Adherence and metal-ion acquisition gene expression increases during infection with Treponema phagedenis strains from bovine digital dermatitis

Abstract

Digital dermatitis (DD) is an ulcerative foot lesion on the heel bulbs of dairy cattle. DD is a polymicrobial disease with no precise etiology, although Treponema spirochetes are found disproportionally abundant in diseased tissue. Within Treponema, several different species are found in DD; however, the species Treponema phagedenis is uniformly found in copious quantities and deep within the skin layers of the active, ulcerative stages of disease. The pathogenic mechanisms these bacteria use to persist in the skin and the precise role they play in the pathology of DD are widely unknown. To explore the pathogenesis and virulence of Treponema phagedenis, newly isolated strains of this species were investigated in a subcutaneous murine abscess model. In the first trial, a dosage study was conducted to compare the pathogenicity of different strains across three different treponemes per inoculum (TPI) doses based on abscess volumes. In the second trial, the expression levels of 11 putative virulence genes were obtained to gain insight into their involvement in pathogenesis. During the RT-qPCR analysis, it was determined that genes encoding for two metal-ion import lipoproteins and two adherence genes were found highly upregulated during infection. Conversely, two genes involved in motility and chemotaxis were found to not be significantly upregulated or utilized during infection. These results were supported by gene expression data from natural M2 lesions of dairy cattle. This gene expression analysis could highlight the preference in strategy for T. phagedenis to persist and adhere in the host rather than engage in motility and disseminate.

Full text

Veterinary Microbiology | Full-Length Text
Adherence and metal-ion acquisition gene expression increases
during infection with Treponema phagedenis strains from bovine
digitaldermatitis
Colton Scott,1 Angelica P. Dias,1 Jeroen De Buck1
AUTHOR AFFILIATION See aliation list on p. 16.
ABSTRACT Digital dermatitis (DD) is an ulcerative foot lesion on the heel bulbs of
dairy cattle. DD is a polymicrobial disease with no precise etiology, although Treponema
spirochetes are found disproportionally abundant in diseased tissue. Within Treponema,
several dierent species are found in DD; however, the species Treponema phagedenis
is uniformly found in copious quantities and deep within the skin layers of the active,
ulcerative stages of disease. The pathogenic mechanisms these bacteria use to persist in
the skin and the precise role they play in the pathology of DD are widely unknown. To
explore the pathogenesis and virulence of Treponema phagedenis, newly isolated strains
of this species were investigated in a subcutaneous murine abscess model. In the rst
trial, a dosage study was conducted to compare the pathogenicity of dierent strains
across three dierent treponemes per inoculum (TPI) doses based on abscess volumes.
In the second trial, the expression levels of 11 putative virulence genes were obtained
to gain insight into their involvement in pathogenesis. During the RT-qPCR analysis, it
was determined that genes encoding for two metal-ion import lipoproteins and two
adherence genes were found highly upregulated during infection. Conversely, two genes
involved in motility and chemotaxis were found to not be signicantly upregulated or
utilized during infection. These results were supported by gene expression data from
natural M2 lesions of dairy cattle. This gene expression analysis could highlight the
preference in strategy for T. phagedenis to persist and adhere in the host rather than
engage in motility and disseminate.
KEYWORDS digital dermatitis, Treponema phagedenis, murine model, bovine, dairy,
lameness
Digital dermatitis (DD) is an ulcerative foot lesion in cattle that develops on the
plantar surface of the skin surrounding the heel bulbs (1). This disease is character
ized by dierent morphological presentations and stages, primarily existing as either
acute, ulcerative granuloma-like lesions or as chronic, hyperkeratotic skin disease (1).
While DD has been reported in beef cattle (2), it predominantly aects dairy cattle, with
the incidence of DD on dairy farms growing worldwide with some regions, including
Canada, being endemic for the disease (3, 4). DD is the leading cause of infectious
lameness and is a principal concern regarding the welfare of animals but it also instigates
signicant economic loss in the form of decreased milk yield, poor reproduction rates,
increased veterinary costs, and culling on dairy farms (5).
Given the prevalence of DD, many studies have investigated the complexity of the DD
microbiome and microbiota to elucidate key etiological members of the disease (6–9).
However, from this work, no precise causative agent of DD has yet been described, and
this has led many researchers to conclude that the disease is polymicrobial in nature
(6, 7). From this work, a group of genera emerges that are disproportionally found in
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 1
Editor Andreas J. Bäumler, University of California,
Davis, Davis, California, USA
Address correspondence to Jeroen De Buck,
jdebuck@ucalgary.ca.
The authors declare no conict of interest.
See the funding table on p. 16.
Received 13 March 2024
Accepted 16 May 2024
Published 28 June 2024
Copyright © 2024 American Society for
Microbiology. All Rights Reserved.
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

diseased tissue, notable members include Treponema, Porphyromonas, Fusobacterium,
and Mycoplasma (6–9). The one genus that is consistently and signicantly found to be
the most abundant is Treponema.
While multiple treponeme phylotypes are found in DD, Treponema phagedenis is the
one treponeme that emerges as the most prevalent species within the active stages
of the disease (6–13). This observation is similarly supported by in situ hybridization
showing the T. phagedenis phylotype is highly invasive within DD lesions, compared to
other phylotypes like T. denticola and T. refringens that were located more supercially
(14, 15). Despite its dominance in lesions, there is a scarcity of literature regarding the
virulence factors of T. phagedenis and how this pathogen could persist, proliferate, and
survive in DD lesions.
To study the pathogenesis of DD microbes, animal models have been developed.
While a bovine calf model exists, reproducing the disease articially in cattle requires the
use of homogenized lesions as an inoculum and cannot be replicated successfully using
strictly pure treponemal cultures for producing lesions (16). Alternatively, a subcutane
ous murine model for DD, originally developed for periodontal disease by Kesavalu
et al. (17) has been employed (17). Two recent studies have been done investigating
the virulence of DD treponemes in this current murine abscess model. In Elliot et al.
(18), T. phagedenis strains isolated from DD were infected into mice, and dierences in
abscess volumes were found between the dierent strains (18). In Arrazuria et al. (19),
combinations of bovine treponemes (T. medium, T. pedis, and T. phagedenis) inoculums
were generated to test the potential synergistic eect of polytreponemal infections (19).
Our objective was to investigate the pathogenicity of several newly isolated T.
phagedenis strains isolated from DD lesions in this murine model and further characterize
the virulence of these pathogens through elucidating the expression levels of selected,
putative virulence genes during infection.
MATERIALS AND METHODS
Biopsy material sourcing for strain isolation
Bovine skin biopsy material was harvested from three Alberta dairy farms recently
described in Dias & De Buck., 2022 (12). Additional biopsy material was originally
collected by Beninger et al. (11) on dairy farms (11) and Caddey et al. (10) from fee
dlot cattle (10) and these samples were revived from −80°C freezer storage. All biopsy
materials belonging to the lesion stages M2 and M4.1 were selected for isolation. The
biopsy pieces were incubated into Oral Treponemal Enrichment Broth (OTEB, Anaerobic
Systems, Morgan Hill, California, USA) supplemented with 10% heat-inactivated fetal
bovine serum (FBS, Avantor, Pennsylvania, USA) and 10% rabbit serum (RS, Gibco,
Waltham, Massachusetts, USA). The broth was also supplemented with the antibiot
ics rifampicin and enrooxacin, both at a concentration of 10 µg/mL. In addition to
culturing, freshly harvested biopsies belonging to M2 were collected for RNA extraction
[detailed in section “RNA extractions (abscesses, cell pellets, and biopsies)”] and RT-qPCR
analysis (detailed in section “Dierential gene expression analysis of virulence genes
during murine infection”). Biopsy processing and culturing occurred in an anaero
bic cabinet (Bactron3000 Sheldon Manufacturing Inc, Cornelius, Oregon, USA), under
anaerobic conditions, with a gas composition mixture of 5% CO2, 5% H2, and 90% N2.
Treponema isolation and purication check
Two methodologies and protocols were employed in parallel to isolate Treponema strains
from these biopsy cultures. All isolation attempts occurred at 24 h after the initial OTEB
culturing of the biopsy pieces.
In the rst methodology based on Evans et al. 2008 (20), 200 µL of OTEB culture was
streaked onto Fastidious Anaerobic Agar (FAA, Neogen Corporation, Lansing, Michigan,
USA) plates. After 2 weeks of growth, treponemal-like colonies (sub-surface, small, hazy
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 2
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

