Spatially Resolved Molecular Analysis of Host Response to Medical Device Implantation Using the 3D OrbiSIMS Highlights a Critical Role for Lipids

Abstract

A key goal for implanted medical devices is that they do not elicit a detrimental immune response. Macrophages play critical roles in the modulation of the host immune response and are the cells responsible for persistent inflammatory reactions to implanted biomaterials. Two novel immune-instructive polymers that stimulate pro- or anti-inflammatory responses from macrophages in vitro are investigated. These also modulate in vivo foreign body responses (FBR) when implanted subcutaneously in mice. Immunofluorescent staining of tissue abutting the polymer reveals responses consistent with pro- or anti-inflammatory responses previously described for these polymers. Three Dimensional OrbiTrap Secondary Ion Mass Spectrometry (3D OrbiSIMS) analysis to spatially characterize the metabolites in the tissue surrounding the implant, providing molecular histology insight into the metabolite response in the host is applied. For the pro-inflammatory polymer, monoacylglycerols (MG) and diacylglycerols (DG) are observed at increased intensity, while for the anti-inflammatory coating, the number of phospholipid species detected decreased, and pyridine and pyrimidine levels are elevated. Small molecule signatures from single-cell studies of M2 macrophages in vitro correlate with the in vivo observations, suggesting potential for prediction. Metabolite characterization by the 3D OrbiSIMS is shown to provide insight into the mechanism of bio-instructive materials as medical devices and to inform on the FBR to biomaterials.

Keywords: 3D OrbiSIMS; Macrophage; immune‐instructive polymers and metabolomics.

Figure 1

Schematic of the in vivo study experimental procedure. Catheters coated with copolymers were implanted subcutaneously into a mouse model of FBR for 28 days. Following implantation, fresh tissue samples were cut and mounted onto glass slides. For histological analysis, tissue sections were stained with H&E to assess tissue structure, MTC to analyze collagen thickness as an indication of fibrosis, and IHC stains to characterize the macrophage phenotype at the catheter‐tissue interface. For 3D OrbiSIMS, tissue slides were washed with ammonia formate, plunged frozen in liquid nitrogen, and then freeze‐dried. The GCIB was rastered across the tissue section with the Orbitrap analyzer collecting the high‐resolution mass spectrum, followed by multivariate analysis to undertake unbiased sample comparison, complemented with targeted analysis.

Figure 2

Principal component analysis (scores and loadings) for different tissue samples. A) Principal component scores plot of PC1 and PC2 for the 3D OrbiSIMS spectra of PDMS, M1 polymer, and M2 polymer tissue section samples on positive polarity. B) Principal component analysis of three different tissue samples, loadings plots of the first (PC1) and (PC2) principal components and peak were assigned to glycerolipids (green) and amino acids (blue). The peak at m/z 326.3781 (didecyldimethyl ammonium), is a commercial surface disinfectant unintentionally introduced to the tissue samples somewhere in the sample handling process.

Figure 3

Relative quantification of characteristic lipid fragments detected by OrbiSIMS in positive and negative ion mode. A) The normalized intensity of four different glycerolipid species in each tissue sample. B) H&E stain image shows the three regions further away from the implant was analyzed, implant (black), mid‐point (pink), and dermis (sky blue). The normalized intensity of glycerolipid as a function of distance from the implant in each sample C) PDMS, D) M1‐polymer, and E) M2‐polymer. F) Venn diagram comparison of the number of phospholipids detected in 3 different tissue samples using 3D OrbiSIMS and unique lipid signature for each sample. G) The list of unique lipid signatures in each sample. H) Normalized intensity of phospholipids in three separate samples.

Figure 4

Characteristic amino acid fragments were detected in the tissue section in positive ion mode. A) The normalized intensity of amino acid fragments in each sample, PDMS (blue), M1‐polymer (orange), and M2‐polymer (red). The normalized intensity of amino acids further away from the implant in each sample B) PDMS, C) M1‐polymer, and D) M2‐polymer, the implant (black), mid‐point (pink), and dermis (sky blue). E‐G) The spectrum of amino acids with RG sequences from each tissue sample.

Figure 5

A) Metabolites detected from in vivo tissue were significantly affected by M1 and M2‐polymers. B) Comparing the compounds pyridine and pyrimidine from in vivo tissue samples and single‐cell macrophage analysis shows simple linear relationships for pyridine and pyrimidine.^{[ 39 ]} C) Histamine from in vivo and D) Purine from in vivo are found predominantly in tissue adjacent to M2‐polymer or PDMS.

Figure 6

Chemical imaging of tissue sample. A,B) Ion images of M1‐plymer and M2‐polymer were acquired (area of 450 × 450 µm), including pyridine, pyrimidine, and histamine metabolite which are divided by total intensity.

Figure 7

Chemical imaging of tissue sample. (A,B) Ion images of the sum of glycerolipids in M1‐polymer and M2‐polymer were acquired (area of 450 × 450 µm), including MG and DG which are divided by total intensity. The blue lines denoted the upper and lower boundaries of the tissue, with the catheter at the bottom of the image.