Characterization of Enterococcus faecalis Isolated from Pullet Layers with Growth Depression and Amyloid Arthropathy Open Access

Enterococcus is a genus of Gram-positive bacteria commonly found in the digestive tract of many animals, including poultry (1,2). These bacteria are known to have a probiotic effect in the gastrointestinal tract (GIT) supporting the gut microbiome, aiding in the digestion of food, and preventing the colonization of harmful bacteria (1,2). However, animals with weakened immune systems or compromised GITs are particularly vulnerable to illness caused by enterococci (1,2,3). The two species most frequently associated with disease in poultry are Enterococcus cecorum and Enterococcus faecalis (1). In broilers, E. cecorum has been described to cause a greater impact than E. faecalis. Enterococcus cecorum is known to cause profound lameness and abnormal positioning of limbs, referred to as enterococcal spondylitis (4,5,6). Enterococcus faecalis can negatively impact a stressed or challenged flock’s health, especially when growth and transmission conditions for the bacterium are optimal. Clinical manifestations of E. faecalis disease can present as septicemia, valvular endocarditis, growth depression, omphalitis, peritonitis, arthritis, or amyloid arthropathy depending on the poultry species affected (6,7,8,9,10,11). In embryos and young birds, E. faecalis can result in late embryonic mortality and weak chicks that fail to thrive in the first week of life (7,8). As with most pathogens, exposure to distinct strains of E. faecalis results in different clinical presentations of varying severity. Previous research discovered that sequence types 82 and 49 are pathogenic strains of E. faecalis and can induce amyloid arthropathy in chickens (7,10,12,13).

Most of the information on E. faecalis inducing amyloid arthropathy in layers or breeders has been described from the European poultry industry (14). Clinical disease induced by E. faecalis is not commonly reported in the North American poultry industry. The goal of this study was to describe genotypes, antimicrobial resistance genes, and virulence factors of E. faecalis isolated from an outbreak in a layer pullet flock. In addition we phenotypically characterized the isolates using an embryo lethality assay.

A pullet breeder layer flock from Ontario, Canada, experienced severe growth depression, poor uniformity, and sporadic lameness throughout rearing and production. Briefly, in January 2023 a flock of 6280 layer chicks were placed on the farm. The farm had fresh softwood shavings on the cement floor. The barn had been washed using detergent and disinfected with a formaldehyde-based disinfectant before placing the birds. The lights were LED bulbs that were managed according to standard practices for leghorn pullets. On day 2 of placement, the chicks were relatively healthy, with adequate uniformity and no obvious health abnormalities. Forty-eight hours postplacement, the mortality rate was increased among female chicks. Poor uniformity, lameness, reluctance to move, and decreased weight gain were noted around day 5 postplacement, and multiple chicks were reported to have red discoloration of the skin surrounding the hock joint. By 2 wk of age, the mortality was less than 1% per week for female chicks. The mortality of all female chicks placed during the first 2 wk was 10.8%. Culling was conducted in the pullet house to remove pullets drastically behind the growth curve and with moderate to severe lameness (Supplemental Video S1). At 5 wk of age, six pullets were submitted to the Animal Health Laboratory (AHL) at the University of Guelph (Canada) for necropsy and bacterial culture of pooled spleen samples.

At approximately 10 wk of age, nearly 20% of the flock exhibited muscle wasting, swollen hock joints, and weakness. At this time, a thorough cull of affected birds was performed. Beginning at 11 wk of age, the live production team placed pullets that developed mild lameness subsequent to the 10-wk cull in isolation pens that were distributed randomly throughout the rearing facilities to determine the proportion of pullets that continued to develop symptoms. Blood samples from two affected pullets were collected. In-field necropsies were performed, and six pullets were submitted to the diagnostic lab for postmortem examination. Complete blood cell counts, blood biochemistry analysis, and bacterial cultures from affected hock joints were performed at the AHL. Isolation pens were used for the remainder of the rearing phase. Pullets continued to exhibit progressively severe lameness, and severely affected pullets were euthanatized. The pullets were transferred to the breeder house at 18 wk of age. At 35 wk of age, hock swab samples from laying hens diagnosed with amyloid arthropathy by gross examination (Fig. 1) and environmental swab samples were submitted to the AHL. Egg production was normal for the flock thereafter, and some hens continued to develop mild signs of lameness throughout the remainder of the production cycle.