cotton-ball morphology) were picked and incubated in a 1 mL OTEB vial, after incubating
for 4–10 days an aliquot of this culture was re-streaked on fresh FAA plates for more
colonies. This sub-culture process continued 2–3 times until the culture was pure, as
evaluated using dark-eld microscopy of the culture.
In the second methodology based on Brodard et al., 2021 (21), 200 µL of OTEB culture
was spotted onto 0.2 µm nitrocellulose lters (MF-Millipore, Oakville, Ontario, Canada)
resting on top of MTGE agar plates (Anaerobic Systems). After 6 h, the lter papers were
carefully and aseptically removed, and then the plates were incubated upside down for
10 days. Colonies appearing under the lter paper spots that resembled the morphology
of treponemal colonies were picked, grown, and checked for purity as described above
for the previous protocol.
Isolates obtained were tested for purity (devoid of non-treponeme contaminants and
single-species) To test for non-treponemes, isolates were grown in supplemented OTEB
for 4 days then streaked onto the non-treponemal Trypticase Soy Agar plates with 5%
sheep blood (TSA, Becton Dickinson and Company, Franklin Lakes, New Jersey USA).
Plates were incubated for 5 days in anaerobic and aerobic conditions. The presence
of CFUs after 5 days indicated impure, contaminated samples. Only samples that did
not yield CFUs were used further. A second purity check to verify isolates were single
species was done by employing a species-specic, multiplex qPCR assay as designed and
described in Beninger et al. (11). Details regarding strain isolation and location are found
in Table 1.
Biopsy material sourcing for gene expression analysis
Six Holstein dairy cows at a commercial farm in Alberta, Canada, underwent tissue biopsy
collection from M2 lesions. Biopsy harvesting procedure was reviewed and approved
by the University of Calgary Veterinary Sciences Animal Care Committee (AC21-0146).
All lactating cows were rst screened for M2 lesions in the parlor, and the lesions were
then conrmed in the trimming chute. Once the cow was restrained in the chute, the
lesion was cleaned with paper towels to remove manure and conrm the active stage
of the lesion. Local anesthesia was administered using 2 mL of lidocaine subcutaneously
(Lidocaine HCl 2%, Zoetis Inc., Kirkland, QC) prior to sampling with a disposable 4 mm
biopsy punch (Integra LifeSciences, Princeton, NJ) from the center of the lesion. All
samples were immediately placed into a 2 mL Eppendorf tube containing 1 mL ice-cold
Trizol (Thermo Scientic) and transported to the laboratory on ice within 2 h after
sampling. Tissue biopsies were briey broken with a pipette tip and then homogenized
using THb Handheld Tissue Homogenizer (Omni International). Homogenized samples
were then stored at −80°C until RNA extraction [detailed in the section “RNA extractions
(abscesses, cell pellets, and biopsies)”].
Inoculum culturing
To prepare the inoculum for infection, 1 mL of glycerol stocks of T. phagedenis isolates
were added into 4 mL New Oral Spirochete (NOS) media (ATCC Medium 1357) supple
mented with 10% FBS & RS, as well as 5 µg/mL rifampicin and enrooxacin. After 4 days
TABLE 1 Treponema phagedenis strains isolated from DD lesions for this study
Strain ID DD lesion stage Cattle type Isolation protocol Biopsy status
11 M2 Dairy (20) Frozen
4457 M2 Beef (20) Frozen
9 M4.1 Dairy (20) Frozen
EB 1900 M4.1 Dairy (21) Fresh
2947 M2 Dairy (21) Fresh
102 M4.1 Dairy (20) Frozen
3133 M2 Dairy (21) Fresh
138-2 M4.1 Dairy (20) Frozen
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 3
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

of culture, an aliquot was extracted to determine OD620 of the culture and the quality of
the culture using darkeld microscopy (40×). Next, 1 mL of this culture was transferred to
4 volumes of freshly supplemented NOS where it was grown and evaluated as above for
another 4 days. This subculturing and evaluation continued while maintaining a uniform
1:5 ratio of old culture to fresh, supplemented NOS, and increasing the amount of culture
volume at each stage until the nal culture volume reached 1L.
A conversion factor of OD620 0.1 = 1.23E8 treponemes/mL obtained from Arrazuria et
al. (22) was used to calculate the culture volume required to reach target inoculum doses.
The appropriate culture volume was then pelleted. Cell pellets were resuspended gently
in 0.4 mL pre-reduced phosphate-buered saline (PBS) pH 7.4. The resuspended doses
were aspirated into syringes.
Formalin-killed (F-K) doses were made by resuspension of cell pellets in 5 mL
10% formalin (4% formaldehyde). Resuspensions were left on the benchtop at room
temperature for 4 days. After incubation, cells were pelleted as above, and the pellet was
washed three more times with 0.5 mL PBS. After the nal wash, the dose was aspirated
into the syringe.
Infection trial design
C57BL/6J (6–8 week old) female mice were purchased from Jackson Laboratories (Bar
Harbor, Maine, USA) and were acclimated to the facility for 1 week prior to infection.
Two infection trials were performed. In trial 1, the goal was to compare three dierent
treponemes per inoculum (TPI) doses at the following cell concentrations: 1.25E9, 5E9,
and 2E10 SPI, and then to investigate the eect these doses have on abscess size.
After trial 1, an appropriate dose of 5 × 109 TPI was selected to move forward with.
This second trial was a gene expression study, where abscesses were harvested and
compared against in vitro cell cultures to deduce the dierential expression levels of
putative virulence genes in in vivo versus in vitro circumstances.
In trial 1 (dosage study), each dose subgroup in the overarching experimental group
contained n = 6 mice for each strain. For eight strains, a total of 48 mice belonged to
each dose level of 3, to yield a total of n = 144 mice. Control infections consisted of
formalin-xed doses and PBS at the same power level of n = 6. In trial 2 (gene expression
study), each strain infection group had a size of n = 10. Across eight strains, this totals n
= 80 mice. Control infections were the same as above, except the group size was lowered
to n = 4.
Murine infections and necropsy
All injections and animal handlings were performed in a Bio-Safety cabinet where mice
were anesthetized with 5% isourane under a 1 L/min oxygen owrate. Mice were then
shaved on the nape of the neck and upper dorsal surface. The injection site was cleaned
with 70% ethanol, and mice were injected subcutaneously along the dorsal midline
between scapulae.
Mice were euthanized via cervical dislocation 7 days post-infection. The site of
injection was exposed during dissection, and the skin ap containing the abscess was
excised using scalpels and scissors. The skin ap was inverted and pinned to a dissection
tray where the abscess was photographed and the volume (mm3) of the abscess was
calculated by measuring the height, width, and length of the abscess using an electronic
caliper gauge.
RNA extractions (abscesses, cell pellets, and biopsies)
Lesions from murine trial 2 were excised from the skin using scalpels and tweezers.
The intact, whole abscess was placed into a 2 mL Eppendorf tube, followed by 1 mL
ice-cold TRIzol (Thermo Scientic). The abscess was briey broken up with a pipette
tip, followed by extensive homogenization using THb Handheld Tissue Homogenizer
(Omni International, Inc, Kennesaw, Georgia, USA) stabilized to a support stand. Tubes
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 4
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

