A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

Lisa Ekman; Elisabeth Bagge; Ann Nyman; Karin Persson Waller; Märit Pringle; Bo Segerman

doi:10.1371/journal.pone.0242880

. 2020 Dec 2;15(12):e0242880. doi: 10.1371/journal.pone.0242880

A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

Lisa Ekman ^1,^2,^*, Elisabeth Bagge ¹, Ann Nyman ^2,³, Karin Persson Waller ^1,², Märit Pringle ¹, Bo Segerman ^4,⁵

Editor: Peter Gyarmati⁶

PMCID: PMC7710049 PMID: 33264351

Abstract

Udder cleft dermatitis (UCD) is a skin condition affecting the fore udder attachment of dairy cows. UCD may be defined as mild (eczematous skin changes) or severe (open wounds, large skin changes). Our aims were to compare the microbiota of mild and severe UCD lesions with the microbiota of healthy skin from the fore udder attachment of control cows, and to investigate whether mastitis-causing pathogens are present in UCD lesions. Samples were obtained from cows in six dairy herds. In total, 36 UCD samples categorized as mild (n = 17) or severe (n = 19) and 13 control samples were sequenced using a shotgun metagenomic approach and the reads were taxonomically classified based on their k-mer content. The Wilcoxon rank sum test was used to compare the abundance of different taxa between different sample types, as well as to compare the bacterial diversity between samples. A high proportion of bacteria was seen in all samples. Control samples had a higher proportion of archaeal reads, whereas most samples had low proportions of fungi, protozoa and viruses. The bacterial microbiota differed between controls and mild and severe UCD samples in both composition and diversity. Subgroups of UCD samples were visible, characterized by increased proportion of one or a few bacterial genera or species, e.g. Corynebacterium, Staphylococcus, Brevibacterium luteolum, Trueperella pyogenes and Fusobacterium necrophorum. Bifidobacterium spp. were more common in controls compared to UCD samples. The bacterial diversity was higher in controls compared to UCD samples. Bacteria commonly associated with mastitis were uncommon. In conclusion, a dysbiosis of the microbiota of mild and severe UCD samples was seen, characterized by decreased diversity and an increased proportion of certain bacteria. There was no evidence of a specific pathogen causing UCD or that UCD lesions are important reservoirs for mastitis-causing bacteria.

Introduction

Udder cleft dermatitis (UCD) is a skin condition that affects the anterior parts of the udder in dairy cows. It has been reported in the UK [1], the USA [2], Sweden [3, 4], Denmark [5], the Netherlands [6] and Norway [7]. The prevalence varies between studies, but in high-prevalence herds, up to 60% of cows may be affected [8]. The UCD lesions vary in appearance and may be classified as mild or severe, based on whether or not skin integrity is breached [4, 8, 9]. The etiology and pathogenesis of the lesions are still largely unknown. Recent studies indicate a multifactorial origin of UCD, associated with both cow- and herd-related risk factors, such as parity, breed, udder conformation, high herd-level production and type of floor in cubicles [4, 6, 8, 9]. In addition, several infectious agents have been implicated in the development of UCD, such as mange mites [10], Treponema spp. [11] and Bovine herpesvirus 4 [12], but the true role of these agents in the etiology of UCD has not been proven. Moreover, culturing of swab samples from UCD lesions has revealed a variety of aerobic and anaerobic bacteria, as well as fungi [2, 3, 13], indicating that the lesions may be a reservoir for pathogens, potentially increasing the risk of infectious diseases such as mastitis. In line with this, a few studies have found associations between UCD, particularly severe cases, and an increased risk of clinical mastitis [4, 14], but it is not known whether mastitis-causing pathogens are a common finding in UCD microbiota. Previous microbiological investigations of UCD lesions have mainly been performed through culturing [3, 13], microscopy [13] or Treponema-specific PCR assays [11, 15]. In recent decades, the use of culture-independent methods to identify the microorganisms present in a sample or an environment has become increasingly common [16]. So far, few studies have been performed on samples from UCD lesions, although a recent study used 16S rRNA-amplicon sequencing to investigate the bacterial microbiota of UCD lesions and compared it with that of healthy skin [17]. They found that certain bacterial genera were more common in samples from UCD lesions, such as Fusobacterium, Helcococcus, Anaerococcus, Trueperella and Porphyromonas, compared to samples from healthy skin. In the 16S rRNA-amplicon sequencing method, specific regions the rRNA gene is PCR amplified and sequenced of from bacteria, to assess the microbiota [18]. Shotgun metagenomic sequencing is an alternative method to analyze the microbiota in which total DNA is sequenced using only a limited number of amplification cycles and this method can detect all types of microbes with improved resolution down to the species and strain level [19, 20]. This method has been used in studies on human gut [21] and bovine ruminal [22] microbiota, as well as in studies on human skin microbiota, for example, in patients with atopic dermatitis [23], and the microbiota of human chronic wounds, such as pressure wounds and venous leg ulcers [24]. We believe that shotgun metagenomic sequencing has the potential to yield additional information on the microbiota of UCD lesions and increase the understanding of the development and clinical course of UCD and give indications how to treat these lesions.

Thus, the main objective of this study was to compare the microbiota of recently developed mild and severe UCD lesions, and healthy skin at the same body site using shotgun metagenomic sequencing to investigate whether specific microbes are associated with UCD lesions. We also wanted to investigate whether common mastitis-causing pathogens are present in UCD lesions, which would indicate that UCD may be a reservoir for udder infections.

Material and methods

Study design and participating cows

Seven Swedish dairy herds with free-stall housing and milking parlors were enrolled in the study. Inclusion criteria were a previous UCD prevalence of 20–40% [8] and that they were located within 200 km of Uppsala, Sweden. The mean herd size was 125 cows (range 87–168 cows), mean herd level production was 10,204 kg milk/cow and year (range 7,680–11,534 kg) and the most common breeds were Swedish Red and Swedish Holstein. Herd visits were performed regularly from April 2018 to April 2019 as part of a longitudinal study of UCD (nine visits per herd at six-week intervals). This study design made it possible to identify and sample cows with recently developed UCD lesions. Ethical approval for this study was issued by a regional Swedish Ethics committee (appointed by the Swedish Board of Agriculture). The herd visits were conducted during one milking and all milking cows were scored for UCD (no, mild or severe). All scoring and sampling were performed by a single researcher. Mild UCD was defined as erythema and small papules or pustules, or small crusts, and severe UCD was defined as a breach of skin integrity, often with large crusts and exudative or bleeding wounds (Fig 1). Cows for sampling were chosen based on their UCD status. The criteria for sampling was a cow with a previous status of no UCD that received a score of mild or severe UCD, as well as a cow with a previous status of mild UCD that received a score of severe UCD. For every cow with a sampled UCD lesion, the aim was to sample the skin from the same body site (fore udder attachment and between the front quarters) of a control cow with no UCD. As far as possible, control cows of the same breed and parity as the UCD cows were selected. At the final herd visit, samples were also obtained from cows with previously registered UCD lesions in order to achieve a total of approximately 10 samples per category (no UCD, mild UCD and severe UCD) from each herd. Thus, cows that had been previously sampled could be sampled again at the final herd visit.

Fig 1 — Samples were taken from (A) healthy control skin at the fore udder attachment, (B) mild and (C) severe udder cleft dermatitis.

Sampling procedure

Sampling was performed in the milking parlor, during milking or just after the milking unit had been removed. Clean disposable gloves were used at all samplings and were changed between cows. If the area for sampling (Fig 1) was visibly dirty, it was cleaned with paper (dry or soaked in water) or sterile gauze compresses (dry or soaked in saline 0.9%, Fresenius Kabi, Bad Homborg, Germany). Severe UCD lesions were always cleaned with sterile gauze compresses soaked in saline to remove loose crusts, necrotic tissue and pus. Finally, the area for sampling was wiped with one dry sterile gauze compress just before sampling. This step was also performed before sampling mild UCD lesions and healthy skin. Each sample was taken using a 50 cm² sponge moistened with saline contained in a sterile Minigrip bag (TS/15-B:NACL, Technical Service Consultants Ltd, Lancashire, UK) according to the manufacturer’s instructions. The area for sampling was wiped using approximately 20 strokes, covering the entire lesion and adjacent skin (approximately 1–5 cm of skin around the lesion, depending on lesion size) or the ventral mid-area of the fore udder attachment and the area between the front quarters for healthy skin samples (Fig 1). Samples were uniquely labeled and immediately put on ice. They were kept cold (at 4°C) during transportation and arrived at the laboratory (Uppsala, Sweden) within 24 hours. A total of 184 samples were taken from cows with no (n = 77), mild (n = 46) or severe (n = 61) UCD. As one herd had very few cases of UCD, the samples from this herd (n = 5) were excluded, leaving 179 samples from 6 herds for further analyses.

Sampling analyses

The samples were processed within a few hours of arrival at the laboratory. First, 50 ml of sterile 0.9% saline (SVA, Uppsala, Sweden) was poured into the Minigrip bag. In order to dislodge microorganisms from the sponge into the fluid, the bag was treated in a stomacher (230 rpm; Stomacher^® 400 Circulator, Seward, West Sussex, UK) for two minutes. The fluid was then poured into a sterile 50 ml plastic tube (Sarstedt, Nümbrecht, Germany) and the tube was centrifuged for 15 minutes at 2,000 g. Most of the supernatant was removed, leaving around 1–2 cm of fluid at the bottom of the tube and the pellet was dissolved in the remaining fluid (approximately 2–5 ml). The solution was then transferred into a 2 ml sterile plastic microtube (Sarstedt, Nümbrecht, Germany) and the samples were kept frozen at -23°C for 1–8 weeks before DNA extraction.

DNA extraction

The microtube samples were thawed at room temperature for 20–40 minutes, briefly vortexed and then centrifuged for two minutes at 2,000 g. The supernatant was removed and the pellet was used for DNA extraction using the DNeasy Powerlyzer Powersoil Kit (12855–100, Qiagen AB, Sollentuna, Sweden) according to the manufacturer’s instructions and with the following additions: solution C1 and solution C6 were heated to 65°C before use to avoid precipitation and the samples were heated to 100°C before the bead-beating step to improve the lysis of cellular structures. The bead-beating step was performed using a FastPrep -24™ homogenizer (MP Biomedicals, Irvine, CA, USA), with the settings 6.5m/s and MP24x2, for 2x60 seconds. After the extraction, the DNA concentration of each sample was measured by fluorometry using Qubit™ 1X dsDNA HS Assay Kit (Q33230, Thermo Fisher Scientific, Waltham, MA, USA) and varied between 0 and 110 ng/μl. The extracted DNA was stored at -23°C until sequenced. From each herd and category (no, mild and severe UCD), 5–6 samples with sufficient DNA concentration were chosen for further analyses–a total of 96 samples. At the sequencing facility (SNP&SEQ Technology Platform, Uppsala, Sweden), the DNA concentration was re-measured with Quant-iT™ (Thermo Fisher Scientific) and DNA fragmentation was analyzed with an Agilent Fragment Analyzer (DNF-467-kit, Santa Clara, CA, USA). Some samples had a high degree of DNA fragmentation. We therefore chose 49 samples with acceptable quality parameters for sequencing, 13 from healthy skin (controls), 17 from mild UCD lesions and 19 from severe lesions.

