Buy Article Permissions and Reprints

Atherosclerotic cardiovascular disease (ASCVD) remains the leading cause of death worldwide.[1] Type 2 diabetes mellitus (T2DM), which is still associated with significant disability and mortality in 2024,[2] is one of the major risk factors for ASCVD development.[3] T2DM prevalence is increasing, primarily due to the ongoing obesity pandemic,[4] with recently published estimates supporting that more than 1.3 billion people will be affected by diabetes mellitus until 2050.[2] Vascular complications, including atherosclerosis and diabetic nephropathy, comprise major sources of disability and mortality among individuals with T2DM,[5] whereas endothelial dysfunction is the main pathophysiological mechanism underlying diabetic vascular disease.[6] Roundabout guidance receptor 4 (Robo4) is an endothelial-specific receptor that regulates endothelial cell (EC) junctional integrity and vascular permeability,[7] [8] as well as EC-mediated inflammatory responses[9] and angiogenesis,[7] [8] all of which are central pathogenetic mechanisms in diabetic vascular complications and ASCVD. Moreover, we have previously shown that mice lacking Robo4 exhibit more severe pancreatic damage and faster development of diabetes when subjected to the multiple low-dose streptozotocin model of diabetes.[10] Based on the above, in the current study, we aimed to explore the potential value of circulating Robo4 (soluble Robo4, sRobo4) as a marker of endothelial dysfunction and vascular disease in individuals with or without T2DM.

Forty-three individuals with T2DM and 27 nondiabetic control individuals (control) of similar age and sex were included in the study. All study participants were free of cardiovascular disease (CVD), chronic renal disease, as well as of any chronic inflammatory disease at the time of inclusion in the study. The study was approved by the Laiko Hospital Ethics Committee and all individuals provided informed consent in accordance with the Declaration of Helsinki. All study participants underwent blood sampling after overnight fasting and a detailed medical history was recorded for every individual, including anthropometric assessment, history of CVD risk factors, and detailed biochemistry profile of liver and renal function markers, markers of systemic inflammation and cholesterol levels. Blood pressure (BP) was also measured twice in both arms in the supine position by a trained physician and estimated glomerular filtration rate (eGFR) was determined with the MDRD (Modification of Diet in Renal Disease) equation.[11] Moreover, subclinical atherosclerosis, as assessed by nonstenotic plaque burden, was measured by standard higher resolution ultrasound in the femoral and carotid arterial bed; all measurements were performed by a single experienced operator blinded to the medical history of the study participants (more details in previous works[12]). sROBO4 was measured in the plasma using a commercially available ELISA kit (EH394RB, Invitrogen) according to manufacturer's instructions. For the statistical analysis of the data, normality of continuous variables was assessed by Kolmogorov–Smirnov and Shapiro–Wilk tests. Presence of outliers was examined with Rout test (Q = 1%). Comparison of continuous variables between two groups was performed by independent samples t-test or Mann–Whitney U test when not normally distributed. Correlation between continuous variables was examined by Spearman's rank test. Finally, to examine the association of sRobo4 levels with higher atherosclerotic burden, we utilized a binary logistic regression model with higher atherosclerotic burden as dependent variable (0–1 atherosclerotic plaques = 0, ≥2 atherosclerotic plaques = 1). Statistical significance was set at p < 0.05.

sROBO4 was below the detection limit of the assay in 7/43 individuals with T2DM and 8/27 control individuals (p = 0.24), resulting in a final cohort of 36 T2DM and 19 control individuals with available sRobo4 measurements. Moreover, two individuals with T2DM and four control individuals were identified as outliers. The final cohort (n = 49) consisting of 34 individuals with T2DM and 15 control individuals was still balanced for age (T2DM: 65.6 ± 8.7 vs. control: 65.7 ± 9.2 years, p = 0.975) and sex (T2DM: 41.2% vs. control: 60.0% male, p = 0.352). Demographics and clinical characteristics of the study cohort are presented in [Table 1]. sROBO4 was comparable between individuals with T2DM and control individuals (T2DM: 1.54 ± 1.19 vs. control: 1.59 ± 0.77 ng/mL, p = 0.46; [Fig. 1A]).