were placed in −20°C ice block and were pulsed for 30–60 s on the lowest RPM set
ting. Homogenized samples were then kept at −80°C until RNA extraction. Similarly to
abscesses, M2 bovine lesions were homogenized as above.
In addition to abscess samples, in vitro samples (cell pellets) were prepared at the
same time as the inoculum preparation for trial 2 from the same 1 L ask. Briey, upon
calculating the dose volume for each strain, 10 culture tubes of identical volume were
pelleted as above. Cell pellets were resuspended in 1 mL ice-cold TRIzol. Resuspension
was transferred to a 2.0 mL Eppendorf tube and was frozen at −80°C until future use.
TRIzol samples from both in vivo and in vitro backgrounds were thawed, and RNA
was extracted via phase separation with 100 µL 1–3 bromochloropropane (BCP) added
to each tube. Samples were shaken vigorously for 15 s by hand. After mixing, samples
were rested for 3 min at room temperature to allow the phases to begin to separate,
this was followed by a centrifugation step of 12,000 × g for 15 min at 4°C. Upon phase
separation, the upper aqueous layer was carefully extracted to a new tube, where the
RNA was collected and puried on-column using RNeasy MinElute Cleanup Kits (Qiagen,
Hilden, Germany) according to the manufacturer’s instructions. The RNA concentration
and purity ratios (A260/230 & A260/280) were measured using a Nanophotometer NP80
UV/Vis spectrophotometer (Implen, California, USA).
Crude RNA was then further depleted of contaminating gDNA using a TURBO
DNA-free kit (Ambion Inc, Austin, Texas, USA). 2.5 µg of RNA was treated in the DNase
reaction according to the manufacturer’s instructions. The sample was then cleaned on a
minElute Spin Column (Qiagen), according to the manufacturer’s instructions.
Dierential gene expression analysis of virulence genes during murine
infection
DNase-treated RNA was used immediately to make cDNA using FastSCRIPT cDNA
synthesis kit (TONBO Biosciences, San Diego, California, USA). Two reactions were set
up for each sample, one served as a no-reverse transcriptase control (NRT), and the
other was a cDNA reaction using reverse transcriptase (RT). 7 µL of RNA was used
for each reaction. Reaction vials were placed into a T100 Thermocycler (Bio-Rad) and
underwent the following reaction steps: 25°C for 10 min, 46°C for 30 min, and 85°C
for 10 min. Reaction tubes were retrieved and the contents were 2× diluted with
nuclease-free water. RT-qPCR assay was performed on a CFX-96 real-time PCR machine
(Bio-Rad, Hercules, California, USA). Each reaction well contained 18 µL of master mix,
comprised of 10 µL of Taqman Fast Advanced Master Mix (Applied Biosystems Waltham,
Massachusetts, USA), 1 µL of 10 mM forward primer, reverse primer and probe, and 5 µL
nuclease-free water. 2 µL of template (either RT or NRT sample) was added to the wells
to bring the total PCR volume to 20 µL. RT-qPCR cycling conditions were 50°C for 2 min,
95°C 20 s, then 40 cycles of 95°C for 10 s and 60°C for 50 s. Three housekeeping genes
(dnaK, tkt, and tpiA) were initially tested on a few plates, ultimately tpiA was selected
as the most stable gene as assessed in RefFinder software (23) and was used for data
normalization. Primer and probe sequences are found in Table 1 (housekeeping genes)
and Table 2 (putative virulence genes) of the supplementary data. Gene expression fold
changes of the target virulence genes of interest were calculated based on the 2−ΔΔCt
relative method and results were reported as mean fold changes.
Putative virulence gene selection and description
Eleven putative virulence genes were selected for gene expression analysis. These 11
genes were initially selected from a pool of putative virulence genes identied through
RNA-seq data in Marcatelli et al. (51). A brief description of the potential virulence roles
these genes possess is given in Table 2, together with the sorting of genes into Clusters
of Orthologous Groups (COG).
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 5
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

Validation of primers and probes of putative virulence genes specicity for T.
phagedenis only
To ensure primers and probes of putative virulence genes are specic for T. phagedenis
but not for the other most abundant treponemes in M2 biopsies, genes were BLASTed
against whole genome sequence data from DD-isolates of T. medium and T. pedis.
Furthermore, qPCR was conducted with primers and probes used for gene expression
with gDNA from T. pedis and T. medium DD-isolates as templates. Real-time qPCR was
performed on a CFX-96 real-time PCR machine. Each reaction well contained 18 µL of
master mix, comprised of 10 µL of Taqman Fast Advanced Master Mix, 1 µL of 10 mM
forward primer, reverse primer and probe, and 5 µL nuclease-free water. 2 µL of template
(T. medium and T. pedis gDNA) was added to the wells to bring the total PCR volume to
20 µL. qPCR cycling conditions were the same as described above. T. phagedenis gDNA
was used as a positive control and non-template control was added as negative control.
Quantication of T. phagedenis from in vivo and in vitro samples
After the upper aqueous phase was removed for RNA extractions, 500 µL of back-extrac
tion buer (BEB) was added to the remaining TRIzol samples. Samples were then
mixed on a horizontal shaker at max speed for 10 min. After mixing, the samples were
centrifuged for 30 min at 12,000 × g. Upon centrifugation, the TRIzol samples were
phase separated once again, and the new upper aqueous layer was extracted and moved
to a new 1.5 mL Eppendorf tube. One volume of 100% isopropanol along with 1–2
µL 20 mg/mL molecular-grade glycogen (Thermo Scientic) was added to the sample.
After mixing, samples were transferred to a −80°C freezer for 1–2 days to improve DNA
precipitation. After this incubation, the samples were thawed and underwent centrifuga
tion for 30 min at 4°C at 12,000 × g to collect the DNA pellet. The DNA pellet was cleaned
with two 70% EtOH wash steps. The same species-specic qPCR designed by Beninger et
al., 2018 (11), was performed on the extracted gDNA samples. Unknown gDNA samples
were quantied into DNA copies/µL using a standard curve.
Whole-genome sequencing of strains
Genomic DNA from all eight isolates was extracted from a 10 mL NOS culture using
a DNeasy Blood and Tissue Kit (Qiagen, Toronto ON, Canada), according to the manufac
turer’s instructions. Upon extraction, DNA quality and quantity were assessed using a
NanoVue plus uorometer (Invitrogen, Burlington, ON, Canada). Each DNA sample was
then diluted to a nal concentration of 0.2 ng/µL, and sent to McMaster Genomics
Facility for sequencing.
TABLE 2 UNIPROT description of 11 putative virulence genes belonging to T. phagedenise
Gene Protein Suspected role in virulence COG family Study source
patB Cystathionine beta-lyase Production of H2S E 24a, 25b, 26a,b
hbpB Hemin-binding lipoprotein Sequestering of heme M 27a, 28c
aB Flagellar lament core protein Essential core protein in agella N 29a, 30b
adhE NADP,1-dependent alcohol dehydrogenase Resistance to oxidative stress E 31a,b
mglA Galactose/methyl galactose import ATP-binding protein Acquisition of the sugars glucose and galactose G 32d, 33d
cheA Chemotaxis protein A Kinase involved in chemotactic signaling N 34a, 35a, 36b, 37c
fbpA Fibronectin-binding protein A Fibronectin binding lipoprotein M 38a, 39a, 40a
msp Major sheath protein Pore-forming β-barrel M 41c, 42a, 43a, 44a
phoU Phosphate-specic transport associated protein Resistance to osmotic stress P 45a, 46a, 47b
troA Periplasmic zinc-binding protein Acquisition of Zn2+, Fe2+, and Mn2+ ions P 48a, 49b
oadB Oxaloacetate decarboxylase beta chain Additional means of ATP synthesis G 50c, 51c
aStudies using in vitro assays to test the virulence of gene.
bStudies using animal models to test the virulence of gene.
cStudies discussing the role of a gene in virulence w/out empirical testing or citing other studies.
dStudies hypothesizing the role of a gene in virulence.
eLetter code: E, amino acid transport/metabolism; P, inorganic ion transport/metabolism; G, carbohydrate transport/metabolism; N, cell motility; M, cell wall/mem
brane/envelope biogenesis.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 6
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

Tapestation (Agilent Technologies, Santa Clara, CA, USA) was used to estimate the
quantity of each library for pooling. DNA libraries for sequencing were then prepared
using a Nextera XT DNA library preparation kit (Illumina, San Diego, CA, USA). Sequenc
ing of DNA samples was performed using the Illumina MIseq platform (Illumina,
San Diego, CA, USA). All sequencing steps, including cluster generation, paired-end
sequencing (2 × 300 bp), and primary data analysis for quality control, were performed
on the instrument.
Sequence reads obtained from the MiSeq platform were checked for poorly
sequenced regions and Illumina adapter sequences were trimmed using Trimmomatic
(52). After trimming sequences, ltered reads were assembled into contigs using the de
novo assembly program Unicycler version 0.5.0 (53), employing built-in error correction
and default parameters (53). Genome annotations and gene predictions were performed
with Prokka 1.13 (54).
A genome-based phylogenetic tree of the isolates was constructed using the
published pipeline PhyloPhlAn version 3.0.67 (55) using MEGA v 11.0.13, with the T.
phagedenis B43.1T reference genome as the outgroup.
Detection of single nucleotide polymorphisms (SNPs) was performed in Geneious
10.2.6. by aligning virulence gene ORFs from the isolates against the same genes from
the type strain B43.1T.
Statistical analysis
The normality of abscess sizes was assessed using the Shapiro-Wilks test. One-way
ANOVA was conducted to assess the statistical signicance between strain infection
groups. Tukey post hoc analysis was performed to elucidate statistical signicance
between individual strain infection groups. Non-parametric results (T. phagedenis
quantities/genome equivalents) were log-transformed to achieve normality. Gene
expression values were calculated from the ΔΔCt method, with tpiA as the reference
marker. Fold changes were calculated via 2ΔΔCt. Condence intervals of fold changes were
calculated assuming unequal variance. P-values of expression data were calculated using
a two-tailed, paired t-test. All gures were generated in GraphPad Prism 9 (GraphPad
Software Inc., La Jolla, San Diego, California, USA). For all the below analysis, P < 0.05 was
considered signicant.
RESULTS
Strains 4457 and 138-2 produce smaller abscesses than other strains at the
lowest and intermediate infectious doses
During infection trial 1, three TPI doses were tested in a murine subcutaneous model.
The lowest dose of 1.25 × 109 was selected based on previous pilot studies (unpub
lished) that determined 109 TPI was the lowest dose possible to achieve consistent and
reproducible lesions in mice. From this point, the dose increased fourfold to reach an
intermediate dose of 5 × 109 TPI and once more to reach our highest dose of 2 ×
1010 TPI. The subcutaneous injections were performed with our eight newly isolated T.
phagedenis strains along with a set of formalin-killed (F-K) doses of strain 11. For the
lowest infectious dose (Fig. 1A), strains 4457 and 138-2 produced the smallest average
abscesses at 14.8 and 16.2 mm3, respectively. Strains 9 and 2947 had signicantly larger
abscesses than these two previous strains, averaging an abscess volume of 49.1 and
53.4 mm3, respectively. The dierences in abscess sizes was observed similarly with the
medium dose (Fig. 1B), where the strains 2947 (P < 0.0001), 102 (P = 0.0051), EB 1900 (P =
0.0053), and 11 (P = 0.0409) were all signicantly larger than 4457 and 138-2. As with the
lowest dose, the strains 4457 and 138-2 produced the smallest average lesion volumes
of 70.6 and 60.9 mm3, respectively with the intermediate dose. These dierences in
abscess volumes between strains were lost during infection with the highest infectious
dose of 2 × 1010 TPI (Fig. 1C), as no strains emerged as signicantly causing larger or
smaller abscesses than others tested. Across all three plots, the F-K doses do not produce
signicantly dierent abscess volumes compared to the other live, strain infections.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 7
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

