Addressing uncertainties in correlative imaging of exogenous particles with the tissue microanatomy with synchronous imaging strategies

Abstract

Exposure to exogenous particles is of increasing concern to human health. Characterizing the concentrations, chemical species, distribution, and involvement of the stimulus with the tissue microanatomy is essential in understanding the associated biological response. However, no single imaging technique can interrogate all these features at once, which confounds and limits correlative analyses. Developments of synchronous imaging strategies, allowing multiple features to be identified simultaneously, are essential to assess spatial relationships between these key features with greater confidence. Here, we present data to first highlight complications of correlative analysis between the tissue microanatomy and elemental composition associated with imaging serial tissue sections. This is achieved by assessing both the cellular and elemental distributions in three-dimensional space using optical microscopy on serial sections and confocal X-ray fluorescence spectroscopy on bulk samples, respectively. We propose a new imaging strategy using lanthanide-tagged antibodies with X-ray fluorescence spectroscopy. Using simulations, a series of lanthanide tags were identified as candidate labels for scenarios where tissue sections are imaged. The feasibility and value of the proposed approach are shown where an exposure of Ti was identified concurrently with CD45 positive cells at sub-cellular resolutions. Significant heterogeneity in the distribution of exogenous particles and cells can be present between immediately adjacent serial sections showing a clear need of synchronous imaging methods. The proposed approach enables elemental compositions to be correlated with the tissue microanatomy in a highly multiplexed and non-destructive manner at high spatial resolutions with the opportunity for subsequent guided analysis.

Keywords: correlative imaging; endogenous elemental imaging; exogenous metal imaging; lanthanide X-ray fluorescence; metal-labelled antibodies; synchrotron X-ray fluorescence spectroscopy; synchrotron confocal X-ray fluorescence spectroscopy.

Graphical Abstract

Candidate elements conjugated to antibodies were used to label biological tissues and have been imaged using X-ray fluorescence spectroscopy.

Fig. 1

Demonstrates the three-dimensional variation in cellular content associated with serial histological sectioning (a and b) and Ti distribution (c and d) from tissues previously associated with a Ti BAH. (a) Tissues were prepared to 3 µm in thickness, counterstained with haematoxylin, and the depths of each section are listed accordingly. (b) Magnified inserts are displayed showing an inflammatory infiltrate. Confocal XRF images displaying the 3D Ti distribution within two tissue sections (c) and (d), which are equivalent tissues imaged in (a) and (b). The distributions were observed at depth increments of 15 µm and are labelled accordingly. The colour bar represents fluorescence Ti counts.

Fig. 2

Analyses of serial murine liver sections. (a) H&E image identifies cellular compositions consisting of diffuse lymphocytes. (b) XRF image panel (1.5 µm resolution). H&E and XRF images do not represent identical regions as these were imaged on serial sections. Scale bar represents 50 µm.

Fig. 3

Analysis of a murine spleen, (a) P XRF intensity map segmented to show CD45 positive cell positions. (b) XRF spectra taken from a CD45 positive cell (red), ECM (blue) and background (black). (c) Zoomed in region within (b) to highlight the Sm peak. (d) XRF image showing the position of exogenous Ti (red) within the P image (green). (e) XANES analysis showing Ti feature highlighted in (d) with anatase and metallic Ti standards.

Fig. 4

A simulated XRF spectrum (a) shows peak positions of endogenous tissue elements Cl, K, Ca, Fe, and Zn with three candidate element tags (Nd, Sm, and Eu). (b) Illustrates which elements tags provide a fluorescence signal that is discrete from the endogenous tissue signal (central white region). (b) Also highlights potential sources of convolution (coloured rings) when additional exogenous elements are present (outer white circle). For example, the presence of V would cause convolution of Pr and Nd, and therefore these elements should be avoided as candidate labels.