Fig. 1 Levels and clinical associations of Robo4. (A) Bar-chart showing the levels of sRobo4 in the plasma of nondiabetic control individuals (control) and individuals with type 2 diabetes mellitus (T2DM). Outliers (control [n = 4] and T2DM [n = 2]) assessed by Rout's test with Q = 1% were removed. (B, C) Scatter-plot showing the correlation of Robo4 with body mass index (B) and with estimated glomerular filtration rate (eGFR, C). (D, E) Bar charts showing the levels of sRobo4 in the plasma of individuals with or without arterial hypertension (D) and with higher (≥2 atherosclerotic plaques) or lower (0–1 atherosclerotic plaques) atherosclerotic burden (E). Lines in bar charts correspond to mean ± standard error of the mean (SEM). p-Values in (A, D, E) were derived by Mann–Whitney U test and correlation coefficient in B, C by Spearman's rank correlation test. (F) Forrest plot showing the odds ratio and 95% confidence intervals for the association of sRobo4 (log10-transformed) with higher atherosclerotic burden. The values were derived by binary logistic regression analysis with higher atherosclerotic burden as the dependent variable and sRobo4 and each variable shown in the respective line as independent variables. The top line corresponds to univariable analysis including only sRobo4 as independent variable. Group refers to T2DM versus control. The x-axis is log2 transformed for visual clarity. p-value for the association of sRobo4 (log10-transformed) with the presence of ≥ 2 atherosclerotic plaques: unadjusted: p<0.05; adjusted separately for group, age, sex, BMI or total cholesterol: p<0.05; adjusted for eGFR: p=0.143. Table 1 Demographics and clinical characteristics of the study cohort

Whole cohort

(n = 49)

Control

(n = 15)

T2DM

(n = 34)

p-Value[a]

Age (years)

65.6 ± 8.7

65.7 ± 9.2

65.6 ± 8.7

0.975

Male sex

23 (46.9)

9 (60.0)

14 (41.2)

0.352

BMI (kg/m2)

31.5 ± 6.1

28.3 ± 4.6

32.9 ± 6.2

0.015

Smoking (active)

6 (12.2)

3 (20.0)

3 (8.8)

0.353

Arterial hypertension

42 (85.7)

13 (86.7)

29 (85.3)

1.000

Hyperlipidemia

38 (77.6)

11 (73.3)

27 (79.4)

0.716

SBP (mm Hg)

137 ± 21

130 ± 16

140 ± 22

0.103

DBP (mm Hg)

76 ± 8

78 ± 8

76 ± 8

0.452

Pulse pressure (mm Hg)

61 ± 18

52 ± 12

65 ± 18

0.021

Total cholesterol (mg/dL)

181 ± 39

196 ± 33

175 ± 40

0.069

LDL (mg/dL)

107 ± 29

119 ± 27

102 ± 29

0.064

HDL (mg/dL)

49 ± 13

54 ± 17

47 ± 10

0.069

Triglycerides (mg/dL)

122 ± 56

115 ± 48

125 ± 59

0.595[b]

HbA1c (%)

N/A

7.1 ± 1.0

N/A

eGFR (mL/min/1.73 m2)

79 ± 24 (n = 47)

79 ± 20 (n = 14)

79 ± 26 (n = 33)

0.967

Therapy

 Insulin

16 (32.7)

0

16 (47.1)

<0.001

 Antihypertensive drugs

38 (77.6)

11 (73.3)

27 (79.4)

0.716

 Lipid-modifying drugs

35 (71.4)

11 (73.3)

24 (70.6)

1.000

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; N/A, not applicable; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus.

Notes: Continuous variables are presented as mean ± standard deviation. Categorical values are presented as absolute count (%).

ap-Value corresponds to the comparison between individuals with T2DM and control individuals. Continuous variables were compared using independent samples t-test unless otherwise specified. Categorical variables were compared using Fisher's exact test.

bp-Value derived by Mann–Whitney U test.

