Label-Free Quantification of Protein Density in Living Cells

Abstract

Intracellular water content, W, and protein concentration, P, are essential characteristics of living cells. Healthy cells maintain them within a narrow range, but often become dehydrated under severe stress; moreover, persistent loss of water (an increase in P) can lead to apoptotic death. It is very likely that protein concentration affects cellular metabolism and signaling through macromolecular crowding (MC) effects, to which P is directly related, but much remains unknown in this area. Obviously, in order to study the biological roles and regulation of MC in living cells, one needs a method to measure it. Simple and accurate measurements of P in adherent cells can be based on its relationship to refractive index. The latter can be derived from two or more (depending on the algorithm) mutually defocused brightfield images processed by the transport-of-intensity equation (TIE) that must be complemented by a determination of volume. Here, we describe the experimental considerations for both TIE imaging and for a particular method of cell volume measurement, transmission-through-dye (TTD). We also introduce an ImageJ plugin for solving TIE. TIE and TTD are fully compatible with each other as well as with fluorescence. A similar approach can be applied to subcellular organelles; however, in this case, the volume must be determined differently.© 2025 The Author(s). Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: Sample preparation for TIE with or without TTD Basic Protocol 2: Acquisition of TIE and TTD images Basic Protocol 3: Calibration of TIE Basic Protocol 4: Measurement of the absorption coefficient of the medium used for TTD Basic Protocol 5: Image processing using Fiji Support Protocol 1: Installation and use of TIE plugin Support Protocol 2: Automation of the double TTD/TIE processing using a Fiji macro.

Keywords: cell volume; macromolecular crowding; protein concentration; refractive index; transmission‐through‐dye microscopy; transport‐of‐intensity equation.

Figure 1

Results of calibration of a dry 20×/0.8 objective using 7.7‐µm silica beads (Spherotech, cat. no. SIP‐60‐10). The three embedded images show a bead immersed in an oil with n = 1.4218 and imaged at –2.5 µm below best focus, 2.5 µm above best focus, and the resultant TIE. The scale bar is 5 µm. The numbers in the table illustrate the calculations: the number of beads in a group (N), the selected area in µm², the average T within the selected area (whose value partly depends on how the boundary was drawn), the average T of the background, and the total T per bead corrected for the background. The results for all beads are averaged and divided by the mean bead volume (239 µm³).

Figure 2

(A) Setup to measure the absorption coefficient of a dye‐containing solution. Both the objective and the condenser are focused on the top of the slide. A 5‐mm, half‐ball lens is immersed in a small volume of the solution; the sample is illuminated through a 635‐nm filter. (B) Image intensity around the touch point. The brighter center results from a smaller depth of the light‐absorbing layer. (C) A circle centered on the touch point is drawn on the image. Correct centering is helped by adjusting the image contrast. (D) Radial intensity profile is obtained by the Radial Profile ImageJ plugin; the results are exported into an appropriate software (e.g., Microsoft Excel) and fitted to a linear regression. The slope of the line gives the value of α in µm⁻¹. For solutions containing 7 mg/ml AB9, α ≈ 0.15‐0.16 µm⁻¹.

Figure 3

The original defocused images BF1 and BF2 and a linear TTD image are cropped to the same area. Small areas containing isolated cells or small groups of cells often produce TIE images with a more uniform background. The brightfield images are processed by TIE, and the TTD image is converted into a logarithmic scale. Note that the TIE computation removes the outer strips of pixels, so that the resultant image is 2 pixels smaller than original ones in both dimensions.

Figure 4

TIE images based on four pairs of blank images (transmitted illumination in the absence of a sample) collected with an ORCA‐FLASH4.0L sCMOS camera (Hamamatsu). The contrast has been digitally enhanced to display random variability. Similar results have been obtained with a cooled CCD camera (Cooke). The size of each field is 1327 µm.

Figure 5

Optimization of the parameters for TIE acquisition. These images of a single HeLa cell have been obtained with a 20×/0.8 dry objective. The upper row shows brightfield images, with each successive image collected after moving the objective by 1 µm toward the cell. The image z = 0 is focused approximately on the level of the coverslip. The second, third, and fourth rows are TIE reconstructions computed at Δz separations of 1 µm, 6 µm, and 10 µm centered on the respective planes specified in the first row; for example, the second image of the third row (the one with T = 91.7 and below z = –2) has been computed from z = 1 (3 µm above z = –2) and z = –5 (not shown). The numbers in TIE images show the computed T in relative units. Small Δz result in an uneven background, and the most consistent numbers require focusing at the level of the coverslip or slightly above.

Figure 6

A HeLa cell with prominent vacuoles that resulted from prolonged incubation in a hypotonic medium. Since vacuoles are contained inside the cell, they produce no contrast in TTD (except from a boundary of brightfield origin). However, they reduce local protein mass, as is obvious from the TIE image. Scale bar, 10 µm.