DNA sequencing

Sequencing libraries were prepared from 10 ng of DNA using the SMARTer ThrupPLEX DNA-Seq kit (R400676, Takara-Clontech, Saint-Germain-en-Laye, France) according to the manufacturer’s preparation guide #080818. Briefly, the DNA was fragmented using a Covaris E220 system (Covaris Inc, Woburn, MA, USA), aiming at 400 bp fragments. The ends of the fragments were end-repaired and stem-loop adapters were ligated to the 5’ ends of the fragments. The 3’ end of the stem loop was subsequently extended to close the nick. Finally, the fragments were amplified and unique index sequences were introduced using seven cycles of PCR followed by purification using AMPure XP beads (Beckman Coulter Inc., Indianapolis, IN, USA). The quality of the library was evaluated using the Agilent Fragment Analyzer system (DNF-910-kit). The adapter-ligated fragments were quantified by qPCR using the Library Quantification Kit for Illumina (KAPA Biosystems/Roche, Wilmington, MA, USA) on a CFX384 Touch instrument (BioRad, Hercules, CA, USA) prior to cluster generation and sequencing. A 400 pM pool of the sequencing libraries in an equimolar ratio was subjected to cluster generation and paired-end sequencing with a 150bp read length in a SP flowcell and the NovaSeq6000 system (Illumina Inc., San Diego, CA, USA), using the v1 chemistry according to the manufacturer’s protocols. Base calling was performed on the instrument by RTA 3.3.4 and the resulting.bcl files were demultiplexed and converted to fastq format with tools provided by Illumina Inc., allowing for one mismatch in the index sequence. Additional statistics on sequence quality were compiled with an in-house script from the fastq files, RTA and CASAVA output files. Sequencing was performed by the SNP&SEQ Technology Platform (Uppsala, Sweden). The raw sequence data has been submitted to the Sequence Read Archive (SRA) and is accessible via the bioproject PRJNA636289. The SRA accessions are listed in S1 Table.

Bioinformatic analyses

The fastq files were first trimmed using Trimmomatic [25]. The parameters for Trimmomatic were "SE -threads 6 ILLUMINACLIP:adaptes.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36 X.fastq.gz X.trimmed.fastq.gz". To remove contaminating cow sequences, the fastq files were then mapped to the Bos taurus genome (ARS-UCD1.2) with Bowtie2 [26] using standard settings. The mapped and unmapped reads were separated using Samtools [27]. Only paired reads where both were unmapped to Bos taurus were kept. A Kraken2 database was built (Sep 2020) with Archaea, Bacteria, fungi, protozoa, viral and UniVec Core sequences according to the instructions in the manual, and used with Kraken2 [28]. The parameters for Kraken2 classifications were "—db krakendb—threads 10—paired X_R1.fastq X_R2.fastq—report X.krakenreport.txt". The Kraken results were then run through Bracken [29], to estimate Species, Genera and Phylum level data. The parameters for bracken-build were "-d krakendb -t 10 -k 35 -l 150" and for Bracken "bracken -d krakendb -i X.krakenreport.txt -o X.bracken.txt -r 150 -l (S or G or P)". The results were visualized using Pavian, a web application for exploring metagenomics classification results [30]. Some of the severe samples showed pronounced elevated levels of the intracellular parasite Babesia, which infects red blood cells. There was also a correlation between the number of Babesia reads and the number of reads mapped to the cow genome in the same sample. Given the association of Babesia with red blood cells and the correlation to cow DNA, the Babesia reads were deemed as contamination due to blood in the sample and were excluded from the analysis.

Statistical analyses

Custom Perl scripts were created to merge Kraken report files into a single table with clade counts for each sample. Counts were expressed as a percentage of all classified reads identified as Bacteria, Archaea, Eukarya (i.e. fungi or protozoa) or virus, and were analyzed descriptively and compared between groups using the Wilcoxon rank sum test. A data dimensionality reduction with principal component analysis (PCA) was performed on the bacterial phylum, genus and species level. Bacterial phyla, genera and species that represented at least 10% in at least one sample were analyzed for differences between control samples and mild and severe UCD samples, respectively, using Wilcoxon rank sum tests and Bonferroni correction to adjust for multiple comparisons. In addition, Fisher’s exact test was used to investigate the distribution of herd, breed and parity between the three groups. The alpha diversity of bacterial species and genera was investigated by calculating Shannon diversity indexes for all samples using $H' = - \sum_{i = 1}^{R} p_{i} \ln p_{i}$ , where p = proportional abundance of each taxa. The diversity was compared between sample types using Wilcoxon rank sum tests. Bacterial species considered to be common mastitis-causing pathogens in Sweden [31] (listed in S2 Table) that represented more than 1% of the classified reads in at least one sample were described and differences in abundance were compared between control samples and mild and severe UCD samples using the Wilcoxon rank sum test. Archaeal phyla and genera were investigated descriptively and by PCA. Fungal reads were investigated descriptively. Two cows were sampled at two different time points. One of these cows was sampled with a mild lesion, and then sampled again when the lesion became severe. The other cow was sampled once when it had a recently developed severe lesion, and then again at the last sampling, three months later, when the lesion was still severe. These samples were investigated descriptively as they could yield information on how the microbiota of UCD lesions can change over time. Statistical calculations were performed using Stata (release 15.1; StataCorp LLC, College Station, TX, USA) and PCA and heatmaps were generated using the TM4 Multiple experiment Viewer (MEV), version 4.8.

Results

An overview of the sequenced samples including cow information, UCD category, quality control results, numbers of sequenced reads and the proportion of reads that mapped to the bovine genome is shown in S1 Table and detailed results from the classification by Kraken2 and Bracken in S2 Table. There were no associations between herd, breed or parity and sample type (P = 0.96, P = 0.42 and P = 0.23, respectively).

Overall microbial abundance

Reads that could be classified to the domains of Bacteria, Archaea and Eukarya and to viruses were used for an analysis of the microbiota composition. Eukarya was further divided into fungal and protozoan groups. All samples were strongly dominated by reads from the Bacteria domain, except for one sample (M6), which had almost 40% fungal reads (Fig 2A). However, apart from this deviating sample, most samples had low (less than 1%) proportions of fungi. The proportion of Archaea classified reads was lower in samples from mild and severe UCD (means 2.8 (SD 1.5) and 0.5% (SD 0.4), respectively) compared to control samples (mean 4.1% (SD 1.0); Fig 2B). The viral reads and the protozoan reads, after filtering out reads from the red blood cell parasite Babesia, each constituted less than 1% of the reads within all samples (Fig 2B) and there was no clear pattern of differences between UCD lesions and control samples. For these reasons, the protozoan and viral reads were not investigated further.

Fig 2 — Distribution of reads classified as Bacteria, Archaea, fungi, protozoa and virus within each sample (A) and sample type (B and C) based on samples from mild (M, n = 17) and severe (S, n = 19) udder cleft dermatitis (UCD) and samples from skin at the fore udder attachment from healthy controls (C, n = 13). The P-values from comparing the abundance between sample types are denoted in the margins of B and C by *** if P≤0.001, ** if P≤0.01 and * if P≤0.05. The exact P-values were for Bacteria: C/M *P =* 0.01, C/S P<0.0001, M/S *P =* 0.02, Archaea: C/M *P =* 0.0007, C/S P<0.0001, M/S *P =* 0.004, fungi: C/M *P =* 0.05, C/S *P =* 0.0007, protozoa: C/M *P =* 0.04, and viruses: C/M *P =* 0.2, C/S *P =* 0.002.

Bacteria

The unsupervised data dimensionality reduction with PCA on the phylum, genus and species level revealed different subgroups of UCD samples (Fig 3). The major driving taxa affecting the PC axes are shown in S1 Fig. On the phylum level (Fig 3A), the first PCA axis (PC1) separated a large subgroup of both mild and severe UCD samples. Both the second (PC2) and third (PC3) PCA axis separated other largely non-overlapping smaller subgroups of UCD samples, one only including severe (PC2) and one mainly including mild (PC3) UCD samples. The third PCA axis also separated the majority of the control samples, together with a few of the mild UCD samples. Most, but not all, UCD samples were separated from the control samples on at least one of the three first PCA axes.

Fig 3 — Data dimensionality reduction with principal component analysis (PCA) performed on the bacterial phylum (A), genus (B) and species level (C) for 49 samples from mild (n = 17) and severe (n = 19) udder cleft dermatitis and samples from skin at the fore udder attachment from healthy controls (n = 13). The (%) given for each PCA axis indicates the variation explained by that specific axis. The fourth PCA axis on genus level separated a subgroup of control samples and is presented in S2 Fig.

On the genus level, the first PCA axis separated a large group of both mild and severe UCD samples including one control sample (C12), while the other PCA axes separated smaller subgroups, mainly including mild or severe UCD samples (Fig 3B). The healthy control samples were more strongly clustered than the UCD samples, except for C12 at PC1, and C9 and C13 at PC 4 and 5 (Fig 3B and S2 Fig). On the species level, the control samples were tightly clustered close to the origin, while the UCD samples were separated into several different subgroups along the PCA axes (Fig 3C). Overall, the PCA analysis suggested that most of the UCD samples were distinctly separable from the control samples, although there appeared to be more than one subgroup of UCD samples. We found no indication that the clustering of samples were related to herd, breed or parity of the animal, or that the extraction month had an association with the results (S3 Fig).

Bacterial abundance

The control samples and the mild UCD samples were dominated by three phyla, Actinobacteria, Firmicutes and Proteobacteria (Fig 4A). Although there was no overall difference between sample types, a substantial number of both mild and severe UCD samples showed a markedly higher proportion (around 80% or higher) of Actinobacteria compared to control samples. The group of samples with a high proportion of Actinobacteria corresponded to the samples separated by the first PCA axis on the phylum level (Fig 3A). A subgroup of the severe UCD samples had a markedly high proportion of the phyla Fusobacteria and Bacteroidetes compared to other samples (Fig 4A). This subgroup largely corresponded to the samples separated by the second PCA axis on the phylum level (Fig 3A). In the most pronounced cases, around one third of the bacterial reads belonged to Fusobacteria or Bacteroidetes. Another group of samples, mainly from mild UCD, was reflected in the third PCA axis on the phylum level and had a relatively high proportion of Firmicutes compared to other UCD samples. The third PCA axis also separated the majority of the control samples, together with a few of the samples from mild UCD and in general, a higher proportion of Proteobacteria was seen in control samples compared to mild and severe UCD samples (Table 1).

Table 1. Comparison of UCD samples and healthy skin.

Rank^a	Taxa	Control	Mild	P (M)	Severe	P (S)
P	Bacteroidetes	2.07	0.72	0.001	2.26	0.54
G	Bacteroides	0.22	0.07	0.04	0.50	0.60
G	Porphyromonas	0.04	0.17	0.07	0.75	0.01
S	Porphyromonas asaccharolytica	0.02	0.01	0.06	0.54	0.006
P	Fusobacteria	0.16	0.07	0.03	0.86	0.009
G	Fusobacterium	0.10	0.04	0.03	0.83	0.005
S	Fusobacterium necrophorum	0.03	0.01	0.03	0.78	0.002
P	Proteobacteria	17.27	8.23	0.002	5.03	<0.0001
P	Actinobacteria	50.55	60.78	0.07	77.04	0.06
G	Brachybacterium	1.43	0.58	0.23	0.17	0.0007
G	Brevibacterium	1.18	1.19	0.90	0.62	0.45
S	Brevibacterium luteolum	0.36	0.25	0.80	0.35	0.79
S	Brevibacterium sp. W0024	0.02	0.02	0.74	0.02	0.25
G	Bifidobacterium	5.52	1.51	0.02	0.32	0.0001
S	Bifidobacterium angulatum	1.01	0.11	0.03	0.05	0.002
G	Corynebacterium	11.68	27.06	0.02	37.96	0.0001
S	Corynebacterium camporealensis	0.33	0.50	0.10	0.60	0.02
S	Corynebacterium frankenforstense	1.18	0.63	0.46	0.33	0.04
S	Corynebacterium jeikeium	0.20	1.25	0.0007	0.62	<0.0001
S	Corynebacterium lactis	0.23	0.25	0.28	4.27	<0.0001
S	Corynebacterium sp. LMM-1652	0.05	0.24	0.001	0.17	<0.0001
S	Corynebacterium resistens	0.03	0.38	<0.0001	0.25	<0.0001
S	Corynebacterium urealyticum	0.10	0.24	0.04	0.77	<0.0001
S	Corynebacterium xerosis	1.10	0.84	0.26	1.69	0.88
G	Trueperella	0.17	0.12	0.17	2.87	<0.0001
S	Trueperella pyogenes	0.13	0.10	0.28	2.87	<0.0001
P	Firmicutes	27.31	23.99	0.68	10.47	0.002
G	Staphylococcus	1.40	2.86	0.14	1.08	0.91
S	Staphylococcus agnetis	0.005	0.008	0.21	0.007	0.27
S	Staphylococcus auricularis	0.12	1.28	0.18	0.11	0.57
S	Staphylococcus capitis	0.03	0.07	0.54	0.04	0.29
S	Staphylococcus chromogenes	0.01	0.03	0.32	0.04	0.36
S	Staphylococcus hominis	0.05	0.06	0.93	0.02	0.03

Open in a new tab

Bacterial phyla, genera and species representing at least 10% of the classified reads in at least one sample of samples from mild (M, n = 17) and severe (S, n = 19) udder cleft dermatitis (UCD) and 13 control (C) samples from cows without UCD. The 49 samples were obtained from 47 cows in six Swedish dairy herds. The median proportion of classified reads for each bacteria and sample type is presented. Differences in abundance between control samples and mild (M) UCD, and between control samples and severe (S) UCD samples were analyzed using the Wilcoxon rank sum test. A P-value ≤0.002 was considered significant due to multiple testing according to the Bonferroni correction.

^aP = Phylum, G = Genus, S = Species.

On the genus and species level, the taxa that represented more than 10% in at least one sample were visualized (Fig 4B and 4C), and differences in abundance of these taxa between sample types are presented in Table 1. In many UCD samples, a single genus represented a larger proportion of the reads compared to control samples (Fig 4B). The specific genus and species that had increased in proportion differed between samples, but some subgroups were visible. The subgroups were also defined by hierarchical clustering (S4 Fig). The largest subgroup had a high proportion of Corynebacterium spp. and corresponded to the samples separated by the first PCA axis on the genus level (Fig 3B and S1 Fig). In addition, this was largely the same group of samples that was dominated by Actinobacteria on the phylum level. Different Corynebacterium species dominated in different samples but, in most cases, one or two species represented a major proportion (>50%) of the Corynebacterium associated reads within each sample (Fig 4C). The first PCA axis on species level separated a group of mainly severe UCD samples with a high proportion of Corynebacterium lactis, whereas Corynebacterium urealyticum, C. xerosis and C. camporealensis contributed to the separation of several mild and severe UCD samples by the third PCA axis (PC3, Fig 3C and S1 Fig). Several of these Corynebacterium spp. differed significantly between sample types (Table 1). A few samples had a higher proportion of Brevibacterium, mainly Brevibacterium luteolum (Fig 4B and 4C), which corresponded to the second PCA axis on the genus and species level (Fig 3B and 3C and S1 Fig). In a third subgroup, mainly including mild UCD lesions, an increased proportion of Staphylococcus spp. was seen (Fig 4B and 4C). This group corresponds to the samples separated by the third PCA axis on the genus level (Fig 3B and S1 Fig) and largely represents the group in which Firmicutes had expanded on the phylum level (Fig 3A). The fourth identified subgroup only comprised severe UCD samples and had a high proportion of anaerobic or facultative anaerobic bacteria, including the genera Trueperella, Fusobacterium and Porphyromonas (Fig 4B and 4C). This group was separated on the fifth PCA axis on the genus level and largely corresponded to the group characterized by Fusobacteria and Bacteroidetes on the phylum level (Fig 3B and S1 Fig). The most dominating species in this group was Porphyromonas asaccharolytica, Trueperella pyogenes, and Fusobacterium necrophorum and the latter two also affected the fourth PCA axis on species level (Fig 3C and S1 Fig). Several control samples had a relatively high proportion of Bifidobacterium spp. and these samples were separated by the fourth PCA axis on genus level (Fig 4C and S2 Fig). The abundance of several genera and species distinguishing these subgroups differed significantly between UCD categories and control samples (Table 1).

Bacterial diversity

On the genus level, the mean Shannon diversity index was significantly higher in controls compared to UCD samples, with a mean of 4.7 (SD 0.8) for control samples, 3.3 (SD 1.0) for mild UCD samples and 2.8 (SD 0.8) for severe UCD samples (Fig 5A). Also, on the species level, there was significantly higher diversity in control samples (mean 6.6, SD 0.7) compared to mild (mean 5.1, SD 1.2) and severe (mean 4.3, SD 0.8)) UCD samples, as well as a higher diversity in mild compared to severe UCD samples (Fig 5B).

Fig 5 — Box plot of the Shannon diversity index for bacterial genera (A) and species (B), in samples from mild (n = 17) and severe (n = 19) udder cleft dermatitis (UCD) and samples from skin at the fore udder attachment from healthy controls (n = 13).

Mastitis-causing bacteria and spirochetes in UCD samples

Among the pathogens that are considered to be common mastitis-causing bacterial species in Sweden, Escherichia coli, Staphylococcus chromogenes, Staph. epidermidis, Staph. haemolyticus, Staph. simulans and Trueperella pyogenes were represented by at least 1% of the reads in at least one sample. There was a higher proportion of Escherichia coli in the control samples compared to mild (P = 0.006) and severe (P<0.0001) UCD samples, although the proportion generally was low in all sample types. For Staph. epidermidis there was a tendency towards higher proportions in the control samples (P = 0.02). For the other Staphylococcus spp. mentioned above, no differences between sample types were seen. Trueperella pyogenes was more frequent in samples from severe lesions (mean 7.1%, SD 9.3, P<0.0001), P = 0.0001) compared to samples from control cows (mean 0.1, SD 0.06), whereas there was no significant difference between controls and mild UCD samples (mean 0.3. SD 0.5, P = 0.28). The Spirochaetes proportion of the classified reads ranged from 0.01 to 0.4% (mean 0.2%) and Treponema spp. reads constituted between 0.002 and 0.2% of the classified reads, with a very low abundance in all sample types.

Other microorganisms

Archaea

As concluded above, there was a lower proportion of sequence reads classified as Archaea in UCD samples compared to control samples (P = 0.007 for mild and P<0.0001 for severe UCD samples; Figs 2B and 6A). Most of the severe UCD samples and a subset of the mild UCD samples had a very low proportion of archaeal classified reads, i.e. less than 0.5% (Fig 6A). These samples were also different in the relative composition of archaeal genera (Fig 6B). In the control samples and some of the UCD samples, reads classified to the genus Methanobrevibacter dominated. In the other UCD samples, two subgroups were distinguishable. One group, including both mild and severe UCD samples, was characterized by a high relative proportion of halophilic archaea (e.g. Halorobrum and Halobacterium) and the other, including severe UCD samples only, by an increased proportion of Methanosarcina (Fig 6B). The Methanosarcina subgroup was largely the same as the subgroup characterized by elevated levels of anaerobic bacteria described above. A PCA on the genus level separated samples with low abundance of archaea on the first PCA axis and the halophilic and Methanosarcina subgroups on the second PCA axis (Fig 6C).

Fig 6 — The archaeal proportion of all classified reads within each sample (A), the distribution of archaeal genera representing ≥5% of the archaeal reads in at least one sample (B) and a principal component analysis (C), based on samples from mild (n = 17) and severe (n = 19) udder cleft dermatitis (UCD) and samples from skin at the fore udder attachment from healthy controls (n = 13).

Fungi

The fungal reads represented a mean of 0.5 (SD 0.4), 2.6 (SD 9.2) and 0.2 (SD 0.1)% of the classified reads in control samples, mild UCD, and severe UCD, respectively. Control samples had a higher proportion of fungal reads, compared to mild (P = 0.05) and severe (P = 0.001), but we found no indication of any specific genera and species that differed between sample types. Fusarium was the most frequently classified genus with a mean of 12,5 (SD 3.2), 10.3 (SD 4.2) and 10.2 (SD 2.3)% of the fungal reads for control, mild and severe samples, respectively. One sample was responsible for the numerically higher mean abundance in samples from mild lesions (Fig 2A). In this specific sample, the majority of the classified fungal reads belonged to the genus Candida, with the species Candida orthopsilosis representing 69% of the fungal reads.

Differences in microbiota between different time points–two examples

In two cases, the data included samples of the same UCD lesion from two different time points. In the first case, the sample (M7) was first identified in December 2018 as a recently developed mild lesion (Fig 7A). Six weeks later, the lesion was scored as a recently developed severe UCD lesion and was therefore sampled again (sample S10; Fig 7B). In both samples, bacteria constituted more than 99% of the classified reads and the proportion of archaeal reads was low. Both samples had a high relative abundance of Actinobacteria and Corynebacterium (Fig 7). However, in the severe UCD sample S10, the proportion of Corynebacterium spp. had increased (60.5 of classified reads) compared to the initial mild UCD sample M7 (37.3% of classified reads), with Corynebacterium camporealensis being mainly responsible for the difference (4.6% of classified reads in M7 compared to 16.3% in S10) (Fig 7). The species level Shannon diversity index was also numerically higher in sample M7 (5.7) compared to sample S10 (4.9). Thus, the microbiota in this UCD lesion shifted over time towards a lesion containing mainly Corynebacterium.

Fig 7 — Sankey visualization (obtained by the metagenomics tool Pavian) of the microbiota in a mild (A) and a severe (B) udder cleft dermatitis lesion in the same cow, sampled in December 2018 and six weeks later in January 2019. The flow diagram illustrates the proportion of bacterial reads assigned to a specific taxon at domain (D), kingdom (K), phylum (P), family (F), genera (G) and species level.

In the second case, a sample (S6) was taken from a cow first identified as having a recently developed severe UCD lesion (Fig 8A), and the cow was then sampled again, around three months later at the final herd visit (sample S7; Fig 8B). As was seen in the previous example, bacteria constituted most of the classified reads, 99.6% of S6 and 97.6% of S7. Among the bacteria, Brevibacterium luteolum was most prevalent in S6, representing 49.2 of the classified reads, but only 0.1% in S7. Corynebacterium was common in both samples, representing 23.7 and 38.0% in S6 and S7, respectively, with the most abundant species identified as Corynebacterium lactis in S6 (6.4% compared to 1.9% in S7) and Corynebacterium urealyticum in S7 (14.9% compared to 0.2% in S6). Apart from these differences, Porphyromonas asaccharolytica (0.1 and 5.0% in S6 and S7, respectively) and Fusobacterium necrophorum (0.6 and 2.7% in S6 and S7, respectively) had increased in S7 compared to S6. Thus, the microbiota in this lesion shifted towards a higher proportion of anaerobic bacteria, and the species level Shannon diversity index shifted from 3.5 in S6 to 5.9 in S7.

Discussion

To our knowledge, this is the first study to investigate the microbiota of UCD lesions using a shotgun metagenomic sequencing approach, and it provides an increased understanding of the microbiological differences in mild and severe UCD lesions compared to healthy skin. Our results show an altered microbiota in both mild and severe UCD lesions compared to the control samples, manifested by decreased bacterial diversity and an increased proportion of certain bacterial genera and species. In line with our results on diversity, Sorge et al. [17] found a lower bacterial diversity in samples from UCD lesions compared to control samples. A similar finding was also reported in a study on digital dermatitis, in which bacterial diversity decreased as the digital dermatitis lesions progressed in severity [32]. In metagenomic studies of the human skin, a high diversity is characteristic of a healthy skin microbiota [33], whereas a loss of diversity and increased proportion of pathogenic or opportunistic bacteria can be defined as dysbiosis [33, 34]. Decreased diversity and dysbiosis are associated with numerous skin conditions, such as atopic dermatitis in humans and dogs [35, 36]. A lower diversity has also been found in diabetic foot ulcers and in the healthy skin of diabetic patients compared to non-diabetic controls [37]. However, it has not been established whether the dysbiosis is a cause or a result of the pathological condition, and further studies within this area are required. We also found that a few of the UCD samples, mainly from mild UCD, had similar microbiota to control samples, indicating that an altered microbiota is not always present in mild lesions.

Bacterial abundance and subgroups of UCD samples

Apart from the decreased diversity, we found an increased proportion of certain bacterial genera and species in the UCD samples compared to the controls. Noticeably, it was not always the same species that increased in different samples. In our dataset, we observed four broad subgroups of UCD samples with different types of dysbiosis. In the largest subgroup, the genus Corynebacterium contributed to a high proportion of reads, although differences in species were seen between samples. In another subgroup of samples, mainly from mild lesions, Staphylococcus spp. represented a relatively high proportion of the classified reads and contributed to the decreased diversity. Several staphylococcal species were found but, in most cases, only one or two staphylococcal species represented more than 1% of the reads within the same sample. Both Corynebacterium spp. and Staphylococcus spp. have previously been identified in UCD lesions through culturing [3, 13]. The genera Corynebacterium and Staphylococcus are also associated with bovine healthy skin from teat apices [38, 39] and, in contrast to our results, Sorge et al. [17] found a higher abundance of Corynebacterium spp. in controls than in UCD samples. These bacteria are also common in human chronic ulcers, as well as being a common finding in healthy skin microbiota in humans [37]. Staphylococcus spp. are also associated with atopic dermatitis in both humans and dogs [23, 35] and are also known to be involved in biofilm formation in human chronic ulcers [40]. A few samples, (three mild and one severe UCD samples) had a high proportion of Brevibacterium (i.e. Brevibacterium luteolum), a coryneform bacteria that, to our knowledge, has not been previously identified in UCD lesions. Brevibacterium is a known skin commensal in humans that, in similarity to corynebacteria, may act as an opportunist, mainly in immunocompromised individuals [41, 42]. The high proportion of common skin bacteria found in UCD samples in the present study implies that, under certain circumstances, skin commensals may increase in relative abundance, resulting in a decreased diversity indicative of an impaired microbiota, or dysbiosis. Such circumstances may include an altered local environment, for example, changes in pH, oxygen levels or humidity. The presence of a dysbiotic microbiota involving common skin bacteria is also seen in human chronic wounds for which there are underlying causes, such as diabetes or venous insufficiencies [24, 37]. In the fourth subgroup indicating dysbiosis, including several severe, but no mild, UCD samples, a more anaerobic microbiota was observed, including bacterial species such as Trueperella pyogenes, Fusobacterium necrophorum and Porphyromonas asaccharolytica. These bacteria have previously been identified in bacteriological studies of UCD [2, 3] and, in line with our results, Sorge et al. [17] found a higher abundance of these bacteria in UCD samples compared to control samples. Finding these opportunistic bacteria in severe UCD lesions is not surprising as they are associated with several bacterial conditions in ruminants, such as interdigital phlegmon, abscesses and wound infections [43, 44]. In chronic wounds, the local environment is associated with low oxygen levels that enable the growth of these bacteria [40]. In addition, Fusobacteria and Porphyromonas are commonly found in combination and, in a laboratory setting, have been shown to form biofilm and could therefore impair the healing of wounds [45]. In line with this, the severe sample (S7) that was re-sampled after a wound duration of around three months showed an increased proportion of these two bacteria. The clinical appearance of the wound that was sampled after three months also showed more characteristics associated with human chronic wounds, such as a lack of granulation tissue, presence of necrotic tissue and fibrin [46], compared to when it had recently developed. The importance of the higher abundance of Bifidobacterium spp. found in control samples is not known, but it could be speculated that these bacteria have a protective effect on the skin barrier. Sorge et al. [17] also found Bifidobacterium spp. to be significantly more common in controls compared to UCD samples. Bifidobacteria produce lactic acid and are commonly used as a probiotic treatment in the restoration of gut microbiota [47]. Although previous studies have suggested an association between UCD and Treponema spp. [1, 15], we found no evidence of Treponema spp. being involved in UCD lesions, as the abundance of this genus was low in all samples, which was also seen in a previous study of UCD [17].

Presence of mastitis-causing bacteria

A low abundance of the bacteria considered to be important mastitis-causing pathogens in Sweden (apart from Trueperella pyogenes) was seen. Thus, our results do not indicate that UCD lesions act as a reservoir for such pathogens. Trueperella pyogenes is a mastitis-causing pathogen [44], typically infecting heifers and cows in the dry period during the summer. In such “summer mastitis”, it is generally believed that flies are responsible for the transmission of the pathogen. This opportunistic bacterium is also common in several other conditions, such as abscesses and pneumonia [44], and we do not believe that the presence of Trueperella pyogenes in severe UCD lesions has any significant effect on the risk of mastitis in dairy herds. However, during the summer, an increased presence of the bacteria in cows with severe UCD could increase the risk of this specific type of summer mastitis if flies are present in the environment. It is not clear, however, whether such a route of transmission could partly explain the association between UCD and mastitis found in previous studies [4, 14].

Abundance of other microorganisms

To our knowledge, the presence of Archaea on bovine skin has not been previously investigated. These single-celled microorganisms have previously been identified in different environments, including the bovine rumen [48] and the human gut and skin [49]. Methanogens, such as Methanobrevibacter, have been found to play a role in ruminal microbial metabolism, by using hydrogen for their growth, and reducing carbon dioxide to methane [48]. The presence of Methanobrevibacter in our samples could be the result of contamination from faeces or the environment, but it is also possible that these archaea are part of the bovine skin microbiota. Archaea have been proposed to play a role in ammonia metabolism in human skin [49] and it could be speculated that they might play a similar role in pH regulation of bovine skin, although this requires further studies. We found differences in archaeal abundance between controls and mild and severe UCD samples, but the importance of these differences is not known as archaeal function is still a poorly explored area of research. Our results suggest that fungi are not associated with UCD lesions in most cases, as the fungal reads represented a low proportion of the majority of samples. However, one mild UCD sample had a high proportion of fungal reads, which indicates that opportunistic fungi may be part of a shift towards a decreased diversity of the microbiota in UCD lesions. In this sample, one specific fungal species was responsible for 69%of the fungal reads. We found low proportions of viral and protozoan reads in all samples, which imply that these agents are not a common part of the UCD microbiota or healthy bovine skin. However, the methods used for preparing samples for DNA extraction might have affected the viral content of the samples, as small virus particles could have been lost prior to DNA extraction. An alternative methodology would probably be needed for specific analysis of the viral content of the samples. Another pathogen that has been associated with UCD is the mange mite, Chorioptes bovis [10]. As Chorioptes bovis is a common finding in Sweden, it would have been interesting to see whether DNA from this pathogen was present in our samples. As no genome sequence was available from the NCBI, no such analysis was performed. In addition, as several studies did not find any evidence of an involvement of mange mites in UCD [2, 5], we do not believe that they are of major interest.

Final remarks

The methods used in this study, regarding both the sampling and laboratory procedures, as well as shotgun metagenomic sequencing, have been scarcely explored for bovine wound microbiota. Shotgun metagenomic sequencing has, however, been shown to enhance the detection of bacterial species compared to 16S amplicon sequencing, in which the selection of primers may affect the results [19, 50]. Contaminating DNA in laboratory reagents may influence the taxonomic classification of metagenomics results [51]. In this study, we did not include negative controls and this constitutes a limitation when interpreting the results. This is especially true when looking at low abundance groups. However, in this study we focus mainly on the most dominating taxa and their proportions. Our results are generally in line with a previous study using 16S amplicon sequencing to investigate UCD lesions in comparison with control samples [17], suggesting that these two studies have increased the understanding of the microbiota of UCD lesions in comparison to healthy skin at the fore udder attachment. In addition, our results demonstrate that mild UCD lesions also display a dysbiotic microbiota and, in combination with the fact that mild UCD lesions often develop into severe lesions [14], this suggests that mild lesions should not be ignored.

Conclusions

The results of this study indicate that the microbiota of UCD lesions is different to that of healthy skin from the same body site, with a dysbiosis manifested as reduced diversity and increased proportions of certain bacteria in mild and severe UCD lesion samples. It is, however, not known if the dysbiosis is a contributing cause to, or a result of, the UCD lesions, and these associations require further investigations. In this study we identified three broad categories of dysbiosis characterized by different groups of bacteria. Although several bacterial species were more frequently identified in the UCD samples, the overall interpretation is that no specific pathogen is involved in the development of UCD, as the bacteria differed between samples. We found no evidence of UCD lesions acting as a reservoir for mastitis-causing bacteria, as such bacteria were found in low proportions in most samples.

Supporting information

S1 Table. Sample overview.

Overview of samples subjected to shotgun metagenomic sequencing in a study of the microbiota of mild (M) and severe (S) udder cleft dermatitis (UCD) in comparison with samples from the same body site from healthy controls (C) in 47 cows in 6 Swedish dairy herds.

(XLSX)

Click here for additional data file.^{(28.1KB, xlsx)}

S2 Table. Raw classification data.

The results of the classification of the sequenced reads (after removal of sequences assigned to the Bos Taurus genome) presented in an Excel table as both clade read counts and normalized into a percentage of classified reads for each taxa.

(XLSX)

Click here for additional data file.^{(11.8MB, xlsx)}

S1 Fig. PCA loadings.

(TIF)

Click here for additional data file.^{(167.8KB, tif)}

S2 Fig. PCA axis 4 on bacterial genus level.

(TIF)

Click here for additional data file.^{(251.8KB, tif)}

S3 Fig. PCA on bacterial genus level colored by herd, breed, parity and extraction month.

(TIF)

Click here for additional data file.^{(190.3KB, tif)}

S4 Fig. Hierarchical clustering of samples.

(TIF)

Click here for additional data file.^{(541.4KB, tif)}

Acknowledgments

Sequencing was performed by the SNP&SEQ Technology Platform in Uppsala. The facility is part of the National Genomics Infrastructure (NGI) Sweden and Science for Life Laboratory. The SNP&SEQ Platform is also supported by the Swedish Research Council and the Knut and Alice Wallenberg Foundation. We would also like to thank Harri Ahola and Karin Ullman at the National Veterinary Institute, Uppsala, Sweden for their help and advice regarding DNA extraction methodology, as well as Maria “Maja” Persson and the staff of “DOA lab” at the National Veterinary Institute for excellent technical assistance. Last but not least, the farmers and cows participating in this study–thank you for your kind hospitality and assistance throughout the study period.

Data Availability

The raw sequence data has been submitted to the Sequence Read Archive (SRA) and is accessible via the bioproject PRJNA636289. All other relevant data are within the paper and its Supporting Information files.

Funding Statement

K.P.W. received funding from The Swedish Research council Formas (grant number 221 -2013-269, www.formas.se) and from Stiftelsen lantbruksforskning - Swedish farmers' foundation for agricultural research (grant number V1430006, www.lantbruksforskning.se). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. A.N. is employed by Växa Sverige, but her salary costs for her work in the study was covered by the grants listed above. Thus, Växa Sverige did not have any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.

References

1.Boyer P, Singleton G. Digital dermatitis, superfoul and severe necrotic enteritis of the udder in dairy cows. Vet Rec. 1998;142(6):147–8. [PubMed] [Google Scholar]
2.Warnick LD, Nydam D, Maciel A, Guard CL, Wade SE. Udder cleft dermatitis and sarcoptic mange in a dairy herd. J Am Vet Med Assoc. 2002;221(2):273–6. 10.2460/javma.2002.221.273 [DOI] [PubMed] [Google Scholar]
3.Persson Waller K. Nekrotisk dermatit framtill på juvret hos mjölkkor [In Swedish]. Svensk Veterinärtidning. 2003;55(8–9):11–6. Available from: https://www.sva.se/media/dwgds2x1/kpw_nekrotisk_dermatit_svt_2003.pdf [Google Scholar]
4.Persson Waller K, Bengtsson M, Nyman AK. Prevalence and risk factors for udder cleft dermatitis in dairy cattle. J Dairy Sci. 2014;97(1):310–8. 10.3168/jds.2013-7186 [DOI] [PubMed] [Google Scholar]
5.Hansen MJ, Nissen MM. A Clinical Study of Udder Cleft Dermatitis on a Danish Dairy Farm. M. Sc. Thesis, University of Copenhagen. 2010. [Google Scholar]
6.Riekerink R, van Amersfort K, Sampimon OC, Hooijer GA, Lam T. Short communication: Prevalence, risk factors, and a field scoring system for udder cleft dermatitis in Dutch dairy herds. J Dairy Sci. 2014;97(8):5007–11. 10.3168/jds.2013-7651 [DOI] [PubMed] [Google Scholar]
7.Eikeland J, Paulsen H. Studie av jurfestesår hos storfe. M. Sc. Thesis, Norwegian University of Life Sciences. 2018. [Google Scholar]
8.Ekman L, Nyman AK, Landin H, Magnusson U, Waller KP. Mild and severe udder cleft dermatitis-Prevalence and risk factors in Swedish dairy herds. Journal of dairy science. 2018;101(1):556–71. 10.3168/jds.2017-13133 [DOI] [PubMed] [Google Scholar]
9.Bouma A, Nielen M, van Soest E, Sietsma S, van den Broek J, Dijkstra T, et al. Longitudinal study of udder cleft dermatitis in 5 Dutch dairy cattle herds. J Dairy Sci. 2016;99:1–9. 10.3168/jds.2015-9445 [DOI] [PubMed] [Google Scholar]
10.Allenstein LC. Mites cause many of the smelly udder sores. Hoard's Dairyman. 1991;136:507. [Google Scholar]
11.Stamm LV, Walker RL, Read DH. Genetic diversity of bovine ulcerative mammary dermatitis-associated Treponema. Vet Microbiol. 2009;136(1–2):192–6. 10.1016/j.vetmic.2008.10.022 [DOI] [PubMed] [Google Scholar]
12.Lyman D. BHV-4 Contributes to Udder Lesions. Bovine Vet. 2019. January 4 [Cited 2020 Mar 23]. Available from: https://www.bovinevetonline.com/article/bhv-4-contributes-udder-lesions [Google Scholar]
13.Beattie KG, Taylor DJ. An investigation into intertrigo (necrotic dermatitis or 'foul udder') in dairy cows. Cattle Pract. 2000;8(4):377–80. [Google Scholar]
14.Ekman L. Udder cleft dermatitis in dairy cows. Doctoral Thesis, Swedish University of Agricultural Sciences. 2020. Available from: https://pub.epsilon.slu.se/16894/ 10.3168/jds.2019-17935 [DOI] [PubMed] [Google Scholar]
15.Evans NJ, Timofte D, Carter SD, Brown JM, Scholey R, Read DH, et al. Association of treponemes with bovine ulcerative mammary dermatitis. Veterinary Record. 2010;166(17):532–3. 10.1136/vr.b4822 [DOI] [PubMed] [Google Scholar]
16.Morey M, Fernandez-Marmiesse A, Castineiras D, Fraga JM, Couce ML, Cocho JA. A glimpse into past, present, and future DNA sequencing. Molecular genetics and metabolism. 2013;110(1–2):3–24. 10.1016/j.ymgme.2013.04.024 [DOI] [PubMed] [Google Scholar]
17.Sorge US, Binger EM, Schefers J, Plummer PJ. Short communication: Metagenomic evaluation of skin biopsies of udder sores in dairy cows. Journal of dairy science. 2019. 10.3168/jds.2018-15863 [DOI] [PubMed] [Google Scholar]
18.Petrosino JF, Highlander S, Luna RA, Gibbs RA, Versalovic J. Metagenomic pyrosequencing and microbial identification. Clinical chemistry. 2009;55(5):856–66. 10.1373/clinchem.2008.107565 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Ranjan R, Rani A, Metwally A, McGee HS, Perkins DL. Analysis of the microbiome: Advantages of whole genome shotgun versus 16S amplicon sequencing. Biochemical and biophysical research communications. 2016;469(4):967–77. 10.1016/j.bbrc.2015.12.083 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Van Rossum T, Ferretti P, Maistrenko OM, Bork P. Diversity within species: interpreting strains in microbiomes. Nature reviews Microbiology. 2020;18(9):491–506. 10.1038/s41579-020-0368-1 . [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Jovel J, Patterson J, Wang W, Hotte N, O'Keefe S, Mitchel T, et al. Characterization of the gut microbiome using 16S or shotgun metagenomics. Frontiers in microbiology. 2016;7:459 10.3389/fmicb.2016.00459 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Neves ALA, Li F, Ghoshal B, McAllister T, Guan LL. Enhancing the resolution of rumen microbial classification from metatranscriptomic data using kraken and mothur. Frontiers in microbiology. 2017;8:2445 10.3389/fmicb.2017.02445 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Byrd AL, Deming C, Cassidy SKB, Harrison OJ, Ng WI, Conlan S, et al. Staphylococcus aureus and Staphylococcus epidermidis strain diversity underlying pediatric atopic dermatitis. Science translational medicine. 2017;9(397). 10.1126/scitranslmed.aal4651 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA, et al. Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC microbiology. 2008;8:43 10.1186/1471-2180-8-43 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics (Oxford, England). 2014;30(15):2114–20. 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome biology. 2009;10(3):R25 10.1186/gb-2009-10-3-r25 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics (Oxford, England). 2009;25(16):2078–9. 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome biology. 2019;20(1):257 10.1186/s13059-019-1891-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Lu J, Breitwieser FP, Thielen P, Salzberg SL. 2017. Bracken: estimating species abundance in metagenomics data. PeerJ Computer Science 3:e104 10.7717/peerj-cs.104 [DOI] [Google Scholar]
30.Breitwieser FP, Salzberg SL. Pavian: interactive analysis of metagenomics data for microbiome studies and pathogen identification. Bioinformatics (Oxford, England). 2020;36(4):1303–4. 10.1093/bioinformatics/btz715 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Sverige Växa. Redogörelse för husdjursorganisationens djurhälsovård 2018/2019 [Report on animal health care within the livestock organisation 2018/2019]. 2020. [cited 2020 Nov 6]. [pp 22–24]. Available from: https://www.vxa.se/globalassets/dokument/statistik/redogorelse-for-husdjursorganisationernas-djurhalsovard-2018-2019.pdf [Google Scholar]
32.Krull AC, Shearer JK, Gorden PJ, Cooper VL, Phillips GJ, Plummer PJ. Deep sequencing analysis reveals temporal microbiota changes associated with development of bovine digital dermatitis. Infection and Immunity. 2014;82(8):3359–73. 10.1128/IAI.02077-14 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Egert M, Simmering R, Riedel CU. The association of the skin microbiota with health, immunity, and disease. Clinical pharmacology and therapeutics. 2017;102(1):62–9. 10.1002/cpt.698 [DOI] [PubMed] [Google Scholar]
34.Structure, function and diversity of the healthy human microbiome. Nature. 2012;486(7402):207–14. 10.1038/nature11234 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Bradley CW, Morris DO, Rankin SC, Cain CL, Misic AM, Houser T, et al. Longitudinal evaluation of the skin microbiome and association with microenvironment and treatment in canine atopic dermatitis. The Journal of investigative dermatology. 2016;136(6):1182–90. 10.1016/j.jid.2016.01.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Williams MR, Gallo RL. The role of the skin microbiome in atopic dermatitis. Current allergy and asthma reports. 2015;15(11):65 10.1007/s11882-015-0567-4 [DOI] [PubMed] [Google Scholar]
37.Gardiner M, Vicaretti M, Sparks J, Bansal S, Bush S, Liu M, et al. A longitudinal study of the diabetic skin and wound microbiome. PeerJ. 2017;5:e3543 10.7717/peerj.3543 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Braem G, De Vliegher S, Verbist B, Heyndrickx M, Leroy F, De Vuyst L. Culture-independent exploration of the teat apex microbiota of dairy cows reveals a wide bacterial species diversity. Vet Microbiol. 2012;157(3–4):383–90. 10.1016/j.vetmic.2011.12.031 [DOI] [PubMed] [Google Scholar]
39.Dahlberg J. Bovine milk microbiota—methods matter. Doctoral Thesis, Swedish University of Agricultural Sciences. 2019. Available from: https://pub.epsilon.slu.se/16433/ [Google Scholar]
40.Wu YK, Cheng NC, Cheng CM. Biofilms in chronic wounds: pathogenesis and diagnosis. Trends in biotechnology. 2019;37(5):505–17. 10.1016/j.tibtech.2018.10.011 [DOI] [PubMed] [Google Scholar]
41.Coyle MB, Lipsky BA. Coryneform bacteria in infectious diseases: clinical and laboratory aspects. Clin Microbiol Rev. 1990;3(3):227–46. 10.1128/cmr.3.3.227 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Magi B, Migliorini L, Sansoni A, Cusi MG. Brevibacterium casei bacteraemia in a port-a-cath carrier patient: a case report. Infez Med. 2018;26(3):263–5. . [PubMed] [Google Scholar]
43.Nagaraja TG, Narayanan SK, Stewart GC, Chengappa MM. Fusobacterium necrophorum infections in animals: pathogenesis and pathogenic mechanisms. Anaerobe. 2005;11(4):239–46. 10.1016/j.anaerobe.2005.01.007 [DOI] [PubMed] [Google Scholar]
44.Ribeiro MG, Risseti RM, Bolanos CA, Caffaro KA, de Morais AC, Lara GH, et al. Trueperella pyogenes multispecies infections in domestic animals: a retrospective study of 144 cases (2002 to 2012). Vet Q. 2015;35(2):82–7. 10.1080/01652176.2015.1022667 [DOI] [PubMed] [Google Scholar]
45.Lockhart JS, Buret AG, Ceri H, Storey DG, Anderson SJ, Morck DW. Mixed species biofilms of Fusobacterium necrophorum and Porphyromonas levii impair the oxidative response of bovine neutrophils in vitro. Anaerobe. 2017;47:157–64. 10.1016/j.anaerobe.2017.05.008 [DOI] [PubMed] [Google Scholar]
46.Lindholm C. Sår [In Swedish]. 3rd ed Lund: Studentlitteratur AB; 2012. [Google Scholar]
47.Musthaq S, Mazuy A, Jakus J. The microbiome in dermatology. Clinics in dermatology. 2018;36(3):390–8. 10.1016/j.clindermatol.2018.03.012 [DOI] [PubMed] [Google Scholar]
48.Janssen PH, Kirs M. Structure of the archaeal community of the rumen. Appl Environ Microbiol. 2008;74(12):3619–25. 10.1128/AEM.02812-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Probst AJ, Auerbach AK, Moissl-Eichinger C. Archaea on human skin. PloS one. 2013;8(6):e65388 10.1371/journal.pone.0065388 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Meisel JS, Hannigan GD, Tyldsley AS, SanMiguel AJ, Hodkinson BP, Zheng Q, et al. Skin Microbiome surveys are strongly influenced by experimental design. the journal of investigative dermatology. 2016;136(5):947–56. 10.1016/j.jid.2016.01.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Salter SJ, Cox MJ, Turek EM, Calus ST, Cookson WO, Moffatt MF, et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014;12:87 10.1186/s12915-014-0087-z [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0242880.r001

Decision Letter 0

Peter Gyarmati

25 Aug 2020

PONE-D-20-18567

Udder cleft dermatitis in dairy cows is associated with an altered and less diverse microbiota compared to the microbiota of healthy skin

PLOS ONE

Dear Dr. Ekman,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 09 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Peter Gyarmati

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you are reporting an analysis of a microarray, next-generation sequencing, or deep sequencing data set. PLOS requires that authors comply with field-specific standards for preparation, recording, and deposition of data in repositories appropriate to their field. Please upload these data to a stable, public repository (such as ArrayExpress, Gene Expression Omnibus (GEO), DNA Data Bank of Japan (DDBJ), NCBI GenBank, NCBI Sequence Read Archive, or EMBL Nucleotide Sequence Database (ENA)). In your revised cover letter, please provide the relevant accession numbers that may be used to access these data. For a full list of recommended repositories, see http://journals.plos.org/plosone/s/data-availability#loc-omics or http://journals.plos.org/plosone/s/data-availability#loc-sequencing.

