Heel skin microbiomes from cattle housed on dairies endemic for and free from clinical digital dermatitis

Introduction
Digital dermatitis (DD) is the most common infectious cause of lameness in dairy cattle and the etiologic agents are not completely understood. One major challenge in identifying etiologic agents is that the vast majority of herds are affected therefore harbor the causative agents. The aim of this study was to compare the heel skin microbiome of cattle from a dairy herd free of clinical DD with those from endemic herds. Our hypothesis was that heel skin from the dairy free of clinical DD was characterized by Treponema spp. that are different to those associated with preclinical skin lesions from dairies that have clinical DD. 

Materials and Methods
All experimental procedures were conducted in accordance with guidelines of the Canadian Council for Animal Care and approved by the University of Saskatchewan Animal Care Committee (project number: 20140031). Dairies willing to participate were identified through a survey distributed by SaskMilk. Seven dairies were identified. One was free of clinical DD. All were in the same geographical area, milked predominately Holsteins and were managed under intensive, commercial systems. Up to ten cows from each endemic dairy with early DD lesions were selected for heel skin biopsy. Animals on the clinical DD-free dairy were randomly selected. Bacterial gDNA was extracted and sequenced using 16S rRNA metagenomics. 

Results
During sampling we noted that several animals from the DD-free dairy had heel skin changes consistent with early DD lesions. On review of hoof trimmer and veterinary records, no lesions progressed to clinical disease. Follow-up examinations over 12 months supported these observations. On the phylum level, most tissue samples were dominated by spirochaetae. On the species level (ID based on closest BLAST match homology to OTU sequence), tissue samples from the endemic dairies were dominated by a combination of Treponema phagedenis (35.6% of all endemic OTU sequences, mean: 11,699 reads/sample), Treponema sp44 (20.8%, mean: 6835 reads/sample), Treponema PT3 (13.3%, mean: 4378 reads/sample), and Treponema PT2 (12.4%, avg. 4093 reads/sample). Tissue samples from the DD-free dairy were dominated by Treponema PT2 (62.7% of all DD-free OTU sequences, avg. 15,561 reads/sample), T. refringens (20.5%, avg. 5,102 reads/sample), and T. sp44 (16.1%, avg. 3,991 reads/sample). Treponema phagedenis, T. medium, T. denticola, T. sp22, and T. pedis made up less than 10 total reads per sample from the DD-free dairy. 

Conclusions
Our results provide strong evidence that a dairy herd can have exclusively early stage lesions of DD that consistently fail to develop into clinical DD lesions. 16-S based phylogenetic evaluation of these lesions demonstrates that the overall composition of the Treponema spp. community differs between farms in which the lesions progress and those where they fail to progress. On farms in which progression occurs the lesions contained higher relative abundance of Treponema phagedenis, Treponema PT1 and Treponema PT3. These outcomes provide a suggestion that the presence of some or all of these species may be necessary for development of clinical digital dermatitis and additional research in this area is warranted. Treponeme communities of these early lesions were consistent with those present in our previous studies of early lesions, demonstrating that the lesion scoring provides consistent outcomes in multiple farms and geographic areas.