Chlamydia gallinacea, not C. psittaci, is the endemic chlamydial species in chicken (Gallus gallus)

To investigate the prevalence and diversity of Chlamydia spp. in domestic birds in China, oral and cloacal swabs of healthy chickens, ducks, geese and pigeons were collected nationwide from live-animal markets and examined by Chlamydia spp. 23 S rRNA gene FRET-PCR followed by high-resolution melting curve analysis and confirmatory sequencing. Overall, 26.2% of the birds (602/2,300) were positive for Chlamydia spp. and five Chlamydia spp. were identified. While occasional detection of C. suis and C. muridarum in poultry is reported here for the first time, the predominant chlamydial agent was C. gallinacea representing 63.8% of all positives (384/602) and 81.2% of positive chickens (359/442). Analysis of the C. gallinacea ompA phylogeny revealed at least 13 well segregated variants (serovars). Seven-month monitoring of C. gallinacea-infected chickens indicated that the infection was persistent. C. gallinacea-infected chickens remained without overt clinical disease, but showed body weight gains significantly reduced by 6.5–11.4% beginning in week 3 post-infection. This study indicates that C. gallinacea is the endemic chlamydial species in chickens, whereas C. psittaci dominates only in pigeons. Further studies are required to address the specific conditions under which C. gallinacea could act as an avian pathogen and possibly also a zoonotic agent.

Following the completion of PCR, the T m of probe hybridization to the targets was determined by highresolution melting (HRM) curve analysis as the peak of the second derivative of the fluorescence released during a temperature increase from 38-85 °C. Based on the unique T m distributions, 11 Chlamydia species are differentiated into 8 distinct groups: 50.6 °C for C. avium and C. caviae; 53.8 °C for C. pneumoniae; 54.9 °C for C. gallinacea; dual peaks of 55.8 °C and 61.4 °C for C. suis, 57.6 °C for C. felis, C. abortus and C. psittaci; a flat peak of 61.0 °C for C. trachomatis; 63.3 °C for C. pecorum; a flat peak of 65.0 °C for C. muridarum. For each Chlamydia species, five concentrations of the targets were used (only 100 and 10 copies of the gene/20 μ l reaction system are shown here) and peaks and curve shape were consistent at all dilutions.
Considerable geographic differences in Chlamydia spp. prevalence were found. There was no Chlamydia spp. detection in five provinces, while PCR prevalence ranged from 1-26% in ten provinces, and was above 30% in nine other provinces (Fig. 3). Overall, the positivity rate of oral swabs (22.0%; 471/2,138) was significantly higher (P < 10 −4 ; Chi-square test) than that of cloacal swabs (16.1%; 293/1,824). In birds from which both oral and cloacal swabs were collected (n = 1,662), 162 were positive in both samples while 185 were only positive in oral swabs and 107 only in cloacal swabs. Of the 162 double-positive birds, 142 harbored the same Chlamydia species in both swabs.
Isolation of C. gallinacea strains and ompA polymorphism. Four C. gallinacea isolates (JXC1-4) were obtained from an oral swab and three cloacal swabs of chickens from Jiangxi province by propagation in Hep-2 cells. Their ompA sequences (GenBank accession number: KT692977) were identical to each other and showed 85.7% similarity to the only complete ompA sequence available in GenBank (strain 08-1274/3; accession number: AWUS01000004). Phylogenetic comparison showed that the ompA gene of the C. gallinacea strains is highly polymorphic, and at least 13 new variants that were deeply separated, and based on the sequence diversity clearly represent serovars 18 , were identified in addition to the published ompA sequences (Fig. 5, Figs S1 and S2). Persistent infection of C. gallinacea in chickens. Monitoring for Chlamydia spp. in naturally-infected chickens in the course of seven months demonstrated persistent C. gallinacea infection (Fig. 6), whereas infections with other chlamydial species were observed only transiently. Thirty-one free-range chickens from a mountain village were naturally infected with C. suis (28/31; 90.3%), C. psittaci (2/31; 6.5%) and C. gallinacea (1/31; 3.2%). As shown in Fig. 6, three weeks after transferring these 31 chickens to a containment animal facility, C. suis and C. psittaci disappeared, while the C. gallinacea carrier status was maintained (93.5%; 29/31) until the end of monitoring 3 months later.
