Ability of Bifidobacterium

Background The composition of the infant microbiome can have a variety of short-and long-term implications for health. It is unclear if maternal probiotic supplementation in pregnancy can impact infant gut microbiome. Objective The aim of our study was to investigate if maternal supplementation of a formulation of Bifidobacterium breve 702258 from early pregnancy until three months postpartum could transfer to the infant gut.


Introduction
The composition of the infant microbiome is important for long-term health 1 .During and after birth, the infant gut is colonised through a variety of means 2 .Members of the genus Bifidobacterium are among the earliest and found in lower levels in infants born via caesarean versus vaginal delivery [3][4][5] .In addition, breast feeding influences the abundance of Bifidobacterium in infant stool 6 .Due to their 'Generally Regarded as Safe' status, the commercial potential of certain Bifidobacterium spp.has driven much of the investigation into their purported benefits, which include positive effects on diarrhoeal disease 7,8 , resistance to microbial infection 9,10 , risk of colorectal cancer 11,12 , and the immune system 13,14 .
It is unclear clear if probiotic supplementation to the mother during pregnancy can impact on the health of her infant 2,4,23 .Previous studies included short durations of maternal antenatal probiotic supplementation and limited antenatal investigation of potential factors including mother-to-infant transfer 23 .The aim of our study was to investigate whether maternal supplementation of B. breve 702258 from early pregnancy until 3-months postpartum, would result in transfer of the strain to the infant gut.This bifidobacterial strain was chosen as it has probiotic potential, given it has been shown to influence fatty acid composition by increasing CLAs.This has the potential to positively influence infant health, given CLAs may influence neurodevelopment and inflammation [24][25][26] .

Study overview
.
The MicrobeMom study was a single-centre, double-blind, placebo-controlled, randomised controlled trial (RCT) of a supplement with B. breve 702258, a strain previously shown to have lipid-modulating potential, in healthy pregnant participants.The trial was prospectively registered with the ISRTCN in August 2016 (ISRCTN53023014, https://doi.org/10.1186/ISRCTN53023014).The study is reported according to the EQUATOR network using the CONSORT guidelines (supplemental file 1) 27 .The study did not include patient or public involvement in trial design.

Ethical approval
The study received institutional ethical approval from the National Maternity Hospital in February 2016 (EC 35.2015).Written informed maternal consent was obtained and study processes were completed in accordance with the Declaration of Helsinki 28 .As this intervention was a nutritional supplement and not a medication, neither approval from Health Products Regulatory Authority, nor a data monitoring committee was mandated.Study recruitment commenced on 15 th of September 2016 and the last visit was completed on 12 th July 2019.

Participant selection
Pregnant individuals were screened for eligibility for the study at their first antenatal visit (between ten-and 15-weeks' gestation), through review of their medical chart.Individuals were eligible if they had a body mass index (BMI) between 18•5 and 35 kg/m 2 , were aged over 18 years, capable of giving informed consent, and had an adequate level of English language to enable study comprehension.Exclusion criteria included history of gestational diabetes (GDM), diabetes mellitus or pre-diabetes, multiple pregnancy, fetal anomaly, previous perinatal death, any medical condition requiring treatment or unwillingness to limit intake of other probiotic food or supplements during the trial (Fig 1).Out of n=3,548 women assessed for eligibility, n=160 were randomised to the intervention (Fig. 1). .

Intervention
At the randomisation visit (approximately 16 weeks' gestation) participants were allocated to either the intervention or control group.Block randomisation was determined by an independent statistician.Randomly permuted block sizes were generated, and study group allocation was placed in an envelope with the corresponding unique participant identifier.The intervention group received a daily oral supplement of B. breve 702258 (all containing a minimum of 1x10 9 colony forming units; designated from here as "probiotic") until three months postpartum, while the second group received an otherwise identical placebo (Fig. 1).
The supplements were provided by the PrecisionBiotics Ltd.Researchers and participants remained blinded to study allocations until all results were analysed.