FIG 1 Subcutaneous abscess volumes following infection challenge with eight Treponema phagedenis
strains and formalin-killed (F-K) control. Abscess sizes were averaged from n = 6 infection groups. For all
three doses, strain 11 was used as the F-K control. (A) Lesion size (mm3) for lowest infectious dose of 1.25
(Continued on next page)
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 8
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

Metal-ion import genes are highly upregulated during infection, while
motility genes have low relative expression
The average gene expression values for six strains were calculated for the 11 putative
virulence genes of interest. From this, eight genes (patB, aB, adhE mglA, phoU, fbpA, msp,
and oadB) had between a 2-fold to 10-fold increase in gene expression when compared
against the in vitro culture. The remaining three genes (patB, cheA, and aB) had <2-fold
change. The two genes relating to metal-ion importing (troA and hbpB) had expression
fold increases of more than 10-fold, with troA being the most highly expressed gene at
17.23 fold increase. The cheA gene was the only gene that was downregulated (<1-fold
change) on average. Only 3 of 11 genes emerged as having non-signicant changes in
relative fold expression when comparing in vivo to in vitro, those genes were patB, aB,
and cheA. The log2 transformed fold changes can be seen in the heatmap (Fig. 2) to
visualize the expression levels as a color gradient. From this depiction of data, not one
strain emerges as having a larger or more profound upregulation or utilization of genes
during the infection. However, when viewing the columns, the genes troA and hbpB
stand out as having high relative expression, while the genes cheA and aB stand out as
having low relative expression.
Quantity of T. phagedenis spirochetes increased 10-fold for strains EB 1900
and 2947 during infections
DNA co-extracted from all samples was used to quantify the number of T. phagede
nis found within in vitro cultures and murine abscesses. DNA was quantied using a
species-specic qPCR. In Fig. 3, the log-transformed genome equivalents (GE) for each
strain in both sample sets is depicted. For strains 102 and 11, there was no signicant
FIG 1 (Continued)
× 109 TPI. (B) Lesion size for an intermediate dose of 5 × 109 TPI. (C) Lesion size for the highest dose of 2
× 1010 TPI. The top of each bar represents the mean average lesion size for each strain and the whiskers
span the standard deviation of each sample set. Individual data points are highlighted as scatter dots.
Statistical signicance lines are presented as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
FIG 2 Gene expression heat map of 11 putative virulence genes from six T. phagedenis strains following infection in a murine model. Expression fold changes
were calculated via 2−ΔΔCt method, and abscess Ct data were compared against in vitro Ct data. Fold changes were then log2 transformed.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-24 9
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

dierence between both in vivo and in vitro populations. Strains 3133 and 9 had a
roughly 10-fold decrease in T. phagedenis numbers once infected in the mice. For strains
EB 1900 and 2947, the opposite was true as the T. phagedenis population increased
roughly 10-fold in the subcutaneous injection site.
Expression fold changes of virulence genes is conrmed in bovine M2 lesions
For the 11 targeted genes, the genes adhE, patB, and oadB had two or fewer data
points (Ct values obtained from the six bovine samples) and were therefore omitted
from further analysis. For the remaining eight genes, fold changes were determined. The
troA gene had the highest fold change in cattle at 56.91. The remaining genes averaged
around a 10-fold increase in gene expression during infection in cattle. The gene with the
lowest fold change was aB.
Despite the signicant alignment of mglA, cheA, and ab genes with sequences from
DD-isolates for both T. medium and T. pedis, troA and hbpB for T. pedis, and msp and fbpA
for T. medium, the virulence genes primers and probes did not recognize and amplify
gDNA from T. pedis and T. medium. Positive amplication was only observed when gDNA
from T. phagedenis was used as a template, conrming their specicity for T. phagedenis.
Genomic sequence analysis of T. phagedenis strains
A genome-based phylogenetic tree of the isolates was constructed using the published
pipeline PhyloPhlAn (Fig. 4). In this tree, the T. phagedenis type strain B43.1T was used as
the outgroup to root the tree. From this phylogeny, we conclude that strains 138-2 and
4457 are identical. This is consistent with the observation in infection trial 1, where both
of these strains induced similarly small abscess sizes compared to the six other strains.
The open reading frame of the 11 putative virulence genes was analyzed for SNPs
when compared against the type strain Treponema phagedenis B43.1T. From this analysis,
FIG 3 Quantities of T. phagedenis strains found in In vitro (cell pellets) and in in vivo (abscesses) samples. Cell quantities are expressed as log10 genome
equivalents (GE) obtained from species-specic qPCR. Bars represent the mean average log10 GE for each sample set and error bars indicate standard deviation.
Statistical signicance was calculated from multiple unpaired t-tests between in vitro and in vivo groups. Statistical signicance lines are presented as follows: *P
< 0.05, **P < 0.01, ****P < 0.0001, ns as non-signicant.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2410
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

it was determined that aB, troA, phoU, and oadB had no SNPs. For the other genes, adhE,
cheA, fbpA, mglA, patB, msp, and hbpB SNPs were detected. The ratio of non-synonymous
to synonymous substitutions or dN/dS was calculated for the seven genes that had SNPs.
dN/dS estimates the selective pressure on a protein-coding gene (Table 3). For the genes,
adhE, fbpB, patB, and msp, the dN/dS was greater than 1 suggesting a positive/diversify
ing selection to these genes. Only the mglA gene had a dN/dS of 1 indicating neutral or
no selection pressure on this gene. For the genes cheA and hbpB, the ratio was less than 1
indicating a negative or purifying selection on these genes.
DISCUSSION
From our analysis of abscess sizes, expression fold changes in putative virulence genes,
and quantities of T. phagedenis during infection, we gained insight into the pathogenesis
and virulence of T. phagedenis strains in a murine model. The subcutaneous injection
with live or dead T. phagedenis into the murine model did not induce a change in
resulting abscess volumes. The determination of the expression fold changes of 11
FIG 4 Phylogenetic tree of the eight isolated T. phagedenis strains. The tree was constructed in PhyloPhlAn and is presented in a circular format. The tree
is rooted to the reference strain B43.1T. The scale bar of 0.10 shows the length of each vertical branch and corresponds to the phylogenetic distance of 0.1
nucleotide substitutions per site or 10% nucleotide sequence deviation.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2411
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

