Molecular Identification of Coxiella burnetii, and Incidence and Risk Factors of Coxiellosis in Bovines of Punjab, Pakistan

Muhammad Zahid Iqbal1, Aneela Zameer Durrani1, Jawaria Ali Khan1, Nisar Ahmad2, Muhammad Usman1,*, Abdul Jabbar1, Saba Usman3, Ahsan Anjum3, Muhammad Husnain1, Nadeem Raza1 and Anwar-ul-Haq4 1Department of Clinical Medicine, University of Veterinary and Animal Sciences, Lahore 2Department of Parasitology, University of Veterinary and Animal Sciences, Lahore 3Department of Pathology, University of Veterinary and Animal Sciences, Lahore 4Livestock and Dairy Development Department, Lahore, Punjab Article Information Received 21 January 2021 Revised 11 March 2021 Accepted 18 April 2021 Available online 10 September 2021


INTRODUCTION
C oxiellosis in ruminants is caused by Coxiella burnetii, an intracellular, Gram-negative, coccobacilli bacterium, that is highly zoonotic in nature and is categorized as type B biological warfare agent (Shabbir et al., 2016). Coxiellosis distributed across the globe except in New Zealand and Antarctica and recent epidemiological studies have shown raising public health apprehension (Cruz et al., 2018;Cardinale et al., 2014).
Pakistan has almost 47.80 million cattle and 40.0 million buffalo and both species are the major contributor in livestock economy (11.2 % share in GDP) of the country (GoP, 2019). In Pakistan, the first case of Coxiellosis was reported in camel in 1955. C. burnetii causes infertility, premature birth, dead or weak calves, sporadic abortions with necrotizing placentitis and sub clinical mastitis in ruminants and resultant gigantic economic impact (Guatteo et al., 2010). Clinically infected cows may also develop infertility, metritis and mastitis. Cows have been known to shed C. burnetii in milk for up to 32 months, which indicates its public health significance (Grist, 1959).
Due to the lack of diagnostic support, the cases of Q fever remain undiagnosed and tend to be clustered with other diseases like fever and abortions in Pakistan (Shabbir et al., 2016). According to literature, the prevalence of coxiellosis ranges from 4.6% to 40% in all livestock species and 10.2% to 26.8% in humans (Ullah et al., 2019). To the best of our knowledge, only two studies regarding C. burnetii in cattle and buffalo are available since 1955 to 2019 and in those studies this antigen was detected serologically (ELISA and CFT). Currently, not a single molecular study on coxiellosis in cattle and buffalo is available from Pakistan.
Keeping in view the previously mentioned paucities and dearth in coxiellosis studies in Pakistan, the current study was designed to investigate molecular evidence of C. burnetii in cows and buffaloes. In addition, different O n l i n e

F i r s t A r t i c l e
associated risk factors were also taken into account. The study will lead to establish diagnosis of coxiellosis in Pakistan and provide a needful insight to policy makers because coxiellosis is not only an issue of livestock sector but also a major public health concern.

Study area and target population
Cattle (n=160) and buffaloes (n=160) were selected from the 09 Union councils (Pattoki I, Pattoki II, Pattoki III, Wan Adhan, Shahikham, Ghumanke, Awan Chak, Baharwal, and Rodhay) located in close proximity to Veterinary Teaching Hospital, University of Veterinary and Animal Sciences, Ravi Campus, Pattoki, District Kasur Pakistan. Most cases of subclinical mastitis, endometritis and abortion were reported from these UCs (Fig. 1).
Inclusion criteria for selecting animals were tick infestation or its history, subclinical mastitis, endometritis, and abortion or recent history of these problems. The animals that fulfilled any of the two criteria were selected.

Collection of samples
A total of 320 blood samples (n=160 cows, n=160 buffaloes) were collected aseptically in plain vacutainers and after proper labeling transported to Medicine Laboratory, University of Veterinary and Animal Sciences, Lahore, where these samples were immediately centrifuged for separation of buffy coat and stored at -20°C until further processing. Sampling for the second phase were conducted from C. burnetii positive (n=40) and negative (n=40) cows and buffaloes after confirmation of coxiellosis during the first phase of study. These samples were collected in EDTA mixed vacutainers and immediately processed for hematology. During sample collection, information about all the potential risk factors was also collected in a questionnaire.

