Evaluating different fixation protocols for spectral cytopathology, part 2: cultured cells

Abstract

Spectral cytopathology (SCP) is a robust and reproducible diagnostic technique that employs infrared spectroscopy and multivariate statistical methods, such as principal component analysis to interrogate unstained cellular samples and discriminate changes on the biochemical level. In the past decade, SCP has taken considerable strides in its application for disease diagnosis. Cultured cell lines have proven to be useful model systems to provide detailed biological information to this field; however, the effects of sample fixation and storage of cultured cells are still not entirely understood in SCP. Conventional cytopathology utilizes fixation and staining methods that have been established and widely accepted for nearly a century and are focused on maintaining the morphology of a cell. Conversely, SCP practices must implement fixation protocols that preserve the sample's biochemical composition and maintain its spectral integrity so not to introduce spectral changes that may mask variance significant to disease. It is not only necessary to evaluate the effects on fixed exfoliated cells but also fixed cultured cells because although they are similar systems, they exhibit distinct differences. We report efforts to study the effects of fixation methodologies commonly used in traditional cytopathology and SCP including both fixed and unfixed routines applied to cultured HeLa cells, an adherent cervical cancer cell line. Data suggest parallel results to findings in Part 1 of this series for exfoliated cells, where the exposure time in fixative and duration of sample storage via desiccation contribute to minor spectral changes only. The results presented here reinforce observations from Part 1 indicating that changes induced by disease are much greater than changes observed as a result of alternate fixation methodologies. Principal component analysis of HeLa cells fixed via the same conditions and protocols as exfoliated cells (Part 1) yield nearly identical results. More importantly, the overall conclusion is that it is necessary that all samples subjected to comparative analysis should be prepared identically because although changes are minute, they are present.

Figure 1

(A) Stacked mean absorbance spectra (overlay shown in the inset), (B) overlay mean second-derivative spectra (nine-point Savitzky-Golay smoothing window), and (C) stacked second-derivative spectra of unfixed (dried) cells 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue) after sample preparation.

Figure 2

(A) PCA scores plot of unfixed (dried) cells 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue) after sample preparation. (B) Loading vectors PC1 and (C) PC2.

Figure 3

Photomicrographs of (top) unfixed (dried), (middle) SurePath- and (bottom) formalin-fixed cells 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue) after sample preparation.

Figure 4

Mean second-derivative spectra (nine-point Savitzky-Golay second-derivative window) for cells fixed and unfixed 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue) after sample fixation of 24 h.

Figure 5

(A) PCA scores plot of cells fixed (24 h) by three different methods and monitored for 1 month: unfixed (dried; red), SurePath- (green) and formalin-fixed (blue). Corresponding loading vectors (B) PC1 and (C) PC2.

Figure 6

(A) PCA scores plot of cultured HeLa cells fixed for 24 h in SurePath and desiccated between each time point (24 h, 1 week, 2 weeks, and 4 weeks). (B, C) Corresponding loading vectors PC1 and PC2, respectively. (D) Photomicrographs of one cell monitored at each time point chronologically from top to bottom.

Figure 7

Mean (A) absorbance (overlay shown in the inset) and (B) second-derivative spectra of cultured HeLa cells fixed in SurePath for 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue).

Figure 8

Mean (A) absorbance (overlay shown in the inset) and (B) second-derivative spectra of cultured HeLa cells fixed in formalin for 24 h (red), 1 week (green), 2 weeks (black), and 4 weeks (blue).

Figure 9

(A) PCA scores plot and loading vectors (B) PC1 and (C) PC2 of exfoliated oral mucosa cells discussed in Part 1 (circles) and cultured HeLa cells discussed thus far (squares) fixed by each method: unfixed (red symbols), SurePath (green), and formalin (blue).