PureMilk: Developing a Test for Identification of Pathogenic Bacteria in Human Breast Milk

Human milk is vital for all infants, but critical for premature and infants that are ill or born with complications. The nutritional value and bioactive components in human milk are essential for immunological and developmental support, and the unique composition helps mitigate the challenges these infants face, consequently contributing to improved outcomes in growth, health, and overall well-being. At the hospital, infants in the neonatal intensive care unit are fed quality assured donor milk if the mother’s milk supply is insufficient during the hospital stay. The donor milk is provided by milk banks, or milk kitchens in hospitals, who are responsible for collecting, handling and storing the milk until use. In addition, they are tasked with ensuring the quality of the milk, among other by assuring it is free of pathogens. No easy decentral test currently exists, which is a substantial pain for the milk bank.

To address the pain, this Ph.D. project aimed to lay the foundation for a simple point-of-use test that accurately and cost-effectively assesses human milk quality. The multidisciplinary effort in the project comprises research on human milk composition and homogenization, bacterial analysis, paper-based sensor development, signal digitalization, and entrepreneurship. Initial efforts included a thorough investigation of human milk for a comprehensive review on the different constituents and how these affects infant health. An investigation of macronutrient stability and a statistical approach to determining trends in macronutritional content was obtained by assessing data from one year of human milk donations. Furthermore, E. coli has been characterized, and divided into toxin secreting and non-toxin secreting to elucidate challenges from treatment in the milk kitchen. Characterization of STEC toxin secretion in human milk has been investigated experimentally and showed that particularly Holder pasteurization increases toxin secretion. Meaning, in a real-world setting, the consequence of Holder pasteurization of milk containing STEC is that the milk becomes dangerous for infant ingestion.

A low-cost decentral analysis method is obtained using lateral flow assay technology for direct analyte identification. Implementing Lateral flow assay technology for milk analysis necessitates sample pretreatment, with the requirement of keeping the analytes of interest intact. In the current format, it can identify E. coli with the ability to differentiate between Shiga-toxin-secreting E. coli and those that do not possess this virulence factor. The sensitivity of 10³ CFU/mL for naked-eye identification was observed, though based on indicative results. The limit of detection for same test with a digital readout was 10 CFU/mL by computational aid, including image analysis followed by a t-test investigation. As lateral flow assays are limited in terms of sensitivity of naked-eye readouts, and as test results are inconvenient to handle as individual analog data points, a digitalization device containing an automated device for signal digitalization and storage is developed. The digitalization device comprises an absorbance reader, relying on the interaction between an LED, the detection area of the lateral flow assay, and an RGB sensor. This enables the identification and characterization of signals originating from the analyte-bound complex down to a colour signal. It was found that cases categorized as negative when read by the naked eye approach, even by a trained eye, was easily detecting with the reader.

The current data generated in this Ph.D. project indicates a great potential for the sensor, but further studies supporting the findings are needed. This includes increasing the number of replicas in each concentration, enabling a sensitivity assessment including lower and upper detection limits. Additionally, the current format needs to be investigated by testing against a larger library of relevant bacteria, determining the test specificity. The reproducibility is also a matter for further investigation, where ii Summary both batch-to-batch assay variations and production methods should be investigated. A commercialization strategy was formulated but should be revised continuously throughout technology maturation.

This dissertation will be interesting to everyone who wants to improve the current working conditions of milk banks and human milk analysis on a global scale, think outside the box for future aspects of human milk analysis, and indirectly improve infants’ health in the future