Label-free atherosclerosis diagnosis through a blood drop of apolipoprotein E knockout mouse model using surface-enhanced Raman spectroscopy validated by machine learning algorithm

Abstract

The direct preventative detection of flow-induced atherosclerosis remains a significant challenge, impeding the development of early treatments and prevention measures. This study proposes a method for diagnosing atherosclerosis in the carotid artery using nanometer biomarker measurements through surface-enhanced Raman spectroscopy (SERS) from single-drop blood samples. Atherosclerotic acceleration is induced in apolipoprotein E knockout mice which underwent a partial carotid ligation and were fed a high-fat diet to rapidly induce disturbed flow-induced atherosclerosis in the left common carotid artery while using the unligated, contralateral right carotid artery as control. The progressive atherosclerosis development of the left carotid artery was verified by micro-magnetic resonance imaging (micro-MRI) and histology in comparison to the right carotid artery. Single-drop blood samples are deposited on chips of gold-coated ZnO nanorods grown on silicon wafers that filter the nanometer markers and provide strong SERS signals. A diagnostic classifier was established based on principal component analysis (PCA), which separates the resultant spectra into the atherosclerotic and control groups. Scoring based on the principal components enabled the classification of samples into control, mild, and severe atherosclerotic disease. The PCA-based analysis was validated against an independent test sample and compared against the PCA-PLS-DA machine learning algorithm which is known for applicability to Raman diagnosis. The accuracy of the PCA modification-based diagnostic criteria was 94.5%, and that of the machine learning algorithm 97.5%. Using a mouse model, this study demonstrates that diagnosing and classifying the severity of atherosclerosis is possible using a single blood drop, SERS technology, and machine learning algorithm, indicating the detectability of biomarkers and vascular factors in the blood which correlate with the early stages of atherosclerosis development.

Keywords: ApoE KO mouse; atherosclerosis; nano‐sized biomarker; principal component analysis; surface‐enhanced Raman spectroscopy.

FIGURE 1

Schematic for nanomarker Raman‐based diagnosis of arteriosclerosis. Atherosclerosis animal model preparation, sample acquisition, histopathology and immunofluorescence staining, pretreatment‐ and label‐free SERS measurement. The red arrows track the process of obtaining a Raman signal based on a nanometer marker for inducing the severity of atherosclerosis, and the blue arrows track the verification process using histopathology. KO, knockout; LCA, left carotid artery; PCA, principal component analysis; RCA, right carotid artery; SERS, surface‐enhanced Raman spectroscopy.

FIGURE 2

Atherosclerotic blood flow validation and histopathology. Confirmation of blood flow and atherosclerosis using MRI‐TOF and atherosclerotic tissue sections from representative (a) mild and (b) severe disease samples. Tissue sections were taken from the atherosclerotic left carotid artery (LCA) and compared with the healthy right carotid artery (RCA). Histopathological images were stained with H&E. The MRI‐TOF images are shown in the horizontal and vertical cross‐sections. H&E, hematoxylin/eosin; MRI‐TOF, time‐of‐flight magnetic resonance imaging.

FIGURE 3

Histopathology and fluorescence of immunohistochemistry. Representative histopathological cross‐sections of blood vessels under accelerated arteriosclerosis (LCA) and in comparative vessels (RCA). (a, c) Movat's pentachrome and H&E staining of (a) RCA and (c) LCA show the atherosclerosis evolution. (b, d) Magnified images for the comparison of the plaque formation and corresponding immunofluorescence confirmation of NF‐κB and VCAM‐1 in (b) RCA and (d) LCA; red arrows: foam cells. (e, f) Movat's pentachrome comparison images of the RCA and LCA in (e) mild and (f) severe disease samples. H&E, hematoxylin/eosin.

FIGURE 4

Fabrication, nanofiltration, and test blood measurement of SERS chip. Morphology of SERS chips before and after deposition of pretreatment‐ and label‐free biological samples. (a, b) Cross‐sectional SEM morphology of ZnO nanorods before gold deposition (a) at a 45° angle and (b) perpendicular to the surface. (c, d) Similar conditions to (a) and (b) but following the gold deposition. (e) The same chip, but following the deposition of a blood sample, and (f) with the region of enhanced Raman signal demonstrating the nanofiltration effect of the substrate. (g) Signal comparison of whole blood drop and plasma between the original drop area and filtrated area. (h) Signal difference among measurement distance from the boundary. The average is plotted as a line and the standard deviation as a lighter shade. SEM, scanning electron microscope; SERS, surface‐enhanced Raman spectroscopy.

FIGURE 5

Raman signals by animal group and labeling by major peaks. Raman spectra of blood sampled from control group C57BL/6 wild‐type mice (black), mice with mild atherosclerotic disease (blue), and mice showing severe atherosclerosis (red). Each solid‐colored line shows the average value of the measured spectra, with lighter shades showing the standard deviation. Spectra were normalized based on the value at 1000 cm⁻¹ (black dotted line). Black dots identify peaks suppressed as atherosclerosis progresses. Green bands indicate peaks that emerge as atherosclerosis progresses.

FIGURE 6

Statistical analysis using principal component analysis (PCA). The principal components of Raman spectra were calculated to form a set of eigenvectors with diagnostic validity for atherosclerosis. The data projected into (a) the three‐dimensional space of the first three principal components and into (b) the two‐dimensional space of the reconstructed principal components. Data are colored to indicate the severity of atherosclerosis: control samples: black, mildly atherosclerotic samples: blue, and severely atherosclerotic samples: red. Samples with clogged blood vessels are shown in purple. Colored bands indicate linear boundaries between classifications based on PC1. (c) Reconstructed graph of PC eigenvectors and derivation formulae. Regions of positive contribution to PC1 are shaded with green bands, whereas negative contributions are indicated by purple bands. (d) Accuracy and significance for each group based on PC1 values. The horizontal line inside the box for each group indicates the mean, and the horizontal line outside the box indicates the deviation.

FIGURE 7

Analysis using machine learning algorithm of PCA‐PLS‐DA. (a) Data distribution when the PLS‐DA machine learning algorithm is used with values up to PC50 as variables. (b) Receiver operating characteristic (ROC) curve and area under the curve (AUC) values for mild and severe disease groups, respectively. (c) Distribution of variability from PC1 to PC50, with the corresponding log scale, inset. (d) Total accuracy curve of PCA‐PLS‐DA according to the number of selected PCs. Inset shows accuracy for each group. (e, f) Confusion matrix for the PCA‐PLS‐DA classification with 5 and 50 PCs as variables, respectively. DA, discriminant(s) analysis; PCA, principal component analysis; PLS, partial least squares regression.

FIGURE 8

Validation of diagnostic criteria using additional Raman data. Diagnostic accuracy was validated by performing a diagnosis on additional data using the diagnostic criteria from Figure 6b. The original training data points used to establish the diagnostic criteria are blurred in the background. For each group, the bar chart represents the severity of atherosclerotic samples, as defined according to PC1. (a) Distribution of Raman data and representative histopathological results from ApoE mice with confirmed blood flow 2 weeks after carotid artery ligation (mild atherosclerosis group). (b) Data from mice with clear atherosclerosis progression after the second week following ligation. (c) Data from atherosclerotic mice at the fourth week following ligation. PC, principal component.