Given that sROBO4 was comparable between individuals with T2DM and control individuals, we next examined whether sROBO4 levels correlated with traditional CVD risk factors in the whole study cohort. Indeed, we observed a positive correlation between sROBO4 and body mass index (BMI; rs = 0.384, p = 0.006; [Fig. 1B]), as well as a negative correlation between sROBO4 and renal function as determined by eGFR (rs = − 0.429, p = 0.003; [Fig. 1C]). Moreover, individuals with arterial hypertension had higher sRobo4 compared with normotensive individuals (p = 0.03, [Fig. 1D]), although a linear correlation between sRobo4 and systolic or diastolic BP was not observed, probably due to the use of antihypertensive medications by 78% of the study cohort ([Table 1]).

Next, given the association of sRobo4 with CVD risk factors including higher BMI, arterial hypertension, and worse renal function, we examined its potential association with the presence and extent of atherosclerosis. Of interest, we observed that individuals (n = 28) with two or more atherosclerotic plaques in their femoral and/or carotid arteries had significantly higher sRobo4 levels compared with individuals (n = 20) with a total of none or one atherosclerotic plaque (p = 0.008, [Fig. 1E]). Finally, we checked whether the association between sRobo4 and higher atherosclerotic burden was affected by age, sex, presence of T2DM, total cholesterol levels, or any of the CVD risk factors that were shown to correlate with sRobo4 levels, namely BMI, eGFR, and arterial hypertension. Of interest, the association between sRobo4 and higher atherosclerotic burden remained significant after separately controlling for group (T2DM vs. control), age, sex, BMI, and total cholesterol ([Fig. 1F], all p < 0.05), while we could not control for the presence of arterial hypertension since all individuals with ≥2 plaques also had arterial hypertension. However, when we controlled for eGFR, the association between sRobo4 and higher atherosclerotic burden became nonsignificant ([Fig. 1F], p = 0.143), suggesting that the association of Robo4 with atherosclerotic disease may be mediated through worse renal function.

Robo4 is a transmembrane receptor almost exclusively expressed in ECs across organs,[13] while its soluble (circulating) form is the product of constitutive ectodomain shedding by ADAM-family metalloproteinases.[14] sRobo4 has been shown to affect angiogenesis in various mouse models[8] [15] and its levels have been shown to increase after cardiac surgery, especially in patients who later developed acute kidney injury.[16] Moreover, a recent study showed that sRobo4 was increased in individuals with pulmonary hypertension and was associated with higher mortality in these patients.[17] In line with this, we observed increased sRobo4 levels in individuals with arterial hypertension, suggesting the potential association of sRobo4 with endothelial damage. Our results did not show an association of sRobo4 with T2DM, although preclinical models suggest an intrinsic role of Robo4 in disease pathogenesis.[10] However, given the small sample size of our study we cannot exclude that such as association may exist. Larger prospective cohort studies are needed to examine whether sRobo4 may be elevated in subgroups of T2DM patients, as well as whether it may have prognostic value for the development of certain complications such as diabetic kidney disease, given the association of sRobo4 with worse renal function in our cohort. Moreover, our analysis revealed an association of sRobo4 with higher atherosclerotic burden, independent of age, sex, total cholesterol levels, BMI, or the presence of T2DM. To the best of our knowledge, this is the first study to describe a possible value of soluble Robo4 as marker of atherosclerosis. Large prospective cohort studies are needed to validate this finding and to determine the possible value of sRobo4 as an independent prognostic factor for the development and progression of ASCVD including major adverse cardiovascular events. Our results also indicate that the association between sRobo4 and atherosclerosis may be mediated through worse renal function. Previous cohort studies have established chronic kidney disease as an independent factor associated with endothelial dysfunction, subclinical atherosclerosis, and progressive ASCVD.[18] [19] [20] [21] Whether sRobo4 may be a useful biomarker for the prediction of progressive atherosclerosis in these patients should be examined in future cohort studies. A limitation of our study is the small sample size, which did not allow us to examine the effect of multiple confounding factors (e.g., inflammation) on the association between sRobo4 and high atherosclerotic burden. In vivo studies utilizing models of atherosclerosis as well as larger prospective cohort studies are warranted to delineate causality and prognostic value of sRobo4 in ASCVD.

Publication History

Accepted Manuscript online:
14 October 2024

Article published online:
11 November 2024

© 2024. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany