Analysis of microbial composition and sharing in low-biomass human milk samples: a comparison of DNA isolation and sequencing techniques

Johanne E Spreckels; Asier Fernández-Pato; Marloes Kruk; Alexander Kurilshikov; Sanzhima Garmaeva; Trishla Sinha; Hiren Ghosh; Hermie Harmsen; Jingyuan Fu; Ranko Gacesa; Alexandra Zhernakova

doi:10.1038/s43705-023-00325-6

Analysis of microbial composition and sharing in low-biomass human milk samples: a comparison of DNA isolation and sequencing techniques

ISME Commun. 2023 Nov 9;3(1):116. doi: 10.1038/s43705-023-00325-6.

Authors

Affiliations

¹ Department of Genetics, University of Groningen and University Medical Center Groningen, Groningen, the Netherlands.
² Medical Center - University of Freiburg, Institute for Infection Prevention and Hospital Epidemiology, Freiburg, Germany.
³ Department of Medical Microbiology, University of Groningen and University Medical Center Groningen, Groningen, the Netherlands.
⁴ Department of Pediatrics, University of Groningen and University Medical Center Groningen, Groningen, the Netherlands.
⁵ Department of Gastroenterology and Hepatology, University of Groningen and University Medical Center Groningen, Groningen, the Netherlands.
⁶ Department of Genetics, University of Groningen and University Medical Center Groningen, Groningen, the Netherlands. a.zhernakova@umcg.nl.

Abstract

Human milk microbiome studies are currently hindered by low milk bacterial/human cell ratios and often rely on 16S rRNA gene sequencing, which limits downstream analyses. Here, we aimed to find a method to study milk bacteria and assess bacterial sharing between maternal and infant microbiota. We tested four DNA isolation methods, two bacterial enrichment methods and three sequencing methods on mock communities, milk samples and negative controls. Of the four DNA isolation kits, the DNeasy PowerSoil Pro (PS) and MagMAX Total Nucleic Acid Isolation (MX) kits provided consistent 16S rRNA gene sequencing results with low contamination. Neither enrichment method substantially decreased the human metagenomic sequencing read-depth. Long-read 16S-ITS-23S rRNA gene sequencing biased the mock community composition but provided consistent results for milk samples, with little contamination. In contrast to 16S rRNA gene sequencing, 16S-ITS-23S rRNA gene sequencing of milk, infant oral, infant faecal and maternal faecal DNA from 14 mother-infant pairs provided sufficient resolution to detect significantly more frequent sharing of bacteria between related pairs compared to unrelated pairs. In conclusion, PS or MX kit-DNA isolation followed by 16S rRNA gene sequencing reliably characterises human milk microbiota, and 16S-ITS-23S rRNA gene sequencing enables studies of bacterial transmission in low-biomass samples.