O n l i n e F i r s t A r t i c l e
Molecular identification and phylogenetic analysis of C. burnetii The genomic DNA from the samples was extracted using Thermo Scientific GeneJET Genomic DNA purification kit (catalogue # K0721) according to manufacturer's guidelines. In 200 μL serum, 400 μL lysis and 20 μL of Proteinase K solution were added and mixed by vortexing. The incubation was done at 56°C for 10 min and then 500 μL wash buffer 1 and 2 were added according to guidelines given. At the end, 200 μL elution buffer was added to elute the genomic DNA. The concentration of the genomic DNA was measured by using spectrophotometer (Nanodrop, USA).
PCR was performed using Thermo Scientific TM DreamTaq Green PCR Master Mix (2X) (catalog number K1081). A total volume of 25 μl reaction mixture was composed of forward primer 1.25 μl, reverse primer 1.25 μl, Master mix 12.5 μl, DNA 2 μl, and water 8 μl. The thermal cycle comprised initial denaturation at 95C for 3 min followed by 32 cycles, each of denaturation at 95C for 30 sec, annealing at 60°C for 30 sec and extension at 72°C for 1 min. It was followed by final extension at 72°C for 7 min. Prior to our sample analysis, the assay was optimized and validated using Amplirun ® Coxiella burnetii DNA Control (catalogue number MBC018).
The resultant PCR product was further evaluated with agarose gel electrophoresis (1.5%) with apposite controls and a 50bp ladder (Invitrogen Co. Carlsbad, CA, USA). The PCR products of IS1111 transpose gene were submitted for DNA sequencing to Applied Biosystems, Foster City, CA, USA. The resulting sequences will be submitted to NCBI gene bank after the publication of this paper. These sequences were edited and aligned using BioEdit software version 7.2.5 (http://www.mbio.ncsu. edu/bioedit/page2.html). If many sequences were found similar in alignment then only one of these sequences was processed for phylogeny. The phylogenetic tree was constructed by comparing with the already reported C. burnetti strains from Pakistan and other countries on GenBank NCBI. The analysis was done using Maximum Likelihood statistical method and Tamura-Nei Model and with the bootstrap analysis for 1000 replicates onto MEGA version 6.0 software (Tamura et al., 2013).

Statistical analysis
Pearson Chi-square test was applied to determine the association with potential risk factors. The odds ratio was also calculated by Fisher's exact test in order to find out the probability of risk within the study area and other risk factors. Hematology results were tested using two sample t-test. Level of significance was accepted at 95% confidence interval in every method. Statistical analysis was performed at IBM SPSS version 23.0 software.

RESULTS
The study reported an overall 22.5% prevalence of coxiellosis in bovines from the study area. The prevalence was higher in cows (32.12%) than in buffaloes (12.5%). Samples were collected from nine UCs and there were significant (p>0.05) differences in the prevalence of coxiellosis in all UCs (Table I). Table I shows the prevalence of coxiellosis in different UCs.

Association of clinical biomarkers with occurrence of coxiellosis
Association of five potential clinical biomarkers of coxiellosis in bovines was studied and results showed that all of these i.e., subclinical mastitis, abortion, post-abortion infertility, subclinical mastitis and abortion, subclinical mastitis and post-abortion infertility, were significantly (p>0.05) associated with coxiellosis (Table II).
Association of these clinical biomarkers was also computed with Fisher's exact test and odds of having coxiellosis were calculated. Risk of occurrence of coxiellosis was 6.3 times more in animals with subclinical mastitis than those without it. Animals having abortion were 5.9 times more likely to be infected with coxiellosis than those without it. Odd ratio of animals with postabortion infertility showed 13 times more likelihood of having coxiellosis than animals, which did not have this problem. The risk of coxiellosis was 11 times more in animals suffering from subclinical mastitis and abortion in comparison with those that did not suffer from these issues.

O n l i n e F i r s t A r t i c l e
The most significant (p>0.05) finding was in bovines which had subclinical mastitis as well as post-abortion infertility. In this case, risk of being C. burnetii positive was 20.8 times more when compared with animals without these problems (Table II).

Association of animal related risk factors with coxiellosis
Association of age, species, parity number, disease history, and body condition score (BCS) as animal related risk factors with occurrence of coxiellosis in bovines was verified (Table III). The occurrence of coxiellosis significantly (p<0.05) increased with age and became 252 times more in animals above ten years of age. Similarly, prevalence of coxiellosis varies significantly (p<0.05) with species and cows were 3.4 times more at risk than buffaloes. Interestingly, parity number also found to be significantly (p<0.05) associated with coxiellosis but only after the second parity and became highest at 5 th parity with 19 times more probability of having coxiellosis. Animals with disease history were also 2.9 times more at risk of having coxiellosis than healthy ones. Similarly, BCS 1 was significantly (p<0.05) associated with coxiellosis however BCS 2 and 3 showed non-significant (p>0.05) association. Chances of coxiellosis were 6.4 times more in BCS 1 animals than those with BCS 3. And, finally, tick infestation or a previous history of it were also significantly (p<0.05) associated with coxiellosis and, interestingly, animals with previous history of tick infestation were 2.3 times more prone to coxiellosis than those which had tick infestation at the time of sampling (Table III).