Data collection and trial management
Demographic information collected at the baseline visit included maternal age, ethnicity, parity, and current smoking status via self-report and electronic medical records.Highest level of educational attainment was assessed through a questionnaire.Economic advantage was assessed using the Pobal Haase-Pratschke (HP Pobal) Deprivation Index 29,30 .Participants were classified as economically advantaged based on a score greater than zero.Baseline weight was measured to the nearest 0.1kg with the mother in light clothing, using a SECA weighing scales (SECA gmbh & co.kg.Hamburg, Germany).Height was measured to the nearest 0.1 cm, using a wall mounted stadiometer, after removal of footwear.These values were used to calculate maternal body mass index (BMI).The trial steering committee met monthly and reviewed study activities.
Maternal venous blood samples were collected at 16-weeks' and 34-weeks' gestation after an overnight fast.Maternal breastmilk samples were collected at 1-month postpartum.Maternal stool samples were collected at 16 weeks' and at 34-weeks' gestation.Participants were asked to provide a whole bowel motion produced within 24 hours before their visit that was stored in a cool dry place until their appointment.A sub-sample of stool (1 g) was taken and added to 7 ml RNAlater.As an environmental control, RNAlater was added to stool-free sample pots and frozen.The remaining sample was stored in containers at -80°C.

Infant
Neonatal stool samples at 1-week, 1-month, and 3-months postpartum were collected from infant nappies.A sub-sample of stool (1 g) was taken and added to 7 ml RNAlater.As an environmental control, RNAlater was added to stool-free sample pots and frozen.The remaining sample was stored in containers at -80°C.PCR analysis was carried out samples to 3-months of age, culture work was performed on limited samples to 3-months of age, and metagenomic sequencing was performed up to 1-month of age.

Primary outcome
The primary outcome was transmission of the bifidobacterial strain to infant stool.
Confirmation of transfer was considered when the results of at least two methods of analysis (detailed below) indicated a positive result.

Nucleic acid extraction
DNA was extracted from all RNAlater fixed stool samples within 8 weeks of sample collection using the AllPrep DNA/RNA kit (Qiagen).Briefly, 100 mg of faecal sample was centrifuged at maximum speed for 10 min and excess RNAlater and supernatant was removed.The pellet was resuspended with 100 μl bacterial lysis buffer (30 mM Tris-HCl, pH 8.0, 1mM EDTA plus 15 mg ml-1 lysozyme) and 10 μl proteinase K (20 mg ml-1)).Following incubation, 1.2 ml Qiagen RLT Plus buffer containing 1% β-mercaptoethanol was added to the sample and vortexed briefly.Samples were transferred to 2 ml sterile bead beating tubes filled with 1 ml of 0.1 mm glass beads and subjected to bead beating for 3 min.700 μl of lysate was added to a QIAshredder spin column (Qiagen) and centrifuged at maximum speed for 2 min.DNA was extracted from the lysed sample using the AllPrep kit (Qiagen) as per manufacturer's instructions.Negative control extraction blanks were included for every 50 samples.DNA was extracted from breast milk samples as previously described 31 .

Metagenomic sequencing and quality control
DNA was quantified and normalized to a concentration of 0.2 ng/ml using the Qubit™ dsDNA HS Assay Kit.Sequencing libraries were prepared following the Nextera XT DNA Library Preparation Kit (Illumina) protocol and then pooled to a concentration of 2 mM.The libraries were sequenced using an Illumina NextSeq platform with a 2x300 bp paired-end kit, generating raw data sets that were subjected to base calling with Illumina's bcl2fastq software (v 2.19).
Adapter removal and quality trimming were performed using TrimGalore (v 0. Community-level microbiome analysis was performed using the vegan package in R with the MetaPhlAn3 species table as input.Alpha (within-sample) diversity was calculated using Shannon, Simpson, and observed species indexes.Statistical difference in diversity between trial groups was tested with multiple comparison using Bonferroni.Beta (between-sample) diversity was calculated using Bray-Curtis dissimilarity with significant differences between trial groups tested with PERMANOVA using vegan's adonis function.