Bacterial culture of pooled spleens aseptically collected from six pullets that died at 5 wk of age was conducted at the AHL using standardized methods. Briefly, the surfaces of spleens were seared with a heated spatula, and an incision was made through the capsule using a sterile blade. Pulp tissue was scraped and a loopful streaked on a Columbia blood agar, MacConkey agar, and phenylethyl alcohol agar (PEA). Spleen samples from six birds were pooled together on each plate type. Blood and PEA agar plates were incubated overnight at 35 C in the presence of 5% CO₂ whereas MacConkey agar plates were incubated at 35 C but in the presence of atmospheric air. The presence of bacterial growth was recorded at 24 and 48 hr of incubation. Bacterial identification was done using MALDI Biotyper Sirius system (Bruker Daltonics Inc., Bremen, Germany) according to the manufacturer’s instructions by smearing individual colonies onto stainless steel target plates and overlaying them with α-cyano-4-hydroxycinnamic acid. Three E. faecalis isolates were collected from tarsometatarsal joints of affected pullets at 35 wk of age during necropsy. Isolates were collected by aseptically swabbing the yellow-orange material and synovial fluid of affected hock joints with Stuart swabs (BD no. 220099, Sparks, MD) followed by culture and identification using the above methods. Two environmental isolates were collected from the litter and egg belt of the affected rearing facility using the same methods. The five isolates were reisolated on trypticase-soy agar slants and shipped on ice to the North Carolina State University College of Veterinary Medicine for sequencing and further genotypic and phenotypic analyses.

Genomic DNA was isolated from overnight broth cultures of each isolate grown for 12–14 hr at 37 C in brain heart infusion broth using the MasterPure Gram-positive DNA purification kit (Lucigen Corporation, Middleton, WI) according to the manufacturer’s instructions. DNA quality and quantity were determined with a Qubit 4.0 Fluorometer and the dsDNA high-sensitivity assay kit (ThermoFisher Scientific, Waltham, MA). Pooled libraries with an insert size of ∼350 bp were prepared with the Nextera DNA Flex Library Prep kit (Illumina, San Diego, CA). The Illumina MiSeq platform was used for sequencing with paired-end 150-nucleotide reads and V2 chemistry using a MiSeq v2 Reagent Kit (600 cycles) according to the manufacturer’s instructions. Contigs with a minimum length of 500 bp were assembled de novo and arranged into scaffolds in CLC Genomics Workbench version 22.0.1 (Qiagen, Germantown, MD). All sequences were annotated using the National Center for Biotechnology Information (NCBI) Prokaryotic Genomes Automatic Annotation Pipeline version 6.3 (15). Default parameters were used for all software unless otherwise specified. Genome coverage ranged from 12× to 28× (average: 21×), and contig N50 values ranged from 20 to 69 (average 50.2). Isolate details, sequencing and assembly statistics, and GenBank accession numbers for raw reads and draft genome assemblies are presented in Table 3. For in silico detection of virulence and antimicrobial resistance genes, the draft genome assemblies were uploaded to the Center for Genomic Epidemiology virulence finder database version 2.0.3 to detect Enterococcal virulence genes using a 98% identity threshold and a minimum gene coverage of 80%. Acquired antimicrobial resistance genes were detected by uploading draft genome sequences to the Center for Genomic Epidemiology ResFinder database version 4.4.1 using a 90% identity threshold and a minimum gene coverage of 60% with the E. faecalis species option selected.

Enterococcus faecalis multilocus sequence types (MLSTs) were determined by uploading draft genome sequences to the Center for Genomic Epidemiology MLST version 2.0.9. Phylogenetic analyses of the draft genome sequences were conducted with a single nucleotide polymorphism (SNP) calling pipeline for WGS, CSIPhylogeny version 1.4 with a minimum Z score of 1.96, heterozygous SNPs ignored, and the altered FastTree option selected. Enterococcus faecalis strain OG1RF (NCBI reference sequence NC_017316.1) and E. cecorum strain SA3 (NCBI reference sequence NZ_CP010064.1) were used as the reference and outgroup strains, respectively, and were included in the final dendrograms, which were visualized using FigTree version 1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/).