No signs of clinical disease were observed in C. gallinacea-inoculated broiler chickens and mock-inoculated chickens. Body weights did not differ significantly between C. gallinacea-and mock-inoculated chickens in the first two weeks pi. However, subsequently in weeks 3-5 pi, C. gallinacea-inoculated chickens showed significantly lower body weights than the mock-inoculated chickens. Overall, C. gallinacea-inoculated chickens showed 8.2%, 11.4% and 6.5% lower weekly weight gains as compared to the control group in weeks 3-5 pi (Fig. 7). were not detected in 5 provinces (green: Shanxi, Tibet, Jilin, Liaoning and Shanghai), the PCR prevalences were between 1-26% in 10 provinces (yellow: Henan, Guangxi, Hebei, Hunan, Zhejiang, Xinjiang, Shaanxi, Inner Mongolia, Hubei and Gansu), and above 30% in 9 provinces (pink: Sichuan, Yunnan, Guangdong, Shandong, Anhui, Jiangxi, Fujian, Hainan and Jiangsu). Grey color was used to show the provinces where specimens were not available for this study. The colors ( for chicken, for duck, for goose, for pigeon) and positions of filled circles indicate different poultry species and the sampling cities. This map was created by YY and CW using Adobe Illustrator CS5 (http://www.adobe.com/products/illustrator.html).

Discussion
There are 11 closely related and distinct Chlamydia species and most animals are susceptible to multiple Chlamydia species, resulting in often asymptomatic infections with low bacterial burdens. While species-specific PCRs have been successfully used to detect individual Chlamydia species such as C. pneumoniae 19 , C. trachomatis 20,21 , C. abortus 22 , C. felis 23 , C. avium 24 , C. gallinacea 14 , C. psittaci and C. pecorum 25 , and C. psittaci and C. abortus 26 , there is a substantial need to establish a genus-specific FRET-qPCR that simultaneously in a single step detects single target copies of all Chlamydia spp. and differentiates the amplification product(s) by species. In this study, a novel FRET-qPCR followed immediately by high-resolution melting analysis of the reaction enables the sensitive detection of all chlamydial species and simultaneous differentiation of 11 species into 8 groups. Discriminatory PCRs for sequencing can be performed on Chlamydia-positive samples if further differentiation between Chlamydia spp. with similar T m (between C. avium and C. caviae; or between C. felis, C. abortus and C. psittaci) is necessary (Fig. 2). This PCR methodology provides a practical and convenient tool for the epidemiological investigation of Chlamydia spp. in a variety of sources.
C. psittaci has long been considered the main chlamydial species in poultry. While C. psittaci was detected in all four avian species investigated in this study, our data showed that C. psittaci was the dominant species only in       pigeons, representing 94.3% (100/106) of Chlamydia-positive pigeons but much lower percentages (9.1-16.3%) in chickens, ducks, or geese (  29 . Furthermore, it was reported that individuals working in a poultry slaughterhouse with C. psittaci-positive ducks presented with cough or flu-like symptoms 30 . Since zoonotic C. psittaci was detected in healthy market birds, health workers and consumers should be aware of the possibility of contracting C. psittaci infection from such birds, particularly from pigeons. The results of this study are consistent with the current idea that the epidemiology of avian chlamydiosis is complex and Chlamydia infections in birds cannot be automatically ascribed to C. psittaci. In agreement with the report from Sachse et al. 10 , C. pecorum was also detected in chickens in this study. Earlier studies also demonstrated the presence of C. abortus, C. avium and C. trachomatis in birds [8][9][10] , whereas these species have not been identified in this study. However, we reported for the first time the detection of C. suis and C. muridarum in chickens, ducks and pigeons. With the exception of the dominance of C. psittaci in pigeons, these chlamydial species represented overall sporadic detections, typically confined to "hotspots". Transmission of these chlamydiae probably occurred when the avian species had close contact to the natural hosts of C. pecorum, C. suis, or C. muridarum. In contrast, our large and China-wide study involving 2,300 domestic birds demonstrates unambiguously that C. gallinacea is the endemic chlamydial species in chickens. The agent was detected in 81.2% of the Chlamydia-positive chickens in all four bird species and in 18 of 24 Chinese provinces investigated in this study (Table 1). Phylogenetic analysis of the variable domains of the OmpA protein of C. gallinacea identified at least 13 well segregated genetic variants (probably serovars), with 10 of them clearly different from known European isolates. This high diversity is consistent with immunoselection of ompA variants in endemic infection conditions 31,32 . For the time being, poultry must be considered the main natural host for C. gallinacea. However, as observed in this study, it is not unusual that Chlamydia spp. cross host barriers and infect new hosts or even humans. This may explain that slaughterhouse workers showed atypical pneumoniae after being exposed to C. gallinacea-carrying chickens 1 .