putative virulence genes in six T. phagedenis isolates, as well as the quantities of T.
phagedenis during infection in a murine abscess model, provides deeper insight into
virulence during infection. There was a high expression of two metal-ion acquisition
genes, troA, a zinc-periplasmic lipoprotein and hbpB, a hemin-binding lipoprotein. In
addition, the gene encoding for the major sheath protein, msp was found highly
expressed during infection. Notably, we uncovered that the motility/chemotaxis genes
of aB and cheA were not signicantly upregulated during infection. We also discovered
that some strains had an increase in the quantities of T. phagedenis from the starting
inoculum, suggesting potential strain-level dierences in the proliferative ability of T.
phagedenis isolates during infection. Although the reasons for this distinction remain
unknown and could warrant future study.
There was a signicant dierence in the abscess volumes between all three doses
for each strain, with the abscess volume increasing as the dose increased. Dierences
between strains’ ability to cause varying abscess volumes were only observed for strains
138-2 and 4457, which produced signicantly smaller abscesses than some of the other
strains, but only at the lowest and intermediate doses (Fig. 1A and B). These strains were
later identied as identical through WGS data and phylogenetic analysis. Ultimately, we
found no strain emerging as more virulent than others, in terms of abscess volume.
In addition to this result, the formalin-killed (F-K) doses are employed as a treatment
group to test the eect of dead spirochete cell structures on the development of
abscesses. F-K doses induced abscesses at every infectious dose level tested. Importantly,
the mechanism of cell killing by formalin/formaldehyde is through peptide cross-link
ing and it does not induce cell lysis; therefore, the integrity of the cell structure is
maintained and intracellular proteins are not exposed/released during treatment (56).
These abscesses were not signicantly smaller than live doses, this is in contrast to other
published literature on murine infections with DD treponemes, where F-K doses either
produced no abscesses at all or lesions of signicantly smaller volume compared to live,
infectious doses (18, 19).
The nature of spirochete infections and pathogenesis could explain the lack of
observed abscess dierences in live versus dead doses. First, spirochete pathogens lack
secretion systems to deliver toxins to damage host cells during infection, which is a
deviation from many other bacterial pathogens in strategy during infection (57). Instead,
treponemes rely heavily on immune system evasion or subversion to persist in their
hosts, consequently adopting a stealth lifestyle to avoid detection and resist killing
by the immune system (58). This unique strategy of subversion means the infectious
doses one selects typically need to be higher in cell quantity compared to other, more
potent pathogens. This is especially true if the desired outcome of an infection trial
is a localized abscess at the infection site, where a high abundance of treponemes is
needed to induce abscess formation. This was demonstrated in murine infection pilot
studies, where it was found for T. phagedenis infections, using doses of less than 109 TPI
did not produce abscesses with any consistency or reliability, suggesting that a 109 TPI
was the minimum dose needed to generate abscesses. Due to the necessity of a high
infectious dose, the abundance of T. phagedenis collectively has high immunogenicity
leading to abscess formation. It is likely the case that formalin-treated doses do not lose
TABLE 3 Average ratio of non-synonymous to synonymous substitutions (dN/dS) for target virulence
genes across all strains that have SNPs
Gene Average dN/dS
adhE 1.25
cheA 0.14
fbpA 1.29
mglA 1
patB 5.14
hbpB 0.93
msp 1.97
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2412
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

this immunogenicity. For example, Sloupenska et al., 2024 (59), found higher immu
nogenicity for formaldehyde-treated whole-cell preparations of dierent B. burgdorferi
morphologies compared to live doses. This reasoning could explain why the immune
responses, in particular the tissue damage induced by the host to respond to and contain
the infection was the same between both types of doses.
Ultimately, this result of insignicance between live and dead T. phagedenis doses
on abscess sizes in mice could be explained in part by the mechanisms of spirochete
pathogenesis. However, the fact that no dierence was observed in abscess sizes
between live and dead doses suggests the passivity and inactivity of strains in hosts.
This conveys a theme of attenuated virulence for T. phagedenis strains from the strict
perspective of abscess sizes alone. Evaluating the virulence of a pathogen based on
abscess sizes alone is a limitation, as T. phagedenis could have engaged in motility and
disseminated, causing dysfunction or impairment in distal organs or even persisting
further away from the site of infection. The lack of known secreted proteases in DD
treponemes could also imply that tissue destruction is primarily a consequence of a
subverted and exaggerated host inammatory response to a litany of bacterial antigens,
with host-initiated damage to infection as likely the principal reason for the pathology
observed during murine abscess trials.
In the RT-qPCR assay, 11 putative virulence genes were targeted, those genes
encoding for metal-ion import lipoproteins, like hbpB and troA had high relative
expression during murine infection (Table 4) and in bovine M2 lesions (Table 5). Hemin-
binding lipoprotein B was found upregulated in T. denticola when it was grown under
iron-limiting conditions (27). Sequence analysis of hbpB revealed that the protein lacks a
TonB-box motif, suggesting a TonB-independent and perhaps novel mechanism of iron
uptake (28). The second metal-ion-binding lipoprotein was troA, a periplasmic zinc-bind
ing protein. The troA gene was found dierentially expressed in response to extracellular
manganese and iron in T. denticola and to manganese and zinc in T. pallidum (29).
This supports the binding capability of troA for multiple and dierent divalent cations.
Mutants of the troA gene have been studied in the swine pathogen Streptococcus suis,
where mutants were found to have attenuated virulence in a murine model compared
to the wild-type strain, conferring the importance of this gene during infection (49).
Spirochetes have a lack of classic iron acquisition mechanisms like the secretion of ferric
siderophores for iron scavenging from the host (28). Iron acquisition in treponemes
seems limited to the import of metal ions via surface, lipoprotein receptors like hbpB and
troA (28). The scarcity of metal ions in hosts and the essentiality of these cation cofactors
explains why both these genes were highly upregulated in both hosts.
In addition to metal-ion import genes, genes encoding outer membrane proteins
involved in adherence were found to have moderately high expression during infection.
The rst gene of focus was the gene encoding bronectin-binding protein A, called
fbpA. Fibronectin-binding proteins were found to be upregulated during intradermal T.
TABLE 4 Average gene expression fold changes for 11 putative virulence genes, during murine infection with 6T. phagedenis strainsa
Gene 3133 9 2947 EB 1900 102 11 Strain average (n = 6) 95% CI of strain averages P-value
patB 0.31 1.34 7.48 1.37 0.37 1.00 1.08 [0.43, 2.70] 0.32
hbpB 4.40 81.27 9.07 47.12 6.88 5.02 13.19 [4.91, 35.46] 1.14 × 10−14
aB 1.74 1.82 0.90 0.61 9.20 0.43 1.38 [0.58, 3.30] 0.40
adhE 4.57 23.35 3.80 11.08 18.41 2.05 7.44 [3.44, 16.08] 9.83 × 10−13
mglA 2.48 2.57 5.20 2.72 2.72 2.57 2.93 [2.33, 3.68] 1.50 × 10−8
cheA 0.31 1.60 3.35 0.76 1.15 0.58 0.97 [0.50, 1.89] 0.43
fbpA 3.44 5.43 16.65 7.88 4.27 2.86 5.57 [3.33, 9.32] 2.64 × 10−11
msp 55.03 3.04 3.97 224.52 0.63 0.77 6.46 [0.98, 42.77] 2.10 × 10−5
phoU 1.96 3.40 2.09 10.65 0.32 3.23 2.44 [0.96, 6.23] 4.30 × 10−5
troA 24.86 62.37 6.34 13.72 77.25 2.51 17.23 [5.96, 49.80] 7.69 × 10−16
adB 2.32 0.68 33.21 3.02 1.28 2.65 2.85 [0.99, 8.23] 1.49 × 10−4
aThe fold changes > 1 indicate upregulation and fold changes < 1 indicate downregulation of genes during infection when compared against the in vitro samples. 95%
condence intervals at an alpha of 0.05 showing [upper and lower] limits of gene expression were calculated.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2413
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

pallidum infection in rabbits, with transcription levels increasing 21–30 days post-infec
tion (30, 39). In addition, T. denticola mutants of fbpA had reduced cytotoxicity to human
gingival broblasts and lowered the growth of these cell types in vitro (40). The second
adherence gene targeted was msp, encoding for the major sheath protein, which is
the most well-studied virulence factor we targeted, owing to its ubiquity and multifunc
tionality (41–44). Msp is the immunodominant antigen in T. denticola and is capable of
adherence to brinogen, type 1 collagen, keratin, and laminin among other structures
(41). Beyond roles in adherence and membrane integrity, msp possesses an even larger
array of functions such as epithelial cell lysis, actin remodeling, impairment of neutrophil
chemotaxis, and phagocytosis activities (42–44).
In addition to elucidating fold changes of putative virulence genes during infection,
the quantity of T. phagedenis present at the end of the infection trial was determined
simultaneously. Through the quantication of T. phagedenis in vivo, we can understand
the dierences in the treponeme population quantity during infection and compare it
against the starting number of cells used in the inoculum. This can shed light on the
extent of the proliferation of strains during the infection. It was determined that strains
3133 and 9 had a roughly 10 fold decrease in the quantity of T. phagedenis during
the infection. The decrease in T. phagedenis numbers for these strains during infection
reects the attenuated ability of these strains to survive in vivo and therefore these cells
were lost via cell death or immune killing. Conversely, strains EB 1900 and 2947 had
a log-fold increase in T. phagedenis during infection. The proliferation of these strains
during infection indicates that the strains were thriving and replicating in the abscess.
Due to the relative nature of expression values obtained and the lack of knowledge
of absolute mRNA transcript quantities during infection, it is not possible to make direct
correlations between the changes in cell quantities to the changes in mRNA transcript
levels. There are patterns and trends in the data that we were able to derive that could
provide some insight into the strategies during infection and could explain why there are
variable quantities of some strains versus others.
One such pattern emerges with troA where it was highly expressed in the stagna
ted/poorer in vivo growing strains (102, 9, 3133) averaging a fold change of 54.83
compared to the more proliferative strains (EB 1900, 2947, 11). The troA gene encodes for
a metal import lipoprotein, which is likely upregulated to improve metal-ion acquisition
and therefore improve survival by being able to supplement enzymes with essential
metal cofactors. Since low-growing strains are struggling to survive in vivo that may be
why troA is signicantly upregulated to scavenge metal ions in the host to improve their
survival. On the ip side, the minimal upregulation by the more proliferative strains could
suggest that their metal-ion homeostasis is more sated and potentially not as dire as the
other strains.
A second pattern emerged for strain 2947 which was the most proliferative/successful
strain during infection. In this strain, two genes (oadB and fbpA) were found to be
signicantly and dierentially upregulated as opposed to the other strains. The oadB
gene is thought to encode a protein that is part of a Na+ symporter channel, which can
TABLE 5 Average fold change of eight putative virulence genes of T. phagedenis found in M2 bovine
lesionsa
Gene Average fold change (n = 4) 95% CI
fbpA 10.21 [3.43, 30.42]
cheA 5.02 [1.02, 24.61]
troA 56.91 [13.83, 234.25]
aB 1.13 [0.13, 9.82]
hbpB 9.71 [1.19, 79.36]
msp 9.59 [0.19, 497.05]
phoU 10.60 [2.14, 52.42]
mglA 16.16 [2.19, 119.45]
aAverage fold changes across n = 4 samples, determined by comparing against in vitro samples. 95% condence
intervals at an alpha of 0.05 showing [upper and lower] limits of gene expression were calculated.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2414
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