Selective isolation of Bifidobacterium spp.
Stool aliquots were retrieved from storage and allowed to thaw at room temperature before resuspension in phosphate buffered saline (PBS; Sigma Aldrich) + 0.06% cysteine-HCL (Sigma Aldrich) at a concentration of 0.1 mg ml -1 .Samples were thoroughly resuspended by vortexing and then serial ten-fold dilutions performed and plated onto modified MRS (Difco™ Lactobacilli MRS Broth, BD) broth containing either lactose (Merck), 2'FL, or LNnT (both of these human milk oligosaccharides were obtained through the donation programme of Glycom A/S, a DSM group company) as sole sugar.Breast milk aliquots were allowed to thaw on ice for 30 min prior inoculation into MRS, supplemented with lactose and 0.06% cysteine-HCl.
. After 24 hours incubation in anaerobic conditions at 37 ºC, the cell suspension was plated in modified MRS broth containing lactose.Individual colonies were screened in both cases and isolated strains were confirmed to belong to the Bifidobacterium spp.by a biochemical assay to detect fructose-6 phosphate phosphoketolase activity as previously described 32 .

Bacterial genomic DNA extraction
Overnight bacterial cultures of a single colonie were grown in mMRS + 0.5% (w/v) lactose and DNA extracted using GenElute bacterial DNA isolation kit (Merck) according to manufacturer's instructions.Briefly, cells were lysed using Gram positive lysis solution supplemented with 2.115 × 10 6 units ml -1 lysozyme from chicken egg white (Merck) and incubated for 30 min at 37℃.RNA was degraded using RNase A provided in the kit.Posterior incubation in lysis solution C plus proteinase K (20 mg ml -1 ) was performed followed by the addition of 96% (v/v) ethanol.The lysate was then transferred to a GenElute Nucleic Acid Binding Columns.Following binding and washing, DNA was eluted in 110 µl of Elution solution (10 mM Tris-HCl, 0.5 mM EDTA, pH 9.0).
Raw sequence reads were assembled into contigs using Spades (v 3.1.1).Prodigal (v 2.6.Strains that had an average nucleotide identity (ANI) >= 99.9 % when compared to the B. breve 702258 genome were deemed to be the same strain Finally, the phylogenetic tree of the core single nucleotide polymorphism (SNP) regions of was examined to identify if these strains formed monophyletic group with B. breve 702258.

PCR detection of B. breve 702258
To

Secondary outcomes
Prespecified secondary outcomes included maternal lipids, glucose, insulin resistance (homeostatic model assessment (HOMA-IR)), and C-reactive protein.Fasting glucose samples were analysed at the clinical site after the shortest possible time interval post sample collection 33,34 .The Beckman Coulter AU680 analyser and the Hexokinase method was employed for this measurement.The rest of the samples taken were centrifuged at 3000 rpm for 10 min.Aliquots were stored at −80°C.Total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglyceride concentrations were analysed on a Roche c702 analyser (Roche Diagnostics).Low density lipoprotein (LDL) cholesterol was estimated using the Friedewald equation 33 .Serum measurements of insulin were determined Insulin was quantified by .automated immunoassay (Roche, E170) with typical coefficients of variance <5%.The HOMA-IR was calculated from fasting glucose and insulin levels using the following equation: (fasting insulin (Mu/l) x fasting glucose (mmol/l))/22•5 34 .
Additional outcomes included maternal and infant microbiome (composition and abundance), inflammatory markers (C3 Complement (C3), Interleukin 6 (IL-6), Tissue-necrosis factor alpha (TNF-Alpha)), lipoproteins (Apo-a, Apo-B100 and Lipoprotein a), Glycosylated Haemoglobin (HbA1c) and pregnancy outcomes.Kits from R&D systems, a bio-techne brand were used to measure TNF-alpha.The electrochemiluminescence immunoassay (ECLIA) was used for the quantitative determination of IL-6 using the Roche cobas e411.Complement C3 was completed using the Abbott Architect c4000.Rate of gestational weight gain per week was calculated and compared to the institute of medicine guidelines 35 .

Adherence
Participants were asked if agreed they took the capsules every-day using a five-point Likert scale, at 34-weeks' gestation, 1-month, and 3-months postpartum, as done previously 38 .
Participants were also asked to return any unused capsules at the end of the study and received fortnightly text messages.