The epidemiological survey of Chlamydia spp. enabled us to identify a flock of Chlamydia-positive chickens in a mountain village that asymptomatically carried C. suis and, to a lesser extent, C. gallinacea, C. psittaci, and C. muridarum. Interestingly, three weeks after transfer of 31 birds from that flock to a containment animal facility, C. suis and C. psittaci completely disappeared while C. gallinacea became the most prevalent species. The latter was also detected in blood, oral and cloacal swabs and 8 other organs. This persistent infection with C. gallinacea in chickens represents a major epidemiological reservoir. Previous carriage of C. suis and C. muridarum in chickens may have been the result of transient infection from continuous contact to pigs and wild mice in the mountain village.
As evident in the chicken flock from the mountain village, chickens carrying C. gallinacea usually do not show clinical signs. To characterize its pathogenic potential, we utilized a C. gallinacea isolate obtained in this study to inoculate chicken embryos and SPF broiler chickens and followed the course of infection. While inoculation of chicken embryos resulted in high mortality, intranasal inoculation of 7-day-old broiler chickens with C. gallinacea did not generate any clinical signs. However, the infected chickens showed significantly lower body weight gain from 3 weeks pi onwards. This is consistent with the report that asymptomatic endemic C. pecorum infections reduce growth rates in calves by up to 48 percent 33,34 . In combination with the ubiquitous endemic nature of C. gallinacea infection in chickens, this growth-suppressing effect of 10% or more may well be of major economic significance.
In conclusion, by using FRET-qPCR, we have demonstrated that Chlamydia spp. are common in healthy domestic birds traded on local live-animal markets in China. Our data confirmed that C. psittaci is the dominant species in pigeons, while the new emerging C. gallinacea is the endemic species in chickens. This investigation supported the notion that C. gallinacea is not a commensal, but a pathogen of moderate pathogenicity, and that persistent infection leads to reduced body weight gain of broilers. It is imperative to further investigate the agent's pathogenicity, host spectrum, transmission mechanisms, and possible zoonotic potential.

Materials and Methods
Oral and cloacal swabs. Between December 2013 and February 2014, a total of 3,962 oral and cloacal swabs were collected from 2,300 apparently healthy poultry (1,791 chickens, 215 pigeons, 179 ducks and 115 geese) in live-poultry markets in 24 provinces that encompass all seven geographical regions of China. Sterile cotton swabs were inserted into the throat and cloaca of the chicken and then turned slowly to absorb the fluid sample. All swabs were immediately placed in sterile tubes containing 400 μ l DNA/RNA stabilization buffer (Roche Molecular Biochemicals, Indianapolis, IN, USA) and then stored at − 80 °C until DNA was extracted as described below. Most of the poultry were free-range raised on family farms, and few animals were originated from industrial poultry operations. Oral and cloacal swabs were obtained from each bird (n = 1,662), but in some cases only oral (n = 476) or cloacal swabs (n = 162) were collected. All work in this study was reviewed and approved by and organs and tissues from chickens, according to the manufacturer's instructions and described before 35 . The extracted DNA was eluted in 200 μ l elution buffer.