be linked to the generation of a sodium motive force (SMF). The SMF could provide
additional means of ATP generation via stronger membrane potential, perhaps one of
the reasons why strain 2947 was eective at proliferation during infection. Strain 2947
expressed fbpA higher than the other strains. The larger magnitude fold change for 2947
could show the improved adherence and therefore persistence/survival of this strain.
The contrast between motility and colonization/adherence strategies in the host is
highlighted by the signicant dierence in gene expression fold changes between these
two groups of genes. The motility genes (aB, cheA) yielded a low expression, while
colonization adherence genes (msp, fbpA) had a relatively high expression. This duality
suggests T. phagedenis prioritizes colonization over motility during murine abscess
infection. This result is drawn strictly from the abscesses at 1 week post-infection and
therefore may not capture the extent to which T. phagedenis ees the site of infection
by engaging motility early on in the infection. This is a stark contrast to other virulent
spirochetes like T. pallidum and B. burgdorferi, whose pathogenesis is crucially linked to
dissemination and immune evasion in the host (44, 58). During DD infection, the extent
to which treponemes disseminate in bovine hosts is unknown. Treponemes are known
to invade deeper into the epithelial layers (51), but beyond this localization, it is unclear
if they persist in other distal organs of the bovine host or what external environmental
reservoirs they may exist in. In a previous murine infection, treponemes were detected
in the kidneys of mice (19). This dissemination of spirochetes during murine infection
might have been because of subcutaneous infection used in his model, as it is currently
unknown if treponemes could breach the dermis and disseminate to distant anatomical
sites in the bovine host. Finally, the treponemes may downregulate their motility genes
after failure to escape the abscess capsule upon formation.
These trends and patterns identied were limited in breadth and therefore it is
dicult to establish a link between the proliferative ability of some strains to those same
strains possessing a more eective and potent virulence than the others. This is similarly
supported by fairly consistent expression proles across all strains. It is then entirely
possible that those strains that had a decrease in quantity or a stagnant population of T.
phagedenis in vivo are equally as virulent and could even be more subversive.
To add to the signicance of the expression data from the murine infections, the
same 11 putative virulence genes were targeted in naturally occurring bovine M2 lesions.
When comparing the expression fold changes between murine and bovine infections,
troA is the gene with the highest fold change in both hosts; however, it is 3× larger in
bovine lesions. The mglA gene had a signicantly larger upregulation in bovine samples,
increasing nearly sixfold compared to the fold change in murine abscesses. This gene is
an ATP-binding protein and is part of an ABC transporter system for glucose, ribose, and
other sugars (33). Treponeme metabolism is centered around the catabolism of peptides
and amino acids, with saccharolytic metabolism being secondary. It is possible that mglA
is being engaged more heavily under more nutrient-starved and stringent conditions
of bovine lesions. The phoU gene was the only gene unique to T. phagedenis and it
had a 5× higher upregulation in bovine lesions. This gene is involved in osmotic stress
resistance, which is likely a more volatile and hypotonic environment in bovine lesions
than murine abscesses, owing to its more substantial upregulation. Interestingly, phoU
has been linked to the development of persisters in Escherichia coli (60) and could play a
role in the persistence of T. phagedenis.
Despite the above-mentioned dierences, the gene expression patterns were fairly
similar in both environments. The similarity in gene expression data between murine and
bovine backgrounds, in addition to conrmation of the qPCR assay being specic only to
DNA from T. phagedenis, further validates the subcutaneous murine infection model for
the study of gene expression patterns in DD pathogens.
From the whole-genome sequence analysis, we determined SNPs belonging to ORFs
of the 11 tested virulence genes when compared against a type strain B43.1T isolated
from Switzerland (61). Within the protein-coding regions of the genes, SNP patterns did
not seem to vary between strains as many strains shared identical SNPs, thus conrming
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2415
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

strain homogeneity. The ratio of non-synonymous to synonymous substitutions provides
an indication of the evolutionary pressures on proteins (62). We observed negative
or purifying selection on the genes cheA and hbpB. Negative selection indicates that
deleterious mutations are purged and the protein sequence is preserved and under
constraint to maintain function (62). Given the essentiality of the auto kinase cheA,
in the chemotaxis pathway, negative selection is expected. For hbpB, the dN/dS of
less than 1 could suggest the importance of this gene for T. phagedenis. The genes
that show a dN/dS of more than 1 were adhE, fbpA, patB, and msp. Positive selection
points to diversifying and adaptive selection taking place to support change in the
protein sequence and possibly function (62). In this group, the notable gene is msp the
major outer sheath protein. Considering its presentation on the cell surface, variations in
protein sequence could confer greater antigenic diversity of strains.
In conclusion, this study delineates the initial characterization of potential virulence
genes and strategies utilized by T. phagedenis to establish and maintain infection within
a murine model. During infection, the metal-ion import genes, troA and hbpB, were
highly upregulated, along with the upregulation of adherance genes, msp and fbpA. This
result was further conrmed and supported by bovine infection data. This conrms the
importance of these genes to the natural and experimental infection in both mamma
lian hosts. Future research could explore the essentiality, mechanisms, or utility of the
metal-ion acquisition genes during infection or in vitro to see if they could serve as
potential, future drug targets against the T. phagedenis phylogroup in DD. Our research
indicates that metal-ion import genes were upregulated during murine infection and
could prove as interesting inhibitory targets to the virulence of T. phagedenis during
infection.
ACKNOWLEDGMENTS
The authors would like to thank Kristen Kalbeish for her assistance with the animal
infection experiment.
AUTHOR AFFILIATION
1Faculty of Veterinary Medicine, University of Calgary, Calgary, Alberta, Canada
AUTHOR ORCIDs
Jeroen De Buck http://orcid.org/0000-0002-4824-9580
FUNDING
Funder Grant(s) Author(s)
Canadian Government | Natural Sciences and
Engineering Research Council of Canada (NSERC)
CRDPJ/536202-2018 Colton Scott
DATA AVAILABILITY
Whole-genome sequence data were obtained for the eight isolated T. phagedenis strains
and deposited in the Sequence Read Archive under accession number PRJNA1065570.
ETHICS APPROVAL
All animal experiment procedures were reviewed and approved by the University of
Calgary Health Sciences Animal Care Committee (AC21-0063).
ADDITIONAL FILES
The following material is available online.
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2416
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

