in

A pioneer calf foetus microbiome

[adace-ad id="91168"]

Experimental design and sample collection

This study was carried out in accordance with the provisions in the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (7th edition, 2004) and all protocols were approved by the Animal Ethics Committee of La Trobe University. Twelve Angus × Friesian cattle foetuses at 5, 6 and 7 months gestation (n = 4 per age) were collected from Radford Warragul Abattoir, Victoria, Australia. Approximately 35–45 min after cows were slaughtered by abattoir staff, the intact uterus (containing the placenta and foetus) was removed. All sampling was conducted at the abattoir using sterile equipment and procedures. The outside surface of the amniotic sac was rinsed three times with sterilised phosphate-buffered saline (PBS; pH 7.0) to remove excess blood. The amnion was cut using sterile scalpels and amniotic fluid was sampled. The amniotic fluid was suctioned using sterile 50-mL syringes with tubing and the amniotic fluid was transferred immediately into 50-mL tubes. Then, the amniotic sac was opened further, the umbilical cord was cut, and the foetus was removed. The abdomen of the foetus was opened using sterilised equipment and the rostral and caudal ends of each GIT compartment were tied with sterile surgical thread to avoid mixing of the contents. The compartments were then separated between the ties. Each compartment was longitudinally incised along the dorsal line. Tissue samples (~ 2 cm2) of the rumen were taken from the dorsal area and caecal tissue samples were taken from the region 5 cm after the ileocaecal valve. Meconium pellets (~ 100 g) were taken by severing the rectum 5 cm from the anus. The fluid, tissue and meconium samples were collected into sterile 15-mL or 50-mL polypropylene centrifuge tubes. All samples were immediately placed into dry ice for transport. All samples were processed within 6 h of collection to extract gDNA.

DNA extraction

Genomic DNA was extracted from 250 mg of ruminal tissue, ruminal fluid, caecal tissue, caecal fluid and meconium. An 8-mL aliquot of amniotic fluid was centrifuged (11,000g, 5 min) to produce sufficient material in the pellet for extraction. DNA was extracted using an Isolate II Genomic DNA kit following the manufacturer’s instructions. Final DNA concentrations and purity were estimated using a P330-Class NanoPhotometer (Implen, München, Germany). All samples were stored at − 80 °C for later analysis.

16S rRNA library preparation and sequencing

Libraries were prepared for sequencing on an Illumina MiSeq following the protocol ‘16S Metagenomic Sequencing Library Preparation’ (Part # 15044223 Rev. B; Illumina, San Diego, CA, USA). The locus-specific primers were the universal 16S rRNA primer pairs S-D-Bact-0341-b-S-17 (5′-CCTACGGGNGGCWGCAG-3′) and S-D-Bact-0785-a-A-21 (5′-GACTACHVGGGTATCTAATCC-3′), Archaea349F (5′-GYGCASCAGKCGMGAAW-3′), and Archaea806R (5′-GGACTACVSGGGTATCTAAT-3′), which target the V3–V4 region of the bacterial and archaeal 16S rRNA genes, respectively. Primers had forward (5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG-3′) and reverse (5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG-3′) Illumina overhang adaptors merged to the 5′ ends.

PCR was performed in 25-µL reactions using 5 µL of each forward and reverse primer (10 µM), 12.5 µL 2 × KAPA HiFi HotStart ReadyMix (Kapa Biosystems, Boston, MA, USA) and 2.5 µL of genomic DNA template (5 µL/ng). PCR cycle settings for the amplification of the bacterial and archaeal V3–V4 region were as follows: denaturation at 95 °C for 3 min, followed by 28 (bacterial) or 30 (archaeal) cycles of 30 s at 95 °C, 30 s at 55 °C and 30 s at 72 °C, followed by an extension step at 72 °C for 5 min. To normalise libraries prior to pooling, the DNA content of PCR reactions was quantified using an Agilent D1000 ScreenTape System (Agilent Technologies, CA, USA). Samples were adjusted to the same molarity (4 nM), pooled, and paired-end sequenced (2 × 300 bp) on an Illumina MiSeq platform. The MiSeq run was performed at La Trobe University Genomics Platform (Melbourne, Australia).

Analysis of sequence data

Raw, de-multiplexed, fastq files were re-barcoded, joined and quality filtered using the UPARSE clustering pipeline (USEARCH version 9.2.64; https://drive5.com/uparse)46. Paired-end reads were merged such that alignments with > 20 bp difference (i.e. approximately more than 10–14% mismatched) were discarded, and merged reads less than 300 bp in length were discarded. Reads that could not be assembled were discarded. Merged reads were quality filtered by discarding reads with total expected errors > 1.0. ESVs were generated with the “unoise3” command47. Taxonomic assignments were performed using the UTAX algorithm. Reference databases were created using the RDP_trainset_15 dataset, available from the UTAX downloads page (https://drive5.com/usearch/manual/utax_downloads.html). A minimum percentage identity of 90% was required for an ESV to be considered a database match hit. ESVs identified as chloroplasts and mitochondrial DNA were removed from the data. After filtering, the average read number (± SD) for each git compartment was: 4342 ± 1594 reads for the AM , 17,276 ± 17,376 reads for the CF, 6182 ± 3918 reads for the CT, 5732 ± 3262 reads for the Mec, 5630 ± 2234 reads for the RF and 5847 ± 4052 reads for the RT. The rarefying threshold of 1000 reads was chosen to maximise the amount of reads included in the analysis whilst minimize the number of samples excluded from the analysis. A total of three bacterial samples resulted in reads below the rarefication threshold (1000 reads) and were excluded from downstream alpha- and beta-diversity analyses. The samples were: Month_5_Mec-2, Month_6_CF-3 and Month_6_Mec-1. DNA extraction or library preparation was unsuccessful for the following samples: n = 0 (5 months cecum fluid), n = 1 (6 and 7 months rumen tissue, 7 months amniotic fluid), n = 2 (5 months rumen tissue, 6 months amniotic fluid), n = 3 (5 and 6 months cecum tissue, 5 months meconium), the remaining GIT compartments and months were n = 4. Raw fastq files for this project and metadata have been deposited with the NCBI SRA database and can be accessed using Bioproject ID: PRJNA421384 or SRA study ID: SRP126299.

Bacterial culture from ruminal fluid samples and identification of bacterial isolates

A 50-mL sample of ruminal fluid was taken from each foetus and maintained under anaerobic conditions (Oxoid AnaeroJar with an AnaeroGen™). A 1-mL aliquot of the ruminal fluid was transferred to anaerobic solid medium and cultured at 37 °C for 48–72 h. The anaerobic solid medium had the following composition (per litre of distilled water): 15 g agar (Oxoid), 10 g peptone (Oxoid), 10 g yeast extract, 8.8 g Oxoid Lab-Lemco beef extract powder, 10 g proteose peptone (Oxoid), 12 g dextrose, 10 g KH2PO4, 12 g NaCl, 20 g soluble starch, 1.2 g l-cysteine hydrochloride and 0.3 g sodium thioglycollate with a pH (at 25 °C) of 7.3 ± 0.1. Colonies were isolated and subcultured 5 times onto new agar media plates, except for the control plates (n = 3) which showed no microbial growth. Colonies were subcultured on fresh media and DNA extracted. The extracts for 5, 6 and 7 months were combined prior to next-generation sequencing of the 16S rRNA genes to characterise the taxonomic structure.

Quantitative PCR

Quantitative PCR (qPCR) was used to enumerate total bacterial and archaeal DNA copy number in each sample type (GIT component and amniotic fluid) as an indicator of abundance. The primer pairs bacF (5′-CCATTGTAGCACGTGTGTAGCC-3′) and bacR (5′-CGGCAACGAGCGCAACCC-3′) were used to amplify bacterial 16S rRNA, and Archaea364F (5′-CCTACGGGRBGCAGCAGG-3′) and Archaea1386R (5′-GCGGTGTGTGCAAGGAGC-3′) were used to amplify archaeal 16S rRNA. PCR reactions were run in triplicate on a CFX Connect Real-Time PCR Detection System (Bio-Rad, CA, USA). The total volume of each reaction mix was 20 μL, comprising 10 μL of SensiFAST SYBR Green Master Mix (Bioline), 0.4 μL of each forward and reverse primer (10 µM), sterile DNA-free water, and 7 ng of DNA. Triplicate control samples (no-DNA templates) were included to verify that no contaminating nucleic acid was introduced into the master mix or into samples. Positive controls contained gDNA extracted from laboratory cultured bacteria (E. coli strain DH5α) and archaea (Methanobrevibacter smithii), respectively. Thermocycling conditions were as follows: initial denaturation for 3 min at 94 °C, followed by 40 cycles of 10 s at 94 °C, 30 s at 60 °C. This was followed by a dissociation protocol (increasing 1 °C every 30 s from 60 °C to 98 °C).

A standard curve was constructed using serial tenfold dilutions from 10−1 to 10−11 of DNA from the bacterium E. coli strain DH5α (Stratagene, CA, USA) or the archaeon M. smithii. Real-time PCR efficiency ranged from 97 to 102%. Copy numbers for each standard curve were calculated based on the following equation: (NA × A × 10−9)/(660 × n), where NA is the Avogadro constant (6.02 × 1023 mol−1), A is the molecular weight of DNA molecules (ng/mol) and n is the length of amplicon (bp).

Control procedures for sample contamination

Potential airborne bacteria were passively sampled to determine if there was a detectable contribution of environmental bacteria contaminating the foetal samples. Sampling tubes containing aerobic or anaerobic medium were opened and exposed to the dissection area in the abattoir for the duration of sampling from each foetus. The exposed media were incubated at 37 °C and samples taken at 72 h and at 2, 3 and 4 weeks. DNA was extracted using an Isolate II Genomic DNA kit. Final DNA concentrations and purity were estimated using a P330-Class NanoPhotometer.

The dissection table and external surfaces of the amniotic sac and intestinal compartments were swabbed to determine if there was a detectable contribution of bacteria contaminating the foetal samples. For each foetus, 9 swabs were taken at six locations using sterile Fisherbrand synthetic-tipped applicator swabs (Thermo Fisher Scientific, MA, USA). The surface of the dissection table (first location) was washed with 70% ethanol and then swabbed prior to dissecting each foetus. The external surface of the amniotic sac (second) was rinsed three times with sterilised PBS to remove excess blood and then swabbed prior to opening the sac. The skin of the foetal abdomen (third) was rinsed three times with sterilised PBS to remove amniotic fluid and then swabbed prior to opening. The external (mesenteric) surfaces of the rumen (fourth), caecum (fifth) and rectum (sixth location) were separately swabbed prior to opening. The 9 swabs for each foetus were tested for the presence of microorganisms, using three swabs for each of the three methods: qPCR, anaerobic culture and aerobic culture. DNA was extracted from three of the swabs using an Isolate II Genomic DNA kit. Final DNA concentrations and purity were estimated using a P330-Class NanoPhotometer. Quantitative PCR was used to more accurately quantify the presence of DNA.

Anaerobic and aerobic liquid mediums were inoculated from three swabs each. Three swabs were placed into a 2.5-L anaerobic Oxoid AnaeroJar with an AnaeroGen sachet (Thermo Fisher Scientific) and three swabs were placed into aerobic Oxoid Nutrient Broth (Thermo Fisher Scientific). All mediums were incubated at 37 °C for 48–72 h. Monitoring for growth during storage at 4 °C was continued for up to one month. The anaerobic liquid medium had the following composition (per litre of distilled water): 10 g peptone (Oxoid), 10 g yeast extract, 8.8 g Oxoid Lab-Lemco beef extract powder, 10 g proteose peptone (Oxoid), 12 g dextrose, 10 g KH2PO4, 12 g NaCl, 20 g soluble starch, 1.2 g L-cysteine hydrochloride and 0.3 g sodium thioglycollate with a pH (at 25 °C) of 7.3 ± 0.1. The aerobic nutrient broth had the following composition (per litre of distilled water): 10 g Lab-Lemco powder, 10 g peptone and 5 g NaCl, with a pH (at 25 °C) of 7.5 ± 0.2.

Controls for kit contamination

Two blank DNA spin columns from two different Isolate II Genomic DNA kits were used as no-template controls to determine if the kits were a source of contamination during DNA extraction and library preparation of samples. The no-template controls were tested using qPCR and Illumina MiSeq next-generation sequencing.

Statistical analysis of the total community

The adequacy of the sampling effort to capture the microbial community richness was examined by generating species rarefaction curves and species accumulation plots using the ‘rarecurve’ and ‘specaccum’ functions in the R library vegan (v. 2.3-4)48 using R 3.2.049. Alpha- and beta-diversity analyses were performed on archaeal and bacterial OTU-matrices rarefied to depths of 2000 and 1000 reads, respectively, using the ‘phyloseq’ (v. 1.16-2)50 and ‘vegan’ (v. 2.4-0)48,51 packages in the R programming language (v. 3.3.1)49. Normality and variance homogeneity of the data were tested using the ‘shapiro.test’ and ‘bartlett.test’ functions. As normality and homogeneity of variance assumptions were not met, Kruskal–Wallis tests were carried out using the ‘kruskal.test’ function with Dunn’s test of multiple comparisons used for post hoc testing. The NMDS ordinations were used to visualise differences between communities from different sample types (GIT components and amniotic fluid). Dissimilarity matrices were generated using the weighted and unweighted UniFrac metrics52. Analysis of similarity (ANOSIM) procedures were implemented to test for significant differences in the mean group centroids53. Differential abundance testing of ESVs was performed using the DESeq2 extension available within the ‘phyloseq’ package50,54. Tests were performed by applying model-fitting normalisation to unrarefied ESV tables as recommended by McMurdie and Holmes for each taxonomic rank50. For differential abundance tests only ESVs with high (> 90%) confidence values for phylum-level taxonomic assignments were considered. All low confidence taxonomic assignment we re-classified as ‘unknown’. Differences were considered significant if Benjamin-Hochberg adjusted p < 0.05. The confidence values for taxonomic assignments of differentially abundant ESVs are listed in supplementary tables.

Statistical analysis of bacterial culture and identification from ruminal fluid samples

gDNA from cultures of ruminal fluid were sequenced alongside community samples on an Illumina MiSeq. Filtering and quality control of isolate sequence data was identical to the process used for community samples. ESVs that contributed < 0.07% to the set of reads associated with cultured isolates were excluded from the final analysis.

Statistical analysis of the quantitative PCR

Means were compared using a two-way factorial ANOVA with post hoc multiple comparisons using Tukey’s honest significant difference test. Due to the unbalanced design, type II sums of squares were calculated using the ‘Anova’ function from the R package ‘car’55.


Source: Ecology - nature.com

A controllable membrane to pull carbon dioxide out of exhaust streams

More than a meal