Two of the E. faecalis strains isolated from the infected joints of pullet breeders were used for the embryo lethality assay. Phosphate-buffered saline (PBS) was used as negative control and OG1RF, a standard reference strain of E. faecalis, was utilized as the positive control. OG1RF is known to have low pathogenicity in embryos. The inoculum was prepared following the protocol used by Blanco et al. (16). The isolates were plated on Trypticase Soy agar with 5% sheep blood plates (BD, Franklin, NJ) and incubated in aerobic conditions for 24 hr at 37 C. Colonies were obtained from the agar plates with sterile cotton swabs and added to 10 ml of sterile PBS. The absorbance of the suspension was determined and adjusted to an optical density at 600 nm of 0.5. Serial dilutions were made in PBS to achieve an inoculum dose of 10² cfu in 0.1 ml.

A total of 104 fertile eggs obtained from a commercial layer hatchery were used for the embryo lethality assay following the method of Walker et al. (17). Briefly, embryos were incubated for 12 days in standard conditions (37 C, 50% relative humidity, and interval turning). Embryo viability was determined by candling before inoculation. Thirty-two embryos were infected for each of the three treatment groups, and eight embryos served as the PBS-injected negative control group. Intra-allantoic injection of 0.1 ml of inoculum for challenged eggs or 0.1 ml of PBS was used for the control groups. Following inoculation, eggs were incubated at 37 C for seven additional days and candled every 24 hr to determine viability. Mortality was recorded daily, and the contents of dead embryos were aerobically cultured to confirm the presence of E. faecalis and rule out contamination. The embryonic death rate was used to determine the pathogenicity of strains. All eggs were cultured at the end of the trial to ensure no contamination had occurred.

Embryo survivability was plotted as Kaplan-Meier survival curves and differences determined by the log rank test of significance. All analyses were conducted with GraphPad Prism version 9.0 software. Unless otherwise stated, default parameters were used for all in silico analyses.

A summary of the laboratory pathologic findings is shown in Table 1. Pullets submitted at 5 wk of age were grossly diagnosed with generalized septicemia. Escherichia coli was isolated from a pooled sample of spleens. No bacteria were isolated from the hocks at this time. At 10 wk of age, the pathologic changes were more indicative of localized hock infections. The in-field necropsy revealed that the swollen hocks were filled with yellow-orange fibrinous material. Culture results resulted in pure growth of E. faecalis from three of five of the hock samples and one of five of the bone marrow samples. No bacteria were isolated from the spleen. Complete blood count and chemistry analysis results are presented in Table 2. The blood analysis showed increased heterophils and globulins, which was consistent with inflammation. The decreased calcium and decreased calcium-to-phosphorus ratio were suggestive of a skeletal disorder. Additionally, there was increased creatine kinase (CK) concentration in the blood that was consistent with muscle injury or lameness. At 35 wk of age, E. faecalis was identified from all hock swab samples of laying hens that demonstrated gait deficiencies and were grossly diagnosed with amyloid arthropathy. Enterococcus faecalis was isolated from the environmental samples in the layer house including the litter floor and the egg belt.

The whole-genome sequences and draft genome assemblies of the E. faecalis isolates are available online (NCBI BioProject PRJNA837978). Of the five E. faecalis strains submitted to NCSU, whole-genome sequencing identified three isolates as sequence type (ST) 82, one as ST49, and one unknown E. faecalis sequence type (Table 3). Phylogenetic comparison of these strains the three ST82 strains MMS_576, MMS_578, and MMS_579 were clonal and more distantly related to the ST49 strain MMS_580 (Fig. 2). The unknown sequence type strain MMS_577 isolated from the egg belt represented a distantly related group from all other strains and was presumed to be an environmental wild-type strain. Strains MMS_578 and MMS_580 were selected for phenotypic characterization using the embryo lethality assay.

All E. faecalis strains isolated from this case contained genes encoding virulence. The virulence factors identified in both ST82 and ST42 included SrtA, ElrA, ebd, gelE, cam, and cyl genes. ST49 had ace (100% identity), hylA (89% identity), and epdB (99.51% identity) instead of cylA, cylM, and tpx, which were identified in ST82 (Table 4). Antimicrobial resistance gene screening indicated all isolates harbored tetM and IsaA genes indicating evidence of tetracycline resistance. In addition to the tetM gene, the two ST49 isolates had the NarA and NarB gene, which encodes for resistance against narasin and salinomycin. MMS580 (ST49) harbored more antimicrobial resistance genes than MMS578 (ST82) and contained acquired antimicrobial resistance genes to ionophores (Table 5).

