# An end-to-end approach for single-cell infrared absorption spectroscopy of bacterial inclusion bodies: from AFM-IR measurement to data interpretation of large sample sets

## TL;DR

This paper introduces an end-to-end method using AFM-IR to study bacterial inclusion bodies at the single-cell level, enabling high-throughput analysis of their structure and variation under stress.

## Contribution

A novel AFM-IR pipeline for high-throughput, label-free analysis of bacterial inclusion bodies, including optimized data collection and interpretation techniques.

## Key findings

- AFM-IR revealed substructure and variation in inclusion bodies at the single-cell level.
- Stress conditions caused distinct clustering in AFM-IR spectra, reflecting stress severity and type.
- Correlation analyses showed complex relationships between IB physical and morphological properties.

## Abstract

Inclusion bodies (IBs) are well-known subcellular structures in bacteria where protein aggregates are collected. Various methods have probed their structure, but single-cell spectroscopy remains challenging. Atomic Force Microscopy-based Infrared Spectroscopy (AFM-IR) is a novel technology with high potential for the characterisation of biomaterials such as IBs.

We present a detailed investigation using AFM-IR, revealing the substructure of IBs and their variation at the single-cell level, including a rigorous optimisation of data collection parameters and addressing issues such as laser power, pulse frequency, and sample drift. An analysis pipeline was developed tailored to AFM-IR image data, allowing high-throughput, label-free imaging of more than 3500 IBs in 12,000 bacterial cells. We examined IBs generated in Escherichia coli under different stress conditions. Dimensionality reduction analysis of the resulting spectra suggested distinct clustering of stress conditions, aligning with the nature and severity of the applied stresses. Correlation analyses revealed intricate relationships between the physical and morphological properties of IBs.

Our study highlights the power and limitations of AFM-IR, revealing structural heterogeneity within and between IBs. We show that it is possible to perform quantitative analyses of AFM-IR maps over a large collection of different samples and determine how to control for various technical artefacts.

The online version contains supplementary material available at 10.1186/s12951-024-02674-3.