Association of coxiellosis with risk factors related to management practices at farms
Data of different risk factors related to management practices at farms were also collected in a predesigned questionnaire. These data were evaluated for computing statistical association with the occurrence of coxiellosis. From different hypothesized risk factors, type of farm i.e., smallholder, commercial, or progressive, presence of goat and dog, and carcass disposal i.e., burning or burial, were found non-significantly (p>0.05) associated with occurrence of coxiellosis in bovines. The rest of risk factors showed mixed results (Table IV). Type of feeding was found significantly (p<0.05) associated with coxiellosis and farms which practiced grazing were 28.7 times more at risk of having C. burnetii positive animals than farms exercising stall-feeding. Similarly, drinking water source was also significantly (p<0.05) associated with our disease of concern and probability of risk was 13.75 times more in farms which offered canal water than those which offered only underground water to their inhabitant animals. The most notable result was significant (p<0.05) association of O n l i n e coxiellosis with the presence of sheep at farm. Animals living in proximity to sheep within the same farms were 7.6 times more at risk than those that lived in solitary farms. Disinfection, manure and parturition management were significantly (p<0.05) associated with coxiellosis and animals residing in farms which lack these practices were 9.7, 3.1, and 4.8 times more prone to coxiellosis, respectively (Table IV).

Phylogenetic analysis
A phylogenetic tree of our isolates and along with those in GenBank database is depicted in Figure 2. When our sequences were aligned based on their characterization from buffaloes and cattle, it was found that sequences isolated from the same species were 100 percent similar. Hence, only one sequence from each species was included to determine phylogeny of C. burnetii isolated from bovines in our study area. The strain C. burnetti Pak/ buffalo/blood isolated from blood of buffalo is labelled with red triangle while the strain C. burnetti Pak/cattle/ blood isolated from blood of cattle is labelled with blue triangle. Our isolate C. burnetti Pak/cattle/blood falls in the same clade with the strains isolated from cattle blood (accession No. MF445016.1) and human blood (accession No. MK078517.1) samples in India. While the isolate C. burnetti Pak/buffalo/blood falls in different clade with the strains isolated from the environmental samples from different regions of Pakistan and with the isolates from goat and bovine milk samples of Indian origin.