Sample size and power
A power analysis determined 60 participants per group were required to provide at least one stool sample up to 3 months of age to detect a difference in the primary outcome at a significance level of 5% and a target power of 80%.In the absence of available literature, this was based on an expected colonisation rate in the placebo of 0% and 15% in the intervention group.

Statistical analysis
Continuous variables were assessed for normality and all non-normally distributed data underwent log 10

Results
The consort flow diagram is shown in Fig. 1.There were 135 infants (65 in the intervention and 70 in the control group) for whom we received a stool sample at either the 1-week, 1month, or 3-month timepoint.Of these, 24 infants had one sample (6 at 1-week, 13 at 1-month and 5 at 3-months), 58 had two samples (12 had 1-week and 1-month samples and 42 had 1month and 3-month samples) and 53 infants had samples at all three timepoints.In total, 300 infant stool samples were analysed (74 at 1-week, 120 at 1-month and 106 at 3-months).

Primary Outcome
The primary outcome of detection of the supplemented strain was confidently identified by at least two methods in two different infants within the probiotic group and none in the placebo group (Table 2).These two mothers both had a previous liveborn infant, were White Irish and had a BMI in the overweight category (BMI between 25-29.9kg/m 2 ) and strongly agreed they took the probiotic capsules every day in late pregnancy.Both infants were male and born at term, one by planned prelabour caesarean delivery and one by spontaneous vaginal delivery.
Both were breastfed on hospital discharge.

Detection of the probiotic strain
Using PCR at any time-point suggestive evidence of the B. breve 702258 strain was found in five infants, three in the probiotic (4•6%) and two in the placebo group (2•9%) (Fig. 2 a).Of these, one infant in the probiotic group had a positive PCR at all three timepoints.Another infant from the probiotic group had two positive samples but lacked a 1-week sample.The third infant in the probiotic group had a positive PCR detection at the 3-month timepoint but not at 1-month and did not have a 1-week sample.The two infants in the placebo group had samples tested at all three time-points, but for one, the positive signal was found in the 3-month sample only and for the other, at 1-month only.Using a targeted culture-based approach, 60 distinct B.
breve isolates were obtained from infant stool, however genome sequencing revealed only two of these were highly similar (>99•99% identity) to the supplemented strain (Fig 2 c).Both were from infants in the probiotic group (detection rate in probiotic group of 3•1%), at either 1month or 3-months.Both samples also had a respective positive PCR result.Culture analysis did not find any positive isolates in the placebo group (0%).There were 37 distinct occurrences of B. breve in infant stool detected.Only one metagenomic assembled genome (MAG) from a single stool sample in the probiotic group was identified with a high similarity to the genome of the supplemented strain (Fig. 2 b).This was in a 1-month infant sample that also had both a positive PCR detection and to which a confirmed isolate had been recovered and sequenced at the 3-month timepoint.

Secondary Outcomes
No differences were found in baseline values of prespecified secondary or additional metabolic outcomes between groups so only the late pregnancy values are reported.No differences were noted in metabolic factors in late pregnancy ( Using PCR analysis, the probiotic strain was detected in nine maternal samples, one at randomisation, four at 34-weeks' gestation, and four at 1-month postpartum (Fig. 2 a).Only two of these (both at 34-weeks and in the intervention group) had cultured bacterial isolates from the same sample with an ANI score of >99.99% to B. breve 702258.The maternal sample with a positive PCR at randomisation did not show evidence of the probiotic by either culture work or metagenomics.A total of 55 B. breve isolates were cultured from maternal stool and breastmilk samples.Of these, four isolates had a genome identity of >99•99% to the probiotic.
Only one of these samples came from a mother that did not have a positive PCR.A single isolate was recovered from breastmilk at 1-month that matched the supplemented strain.This was from a mother/infant dyad which had a positive detection in the infant stool at the same timepoint.This infant was fed some breastmilk (and formula) from delivery but had moved to formula only by three months.No B. breve MAG with a similarity to the supplemented strain were recovered from maternal stool metagenomic sequencing.Taken together, only two mothers from the probiotic group (2/65, 3•1%) and none in the control group had evidence of the probiotic strain confirmed in their stool through two methods p=0•230.Whilst PCR analysis determined the presence of B. breve 702258 in two different mother/infant dyads, this was not reproduced by the other methods at the same timepoint.Independent of time point and sample site however, there was evidence for the strain in a single dyad by both PCR and culture.
There was no difference in microbial community diversity in both maternal and infant samples (Fig. 3 a & b).Supplementation also did not influence the community structure as measured by Bray-Curtis (Fig. 3 c).In the maternal samples, the overall relative abundance of Bifidobacterium (including B. breve) remained stable across all timepoints and there was no difference between groups (Fig. 3 d).The relative abundance of total Bifidobacterium in infant samples increased over time but this was not influenced by the probiotic (Fig. 4 d & e).

