance. The help of Dr. L.D. Dikkeschei, Ph.D., Laboratory of Clinical Chemistry, Isala Clinics, Zwolle, The Netherlands, is greatly appreciated. References [1] Hu YW, Zheng L, Wang Q. Characteristics of apolipoprotein M and its relation to atherosclerosis and diabetes. Biochim Biophys Acta 2010;1801:100–5. [2] Nielsen LB, Christoffersen C, Ahnström J, Dahlbäck B. ApoM: gene regulation and effects on HDL metabolism. Trends Endocrinol Metab 2009;20:66–71. [3] Christoffersen C, Jauhiainen M, Moser M, et al. Effect of apolipoprotein M on high density lipoprotein metabolism and atherosclerosis in low density lipoprotein receptor knock-out mice. J Biol Chem 2008;283:1839–47. [4] Axler O, Ahnström J, Dahlbäck B. An ELISA for apolipoprotein M reveals a strong correlation to total cholesterol in human plasma. J Lipid Res 2007;48: 1772–80. [5] Dullaart RPF, Plomgaard P, de Vries R, Dahlbäck B, Nielsen LB. Plasma apolipoprotein M is reduced in metabolic syndrome but does not predict intima media thickness. Clin Chim Acta 2009;406:129–33. [6] Plomgaard P, Dullaart RPF, de Vries R, et al. Apolipoprotein M predicts pre-HDL formation: studies in type 2 diabetic and nondiabetic subjects. J Intern Med 2009;266:258–67.
Björn Dahlbäck University of Lund, Department of Laboratory Medicine, University Hospital, Malmö, Sweden a
Lars Bo Nielsen a,b Department of Clinical Biochemistry, Rigshospitalet, Copenhagen, Denmark b Department of Biomedical Sciences, University of Copenhagen, Denmark Robin P.F. Dullaart ∗ Department of Endocrinology, University Medical Center Groningen and University of Groningen, P.O. Box 30001, 9700 RB Groningen, The Netherlands ∗ Corresponding
author. Tel.: +31 50 361 3731; fax: +31 50 361 9392. E-mail address: [email protected] (R.P.F. Dullaart) 1
Contributed equally to the manuscript. 23 September 2010 Available online 3 November 2010
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Letter to the Editor Relationship of plasma apolipoprotein M with proprotein convertase subtilisin–kexin type 9 levels in non-diabetic subjects
a r t i c l e Keywords: Apolipoprotein M Apolipoprotein B LDL cholesterol Obesity PCSK9
i n f o
a b s t r a c t Purpose: Apolipoprotein M (apoM) retards atherosclerosis development in murine models, and may be regulated by pathways involved in LDL metabolism. Proprotein convertase subtilisin–kexin type 9 (PCSK9) plays a key role in LDL receptor processing. We determined the extent to which plasma apoM is related to PCSK9 levels in subjects with varying degrees of obesity. Methods: We sought correlations between plasma apoM and PCSK9, measured using recently developed ELISAs, in 79 non-diabetic subjects. Results: ApoM and PCSK9 levels were both correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB (P < 0.05 to P < 0.001). ApoM correlated positively with PCSK9 in lean individuals (n = 37, r = 0.337, P = 0.041), but not in overweight subjects (n = 32, r = 0.125, P = 0.50) and in obese subjects (n = 10, r = −0.055, P = 0.88). Conclusions: The PCSK9 pathway may contribute to plasma apoM regulation in humans. The influence of PCSK9 on circulating apoM appears to be modified by adiposity. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Interest in the regulation of apolipoprotein (apo) M comes from the observation that apoM retards atherosclerosis development in murine models [1–3]. Its plasma level is related positively with HDL cholesterol and apoA-I [4,5]. Remarkably, apoM is also correlated positively with plasma total cholesterol and LDL cholesterol [4–8], although apoM is mainly associated with HDL with only a minor proportion being located in the LDL fraction [7]. ApoM deficiency decreases whereas apoM overexpression increases plasma cholesterol [3]. This effect is more pronounced in the setting of LDL receptor deficiency [9]. Conversely, in murine models of LDL receptor deficiency plasma apoM is increased, which in turn may impair LDL and VLDL removal from plasma [3]. Recently, plasma apoM was found to decrease after statin treatment in humans [10]. Taken together, these findings favour an intricate relation between apoM regulation and LDL cholesterol homeostasis. Proprotein convertase subtilisin–kexin type 9 (PCSK9) provides a key regulatory pathway for LDL receptor processing [11,12]. The mechanisms whereby PCSK9 promotes LDL receptor degradation involve intracellular targeting of internalized LDL receptors towards the lysosomal compartment. PCSK9 is a secreted protease and binds to the extracellular domain of the LDL receptor, where it acts as a chaperone targeting the LDL receptor towards intracellular degradation and preventing LDL receptor recycling to the cell surface. This implies that circulating PCSK9 is physiologically relevant [11,12]. Indeed, the LDL apoB fractional catabolic rate correlates inversely with plasma PCSK9 levels in humans [13]. Human studies have now shown that plasma apoB-containing lipoproteins levels are correlated positively with the PCSK9 concentration [13,14]. 0021-9150/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2010.10.041
Against this background we hypothesized that plasma apoM is positively related to the PCSK9 level. Here we determined the extent to which plasma apoM correlates with PCSK9 levels in nondiabetic subjects with varying degrees of obesity. 2. Materials and methods 2.1. Subjects The medical ethics committee of the University Medical Center Groningen, The Netherlands, approved the study. Participants (aged >18 years) were recruited by advertisement and provided written informed consent. Physical examination did not reveal abnormalities. Previously diagnosed diabetes mellitus, cardiovascular disease, renal insufficiency, hypertension, thyroid disorders, liver disease and pregnancy were exclusion criteria. Smokers and subjects who used >3 alcoholic drinks per day were also excluded. BMI was calculated as weight divided by height squared (kg/m2 ). Overweight and obesity were defined as BMI ≥ 25 kg/m2 and ≥30 kg/m2 , respectively. Homeostasis model assessment (HOMAir ) was used to assess insulin sensitivity, and was calculated with the formula: fasting plasma insulin × glucose/22.5. The subjects were studied after an overnight fast. 2.2. Laboratory analyses Venous blood samples were collected into EDTA-containing tubes (1.5 mg/mL). Plasma samples were stored at −80◦ C until analysis. Glucose was measured shortly after blood collection. Cholesterol and triglycerides were assayed by routine enzymatic
Letter to the Editor / Atherosclerosis 214 (2011) 492–494
2.3. Statistical analysis Data are given in mean ± SD or in median (interquartile range). Univariate relationships were calculated using Spearman’s rank correlation. Multivariable regression analysis was performed to the independent contribution of variables. Two-sided P-values <0.05 were considered significant.