Supplemental Material
Supplemental tables (IAI00117-24-s0001.docx). Tables S1 and S2.
REFERENCES
1. Cheli R, Mortellaro C. 1974. La dermatite digitale del bovino Proceedings
of the 8th International Conference on Diseases of Cattle, p 208–213
2. Sullivan LE, Carter SD, Blowey R, Duncan JS, Grove-White D, Evans NJ.
2013. Digital dermatitis in beef cattle. Vet Rec 173:582. https://doi.org/
10.1136/vr.101802
3. Jacobs C, Orsel K, Barkema HW. 2017. Prevalence of digital dermatitis in
young stock in Alberta, Canada, using pen walks. J Dairy Sci 100:9234–
9244. https://doi.org/10.3168/jds.2017-13044
4. Solano L, Barkema HW, Pickel C, Orsel K. 2017. Eectiveness of a
standardized footbath protocol for prevention of digital dermatitis. J
Dairy Sci 100:1295–1307. https://doi.org/10.3168/jds.2016-11464
5. Palmer MA, O’Connell NE. 2015. Digital dermatitis in dairy cows: a review
of risk factors and potential sources of between-animal variation in
susceptibility. Animals (Basel) 5:512–535. https://doi.org/10.3390/
ani5030369
6. Nielsen MW, Strube ML, Isbrand A, Al-Medrasi W, Boye M, Jensen TK,
Klitgaard K. 2016. Potential bacterial core species associated with digital
dermatitis in cattle herds identied by molecular proling of interdigital
skin samples. Vet Microbiol 186:139–149. https://doi.org/10.1016/j.
vetmic.2016.03.003
7. Bay V, Griths B, Carter S, Evans NJ, Lenzi L, Bicalho RC, Oikonomou G.
2018. 16S rRNA amplicon sequencing reveals a polymicrobial nature of
complicated claw horn disruption lesions and interdigital phlegmon in
dairy cattle. Sci Rep 8:15529. https://doi.org/10.1038/s41598-018-33993-
9
8. Zinicola M, Lima F, Lima S, Machado V, Gomez M, Döpfer D, Guard C,
Bicalho R. 2015. Altered microbiomes in bovine digital dermatitis lesions,
and the gut as a pathogen reservoir. PLoS One 10:e0120504. https://doi.
org/10.1371/journal.pone.0120504
9. Wilson-Welder JH, Alt DP, Nally JE. 2015. Digital dermatitis in cattle:
current bacterial and immunological ndings. Animals (Basel) 5:1114–
1135. https://doi.org/10.3390/ani5040400
10. Caddey B, Orsel K, Naushad S, Derakhshani H, De Buck J. 2021.
Identication and quantication of bovine digital dermatitis-associated
microbiota across lesion stages in feedlot beef cattle. mSystems
6:e0070821. https://doi.org/10.1128/mSystems.00708-21
11. Beninger C, Naqvi SA, Naushad S, Orsel K, Luby C, Derakhshani H,
Khapour E, De Buck J. 2018. Associations between digital dermatitis
lesion grades in dairy cattle and the quantities of four Treponema
species. Vet Res 49:111. https://doi.org/10.1186/s13567-018-0605-z
12. Dias AP, De Buck J. 2022. Detection and quantication of bacterial
species DNA in bovine digital dermatitis lesions in swabs and ne-
needle aspiration versus biopsies. Front Vet Sci 9:1040988. https://doi.
org/10.3389/fvets.2022.1040988
13. Dias AP, Orsel K, De Buck J. 2024. Quantifying and mapping digital
dermatitis-associated bacteria in lesion and nonlesion body sites and
dairy farm environment. J Dairy Sci 107:3252–3268. https://doi.org/10.
3168/jds.2023-24160
14. Klitgaard K, Boye M, Capion N, Jensen TK. 2008. Evidence of multiple
Treponema phylotypes involved in bovine digital dermatitis as shown by
16S rRNA gene analysis and uorescence in situ hybridization. J Clin
Microbiol 46:3012–3020. https://doi.org/10.1128/JCM.00670-08
15. Moreira TF, Facury Filho EJ, Carvalho AU, Strube ML, Nielsen MW,
Klitgaard K, Jensen TK. 2018. Pathology and bacteria related to digital
dermatitis in dairy cattle in all year round grazing system in Brazil. PLoS
One 13:e0193870. https://doi.org/10.1371/journal.pone.0193870
16. Gomez A, Cook NB, Bernardoni ND, Rieman J, Dusick AF, Hartshorn R,
Socha MT, Read DH, Döpfer D. 2012. An experimental infection model to
induce digital dermatitis infection in cattle. J Dairy Sci 95:1821–1830.
https://doi.org/10.3168/jds.2011-4754
17. Kesavalu L, Holt SC, Ebersole JL. 1998. Virulence of a polymicrobic
complex, Treponema denticola and Porphyromonas gingivalis, in a murine
model. Oral Microbiol Immunol 13:373–377. https://doi.org/10.1111/j.
1399-302x.1998.tb00694.x
18. Elliott MK, Alt DP, Zuerner RL. 2007. Lesion formation and antibody
response induced by papillomatous digital dermatitis-associated
spirochetes in a murine abscess model. Infect Immun 75:4400–4408.
https://doi.org/10.1128/IAI.00019-07
19. Arrazuria R, Knight CG, Lahiri P, Cobo ER, Barkema HW, De Buck J. 2020.
Treponema spp. isolated from bovine digital dermatitis display dierent
pathogenicity in a murine abscess model. Microorganisms 8:1507. https:
//doi.org/10.3390/microorganisms8101507
20. Evans NJ, Brown JM, Demirkan I, Murray RD, Vink WD, Blowey RW, Hart
CA, Carter SD. 2008. Three unique groups of spirochetes isolated from
digital dermatitis lesions in UK cattle. Vet Microbiol 130:141–150. https://
doi.org/10.1016/j.vetmic.2007.12.019
21. Brodard I, Alsaaod M, Gurtner C, Jores J, Steiner A, Kuhnert P. 2021. A
lter-assisted culture method for isolation of Treponema spp. from
bovine digital dermatitis and their identication by MALDI-TOF MS. J Vet
Diagn Invest 33:801–805. https://doi.org/10.1177/10406387211008511
22. Arrazuria R, Caddey B, Cobo ER, Barkema HW, De Buck J. 2021. Eects of
dierent culture media on growth of Treponema spp. isolated from
digital dermatitis. Anaerobe 69:102345. https://doi.org/10.1016/j.
anaerobe.2021.102345
23. Xie F, Xiao P, Chen D, Xu L, Zhang B. 2012. miRDeepFinder: a miRNA
analysis tool for deep sequencing of plant small RNAs. Plant Mol Biol 80.
https://doi.org/10.1007/s11103-012-9885-2
24. Kelley BR, Callahan SM, Johnson JG. 2021. Transcription of cystathionine
β-lyase (MetC) is repressed by HeuR in Campylobacter jejuni, and
methionine biosynthesis facilitates colonocyte invasion. J Bacteriol
203:e0016421. https://doi.org/10.1128/JB.00164-21
25. Ejim LJ, D’Costa VM, Elowe NH, Loredo-Osti JC, Malo D, Wright GD. 2004.
Cystathionine beta-lyase is important for virulence of Salmonella enterica
serovar Typhimurium. Infect Immun 72:3310–3314. https://doi.org/10.
1128/IAI.72.6.3310-3314.2004
26. Chu L, Wu Y, Xu X, Phillips L, Kolodrubetz D. 2020. Glutathione
catabolism by Treponema denticola impacts its pathogenic potential.
Anaerobe 62:102170. https://doi.org/10.1016/j.anaerobe.2020.102170
27. Chu L, Song M, Holt SC . 1994. Eect of iron regulation on expression and
hemin-binding function of outer-sheath proteins from Treponema
denticola. Microb Pathog 16:321–335. https://doi.org/10.1006/mpat.
1994.1033
28. Cullen PA, Haake DA, Adler B. 2004. Outer membrane proteins of
pathogenic spirochetes. FEMS Microbiol Rev 28:291–318. https://doi.
org/10.1016/j.femsre.2003.10.004
29. Li C, Wolgemuth CW, Marko M, Morgan DG, Charon NW. 2008. Genetic
analysis of spirochete agellin proteins and their involvement in motility,
lament assembly, and agellar morphology. J Bacteriol 190:5607–5615.
https://doi.org/10.1128/JB.00319-08
30. De Lay BD, Cameron TA, De Lay NR, Norris SJ, Edmondson DG. 2021.
Comparison of transcriptional proles of Treponema pallidum during
experimental infection of rabbits and in vitro culture: highly similar, yet
dierent. PLoS Pathog 17:e1009949. https://doi.org/10.1371/journal.
ppat.1009949
31. Kavanaugh DW, Porrini C, Dervyn R, Ramarao N. 2022. The pathogenic
biomarker alcohol dehydrogenase protein is involved in Bacillus cereus
virulence and survival against host innate defence. PLoS One
17:e0259386. https://doi.org/10.1371/journal.pone.0259386
32. Ruby JD, Lux R, Shi W, Charon NW, Dasanayake A. 2008. Eect of glucose
on Treponema denticola cell behavior. Oral Microbiol Immunol 23:234–
238. https://doi.org/10.1111/j.1399-302X.2007.00417.x
33. Porcella SF, Popova TG, Hagman KE, Penn CW, Radolf JD, Norgard MV.
1996. A mgl-like operon in Treponema pallidum, the syphilis spirochete.
Gene 177:115–121. https://doi.org/10.1016/0378-1119(96)00286-7
34. Lux R, Miller JN, Park NH, Shi W. 2001. Motility and chemotaxis in tissue
penetration of oral epithelial cell layers by Treponema denticola. Infect
Immun 69:6276–6283. https://doi.org/10.1128/IAI.69.10.6276-6283.2001
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2417
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.