Twenty-four hours after embryo infection, ST82 exhibited a survivability rate of 52.6%, while the survivability rate of ST42 was 89%. Five days postinfection, ST82 and ST49 demonstrated survivability rates of 37.7% and 84.4%, respectively. The ST49 group survivability was comparable to the reference standard strain (OG1RF) at 81.6%. Based on the ELA findings, ST82 was more pathogenic than ST49 (P ≤ 0.0001), although this strain lacked the number of virulence factors and antimicrobial resistance genes (Fig. 3).

Enterococcus faecalis was isolated from three tarsometatarsal joint swabs multiple weeks after the initial onset of clinical signs in the flocks. The two major impacts on the flock were growth depression and lameness. Necropsy and bacteriology testing from 1–2-wk-old chicks was not pursued because Enterococcus spp. infection was not considered to rule out initially. In retrospect, it might have been possible that an Enterococcus spp. infection was responsible for the initial clinical signs. According to Pillar et al. (2), early contamination of Enterococcus results in overcolonization of the GIT, and impaired absorption due to local inflammation in the small intestines could result in decreased nutrient absorption for proper growth and development. Overall inflammation would have caused chicks to be less active and consequently consume less food resulting in decreased weight gain.

The necropsy at 5 wk of age indicated that pullets were experiencing septicemia characterized by inflammation of the spleen and femoral head necrosis. Only E. coli was isolated from enlarged spleens. This is not uncommon, as E. coli is frequently co-isolated with Enterococcus from diseased poultry (17,18). The bacterial cultures collected at 10 wk yielded moderate E. faecalis growth from bone marrow and joint samples. Taking into consideration the flock’s clinical history and bacterial culture of spleens and joint lesions, E. faecalis complicated by early septicemic disease caused by E. coli was the suspected etiology of lameness and decreased weight gain in the flock.

Before the 10-wk bacterial culture was reported by the diagnostic laboratory, blood was collected from two 10-wk-old pullets for biochemical analysis (Table 2). The abnormally low calcium raised concerns about calcium, phosphorus, or vitamin D3 deficiency (rickets). This may have been a primary deficiency from an imbalanced diet or a secondary deficiency caused by decreased feed intake (19,20,21). Considering the flock presentation, it is more likely related to decrease feed intake from lame hens. The increased CK aligns with the gross pathology findings of prominent keel and muscle wasting (22). Finally, an inflammatory leukogram characterized by a marked heterophilia supported the gross diagnosis of septicemia. In this case, the blood biochemical analysis served as a useful preliminary test that supported subsequent pathological findings.

Amyloid arthropathy was presumptively confirmed via gross identification of yellow-orange material in the tarsometatarsal joints of lame hens. This was supported by isolation of E. faecalis from the joint material. Multiple researchers have previously described amyloid arthropathy in layers and breeder flocks characterized by swollen hock joints filled with orange material yielding cultures of E. faecalis (7,12,13). This is interesting, as white chickens have been reported to be more resistant to amyloid arthropathy induced by E. faecalis (23).

Multilocus sequence typing revealed three out of five strains isolated from this case were MLST type ST82 and ST49 and one unknown MLST (Table 3). Only certain pathogenic strains of E. faecalis can produce amyloid arthropathy. The ST82 and ST49 strains isolated from this flock have been confirmed to cause amyloid arthropathy in layers (7,24). These two sequence types have also been described as pathogenic strains capable of inducing severe arthritis (7,13,25). ST82 is a common strain used in previous research to replicate the disease and has also been used to discover virulence genes linked to amyloid arthropathy (7,26). Further, ST82 is the prominent strain identified in the European poultry industry to cause amyloid arthropathy in layers and broilers (6,26). Phylogenetic analysis was conducted following whole-genome sequencing of the draft genome assembly (Fig. 3). ST82 and ST49 were closely related and more distantly related to the unknown ST strain. This analysis revealed that two hock joint isolates were clonally related to a house floor isolate, which demonstrated the presence of the causative agent in the housing environment presumably being shed in feces. It is unlikely that the initial infection originated from the litter, as the barn was cleaned and disinfected and new shavings were placed over the concrete floor.