Seventy-nine non-diabetic Caucasian subjects participated (Table 1). BMI ranged from 19.4 to 40.4 kg/m2 . Forty-two subjects (53%) were overweight (n = 32) or obese (n = 10). Plasma insulin and HOMAir were higher (P < 0.001), but glucose was similar (P = 0.81) in subjects with BMI ≥ 25 kg/m2 vs. lean subjects (not shown). In the whole group, plasma apoM ranged from 0.53 to 1.95 mol/L, and PCSK9 ranged 43 from to 339 g/L. ApoM was lower in subjects with BMI ≥ 25 kg/m2 (1.04 ± 0.26 mol/L) compared to lean individuals (1.14 ± 0.22 mol/L, P = 0.046). In contrast, PCSK9 levels were not decreased in overweight or obese subjects (BMI ≥ 25 kg/m2 : 162 ± 62 g/L vs. BMI < 25 kg/m2 : 146 ± 52 g/L, P = 0.22). Plasma apoM was not different between men (1.06 ± 0.18 mol/L) and women (1.13 ± 0.31 mol/L, P = 0.26), but PCSK9 levels were lower in the men (141 ± 54 g/L) than in the women (172 ± 59 g/L, P = 0.023). Table 1 Clinical characteristics, plasma glucose, insulin, homeostasis model assessment (HOMAir ), plasma lipids, apolipoproteins (apos), proprotein convertase subtilisin–kexin type 9 (PCSK9) levels, and correlation coefficients with apoM and PCSK9 in 79 non-diabetic subjects. Variable
Correlation coefficient of variable with apoM
PCSK9
55 ± 10 44/35 131 ± 19
0.028
−0.120
−0.115
0.022
82 ± 11
−0.133
−0.079
25.9 ± 3.9 5.6 ± 0.7 6.7 (4.3–8.4) 1.57 (1.03–2.20)
−0.298** 0.174 0.014 0.042
−0.021 −0.141 0.245* 0.219*
5.72 ± 1.00 4.25 ± 1.06 3.54 ± 0.88 1.46 ± 0.40 1.27 (0.88–1.91) 0.96 ± 0.24 1.42 ± 0.22 1.09 ± 0.25 155 ± 58
0.365*** 0.303** 0.273** 0.132 0.210* 0.253* 0.216*
0.357*** 0.398*** 0.292** −0.130 0.337** 0.278** 0.034 0.149
Data in mean (SD) or in median (interquartile range). BMI, body mass index. Spearman’s correlation coefficients are shown. * P ≤ 0.05. ** P ≤ 0.01. *** P ≤ 0.001.
lean overweight obese
1.5
1.0
0.5
0.0
3. Results
Age (years) Sex (m/f) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) BMI (kg/m2 ) Glucose (mmol/L) Insulin (mU/L) HOMAir (mU × mmol/(l2 × 22.5)) Total cholesterol (mmol/L) Non-HDL cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) Triglycerides (mmol/L) ApoB (g/L) ApoA-I (g/L) ApoM (mol/L) PCSK9 (g/L)
2.0
ApoM (µmol/L)
methods. HDL cholesterol was measured with a homogeneous enzymatic colorimetric test. Non-HDL cholesterol was calculated as the difference between total cholesterol and HDL cholesterol. LDL cholesterol was calculated with the Friedewald formula. ApoAI and apoB were measured by immunoturbidimetry. ApoM was assayed using a sandwich ELISA based on two highly specific monoclonal antibodies as described [4]. Plasma PCSK9 was measured using a sandwich ELISA exactly as described [13]. The inter-assay coefficients of variation were 4.9% and 5.6%, respectively
493
0
100
200
300
pcsk9 (µg/L) Fig. 1. Scatterplot showing relationship of plasma apolipoprotein M (apoM) with proprotein convertase subtilisin–kexin type 9 (PCSK9) concentrations in 37 lean subjects, 32 overweight subjects and 10 obese subjects. Spearman’s correlation coefficients: lean subjects r = 0.337, P = 0.041; overweight subjects r = 0.125, P = 0.50; obese subjects r = −0.055, P = 0.88.
Plasma apoM was correlated inversely with BMI, but was unrelated to insulin and HOMAir . PCSK9 was correlated positively with insulin and HOMAir but was not correlated with BMI (Table 1). ApoM as well as PCSK9 levels were correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB. Additionally, apoM was related positively with apoA-I and PCSK9 correlated positively with triglycerides. Despite relationships of both apoM and PCSK9 with apoB-containing lipoproteins, the correlation of apoM with PCSK9 was not significant in the whole group (P = 0.19; Table 1). Since the presence of overweight could confound the putative relationship between PCSK9 and apoM, we determined relationships of apoM with PCSK9 in lean, overweight and obese subjects separately. As shown in Fig. 1, apoM did correlate positively with PCSK9 in lean subjects (r = 0.337, P = 0.041). However, apoM was not correlated with PCSK9 in overweight subjects (n = 32, r = 0.125, P = 0.50), in obese subjects (n = 10, r = −0.055, P = 0.88) and in overweight + obese subjects combined (n = 42, r = 0.099, P = 0.53). Multivariable regression analysis showed that in lean subjects apoM remained correlated positively with PCSK9 (ˇ = 0.413, P = 0.021) after adjustment for BMI and apoB. In contrast in overweight + obese subjects combined, apoM was unrelated to PCSK9 (ˇ = 0.011, P = 0.94) after controlling for BMI and apoB. 4. Discussion Here we demonstrate that plasma apoM is correlated positively with the PCSK9 level in lean subjects without diabetes mellitus. No such relationship was observed in overweight and obese individuals. Our findings support the hypothesis that the PCSK9 pathway may be involved in apoM regulation in humans, and suggest that the possible role of the PCSK9 pathway on apoM homeostasis may be modified by adiposity. Both plasma apoM and PCSK9 concentrations were correlated positively with total cholesterol, non-HDL cholesterol, LDL cholesterol and apoB, as expected [4–8,13,14]. PCSK9 facilitates LDL receptor degradation [11,12]. In lean individuals, the positive relationship of apoM with PCSK9 levels was independent of BMI and apoB-containing lipoproteins. This result, therefore, agrees with the
494
Letter to the Editor / Atherosclerosis 214 (2011) 492–494
concept that the clearance of circulating apoM is affected by the LDL receptor-pathway in humans [3,9]. Relevant differences were found for the association of apoM and PCSK9 with obesity and insulin sensitivity. ApoM was correlated inversely with BMI, in accord with other data [5,6,8]. The relationship of apoM with insulin and insulin sensitivity was not significant. Plasma PCSK9 was not significantly related to BMI, but was correlated positively with plasma insulin and HOMAir . Positive relationships of plasma PCSK9 with insulin and HOMAir were reported recently using a different ELISA [14]. Plasma PCSK9 levels varied 100-fold in that study [14], exceeding the range of values that was currently observed. That report also showed a positive relationship of PCSK9 with BMI, but this correlation was weaker than that with insulin and HOMAir [13]. In vitro studies have suggested that insulin decreases apoM [15], but may increase PCSK9 expression [16]. These apparently contrasting effects of insulin could in part be responsible for the differences in associations of apoM compared to PCSK9 with insulin and insulin sensitivity. Although the precise mechanisms responsible for a possible modifying effect of adiposity on the contribution of PCSK9 to apoM regulation are unknown and our findings need to be replicated in a larger cohort, metabolic factors related to insulin-regulated pathways could be involved. In conclusion, the present study suggests that PCSK9, which plays a key role in LDL receptor degradation, may affect apoM regulation in lean individuals.
[7] Christoffersen C, Nielsen LB, Axler O, et al. Isolation and characterization of human apolipoprotein M-containing lipoproteins. J Lipid Res 2006;47:1833–44. [8] Ooi EM, Watts GF, Chan DC, et al. Association of apolipoprotein M with high-density lipoprotein kinetics in overweight-obese men. Atherosclerosis 2010;210:326–30. [9] Christoffersen C, Pedersen TX, Gordts PL, et al. Opposing effects of apolipoprotein M on catabolism of apolipoprotein B-containing lipoproteins and atherosclerosis. Circ Res 2010;106:1624–34. [10] Kappelle PJWH, Ahnström J, Dikkeschei LD, et al. Plasma apolipoprotein M responses to statin and fibrate administration in type 2 diabetes mellitus. Atherosclerosis. doi:10.1016/j.atherosclerosis.2010.07.048. [11] Horton JD, Cohen JC, Hobbs HH. PCSK9: a convertase that coordinates LDL catabolism. J Lipid Res 2009;50(Suppl.):S172–7. [12] Lambert G, Charlton F, Rye KA, Piper DE. Molecular basis of PCSK9 function. Atherosclerosis 2009;203:1–7. [13] Chan DC, Lambert G, Barrett PH, et al. Plasma proprotein convertase subtilisin/kexin type 9: a marker of LDL apolipoprotein B-100 catabolism? Clin Chem 2009;55:2049–52. [14] Lakoski SG, Lagace TA, Cohen JC, Horton JD, Hobbs HH. Genetic and metabolic determinants of plasma PCSK9 levels. J Clin Endocrinol Metab 2009;94:2537–43. [15] Wolfrum C, Howell JJ, Ndungo E, Stoffel M. Foxa2 activity increases plasma high density lipoprotein levels by regulating apolipoprotein M. J Biol Chem 2008;283:16940–9. [16] Costet P, Cariou B, Lambert G, et al. Hepatic PCSK9 expression is regulated by nutritional status via insulin and sterol regulatory element-binding protein 1c. J Biol Chem 2006;281:6211–8.
Paul J.W.H. Kappelle 1 Department of Endocrinology, University Medical Center Groningen and University of Groningen, Groningen, The Netherlands
Acknowledgements
Gilles Lambert a,b,1 The Heart Research Institute, Sydney, Australia b Université de Nantes, Faculté de Médecine, Nantes, France a
This study is in part supported by grant 2001.00.012 from the Dutch Diabetes Research Foundation (to RPFD), the Danish National Research Council (217-07-0352), the Novo Nordisk Foundation and the Rigshospitalet, Copenhagen (to LBN), the Swedish Research Council (#07143), the Swedish Heart-Lung Foundation, the A. Påhlsson Foundation and the Wallenberg Foundation (to BD) and a grant in aid G-08S-3700 from the Heart Research Foundation of Australia (to GL). We thank Ms. Francine Petrides for excellent technical assistance. The help of Dr. L.D. Dikkeschei, Ph.D., Laboratory of Clinical Chemistry, Isala Clinics, Zwolle, The Netherlands, is greatly appreciated. References [1] Hu YW, Zheng L, Wang Q. Characteristics of apolipoprotein M and its relation to atherosclerosis and diabetes. Biochim Biophys Acta 2010;1801:100–5. [2] Nielsen LB, Christoffersen C, Ahnström J, Dahlbäck B. ApoM: gene regulation and effects on HDL metabolism. Trends Endocrinol Metab 2009;20:66–71. [3] Christoffersen C, Jauhiainen M, Moser M, et al. Effect of apolipoprotein M on high density lipoprotein metabolism and atherosclerosis in low density lipoprotein receptor knock-out mice. J Biol Chem 2008;283:1839–47. [4] Axler O, Ahnström J, Dahlbäck B. An ELISA for apolipoprotein M reveals a strong correlation to total cholesterol in human plasma. J Lipid Res 2007;48: 1772–80. [5] Dullaart RPF, Plomgaard P, de Vries R, Dahlbäck B, Nielsen LB. Plasma apolipoprotein M is reduced in metabolic syndrome but does not predict intima media thickness. Clin Chim Acta 2009;406:129–33. [6] Plomgaard P, Dullaart RPF, de Vries R, et al. Apolipoprotein M predicts pre-HDL formation: studies in type 2 diabetic and nondiabetic subjects. J Intern Med 2009;266:258–67.
Björn Dahlbäck University of Lund, Department of Laboratory Medicine, University Hospital, Malmö, Sweden a
Lars Bo Nielsen a,b Department of Clinical Biochemistry, Rigshospitalet, Copenhagen, Denmark b Department of Biomedical Sciences, University of Copenhagen, Denmark Robin P.F. Dullaart ∗ Department of Endocrinology, University Medical Center Groningen and University of Groningen, P.O. Box 30001, 9700 RB Groningen, The Netherlands ∗ Corresponding
author. Tel.: +31 50 361 3731; fax: +31 50 361 9392. E-mail address: [email protected] (R.P.F. Dullaart) 1
Contributed equally to the manuscript. 23 September 2010 Available online 3 November 2010