35. Lux R, Sim JH, Tsai JP, Shi W. 2002. Construction and characterization of a
cheA mutant of Treponema denticola. J Bacteriol 184:3130–3134. https://
doi.org/10.1128/JB.184.11.3130-3134.2002
36. Sze CW, Zhang K, Kariu T, Pal U, Li C. 2012. Borrelia burgdorferi needs
chemotaxis to establish infection in mammals and to accomplish its
enzootic cycle. Infect Immun 80:2485–2492. https://doi.org/10.1128/IAI.
00145-12
37. Lin T, Gao L, Zhang C, Odeh E, Jacobs MB, Coutte L, Chaconas G, Philipp
MT, Norris SJ. 2012. Analysis of an ordered, comprehensive STM mutant
library in infectious Borrelia burgdorferi: insights into the genes required
for mouse infectivity. PLoS One 7:e47532. https://doi.org/10.1371/
journal.pone.0047532
38. Ke W, Molini BJ, Lukehart SA, Giacani L. 2015. Treponema pallidum subsp.
pallidum TP0136 protein is heterogeneous among isolates and binds
cellular and plasma bronectin via its NH2-terminal end. PLoS Negl Trop
Dis 9:e0003662. https://doi.org/10.1371/journal.pntd.0003662
39. Djokic V, Giacani L, Parveen N. 2019. Analysis of host cell binding
specicity mediated by the Tp0136 adhesin of the syphilis agent
Treponema pallidum subsp. pallidum. PLoS Negl Trop Dis 13:e0007401.
https://doi.org/10.1371/journal.pntd.0007401
40. Xu X, Steensen B, Robichaud TK, Mikhailova M, Lai V, Montgomery R,
Chu L. 2015. Fibronectin-binding protein TDE1579 aects cytotoxicity of
Treponema denticola. Anaerobe 36:39–48. https://doi.org/10.1016/j.
anaerobe.2015.09.010
41. Edwards AM, Jenkinson HF, Woodward MJ, Dymock D. 2005. Binding
properties and adhesion-mediating regions of the major sheath protein
of Treponema denticola ATCC 35405. Infect Immun 73:2891–2898. https:/
/doi.org/10.1128/IAI.73.5.2891-2898.2005
42. Jones MM, Vanyo ST, Visser MB. 2019. The Msp protein of Treponema
denticola interrupts activity of phosphoinositide processing in
neutrophils. Infect Immun 87:e00553-19. https://doi.org/10.1128/IAI.
00553-19
43. Puthengady Thomas B, Sun CX, Bajenova E, Ellen RP, Glogauer M. 2006.
Modulation of human neutrophil functions in vitro by Treponema
denticola major outer sheath protein. Infect Immun 74:1954–1957. https:
//doi.org/10.1128/IAI.74.3.1954-1957.2006
44. Radolf JD, Caimano MJ, Stevenson B, Hu LT. 2012. Of ticks, mice and
men: understanding the dual-host lifestyle of lyme disease spirochaetes.
Nat Rev Microbiol 10:87–99. https://doi.org/10.1038/nrmicro2714
45. Staton GJ, Clegg SR, Ainsworth S, Armstrong S, Carter SD, Radford AD,
Darby A, Wastling J, Hall N, Evans NJ. 2021. Dissecting the molecular
diversity and commonality of bovine and human treponemes identies
key survival and adhesion mechanisms. PLoS Pathog 17:e1009464.
https://doi.org/10.1371/journal.ppat.1009464
46. de Almeida LG, Ortiz JH, Schneider RP, Spira B. 2015. phoU inactivation in
Pseudomonas aeruginosa enhances accumulation of ppGpp and
polyphosphate. Appl Environ Microbiol 81:3006–3015. https://doi.org/
10.1128/AEM.04168-14
47. Buckles EL, Wang X, Lockatell CV, Johnson DE, Donnenberg MS. 2006.
PhoU enhances the ability of extraintestinal pathogenic Escherichia coli
strain CFT073 to colonize the murine urinary tract. Microbiology
(Reading) 152:153–160. https://doi.org/10.1099/mic.0.28281-0
48. Saraithong P, Goetting-Minesky MP, Durbin PM, Olson SW, Gherardini
FC, Fenno JC. 2020. Roles of TroA and TroR in metalloregulated growth
and gene expression in Treponema denticola. J Bacteriol 202:e00770-19.
https://doi.org/10.1128/JB.00770-19
49. Wichgers Schreur PJ, Rebel JMJ, Smits MA, van Putten JPM, Smith HE.
2011. TroA of Streptococcus suis is required for manganese acquisition
and full virulence. J Bacteriol 193:5073–5080. https://doi.org/10.1128/JB.
05305-11
50. Dahinden P, Auchli Y, Granjon T, Taralczak M, Wild M, Dimroth P. 2005.
Oxaloacetate decarboxylase of Vibrio cholerae: purication, characteriza
tion, and expression of the genes in Escherichia coli. Arch Microbiol
183:121–129. https://doi.org/10.1007/s00203-004-0754-5
51. Marcatili P, Nielsen MW, Sicheritz-Pontén T, Jensen TK, Schafer-Nielsen C,
Boye M, Nielsen M, Klitgaard K. 2016. A novel approach to probe host-
pathogen interactions of bovine digital dermatitis, a model of a complex
polymicrobial infection. BMC Genomics 17:987. https://doi.org/10.1186/
s12864-016-3341-7
52. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a exible trimmer for
Illumina sequence data. Bioinformatics 30:2114–2120.
https://doi.org/10.1093/bioinformatics/btu170
53. Wick RR, Judd LM , Gorrie CL, Holt KE. 2017. Unicycler: resolving bacterial
genome assemblies from short and long sequencing reads. PLoS
Comput Biol 13:e1005595. https://doi.org/10.1371/journal.pcbi.1005595
54. Seemann T. 2014. Prokka: rapid prokaryotic genome annotation.
Bioinformatics 30:2068–2069. https://doi.org/10.1093/bioinformatics/
btu153
55. Segata N, Börnigen D, Morgan XC, Huttenhower C. 2013. PhyloPhlAn is a
new method for improved phylogenetic and taxonomic placement of
microbes. Nat Commun 4:2304. https://doi.org/10.1038/ncomms3304
56. Thavarajah R, Mudimbaimannar VK, Elizabeth J, Rao UK, Ranganathan K.
2012. Chemical and physical basics of routine formaldehyde xation. J
Oral Maxillofac Pathol 16:400–405. https://doi.org/10.4103/0973-029X.
102496
57. Huang J, Chen J, Xie Y, Liu Z. 2022. Subversion of the immune response
of human pathogenic spirochetes. J Clin Lab Anal 36:e24414. https://doi.
org/10.1002/jcla.24414
58. Radolf JD, Deka RK, Anand A, Šmajs D, Norgard MV, Yang XF. 2016.
Treponema pallidum, the syphilis spirochete: making a living as a stealth
pathogen. Nat Rev Microbiol 14:744–759. https://doi.org/10.1038/
nrmicro.2016.141
59. Sloupenska K, Koubkova B, Horak P, Dolezilkova J, Hutyrova B, Racansky
M, Miklusova M, Mares J, Raska M, Krupka M. 2024. Antigenicity and
immunogenicity of dierent morphological forms of Borrelia burgdorferi
sensu lato spirochetes. Sci Rep 14:4014. https://doi.org/10.1038/s41598-
024-54505-y
60. Li Y, Zhang Y. 2007. PhoU is a persistence switch involved in persister
formation and tolerance to multiple antibiotics and stresses in
Escherichia coli. Antimicrob Agents Chemother 51:2092–2099. https://
doi.org/10.1128/AAC.00052-07
61. Kuhnert P, Brodard I, Alsaaod M, Steiner A, Stoel MH, Jores J. 2020.
Treponema phagedenis (ex Noguchi 1912) Brumpt 1922 sp. nov., nom.
rev., isolated from bovine digital dermatitis. Int J Syst Evol Microbiol
70:2115–2123. https://doi.org/10.1099/ijsem.0.004027
62. Jeares DC, Tomiczek B, Sojo V, dos Reis M. 2015. A beginners guide to
estimating the non-synonymous to synonymous rate ratio of all protein-
coding genes in a genome. Methods Mol Biol 1201:65–90. https://doi.
org/10.1007/978-1-4939-1438-8_4
Full-Length Text Infection and Immunity
Month XXXX Volume 0 Issue 0 10.1128/iai.00117-2418
Downloaded from https://journals.asm.org/journal/iai on 28 June 2024 by 136.159.213.222.