Sequence types associated with human and animal E. faecalis infections commonly contain hyl, efasAfts, cyl, and gelE virulence factors (27,28,29,30,31,32,33,34). It has been reported that the presence or lack of certain virulence factors does not correlate to the pathogenicity of sequence types (30). The pathogenicity of sequence types is influenced by a combination of virulence factors and host immunity. For instance, the gelE virulence factors encode for gelatinase, which assists in Enterococci biofilm formation (30,31). Research conducted by Ciftci et al. reported that E. faecalis containing the gelE virulence factor is more likely to induce amyloid arthropathy compared to sequence types without this virulence factor (29). The thiol peroxidase (tpx) virulence factor aids E. faecalis in the invasion of macrophages by inhibiting oxidative stress and free-radical chain reactions (33,34). Only the ST82 isolates contained the tpx virulence factor, which may play a role in ST82 pathogenicity. The cyl virulence factor encodes for cytolysin, which gives E. faecalis hemolytic and cytotoxic properties (30,31). The entire cyl operon or multiple cyl genes have been identified in highly pathogenic E. faecalis sequence types capable of causing mortality. Enterococcus faecalis strains containing few cyl genes have been reported to be less pathogenic (30,35). Both ST82 and ST49 contained the gelE gene, but only ST82 contained cyl and tpx.

The variation between ST82 and ST49 could explain the results of the ELA, which indicated ST82 was more virulent than ST49. However, ST49 had more acquired antimicrobial resistance (AMR) genes. ST49 had AMR genes to antimicrobials in the tetracycline, lacosamide, streptogramin, and fluoroquinolone families, and to some ionophores commonly used in poultry production. These AMR genes have previously been described from infections from human hospitals, poultry farms, and meat samples from slaughter plants (36,37,38,39). Commensal Enterococci isolated from commercial poultry systems had elevated resistance to tetracyclines imparted by the tetM and/or tetL genes (38). There is a correlation between AMR gene resistance related to point mutation in gyrA and parC (39). It is important to note that while the isolates harbor the specified AMR genes, it cannot be assumed that these genes are actively expressed. Consequently, the phenotypic analysis may not consistently reflect the expression of these genes across the strains. Although ST82 and ST49 have been isolated from infected poultry species in European countries, there has only been one isolated and reported case from broiler breeders and no cases from layers in North America (14). These pathogenic E. faecalis may be widely present in commercial poultry in North America but unable to be isolated due to unsensitive culture techniques or lack of surveillance.

Enterococcus faecalis–induced amyloid arthropathy is a rare condition in North America. The Enterococcus pathogens identified in the case have genetic sequences and other characteristics similar to isolates found in Europe. Both ST82 and ST49 identified in the current outbreak were pathogenic strains of E. faecalis and capable of causing amyloid arthropathy. ST82 lacked virulence factors and had fewer antimicrobial resistance genes compared to ST49 but caused greater mortality in the embryos.

The focus of this study was to characterize a case of amyloid arthropathy in layer breeders and examine the genotypic characteristics of the causative E. facealis isolates. However, it is essential to conduct additional phenotypic analyses to better understand the pathogenic potential of each sequence type. This analysis will provide insight into whether the genotypic properties of the isolates are actively expressing the observed characteristics. The combination of genotypic and phenotypic analyses will offer a comprehensive understanding of amyloid arthropathy caused by E. faecalis in poultry.

It is also important to conduct further research to understand the prevalence of E. faecalis in North American hatcheries and production, including investigating flocks that experience early mortality, runting/stunting, and lameness without a definitive diagnosis. This investigation will also require evaluating hatcheries and breeder flocks to identify the source of infection. Once the source of infection is identified, improvements in sanitation and biosecurity measures can be evaluated to control the extent of E. faecalis infections.

This Whole Genome Shotgun project has been deposited at GenBank under the assembly accession numbers GCA_032870015.1 (Enterococcus faecalis MMS576), GCA_032871015.1 (Enterococcus faecalis MMS577), GCA_032870055.1 (Enterococcus faecalis MMS578), GCA_032870075.1 (Enterococcus faecalis MMS579), and GCA_032869765.1 (Enterococcus faecalis MMS580). The versions described in this paper are the first versions.

Supplemental data associated with this article can be found at https://doi.org/10.1637/aviandiseases-D-24-00064.s1.

AHL =

Animal Health Laboratory;

AMR =

antimicrobial resistance;

CK =

creatine kinase;

ELA =

embryo lethality assay;

GIT =

gastrointestinal tract;

MLST =

multilocus sequence type;

PBS =

phosphate-buffered saline;

PEA =

phenylethyl alcohol agar;

SNP =

single-nucleotide polymorphism;

ST =

sequence type