DISCUSSION
This is the first study in Pakistan that provided a comprehensive account of molecular prevalence of coxiellosis and its associated risk factors in bovines. Owing to the lack of studies in humans and animals and thereby unavailability of genetic information related to the strains of C. burnetii circulating in Pakistan, we used the primer set which targeted IS1111 gene that is present abundantly in the entire genome of C. burnetii. Thus, targeting IS1111 gene enhanced the sensitivity of our PCR as compared to single copy target gene PCR (Torez et al., 2014). We hypothesized that in Pakistan, coxiellosis has been causing huge damage to bovines since long. Because it is not reported yet so, its cases tend to mix with other diseases and thereby managed improperly.
According to this study, the overall prevalence of coxiellosis in bovines was 22.5%, slightly higher than the previous studies in which prevalence ranged from 7% to 19.3% (Keshavamurthy et al., 2019;Klemmer et al., 2019). This trivial surge in prevalence is probably due to different sampling strategies: selective sampling technique was adopted in this study while other studies used random sampling method. Apart from this, prevalence at species level was nearly similar to erstwhile studies conducted by De Biase et al. (2018) and Gache et al. (2017) who reported 25% and 26% prevalence of coxiellosis in cows, respectively. However, this approximate similarity is due to the similar sampling strategies. As far as prevalence of coxiellosis in buffaloes is concerned, there are very few studies available and these studies documented the prevalence range from 4.8 to 17.5% and results of our study lies within this range (Klemmer et al., 2018). The low prevalence rate of coxiellosis in buffaloes might be due to less susceptibility of buffaloes to infection as compared to cow (Dua, 2003;Rashid et al., 2019).
We expected that animals having the issues of subclinical mastitis, abortion, post-abortion infertility, subclinical mastitis and abortion, subclinical mastitis and post-abortion infertility are likely to be C. burnetii positive. This is confirmed by the results of this study and all of these clinical biomarkers were found significantly (p<0.05) associated with coxiellosis. De Biase et al. (2018) and Gache et al. (2017) also reiterated that coxiellosis is significantly associated with mastitis and reproductive problems; whereas Agerholm (2013)  Association of the different animal-related risk factors with occurrence of coxiellosis was also verified in our study. The occurrence of the coxiellosis was higher in older animals and those with high parity number. This might be due to the fact that with increasing age exposure time to infection increases. Our results coincide with those of Ullah et al. (2019) who also confirmed the higher prevalence of coxiellosis in small ruminants with high parity number. Besides, animals having previous history of any disease or those with low BCS also become vulnerable to C. burnetii due to their compromised immune system compared with healthy animals having intact immune system. These findings are consistent with those of Dhaka et al. (2019) and Ullah et al. (2019). We hypothesized that tick infestation is the most crucial factor in the occurrence of coxiellosis in bovines due to the role of ticks in the lifecycle of C. burnetii (Duron et al., 2015). Our findings confirmed that C. burnetii was found significantly (p<0.05) higher in animals which had tick infestation at the time of sampling rather than those with previous history of tick infestation. This finding is compatible with those of Knobel et al. (2013) and Galay et al. (2020).
The management practices at farm were also taken into account to establish their association with occurrence of coxiellosis. The disease incidence was significantly (p<0.05) higher in smallholder farms, and those which lack manure management, disinfection, outside carcass disposal and separate calving area. To the best of our knowledge, no previous study reported the association of these factors with the occurrence of coxiellosis. Likewise, the incidence of the disease was found significantly (p<0.05) higher in grazing animals than in stall-fed animals. The previous studies also reiterated that grazing is associated with acquiring and transmission of infection among livestock (Rizzo et al., 2016). We also sought to inquire the role of other animals; namely, dog, sheep, and goat, in the occurrence of coxiellosis in bovines. Of these animals, the presence of sheep in the farm were found significantly (p<0.05) associated with the occurrence of this disease in bovines. Our results are partially compatible with those of Rashid et al. (2019) who reported that mixing of small ruminants with bovines increased the incidence of coxiellosis in the latter species. However, our study did not confirm that goat companionship led to the significant (p<0.05) disease incidence in bovines. Regarding the role of dogs, our results differ from those of Porter et al. (2016) who reported that dog was a source of transmission of coxiellosis in bovines.
The distinct clustering of C. burnetii sequences from this study indicates potential genetic variability in the isolates from different species in the same geographical area.
The findings of C.burnetti Pak/cattle/blood are particularly striking which clades with Indian isolates contrary to C.burnetti Pak/buffalo/blood of the same area that shows genetic similarity with environmental isolates of the same geographical origin. A previous study confirmed that the genetic variability is associated with the pathogenicity of different strains in domesticated ruminants (Martinov, 2007). So, the genetic variability reported in the current study not only confirms the circulation of different strains on C. burnetii in cows and buffaloes but also emphasizes the requirement of further studies to explore these strains and their pathogenesis.
Following the confirmation of coxiellosis in bovines, hematology profile of bovines was also investigated. This hematology profile describes that C. burnetii affects various parameters differently in cows and buffaloes. This variation among cows and buffaloes may be partly due to different strains and partly due to some other factors. However, due to non-availability of any study related to exact pathogenesis of C. burnetii and its correlation with hematology in bovines, it is premature and erroneous to describe any reason of this change. To the best of our knowledge, this is the first study that reported the hematology profile of C. burnetii positive cows and buffaloes. In small ruminants, one study is available which described that hematology remains unchanged in C. burnetii positive animals but these results are dissimilar with our results (Čebulj-Kadunc et al., 2014).
Further studies are required to characterize the strains of C. burnetii by strain-specific genes, validate hematology of C. burnetii positive bovines, and evaluate the potential impacts of coxiellosis on animals and humans in Pakistan.

ACKNOWLEDGMENT
The study was conducted by the support of PAK-US Science and Technology Project number: 1122 titled "Capacity Building for Vector-borne neglected diseases of Livestock" funded by Higher Education Commission (HEC), Pakistan.

Ethics approval and consent to participate
The study was not based on any clinical trial. Institutional Review Committee for Biomedical Research, University of Veterinary and Animal Sciences, Lahore approved the sample collection methodology. Moreover, permission of animal owners was sought before sample collection.

Statement of conflict of interest
The authors declare no conflict of interest regarding publication of this manuscript.