Comment Principal findings
Our data suggests mother to child transmission in two cases.There were no differences in prespecified secondary outcomes.The placebo group had higher incidence of GDM and more LGA infants, but this may be explained by higher age, multiparity and baseline HbA1c in the control group.

Results in context of what is known
We hypothesised that maternal probiotic supplementation would support transfer of the probiotic to the infant.There is some other albeit limited evidence to suggest that this is possible, and the mechanisms are yet to be determined 23,39 .In our study, there was a confirmed culture of B. breve 702258 from a maternal breastmilk sample for which there was a detection and isolation of the strain from the respective infant.Whilst tempting to consider this as a mode of transmission, the rare occurrence limits any certainty.The overall detection rate of transmission was lower than expected.
In their study of mothers of infants at high risk of eczema, Lahtinen et al were the first to show that an antenatal probiotic (Lactobacillus rhamnosus GG) given to mothers from 36-weeks' gestation can alter the bifidobacterial composition of infants 39 .Although they did not find transfer of the probiotic from mothers to children, they found infants born to mothers in the intervention group had higher colonisation with B. longum.This suggests a potential favourable effect 40 .In contrast, we did not see an impact on infant, or maternal, microbial composition, although our study was not powered to any subtle effects.
. We provided a single probiotic strain whilst some others have supplemented with multiple strains 22,41 .In a study of antenatal probiotic milk (containing L. rhamnosus GG, Lactobacillus acidophilus La-5, and Bifidobacterium animalis subsp.lactis Bb-12), Dotterud et al. found that only L. rhamnosus GG was detected in infant faecal samples at ten-days and three-months of age 42 .This suggests effects on the maternal and infant microbiome may be both strain-specific and time-dependent.The selection of an alternative strain that has a high transfer efficiency and can establish in the maternal microbiome will likely provide greater transfer from mothers to infants.
We found that maternal intake of B. breve 702258 did not alter maternal biomarkers.The dose of the probiotic (10 9 CFU) likely does not explain the lack of findings as a dose of at least 10 7 CFU is considered large enough to confer a benefit and doses from 10 5 -10 14 CFU have been used in pregnancy 43 .Although there are plausible mechanistic pathways, such as production of short-chain fatty acids, several systematic reviews suggest no conclusive benefit of probiotic intake on metabolic and inflammatory markers in healthy populations, including preconception and pregnancy [44][45][46] .A recent meta-analysis of 16 studies of pregnancy probiotic supplementation in relation to GDM specifically however, found positive effects on metabolic status including insulin, fasting glucose, and HOMA-IR 43 .One included study in this review found attenuated increases in LDL cholesterol in women with a new diagnosis of GDM or impaired glucose tolerance with daily probiotic supplementation (Lactobacillus salivarius UCC118) until delivery 47 .
In line with previous research, we did not find any effect on mode of delivery or preterm birth 48 .
This finding further suggests the safety of probiotic supplementation in pregnancy, although we were not powered to see differences in this outcome.Previous research highlights the potential impact of probiotics on the neonatal gut is influenced by a myriad of factors such as maternal gut, mode of delivery, infant feeding, and commensal bacteria 49,50 .Our study was an intention to treat analysis therefore all infants were included with data on the primary outcome, regardless of these factors.Future work could explore the impact of probiotic supplementation on transmission in select groups of mother-child dyads.

Research implications
A major strength of our study is the duration of supplementation 23,51 .This provides crucial insight into the potential safety of longer-term supplementation during pregnancy.Another strength is the use of multiple robust methods to detect the strain, particularly the use of shotgun sequencing, which is lacking in the previous literature 23,50,52 .Since positive detections were obtained during PCR for both randomisation stage and placebo samples, the specificity of PCR could not be relied upon.Incorporation of both culture work and metagenomics enabled a robust analysis, and subsequently only samples with a positive detection by at least two methods were considered when confirming detection.Whilst metagenomic sequencing can be limited by sequencing read depth, and may not detect bacteria present in low abundance, the use of culture and PCR overcomes this.Future work could continue to employ multiple methods of analysis with even greater read depth.

Clinical implications
Whilst we detected the probiotic strain in infant stool, the low rates suggest that maternal probiotic supplementation with this strain may not be a scalable intervention to influence the infant gut.The mechanisms of colonisation of the infant gut through maternal probiotic supplementation remain unclear and requires more research.Most previous studies have included the antenatal period only, unlike our study where we supplemented up to 3 months postpartum 23,41 .Although our study was not powered to identify adverse effects and our population was relatively homogenous, clinicians should note the suggested safety of maternal probiotic supplementation from early pregnancy.Given the potential to positively influence .the health of infants throughout the life-course, similar trials with alternate or multi-strain probiotics, or synbiotics, are warranted.The role of other factors such as parity, diet and lifestyle should be explored 53,54 .

Conclusion
Direct strain transfer from mothers to infants is possible, though at a low level with this strain.
This highlights potential for maternal probiotic supplementation as an intervention to introduce beneficial microbial strains into the infant gut microbiome.
.  Of these, 24 infants had one sample (six at 1-week, 13 at 1-month and five at 3-months).

Table and Figure legend
An additional 58 infants had two samples (12 with both 1-week and 1-month samples, and 42 with both 1-month and 3-month samples).Finally, a further 53 infants had stool samples at all three timepoints.In total, 300 infant stool samples were analysed (74 at 1-week, 120 at 1-month and 106 at 3-months).the intervention and 70 in the control group) for whom we received a stool sample for analysis of the primary outcome at either the 1-week, 1-month, or 3-month timepoint.Of these, 24 infants had one sample (six at 1-week, 13 at 1-month and five at 3-months).An additional 58 infants had two samples (12 with both 1-week and 1-month samples, and 42 with both 1-month and 3-month samples).Finally, a further 53 infants had stool samples at all three timepoints.In total, 300 infant stool samples were analysed (74 at 1-week, 120 at 1-month and 106 at 3-months).    .

3 SLIDINGWINDOW: 5 :
3) wasused to predict open reading frames (ORF) that were then aligned to a combined bifidobacterial genome-based database with a BLASTP v2.6.0 (cut-off E-value of 0.0001) and a MySQL relational database used to assign annotations.Transfer RNA genes were identified employing tRNAscan-SE v1.3.1 and ribosomal RNA genes were detected based on the software package Rnammer v1.2 supported by BLASTN v2.6.0.Strain classification and SNP region phylogenetic reconstructions.Taxonomical classification of strains was based on whole genome comparison to a local database of genomes of 66 Bifidobacterium species using fastANI v1.3.A threshold of 98% average nucleotide identity was used to classify strains.A Phylogenetic tree based on core single nucleotide polymorphisms was SNP generated as follows.Raw paired end reads were trimmed with trimmomatic (v.0.36) with the following options "-phred33 ILLUMINACLIP:TruSeq3-PE.fa:2:30:10 LEADING:3 TRAILING:20 MINLEN:50".Snippy (v.4.4.5) was used to predict SNPs for each strain compared to a species-specific reference genome and to generate an alignment of core SNPs.Gubbins (v.3.0.0)was used to identify and remove putative areas of recombination with the following options "-m 4 -b 4000 --first-tree-builder fasttree".RAxML was used to generate a phylogenetic tree with RAxML as follows: 20 maximum likelihood (ML) trees (parameters: -m ASC_GTRGAMMA --asc-corr=lewis -p 12345 -N 20) and 100 bootstrap trees (parameters: -m ASC_GTRGAMMA --asc-corr=lewis -p 12345 -x 12345 -N 100) were used to generate the best tree (parameters: -m ASC_GTRGAMMA --asc-corr=lewis -p 12345 -f b).Phylogenetic trees were visualised and annotated in iTOL (https://itol.embl.de/).All genomes of isolated Bifidobacterium breve strains were compared to the genome of B. breve 702258 with fastANI.

16 weeksFigure 2 |
Figure 2 | Detection of supplemented B. breve 702258.(a) PCR analysis to detect the strain-specific marker gene of the probiotic strain.Bar charts illustrate the percentage of samples in each trial group that had a positive detection of B. breve 702258 for each sample type.(b) Maximum likelihood phylogenetic tree showing the relatedness of metagenome-assembled genomes (MAGs) recovered from the infant stool to the genome of B. breve 702258.(c) Phylogenetic tree showing the genomic relatedness of cultured B. breve isolates from infant stool to the B. breve 702258 genome.Phylogenetic trees were based on the alignment of core SNP regions of all genomes shown.

Figure 3 |
Figure 3 | Microbiome composition analysis of the maternal & infant stool microbiome.Alpha diversity of the maternal stool (a) & infant stool (b) microbiome as determined by Simpson, Shannon, and Observed species.Beta diversity of both the maternal & infant stool microbiome (c) as determined using Bray-Curtis dissimilarity measure.The relative abundance of overall Bifidobacterium and the top 4 individual Bifidobacterium species in the maternal (d) and infant (e) stool compared between trial groups.
detect the presence of B. breve 702258 in samples, a primer set specific to a distinct CRISPR Primer efficiency was determined to be >95% with an R 2 of 0.9996.Briefly, DNA was normalized to 0.2 ng µl -1 and 3 µl used as template with 0.8 µl of each primer (10 µM), 0.4 µl probe (10 µM), 10 µl Luna Universal probe qPCR master mix (NEB), and 5 µl water.Reactions were carried out in 96-well plates using Lightcycler 480 (Roche) with initial denaturation for 60 s at 95 o C, followed by 45 cycles of denaturation for 15 s at 95 o C and extension for 30 s at 60 o C with signal detection using LC480 FAM filter.Positive signal was regarded as ct value less than negative control of sterile water.
A positive template DNA control (genomic DNA extracted from B. breve 702258) was included for all plate runs.All suspected positive detections of 702258 determined using qPCR were confirmed using regular PCR with product visualization on 1.5% agarose gel.For this PCR, a second set of primers targeting the same CRISPR spacer region were used (ION-F | ATTCGTATACAGGGCGGTTG; ION-R | GCTCGAATGCGGTTGAGTAT).Briefly, 1 µl DNA was mixed with 0.5 µl of each primer, 12.5 µl Accuzyme (Bioline), and 10.5 µl water and cycled with initial denaturation for 2 min at 94 o C, and 30 cycles of 1 min at 94 o C, 25 s at 58 o C, 20 s at 72 o C, followed by final elongation for 1 min at 72 o C.
In the case of significant findings, relevant baseline information between study groups were compared.Results for the primary outcome include relative risk (RR) and 95% or exponential transformation prior to analysis.If normality was not achieved, the appropriate non-parametric test was used.Continuous variables are presented as mean (standard deviation) or median and interquartile range (25th, 75th centile) for non-normally distributed data.Categorical variables are presented as number and frequency (%).Primary and secondary outcomes were compared between groups.Comparison statistics were generated through independent sample t-tests using parametric or normally distributed log-transformed data.Mann-Whitney U test was used for non-parametric data.Chi-square (χ2) tests were used for categorical variables, and Fishers' exact tests were used when assumptions for chi-square were violated.

Table 1 :
Maternal and infant characteristics in the MicrobeMom Randomised

Table 2 :
Primary and secondary outcomes after the intervention in the MicrobeMom Study