Interview: Jannie Henderickx on how mode of delivery affects fungal colonisation of the gut in preterm infants
A study entitled “The first fungi: mode of delivery determines early life fungal colonization in the intestine of preterm infants”
was published recently in Microbiome Research Reports, co-authored by Jannie Henderickx, former PhD candidate at the laboratory of Microbiology at Wageningen University & Research in the Netherlands and a customer of Novogene. Jannie’s PhD research focused on the development of the gastrointestinal (GI) tract of preterm infants from a microbiological perspective. The start of life for preterm infants is very different than that of full-term infants, as they are usually exposed to a lot of clinical and environmental variables that influence the microbiome heavily, such as antibiotics, adapted feeding modes, and respiratory support. It is already widely known that there is a huge amount of variability within the microbiome, so many scientists are interested in studying the effects of the start of life on the mycobiome. Up until now, this had not been studied in preterm infants, and it would be logical that there would be a fungal community in the gut of the infants.
Jannie undertook this study in conjunction with one of the ten NICUs in the Netherlands at the Isala Women and Children’s Hospital in Zwolle and developed a close collaboration with doctors and research nurses at the facility. The samples obtained for analysis derived from a previous study, the EIBER study, that was carried out from 2012 to 2014, with the aim of investigating the effects of antibiotic treatment on microbiome development within the first 6 weeks of life. These samples were still available so were used by Jannie to characterise the mycobiome of the same infants. The outcomes are interesting, and comparative to those of James et al., 2020. It is the second study of its kind and concluded with similarities and notable differences to the findings of James et al., which Jannie discussed in detail with us in a recent interview.
The study was conducted on a final amount of 109 samples from 56 participants. Because the samples had been used as part of a previous study, there were some limitations in availability and amount. Many more samples were initially intended to be included in the study, but DNA extraction yielded quite a low level of genetic material in many of the samples. Preterm infant faecal samples have a low biomass anyway, but as the mycobiome makes up only ~0.1% of the total genetic content of the microbiome, fungal biomass was even more difficult to extract in sufficient quantities, resulting in many samples failing quality control. These limited sample numbers were also a reason for categorising the infants based on gestational age, allowing for more samples per group for a better picture from a statistical perspective.
A variable not accounted for within the study was the ration of human milk to formula that the infants received. This was because they found no evidence that the amount of human milk consumed had any effect on the mycobiome. Also, sample number was already a limitation, so sub-dividing the groups further based on feeding mode would have created smaller groups with little data resulting from the analysis of each. The type of milk consumed by the infants in this study was very variable and if any evidence of its effect been identified on the mycobiome, it would have been investigated further.
Throughout the sample processing and preparation, Jannie also subjected DNA mock communities to the same protocols, ensuring they would work as a viable standard for quality assessment of the sequencing procedures and library preparations. Within these mock communities the fungal composition and proportions are known. This is helpful during data annotation and taxonomic assignment as it ensures the correct genera and species are being annotated if the data deviates from the normal pipeline, as occurred in this study. The mock community used in this study was a commercially available one containing species related to the GI tract of adults, which served well as a control for the purposes of this research.
Initial results showed a high relative abundance of single phyla in both preterm and full-term infants, leading to a very homogenous mycobiome. This correlated to previous data and was hypothesised as a finding, particularly due to the fact that all infants included in the study had undergone antibiotic treatment (the aim of the EIBER study from which the samples derived). Treatment with antibiotics is known to stimulate growth of Candida species, and dominance of Candida was found in most cases within this study. There was still a high level of individuality however, likely due to the myriad of clinical interventions preterm infants are exposed to which are known to influence the microbiome and therefore probably the mycobiome.
Candida spp. relative abundance was prevalent in preterm infants and full-term infants, with an observed trend of increasing relative abundance from preterm to full-term infants. It could be hypothesised that this was due to antibiotic treatment, however, due to a small sample size, this is not known for certain. A much more targeted approach would be required to investigate this. It would make sense to see Candida enrichment due to antibiotic treatment, however there are many other factors that could be responsible for this high abundance, such as mode of delivery.
Most preterm infants are delivered via caesarean section (c-section), and this has a significant impact on the development of the mycobiome when compared with vaginal birth. Infants born via c-section are found to have a skin-like microbiome, while vaginally delivered infants have a vaginal-like microbiome. The type of c-section can also have an impact on microbiome development, whether it is a planned or emergency procedure. The main distinction here is that the mother will have typically undergone labour when an emergency c-section is decided upon, meaning it likely is much more similar to a vaginal delivery than a planned c-section, where often no labour has occurred yet. However, this hypothesis could not be confirmed by the data. It is difficult to make a conclusion or hypothesis about why this was the case as there is commonly no distinction made between the types of c-section in previous literature.
As mentioned, fungi are estimated to comprise only 0.1% of the total genetic content of the microbiome in the GI tract. To be able to make interkingdom inferences about interactions between bacteria and fungi it is essential to include absolute, as well as relative, abundances to determine how bacteria influence the fungi and vice versa. In this study they propose qPCR of samples as a piece of future work related to the work in this study, to provide quantitative data about the fungal populations within the gut. Also required for the general future of mycobiome studies is the development of databases, as currently many species remain as “unknown” during taxonomic annotation. This issue may also relate to “fungi being fungi” and existing in multiple forms. Often, two forms of the same species will be identified as two different species, or even genera. Mycobiome is gaining traction however, so Jannie is hopeful that within the next few years, these resources will be developed and built out to increase our understanding of what a normal mycobiome composition looks like.
As part of her PhD research, Jannie was involved in establishing a new cohort of preterm infants in the “From Mum to Bum” study, and research into this cohort is currently underway. This study is being conducted as a follow-up to the EIBER study and the main aim is to investigate the effect of preterm microbiome development on human milk digestion. With regards to future work on investigations into the development of the mycobiome, this new cohort may provide access to a larger sample size with more longitudinal data, as samples are taken in the first 6 weeks of life with follow ups at 3 and 6 months of age. As the size of the cohort in the current study was a limitation, this new cohort of preterm infants may provide more data over a longer time period, allowing for the analysis of different parameters and a more comprehensive comparison with previous studies.
For this study, Novogene’s ITS amplicon-based sequencing service was used. ITS2 has been determined to be the universal barcode for fungi. Between ITS1 and ITS2 there are differences in the quality of data produced, or how precise they are for data annotation, however for now ITS2 is the industry standard. When determining the best protocol for this study, previous literature to refer to was limited, as only the aforementioned James et al. study had been carried out with regards to the preterm infant mycobiome. In conjunction with Novogene Technical Support, Jannie and her colleagues determined the best protocol and sequencing strategy for her research goals and produced a great piece of work with interesting outcomes and lots of future potential. Novogene would like to wish Jannie the best of luck with her future career!