3. Thank you for stating the following in the Competing Interests section:

"The authors have declared that no competing interests exist."

We note that one or more of the authors are employed by a commercial company: Växa Sverige.

3.1. Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

3.2. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

4. We noted in your submission details that a portion of your manuscript may have been presented or published elsewhere.

"The manuscript is a part of a doctoral thesis and is thus published as a manuscript in the print version of the thesis (ISBN 978-91-7760-564-5). The thesis is available at https://pub.epsilon.slu.se/16894/, this manuscript is not included in the online publication. "

Please clarify whether this publication was peer-reviewed and formally published. If this work was previously peer-reviewed and published, in the cover letter please provide the reason that this work does not constitute dual publication and should be included in the current manuscript.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Ekman et al. investigate the skin microbiome of healthy and diseased cows using a shotgun metagenomic sequencing approach. Although I believe the study to have merit, it does contain a number of methodological flaws that make it difficult to evaluate in its current form.

Major

DNA extraction, library preparation and sequencing

Negative controls are an important component for any type of experiment where measurements are being made. The importance of these and their impact have been documented extensively throughout the microbiome literature (see https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-014-0087-z). Because they were not included in this experiment, it is an obvious limitation that should be discussed, but I also think that their impact depends on how you analyze and present results. For example, if you make claims about specific OTUs, then you should be confident that what you are describing is real, but if you make claims about general trends at a specific taxonomic level (e.g., phylum, genus, species, etc..), as you do here, then contamination may have less of an impact on the overall results, especially when DNA yield is high and contaminants are in low abundance. Please address the importance of negative controls in microbiome research and discuss the limitations of not including them in your study.

It is also not clear how putative batch effects were handled in this study. Were all samples extracted on the same day? Were they sequenced on the same date, on the same machine?

Bioinformatics

Overall, I think that the methods describing your bioinformatic methods is very well done; however, I do have a concern regarding the omission of how adapter sequences and low quality sequences were handled. Adapter contamination is known to affect alignment statistics for commonly used sequence aligners like BWA and Bowtie2 (https://academic.oup.com/bioinformatics/article/30/15/2114/2390096), resulting in reads that are incorrectly classified as “unmapped”. This does appear to have affected the reported proportions of “host” versus “non-host” DNA for severe UCD samples and others, as a large proportion of your sequence reads were classified to the Babesia bigemina genome, which I know to suffer from Bos taurus DNA contamination. Failure to remove these adapter sequences may have also affected classification statistics for other sequences processed with Kraken2. Please rerun your raw sequence data through a read trimming tool like Trimmomatic, cutadapt or other relevant tool before proceeding with the remainder of your analysis. Please also add a column to Supplementary Table 1 describing the number of sequence reads that survived this trimming and other quality control steps.

Statistics

Breed, herd, and parity are likely confounders that should be included in your statistical analysis. Please consider accounting for these potential confounders and update your methods.

Minor

Title: Please consider changing the title, as this seems to be a study comparing the microbiome of healthy and diseased skin (i.e., UCD).

Line 66: Sequencing methods (e.g., 16S amplicon sequencing, shotgun sequencing, etc..) are all biased in some way, as are the sequence databases used to describe the underlying results. You do an excellent job of discussing the tradeoffs of each approach later in the manuscript, but I recommend removing this sentence as it is inaccurate.

Line 77: Shotgun metagenomics can also reveal strain-level information. Please add this and a relevant citation.

Line 102: Change “All scoring and sampling were performed by the first author” to “All scoring and sampling was performed by a single researcher”.

Line 119: Was a pre-dipping solution applied prior to milking? If so, please add and specify relevant product information.

Line 215: Please clarify what you mean by “complexity”. What does it mean for the complexity of a sample to be estimated and why was this done?

Line 220: Please list the bacterial species that were included in this analysis.

Line 248: Please clarify what is meant by “... there was no convincing trend …” Was this based on a pre-determined p-value or “eyeball method”?

Line 270: How are the samples labeled on the ordination plots? It seems that some of the points/samples have labels, but others do not. What is the criterion for labeling samples?

Lines 476-477: The design of this study was probably not sufficient to determine “causation” of dysbiosis; consider re-wording this sentence and others (e.g., 567-569).

Line 555: Delete “Unbiased”. I know what you mean here, but there are numerous other sources of bias in microbiome studies. The choice of sampling device, the choice of DNA extraction kit, the choice of DNA shearing method (e.g., enzymatic versus mechanical), the choice of sequencing depth (e.g., shallow versus deep), and the choice of reference sequence database can all bias shotgun metagenomic experiments.

Reviewer #2: The primary goal of this manuscript is to compare the cutaneous microbiome of sites with mild/severe dermatitis lesions to healthy skin of the fore udder attachment. To accomplish this, the authors generated shotgun metagenomic sequencing data for 49 samples. Analysis of these samples with the k-mer approach Kraken revealed differential communities between controls and samples with UCD. Notably, subgroups existed within the UCD samples and few known mastitis-causing pathogens were identified. Refreshingly, they looked across kingdoms and not only at bacteria. While authors have generated a valuable dataset and the overall methodology is technically sound, there are a few areas that could be improved.

Methods:

- The authors utilized Kraken to classify their metagenomic sequencing reads and then directly used these results to draw conclusions about which genera and species were differential between groups. Because Kraken classifies individual reads to their best matching location in the taxonomic tree and does not actually estimate the abundance of species, the program Bracken (https://ccb.jhu.edu/software/bracken/) should be applied to the Kraken results and then the Bracken species abundances should be compared between groups.

Figure 3:

- Are any of the UCD subgroups, driven by the herd of the animal?

- In the "Bacterial abundance" section, the authors discuss several of the taxa that are driving these ordinations. Can the authors add biplot lines for these driver taxa to the PCA plots?

-For the species based results (Fig 3C), are similar conclusions drawn if the authors perform ordination analysis on a distance matrix, calculated with Bray Curtis, on the samples? It's unclear how there is so little variance at the species level between the control samples.

Figure 4:

- The authors frequently describe different UCD subgroups. To complement these descriptions, can the authors utilize a hierarchical clustering based method to more robustly define them?

Minor comments:

- In Fig 2A, the red/green coloring of Archaea and Fungi will be difficult for colorblind individuals to distinguish. Can the authors please update?

- To make Fig 2B more informative, can the authors use individual scales for the different kingdoms and denote p-values for those comparisons that are significant?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: A Byrd

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Dec 2;15(12):e0242880. doi: 10.1371/journal.pone.0242880.r002

Author response to Decision Letter 0

9 Oct 2020

Response to editor comments:

1. The revised manuscript, including figures and supplemental material are updated according to the style requirements and templates.

2. The raw sequence data can be found in the NCBI Sequence Read Archive, with the accession number series SRR11913436–SRR11913484.

3. 3.1. K.P.W. received funding from The Swedish Research council Formas (grant number 221 - 2013-269, www.formas.se) and from Stiftelsen lantbruksforskning - Swedish farmers' foundation for agricultural research (grant number V1430006, www.lantbruksforskning.se). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. A.N. is employed by Växa Sverige, but her salary costs for her work in the study was covered by the grants listed above. Thus, Växa Sverige did not have any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.

3.2. The authors have declared that no competing interests exist. A.N. is employed by the commercial company Växa Sverige. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

4. The manuscript has been published as a part of a doctoral thesis: However, this publication is not peer reviewed, and it does not constitute dual publication.

5. We also discovered an error in the Ethics Statement, as an older approval number was used by mistake. The correct statement should read: Ethical approval was obtained from a regional Swedish ethics committee appointed by the Swedish board of Agriculture, approval number 5.8.18-06335/2018. Also updated in the box during the submission of the revision.

Response to reviewers:

Major

DNA extraction, library preparation and sequencing

We have added a short discussion on this limitation in the final remarks.

It is also not clear how putative batch effects were handled in this study. Were all samples extracted on the same day? Were they sequenced on the same date, on the same machine?

We have added information about extraction day in S1 Table. All sequencing was done in a single novaseq-run. To roughly determine that DNA extraction date did not interfer with the identified subgroups of samples we colored the samples by extraction month (which is also related to sampling date) in the PCA plot for the bacterial genus level and found no evidence of such interference (Fig S3).

Bioinformatics

Indeed, there was a fraction of the reads that had adaptor content. We have run all fastq files through trimmomatic to remove them. There were very few reads dropped during this procedure, but up to ~15% of the reads were shortened by the trimming. The kraken analysis was then updated. We used a new fresh build of the kraken database in this step and added a Bracken analysis. We have updated the bioinformatic description in the method section and the information in table S1.

Statistics

Breed, herd, and parity are likely confounders that should be included in your statistical analysis. Please consider accounting for these potential confounders and update your methods.

We agree that especially herd, but also breed and parity, may have an effect on the microbiota. To investigate the distribution of these factors over the control and UCD groups we performed Fisher’s exact test and found that these factors did not differ significantly between the groups (this information has been updated in the manuscript). As we wanted to compare the bacterial abundance between groups and as the number of observations were limited (and the bacteria far from normally distributed within samples), we believe that a multivariable analysis is less useful for this dataset. As a spot check, we did test a mixed effect ordinal logistic regression model with bacterial abundance as an ordinal outcome for analyzing some of the most abundant species and found that the estimates (and significance levels) for UCD status was largely unaffected when these potential confounders were included (although herd was significantly associated with e.g. Bifidobacterium abundance). Thus, we would like to keep the Wilcoxon rank sum tests in the manuscript.

In addition, to visually interpret if any of the above mentioned variables had any substantial impact on the clustering of samples in the PCA, we colored the samples by herd, breed and parity in the PCA plot for the bacterial genus level, and found no evidence of such associations (FigS3).

Minor

Title: Please consider changing the title, as this seems to be a study comparing the microbiome of healthy and diseased skin (i.e., UCD).

We have changed the title.

We have removed this sentence.

Line 77: Shotgun metagenomics can also reveal strain-level information. Please add this and a relevant citation.

We have added this and included a recent review on this topic to the citations.

Line 102: Change “All scoring and sampling were performed by the first author” to “All scoring and sampling was performed by a single researcher”.

Changed according to suggestion.

Line 119: Was a pre-dipping solution applied prior to milking? If so, please add and specify relevant product information.

The majority of herds did not use any pre-dipping solution, although all of them used some kind of teat spray or dip post-milking. We do not have any information on the products that were used, but since the UCD lesions are located rather far from the teats, and as all cows within a herd (controls and UCD cows) received the exact same products and procedures during milking, we do not believe that this had any effect on the results of the study.