Thermal cycling and melting curve analysis. The FRET-qPCR for Chlamydia spp. was performed in a LightCycler 480-II real-time PCR platform using a high-stringency 18-cycle step-down temperature protocol without fluorescence acquisition followed by 30 fluorescence acquisition cycles with a hybridization temperature of 51 °C 25  Sensitivity. To test the sensitivity of the FRET-qPCR, PCR amplification products from 11 Chlamydia species were gel purified with the QIAquick Gel Extraction Kit (Qiagen, Valencia, CA, USA) and quantified using the PicoGreen DNA fluorescence assay (Molecular Probes, Eugene, OR, USA). To obtain quantitative standards, the molarity of the Chlamydia DNA was determined using the calculated molecular mass of the PCR products, which then were adjusted to provide solutions containing 10,000, 1,000, 100, 10, 1 gene copies per PCR reaction in T 10 E 0.1 buffer as described previously 37 . Confirmatory sequencing. Chlamydia-positive samples based on FRET-qPCR and high-resolution melting curve analysis were further amplified for DNA sequencing to confirm the identified Chlamydia spp. PCRs were designed to target a variable region of 16S rRNA gene (697 bp) and a highly variable region of 23S rRNA gene (329 bp) for all 11 Chlamydia spp. (Table S1). To investigate the polymorphism of ompA gene in the C. gallinacea isolates and clinical specimens, the ompA PCR-1 was designed to amplify the complete ompA gene of C. gallinacea (1,188 bp) while ompA PCR-2 and ompA PCR-3 amplify the regions of variable domains (VD) 1-2 (435 bp) and VD 3-4 (421 bp) of C. gallinacea (Table S1). These PCRs were performed as described above for FRET-qPCR. The PCR products were electrophoresed through 2% agarose gel (BIOWEST ® , Hong Kong, China) and purified for automated DNA sequencing (GenScript, Jiangsu, Nanjing, China).
Monitoring of Chlamydia spp. naturally-infected chickens. Thirty-one locally-bred, free-range chickens, around 42 weeks of age, were transported from a mountain village in Jiangxi Province to a containment animal facility at the Poultry Institute in Jiangsu province. This mountain village has several swine farms and many free-range chickens. The chickens appeared healthy on arrival and were housed on the floor with free access to antibiotics-free food and water. Oral and cloacal swabs were collected aseptically and placed into tubes with 600 μ l SPG buffer for isolation of Chlamydia spp. and for DNA extraction. Two months after being moved into the animal facility, five randomly-selected chickens were euthanized to test for Chlamydia DNA in their blood samples, oral and cloacal swabs, and organs (tracheas, heart, liver, spleen, lung, kidney and pancreas) by FRET-qPCR.
Scientific RepoRts | 6:19638 | DOI: 10.1038/srep19638 Chicken embryos, broilers and C. gallinacea infection. White Leghorn chicken embryos purchased from Beijing Merial Vital Laboratory Animal Technology Co., Ltd (Beijing, China) were incubated with rotation at 37.8 °C and 60% relative humidity. The yolk sac was inoculated at embryonic development day 7 with 200 μ l C. gallinacea suspension (2 × 10 6 genomes) or SPG buffer (as mock inoculation) as described 38,39 . Candling was performed once daily to determine the vitality of the chicken embryos.
One-day-old SPF AA broiler chickens obtained from Sandeli Animal Husbandry Development Co., Ltd (Zhenjiang, China) were individually tagged and housed in a containment level 2 facilities with free access to antibiotics-free food and water. After one week, the chickens were separated into two groups that were inoculated intranasally with 20 μ l of 2 × 10 6 genomes of C. gallinacea diluted in SPG buffer (n = 15) or an equal volume of SPG buffer as control group (n = 15). The body weight of each chicken was recorded weekly before the inoculation and up to 5 weeks post inoculation while the oral and cloacal swabs were collected to detect the Chlamydia DNAs by FRET-qPCR.
Statistical analysis. Comparisons of the PCR prevalences of Chlamydia spp. in oral and cloacal swabs were analyzed by the Chi-Square Test. The two-tailed Tukey honest significant difference (HSD) test (Statistica, StatSoft, Tulsa, USA) was performed to compare the means of the body weight in mock-and C. gallinacea-inoculated chickens. Differences at P ≤ 0.05 were considered significant.