Al-Khaz'Aly Ali, Ghandorah Salim, Topham Jared J, Osman Nasir, Louie Taye, Farshidfar Farshad, Amrein Matthias
Department of Medical Sciences, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada; Department of Computer Science, Faculty of Science, University of Calgary, Calgary, AB, Canada.
Department of Medical Sciences, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada.
Biophys J. 2025 Jan 7;124(1):77-92. doi: 10.1016/j.bpj.2024.11.003. Epub 2024 Nov 6.
All living cells vibrate depending on metabolism. It has been hypothesized that vibrations are unique for a given phenotype and thereby suitable to diagnose cancer type and stage and to pre-assess the effectiveness of pharmaceutical treatments in real time. However, cells exhibit highly variable vibrational signals, can be subject to environmental noise, and may be challenging to differentiate, having so far limited the phenomenon's applicability. Here, we combined the sensitive method of force spectroscopy using optical tweezers with comprehensive statistical analysis. After data acquisition, the signal was decomposed into its spectral components via fast Fourier transform. Peaks were parameterized and subjected to principal-component analysis to perform an unbiased multivariate statistical evaluation. This method, which we term cell vibrational profiling (CVP), systematically assesses cellular vibrations. To validate the CVP technique, we conducted experiments on five U251 glioblastoma cells, using 8- to 10-μm polystyrene beads as a control for comparison. We collected raw data using optical tweezers, segmenting into 150+ 5-s intervals. Each segment was converted into power spectra representing a frequency resolution of 10,000 Hz for both cells and controls. U251 glioblastoma cells exhibited significant vibrations at 402.6, 1254.6, 1909.0, 2169.4, and 3462.8 Hz (p < 0.0001). This method was further verified with principal-component analysis modeling, which revealed that, in cell-cell comparisons using the selected frequencies, overlap frequently occurred, and clustering was difficult to discern. In contrast, comparison between cell-bead models showed that clustering was easily distinguishable. Our paper establishes CVP as an unbiased, comprehensive technique to analyze cell vibrations. This technique effectively differentiates between cell types and evaluates cellular responses to therapeutic interventions. Notably, CVP is a versatile, cell-agnostic technique requiring minimal sample preparation and no labeling or external interference. By enabling definitive phenotypic assessments, CVP holds promise as a diagnostic tool and could significantly enhance the evaluation of pharmaceutical treatments.
所有活细胞都会根据新陈代谢而振动。据推测，振动对于给定的表型是独特的，因此适合于诊断癌症类型和阶段，并实时预评估药物治疗的有效性。然而，细胞表现出高度可变的振动信号，可能会受到环境噪声的影响，并且可能难以区分，这限制了该现象目前的适用性。在此，我们将使用光镊的力谱敏感方法与全面的统计分析相结合。数据采集后，通过快速傅里叶变换将信号分解为其频谱成分。对峰值进行参数化，并进行主成分分析以进行无偏多元统计评估。我们将这种方法称为细胞振动分析（CVP），它系统地评估细胞振动。为了验证CVP技术，我们使用8至10μm的聚苯乙烯珠作为对照，对五个U251胶质母细胞瘤细胞进行了实验。我们用光镊收集原始数据，将其分割成150个+5秒的间隔。每个片段都转换为功率谱，代表细胞和对照的频率分辨率均为10,000Hz。U251胶质母细胞瘤细胞在402.6、1254.6、1909.0、2169.4和3462.8Hz处表现出明显的振动（p<0.0001）。通过主成分分析模型进一步验证了该方法，结果表明，在使用选定频率进行细胞间比较时，经常会出现重叠，并且聚类难以辨别。相比之下，细胞-珠模型之间的比较表明聚类很容易区分。我们的论文将CVP确立为一种无偏、全面的分析细胞振动的技术。该技术有效地区分细胞类型，并评估细胞对治疗干预的反应。值得注意的是，CVP是一种通用的、与细胞无关的技术，只需要最少的样品制备，无需标记或外部干扰。通过能够进行明确的表型评估，CVP有望成为一种诊断工具，并可显著增强药物治疗的评估。