Line 215: Please clarify what you mean by “complexity”. What does it mean for the complexity of a sample to be estimated and why was this done?

We agree that this analysis is somewhat difficult to interpret. We have removed this from the manuscript (including Fig 5C).

Line 220: Please list the bacterial species that were included in this analysis.

We have indicated these bacterial species in the S2 Table (Bracken results presented in percentages).

Line 248: Please clarify what is meant by “... there was no convincing trend …” Was this based on a pre-determined p-value or “eyeball method”?

We have updated the information of fungal reads in the manuscript. There were no significant differences between sample types in the proportion of classified reads for protozoa and viruses, we have now added this information in Figure 2. In addition, a brief “eyeball method” of the table of classified viral and protozoan reads gave no indication that any taxa were more abundant in any specific sample type.

Line 270: How are the samples labeled on the ordination plots? It seems that some of the points/samples have labels, but others do not. What is the criterion for labeling samples?

To make a more unbiased presentation of the data, all samples have now been indicated in the plots.

Lines 476-477: The design of this study was probably not sufficient to determine “causation” of dysbiosis; consider re-wording this sentence and others (e.g., 567-569).

We have changed the wording in this sentence to clarify. In our understanding, a decreased diversity is indicative of an unbalanced microbiota, and thus a dysbiosis, so it was not our intention to assume any causation. We also added a sentence in the Conclusions section to clarify this.

We have deleted the word unbiased.

Methods:

We have run the data through Bracken to estimate species, genera and phylum level data. We have updated the manuscript to use Bracken data instead of Kraken data. We had to build a new Kraken database and reanalyze the data, since the previously used database lacked files required by Bracken.

Figure 3:

- Are any of the UCD subgroups, driven by the herd of the animal?

We agree that herd (and also breed and parity), may have an effect on the microbiota. To visually interpret if any of the these variables had any substantial impact on the clustering of samples in the PCA, we colored the samples by herd, breed and parity in the PCA plot for the bacterial genus level, and found no evidence of such associations (FigS3).

In addition, to investigate the distribution of herd, breed and parity over the control and UCD groups we performed Fisher’s exact test and found that these factors did not differ significantly between the groups (this information has been updated in the manuscript). As we wanted to compare the bacterial abundance between groups and as the number of observations were limited (and the bacteria far from normally distributed within samples), we believe that a multivariable analysis is less useful for this dataset. As a spot check, we did test a mixed effect ordinal logistic regression model with bacterial abundance as an ordinal outcome for analyzing some of the most abundant species, with herd as random factor, and found that the estimates (and significance levels) for UCD status was largely unaffected when these potential confounders were included (although herd was significantly associated with e.g. Bifidobacterium abundance).

- In the "Bacterial abundance" section, the authors discuss several of the taxa that are driving these ordinations. Can the authors add biplot lines for these driver taxa to the PCA plots?

In our opinion the figure got too dense with biplot lines, but we have added a supplemental figure with loadings for the PC axis used (S2 Fig).

The control samples did not show a large variability for the taxa which were the main drivers for the displayed PC axes. However, the control samples do spread out in projections with higher PCs. We do not have the Bray Curtis distance option in the PCA tool used. We have tried to plot a Bray Curtis based ordination in R. The control samples cluster in there as well, although not fully as tight as in the PCA, and UCD samples cluster in roughly similar patterns as in Fig 3C.

Figure 4:

- The authors frequently describe different UCD subgroups. To complement these descriptions, can the authors utilize a hierarchical clustering based method to more robustly define them?

We have added a hierarchical clustering analysis to the supplement (S4 Fig).

Minor comments:

- In Fig 2A, the red/green coloring of Archaea and Fungi will be difficult for colorblind individuals to distinguish. Can the authors please update?

We have changed the colors.

- To make Fig 2B more informative, can the authors use individual scales for the different kingdoms and denote p-values for those comparisons that are significant?

We have made an enlargement for the minor taxa and added the p-values.

Attachment

Submitted filename: Response_to_Reviewers.docx

Click here for additional data file.^{(26.8KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0242880.r003

Decision Letter 1

Peter Gyarmati

30 Oct 2020

PONE-D-20-18567R1

A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

PLOS ONE

Dear Dr. Ekman,

Please submit your revised manuscript by Dec 14 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Peter Gyarmati

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Reviewer #1: Thank you for responding to our comments in the previous round. There are just a few minor issues noted on most recent version, which would improve the readability and interpretability of the manuscript:

Line 144: Add catalog numbers for this product and others where applicable.

Line 161: Change “Covairs E220” to “Covaris E220” and add manufacturer information as you do elsewhere.

Lines 181-182: Add in-text citation for Trimmomatic and in reference list. Change trimmomatics to “Trimmomatic” or “trimmomatic”.

Line 252: Remove sample identifiers from PCA plots in Figure 3 and elsewhere. This will make it easier for your readers to see clustering by groups.

Line 412 and 428: Provide a more detailed description of these visualizations, including what they are meant to convey, axes descriptions (e.g., D = Domain, K = Kingdom, etc..), and label descriptions (e.g., integer labels above each taxonomic unit denote the number of reads classified). Consider replacing integer labels above each taxonomic unit with relative abundances, as I think this would do a better job of highlighting changes in the microbiome between each time point.

Line 544: Specify the kind of contamination you are referring to (e.g., DNA).

Reviewer #2: I thank the authors for taking the time to address my concerns, particularly those around the analysis approach and the PCA plots. I do have several additional minor comments

- Fig S2: There should be a legend for the point colors.

- Fig S3: The axes should be labeled.

- Bioinformatics analyses section:

o trimmomatics should be Trimmomatic and appropriately cited

o Parameters used for Trimmomatic, Kraken, Bracken should be specified.

- Reference should be given for mastitis-causing pathogens

- Mastitis-causing bacteria and spirochetes in UCD samples section: Epidermidis should not be capitalized.

- Fig 6B: Boxplots similar to Fig2B would allow easier visual comparison between groups.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

PLoS One. 2020 Dec 2;15(12):e0242880. doi: 10.1371/journal.pone.0242880.r004

Author response to Decision Letter 1

9 Nov 2020

Line 144: Add catalog numbers for this product and others where applicable.

We have added this information where available.

Line 161: Change “Covairs E220” to “Covaris E220” and add manufacturer information as you do elsewhere.

We have corrected spelling and added the manufacturer information.

Lines 181-182: Add in-text citation for Trimmomatic and in reference list. Change trimmomatics to “Trimmomatic” or “trimmomatic”.

We have corrected the text and added a reference.

Line 252: Remove sample identifiers from PCA plots in Figure 3 and elsewhere. This will make it easier for your readers to see clustering by groups.

We have removed the sample identifiers from the PCA plots in Fig 3, 6C and S2.

We have added a further description of Fig 7 & 8. We did consider changing the integer labels to relative abundance as we agree that that would be more of interest here. However, this was not possible within the Pavian visualization tool, and as we believe that the label text is on the verge of too small, we decided to remove the integer labels from the flow diagrams.

Line 544: Specify the kind of contamination you are referring to (e.g., DNA).

We have specified the type of contamination.

Reviewer #2: I thank the authors for taking the time to address my concerns, particularly those around the analysis approach and the PCA plots. I do have several additional minor comments

- Fig S2: There should be a legend for the point colors.

- Fig S3: The axes should be labeled.

We have added this information in Fig S2 and Fig S3.

- Bioinformatics analyses section:

o trimmomatics should be Trimmomatic and appropriately cited

We have corrected this information.

o Parameters used for Trimmomatic, Kraken, Bracken should be specified.

We have included the parameters used for these tools.

- Reference should be given for mastitis-causing pathogens

We have added this information.

- Mastitis-causing bacteria and spirochetes in UCD samples section: Epidermidis should not be capitalized.

We have corrected the text here.

- Fig 6B: Boxplots similar to Fig2B would allow easier visual comparison between groups.

As the proportion of archaeal reads differed a lot between samples, with a generally lower abundance in the UCD samples (especially the severe samples) compared to the controls we believe that box plots might lead to over-interpretation of differences between sample types that in reality might be more related to the amount, and thus we would like to keep Fig 6 in its current, more descriptive, form.

PLoS One. doi: 10.1371/journal.pone.0242880.r005

Decision Letter 2

Peter Gyarmati

11 Nov 2020

A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

PONE-D-20-18567R2

Dear Dr. Ekman,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Peter Gyarmati

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

PLoS One. doi: 10.1371/journal.pone.0242880.r006

Acceptance letter

Peter Gyarmati

19 Nov 2020

PONE-D-20-18567R2

A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

Dear Dr. Ekman:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Peter Gyarmati

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Sample overview.

(XLSX)

Click here for additional data file.^{(28.1KB, xlsx)}

S2 Table. Raw classification data.

(XLSX)

Click here for additional data file.^{(11.8MB, xlsx)}

S1 Fig. PCA loadings.

(TIF)

Click here for additional data file.^{(167.8KB, tif)}

S2 Fig. PCA axis 4 on bacterial genus level.

(TIF)

Click here for additional data file.^{(251.8KB, tif)}

S3 Fig. PCA on bacterial genus level colored by herd, breed, parity and extraction month.

(TIF)

Click here for additional data file.^{(190.3KB, tif)}

S4 Fig. Hierarchical clustering of samples.

(TIF)

Click here for additional data file.^{(541.4KB, tif)}

Attachment

Submitted filename: Response_to_Reviewers.docx

Click here for additional data file.^{(26.8KB, docx)}

Data Availability Statement

[pone.0242880.ref001] 1.Boyer P, Singleton G. Digital dermatitis, superfoul and severe necrotic enteritis of the udder in dairy cows. Vet Rec. 1998;142(6):147–8. [PubMed] [Google Scholar]

[pone.0242880.ref002] 2.Warnick LD, Nydam D, Maciel A, Guard CL, Wade SE. Udder cleft dermatitis and sarcoptic mange in a dairy herd. J Am Vet Med Assoc. 2002;221(2):273–6. 10.2460/javma.2002.221.273 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref003] 3.Persson Waller K. Nekrotisk dermatit framtill på juvret hos mjölkkor [In Swedish]. Svensk Veterinärtidning. 2003;55(8–9):11–6. Available from: https://www.sva.se/media/dwgds2x1/kpw_nekrotisk_dermatit_svt_2003.pdf [Google Scholar]

[pone.0242880.ref004] 4.Persson Waller K, Bengtsson M, Nyman AK. Prevalence and risk factors for udder cleft dermatitis in dairy cattle. J Dairy Sci. 2014;97(1):310–8. 10.3168/jds.2013-7186 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref005] 5.Hansen MJ, Nissen MM. A Clinical Study of Udder Cleft Dermatitis on a Danish Dairy Farm. M. Sc. Thesis, University of Copenhagen. 2010. [Google Scholar]

[pone.0242880.ref006] 6.Riekerink R, van Amersfort K, Sampimon OC, Hooijer GA, Lam T. Short communication: Prevalence, risk factors, and a field scoring system for udder cleft dermatitis in Dutch dairy herds. J Dairy Sci. 2014;97(8):5007–11. 10.3168/jds.2013-7651 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref007] 7.Eikeland J, Paulsen H. Studie av jurfestesår hos storfe. M. Sc. Thesis, Norwegian University of Life Sciences. 2018. [Google Scholar]

[pone.0242880.ref008] 8.Ekman L, Nyman AK, Landin H, Magnusson U, Waller KP. Mild and severe udder cleft dermatitis-Prevalence and risk factors in Swedish dairy herds. Journal of dairy science. 2018;101(1):556–71. 10.3168/jds.2017-13133 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref009] 9.Bouma A, Nielen M, van Soest E, Sietsma S, van den Broek J, Dijkstra T, et al. Longitudinal study of udder cleft dermatitis in 5 Dutch dairy cattle herds. J Dairy Sci. 2016;99:1–9. 10.3168/jds.2015-9445 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref010] 10.Allenstein LC. Mites cause many of the smelly udder sores. Hoard's Dairyman. 1991;136:507. [Google Scholar]

[pone.0242880.ref011] 11.Stamm LV, Walker RL, Read DH. Genetic diversity of bovine ulcerative mammary dermatitis-associated Treponema. Vet Microbiol. 2009;136(1–2):192–6. 10.1016/j.vetmic.2008.10.022 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref012] 12.Lyman D. BHV-4 Contributes to Udder Lesions. Bovine Vet. 2019. January 4 [Cited 2020 Mar 23]. Available from: https://www.bovinevetonline.com/article/bhv-4-contributes-udder-lesions [Google Scholar]

[pone.0242880.ref013] 13.Beattie KG, Taylor DJ. An investigation into intertrigo (necrotic dermatitis or 'foul udder') in dairy cows. Cattle Pract. 2000;8(4):377–80. [Google Scholar]

[pone.0242880.ref014] 14.Ekman L. Udder cleft dermatitis in dairy cows. Doctoral Thesis, Swedish University of Agricultural Sciences. 2020. Available from: https://pub.epsilon.slu.se/16894/ 10.3168/jds.2019-17935 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref015] 15.Evans NJ, Timofte D, Carter SD, Brown JM, Scholey R, Read DH, et al. Association of treponemes with bovine ulcerative mammary dermatitis. Veterinary Record. 2010;166(17):532–3. 10.1136/vr.b4822 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref016] 16.Morey M, Fernandez-Marmiesse A, Castineiras D, Fraga JM, Couce ML, Cocho JA. A glimpse into past, present, and future DNA sequencing. Molecular genetics and metabolism. 2013;110(1–2):3–24. 10.1016/j.ymgme.2013.04.024 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref017] 17.Sorge US, Binger EM, Schefers J, Plummer PJ. Short communication: Metagenomic evaluation of skin biopsies of udder sores in dairy cows. Journal of dairy science. 2019. 10.3168/jds.2018-15863 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref018] 18.Petrosino JF, Highlander S, Luna RA, Gibbs RA, Versalovic J. Metagenomic pyrosequencing and microbial identification. Clinical chemistry. 2009;55(5):856–66. 10.1373/clinchem.2008.107565 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref019] 19.Ranjan R, Rani A, Metwally A, McGee HS, Perkins DL. Analysis of the microbiome: Advantages of whole genome shotgun versus 16S amplicon sequencing. Biochemical and biophysical research communications. 2016;469(4):967–77. 10.1016/j.bbrc.2015.12.083 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref020] 20.Van Rossum T, Ferretti P, Maistrenko OM, Bork P. Diversity within species: interpreting strains in microbiomes. Nature reviews Microbiology. 2020;18(9):491–506. 10.1038/s41579-020-0368-1 . [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref021] 21.Jovel J, Patterson J, Wang W, Hotte N, O'Keefe S, Mitchel T, et al. Characterization of the gut microbiome using 16S or shotgun metagenomics. Frontiers in microbiology. 2016;7:459 10.3389/fmicb.2016.00459 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref022] 22.Neves ALA, Li F, Ghoshal B, McAllister T, Guan LL. Enhancing the resolution of rumen microbial classification from metatranscriptomic data using kraken and mothur. Frontiers in microbiology. 2017;8:2445 10.3389/fmicb.2017.02445 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref023] 23.Byrd AL, Deming C, Cassidy SKB, Harrison OJ, Ng WI, Conlan S, et al. Staphylococcus aureus and Staphylococcus epidermidis strain diversity underlying pediatric atopic dermatitis. Science translational medicine. 2017;9(397). 10.1126/scitranslmed.aal4651 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref024] 24.Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA, et al. Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC microbiology. 2008;8:43 10.1186/1471-2180-8-43 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref025] 25.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics (Oxford, England). 2014;30(15):2114–20. 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref026] 26.Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome biology. 2009;10(3):R25 10.1186/gb-2009-10-3-r25 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref027] 27.Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics (Oxford, England). 2009;25(16):2078–9. 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref028] 28.Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome biology. 2019;20(1):257 10.1186/s13059-019-1891-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref029] 29.Lu J, Breitwieser FP, Thielen P, Salzberg SL. 2017. Bracken: estimating species abundance in metagenomics data. PeerJ Computer Science 3:e104 10.7717/peerj-cs.104 [DOI] [Google Scholar]

[pone.0242880.ref030] 30.Breitwieser FP, Salzberg SL. Pavian: interactive analysis of metagenomics data for microbiome studies and pathogen identification. Bioinformatics (Oxford, England). 2020;36(4):1303–4. 10.1093/bioinformatics/btz715 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref031] 31.Sverige Växa. Redogörelse för husdjursorganisationens djurhälsovård 2018/2019 [Report on animal health care within the livestock organisation 2018/2019]. 2020. [cited 2020 Nov 6]. [pp 22–24]. Available from: https://www.vxa.se/globalassets/dokument/statistik/redogorelse-for-husdjursorganisationernas-djurhalsovard-2018-2019.pdf [Google Scholar]

[pone.0242880.ref032] 32.Krull AC, Shearer JK, Gorden PJ, Cooper VL, Phillips GJ, Plummer PJ. Deep sequencing analysis reveals temporal microbiota changes associated with development of bovine digital dermatitis. Infection and Immunity. 2014;82(8):3359–73. 10.1128/IAI.02077-14 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref033] 33.Egert M, Simmering R, Riedel CU. The association of the skin microbiota with health, immunity, and disease. Clinical pharmacology and therapeutics. 2017;102(1):62–9. 10.1002/cpt.698 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref034] 34.Structure, function and diversity of the healthy human microbiome. Nature. 2012;486(7402):207–14. 10.1038/nature11234 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref035] 35.Bradley CW, Morris DO, Rankin SC, Cain CL, Misic AM, Houser T, et al. Longitudinal evaluation of the skin microbiome and association with microenvironment and treatment in canine atopic dermatitis. The Journal of investigative dermatology. 2016;136(6):1182–90. 10.1016/j.jid.2016.01.023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref036] 36.Williams MR, Gallo RL. The role of the skin microbiome in atopic dermatitis. Current allergy and asthma reports. 2015;15(11):65 10.1007/s11882-015-0567-4 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref037] 37.Gardiner M, Vicaretti M, Sparks J, Bansal S, Bush S, Liu M, et al. A longitudinal study of the diabetic skin and wound microbiome. PeerJ. 2017;5:e3543 10.7717/peerj.3543 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref038] 38.Braem G, De Vliegher S, Verbist B, Heyndrickx M, Leroy F, De Vuyst L. Culture-independent exploration of the teat apex microbiota of dairy cows reveals a wide bacterial species diversity. Vet Microbiol. 2012;157(3–4):383–90. 10.1016/j.vetmic.2011.12.031 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref039] 39.Dahlberg J. Bovine milk microbiota—methods matter. Doctoral Thesis, Swedish University of Agricultural Sciences. 2019. Available from: https://pub.epsilon.slu.se/16433/ [Google Scholar]

[pone.0242880.ref040] 40.Wu YK, Cheng NC, Cheng CM. Biofilms in chronic wounds: pathogenesis and diagnosis. Trends in biotechnology. 2019;37(5):505–17. 10.1016/j.tibtech.2018.10.011 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref041] 41.Coyle MB, Lipsky BA. Coryneform bacteria in infectious diseases: clinical and laboratory aspects. Clin Microbiol Rev. 1990;3(3):227–46. 10.1128/cmr.3.3.227 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref042] 42.Magi B, Migliorini L, Sansoni A, Cusi MG. Brevibacterium casei bacteraemia in a port-a-cath carrier patient: a case report. Infez Med. 2018;26(3):263–5. . [PubMed] [Google Scholar]

[pone.0242880.ref043] 43.Nagaraja TG, Narayanan SK, Stewart GC, Chengappa MM. Fusobacterium necrophorum infections in animals: pathogenesis and pathogenic mechanisms. Anaerobe. 2005;11(4):239–46. 10.1016/j.anaerobe.2005.01.007 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref044] 44.Ribeiro MG, Risseti RM, Bolanos CA, Caffaro KA, de Morais AC, Lara GH, et al. Trueperella pyogenes multispecies infections in domestic animals: a retrospective study of 144 cases (2002 to 2012). Vet Q. 2015;35(2):82–7. 10.1080/01652176.2015.1022667 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref045] 45.Lockhart JS, Buret AG, Ceri H, Storey DG, Anderson SJ, Morck DW. Mixed species biofilms of Fusobacterium necrophorum and Porphyromonas levii impair the oxidative response of bovine neutrophils in vitro. Anaerobe. 2017;47:157–64. 10.1016/j.anaerobe.2017.05.008 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref046] 46.Lindholm C. Sår [In Swedish]. 3rd ed Lund: Studentlitteratur AB; 2012. [Google Scholar]

[pone.0242880.ref047] 47.Musthaq S, Mazuy A, Jakus J. The microbiome in dermatology. Clinics in dermatology. 2018;36(3):390–8. 10.1016/j.clindermatol.2018.03.012 [DOI] [PubMed] [Google Scholar]

[pone.0242880.ref048] 48.Janssen PH, Kirs M. Structure of the archaeal community of the rumen. Appl Environ Microbiol. 2008;74(12):3619–25. 10.1128/AEM.02812-07 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref049] 49.Probst AJ, Auerbach AK, Moissl-Eichinger C. Archaea on human skin. PloS one. 2013;8(6):e65388 10.1371/journal.pone.0065388 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref050] 50.Meisel JS, Hannigan GD, Tyldsley AS, SanMiguel AJ, Hodkinson BP, Zheng Q, et al. Skin Microbiome surveys are strongly influenced by experimental design. the journal of investigative dermatology. 2016;136(5):947–56. 10.1016/j.jid.2016.01.016 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0242880.ref051] 51.Salter SJ, Cox MJ, Turek EM, Calus ST, Cookson WO, Moffatt MF, et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014;12:87 10.1186/s12915-014-0087-z [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

A shotgun metagenomic investigation of the microbiota of udder cleft dermatitis in comparison to healthy skin in dairy cows

Lisa Ekman

Elisabeth Bagge

Ann Nyman

Karin Persson Waller

Märit Pringle

Bo Segerman

Roles

Abstract

Introduction

Material and methods

Study design and participating cows

Fig 1. Illustration of sampling site.

Sampling procedure

Sampling analyses

DNA extraction

DNA sequencing

Bioinformatic analyses

Statistical analyses

Results

Overall microbial abundance

Fig 2. Microbial abundance.

Bacteria

Fig 3. Unsupervised analysis.

Bacterial abundance

Fig 4. Bacterial abundance.

Table 1. Comparison of UCD samples and healthy skin.

Bacterial diversity

Fig 5. Bacterial diversity.

Mastitis-causing bacteria and spirochetes in UCD samples

Other microorganisms

Archaea

Fig 6. Archaeal abundance.

Fungi

Differences in microbiota between different time points–two examples

Fig 7. Example 1.

Fig 8. Example 2.

Discussion

Bacterial abundance and subgroups of UCD samples

Presence of mastitis-causing bacteria

Abundance of other microorganisms

Final remarks

Conclusions

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Peter Gyarmati

Roles

Author response to Decision Letter 0

Decision Letter 1

Peter Gyarmati

Roles

Author response to Decision Letter 1

Decision Letter 2

Peter Gyarmati

Roles

Acceptance letter

Peter Gyarmati

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases