In the past decade, new circoviruses have been found in humans, bobcats, giant pandas, and domestic canines and felines, among other species [1-5]. Circoviruses are grouped into the order Cirlivirales, composed of circular viruses (15–25 nm in diameter) with positive-sense single-stranded DNA of 1.7–3.8 kb. Within this order, the family Circoviridae is divided in two genera: Circovirus, which are found in mammals, birds and fish; and Cyclovirus, which affect mammals, birds and arthropods [6]. In domestic cats, two circoviruses have been described: feline stool-associated circular DNA virus (FeSCV) [7] and feline-associated cyclovirus (FECV) [8].
First identification of FeSCV occurred in 2018 in diarrheic and healthy cats from the same cattery [7]. Phylogenetic analysis showed that it didn’t belong to the genus Circovirus nor Cyclovirus, and was considered an unclassified circular DNA virus [7, 9]. Recently, FeSCV was classified in the ICTV as phylum Cressdnaviricota, class Arfiviricetes, order Cirlivirales, family Vilyaviridae, genus Andurilvirus, species Andurilvirus erebor.
In 2014, FeCV (order Cirlivirales, family Circoviridae, genus Cyclovirus, species Cyclovirus gato) was identified in a virome study of domestic cat feces in the United States [8]. However, there was no data on its pathogenesis or prevalence. In this paper we report the first description of FeSCV in Brazil and two novel cycloviruses detected in cats’ fecal samples. We also analyzed risk factors associated with FeSCV infection.
DNA samples (137) from rectal swabs or stool from 83 cats (Supplementary Information 1) were analyzed. Out of these samples, 29 were from single sampling and 108 were serial sampling taken from 54 cats, 3 to 12 months apart from each other. Cats were classified by age according to the American Association of Feline Practitioners (AAFP) and American Animal Hospital Association (AAHA) guidelines (2021) [10]. Previous testing for feline immunodeficiency virus (FIV - Lentivirus felimdef) and feline leukemia virus (FeLV - Gammaretrovirus felleu) was carried out in 87.9% (73/83) of the cats using SNAP FIV/FeLV combo test (IDEXX), and 96.4% (80/83) of rectal swabs were tested for feline coronavirus (FCoV - Alphacoronavirus suis) with RT-qPCR targeting the E gene.
A clarification protocol with three cycles of freezing/thawing were performed in all samples, prior to DNA extraction with RiboPure™ RNA Purification kit (Ambion®), in 123 swab samples from a previous research project, or PureLink™ Genomic DNA Mini Kit (Invitrogen) in 14 fecal samples, as recommended by the manufacturers. The extracted DNA was stored at −20 °C until use. The integrity of nucleic acids was confirmed with 16S gene PCR [11], with 1× DreamTaq Green PCR Master Mix (ThermoFisher Scientific), 400 nM of primers 519A (sense) and 907B (antisense) and 2 μL of extracted DNA, in a 20 μL reaction. Amplification conditions were: 94 ºC/5 min; 35 cycles of 94 ºC/30 s, 50 ºC/30 s and 72 ºC/30 s; and 72ºC/5 min. Positive and negative control were, respectively, DNA from a culture of Leptospira interrogans serovar Canicola and ultrapure water.
Samples were screened for circoviruses using a hemi-nested PCR targeting the Rep gene [1]. The 20 μL reaction contained 1× DreamTaq Green PCR Master Mix (ThermoFisher Scientific), 500 nM of each primer pair (CV-F1/CV-R1 or CV-F1/CV-R2) and 2 μL of extracted DNA or 1 μL of PCR product, respectively for the PCR and hemi-nested PCR reaction. The PCR conditions were: 94 ºC/5 min; 14 cycles of 94 ºC/30 s, 48 ºC/30 s (with an increase of 0.5 ºC per cycle) and 72 ºC/1 min s; 31 cycles of 94 ºC/30 s, 55 ºC/30 s and 72 ºC/1 min; and 72 ºC/5 min; and hemi-nested PCR were: 94 ºC/5 min; 40 cycles of 94 ºC/20 s, 58 ºC/20 s and 72 ºC/45 s and 72 ºC/5 min. Positive and negative control were PCV-2 DNA sample and ultrapure water, respectively.
After, samples were tested with specific primers for FeSCV targeting the Rep gene [7], with some modifications: FeSCV1F-alt (5’ GCT AAG GTC TGG CTC AG 3’) and FeSCV1R-alt (5’ CTA TRT CCA GGT CGG GAG 3’). PCR used a total volume of 20 μL, with 1× DreamTaq Green PCR Master Mix (ThermoFisher Scientific), 400 nM of each primer, and 2 μL of extracted DNA. Amplification conditions were: 94 ºC/5 min; 50 cycles of 94 ºC/20 s, 53 ºC/20 s and 72 ºC/30 s and 72 ºC/5 min. A positive sample from circoviruses hemi-nested PCR was used as positive control and ultrapure water as negative control.
All amplicons were analyzed by electrophoresis in a 1.5% agarose gel stained with SYBR® Safe DNA gel Stain (Invitrogen™) compared with GeneRuler 100 bp DNA ladder (Thermo Scientific™), and were purified with PureLink™ Quick Gel Extraction Kit (Invitrogen) or ExoSAP-IT Express PCR Product Cleanup Reagent (ThermoFisher Scientific™), as the manufacturer's instructions. Bidirectional sequencing was performed by BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems), using EDTA and ethanol purification protocol, and the ABI 3500 sequencer (Applied Biosystems), following manufacturer's recommendations. The reliability of resulting sequences was analyzed with Phred Electropherogram Quality Analysis software [12], checked and edited manually on MEGA X software [13], and the consensus sequences were assembled with Cap-Contig program available in Bioedit software [14].
Seven sequences were aligned with GenBank database reference sequences of FeCV, FeSCV and other cycloviruses, circoviruses and circular Rep-encoding ssDNA virus (CRESS DNA virus), using codon alignment with Muscle tool implemented in MEGA X software [13]. The unrooted nucleotide phylogenetic tree was constructed using Maximum Likelihood (ML) method and Tamura-Nei 93 (TN93) evolutionary model, with 1,000 replicates (bootstrap), gamma distribution and invariable sites [13]. The nucleotide and amino acid identity of sequences were calculated using Bioedit software [14].
The association of FeSCV infection regarding risk factors, such as age, sex, breed, coinfection with FIV, FeLV, FCoV with and without sign of feline infectious peritonitis (FIP) were statistically analyzed using Fisher's exact test and Qui-square ran by R v4.5.
Eight cats (9.6% - 8/83) were positive for circoviruses hemi-nested PCR, but only seven sequences were submitted to phylogenetic analyses (Figure 1). Sample 6 shared an identity of 97.8% for nucleotides (nt) and 98.3% for amino acids (aa) with FeSCV KU14 (NC040381). Sequences from samples 32, 34, 36, 38 and 96 were tightly related to each other (99.7% - 100% nt and 100% aa identity) and segregated with bat-associated cyclovirus (Cyclovirus bastao, isolate POA/2012/VI - NC025792), showing 95.6 - 95.9% nt and 93.9% aa of identity. These five samples were from the same cattery, São Paulo 7, and despite this virus being clustered with bat associated cyclovirus, found in insectivorous bats from South Brazil [15], it has a low amino acid identity with our samples, which could suggest a new viral species. Sample 54 segregated with army ant cyclovirus (Cyclovirus fourmi - ON324073), with identity of 98.2% (nt) and 100% (aa). They may be feline specific viruses or a species hosted by insects, as cats hunt these animals and cycloviruses are likely to have them as hosts. Regarding FeCV (NC024700), all sequences from this study had low identity, with only 39.2 - 48.1% (nt) and 32.5 - 41.5% (aa) of identity, which shows that they are not the same virus.
Concerning the risk factors associated with circoviruses (Table 1), females (15.4% (8/52) were more likely to have circoviruses (p-value = 0.02275). However, the study had more females than males. Most of the Cyclovirus positive cats were breeding animals, who frequently experienced heat, gestation and lactation, which can be stressful and could lead to immunosuppression. Besides, cattery 7, where the majority of the felines positive for Cyclovirus were housed (5/7), had inadequate welfare and sanitary conditions, another contributing factor to increased stress.
When screening for FeSCV specifically, 19.3% (16/83) of the cats were positive. Only one cat remained positive in the second collection 10 months later, while all other cats showed positive results only at first or second sampling. Analysis of different risk factors (Table 1), showed that regarding age, young adults and kittens together (14/52 - 26.9%), were more likely to be positive to FeSCV (p-value = 0.00912). Young cats tend to interact more with other cats, thus they would be more prone to virus exposure, furthermore young animals are developing immune responses to most pathogens, which could make them more susceptible when exposed to pathogens.
Table 1. Risk factors associated with circoviruses and FeSCV
|
Risk factors
|
Number of animals
|
Circoviruses
|
FeSCV
|
|
Positive
|
Negative
|
p-value
|
Positive
|
Negative
|
p-value
|
|
Sex
|
Females
|
52 (62.6%)
|
8 (15.4%)
|
44 (84.6%)
|
0.02275*
|
8 (15.4%)
|
0.2443
|
0.2443
|
|
Males
|
31 (37.4%)
|
0
|
31 (100%)
|
8 (25.8%)
|
23 (74.2%)
|
|
Age
|
Kitten
|
14 (16.9%)
|
2 (14.3%)
|
12 (85.7%)
|
0.2479
|
6 (42.9%)
|
8 (57.1%)
|
0.00912**,a
|
|
Young Adult
|
38 (45.8%)
|
5 (13.2%)
|
33 (86.8%)
|
8 (21.0%)
|
30 (78.9%)
|
|
Mature Adult
|
13 (15.7%)
|
0
|
13 (100%)
|
1 (7.7%)
|
12 (92.3%)
|
|
Senior
|
16 (19.3%)
|
1 (6.3%)
|
15 (93.7%)
|
0
|
16 (100%)
|
|
No data
|
2 (2.4%)
|
0
|
2 (100%)
|
1 (50.0%)
|
1 (50.0%)
|
|
Breed
|
DSH
|
52 (62.6%)
|
3 (5.8%)
|
49 (94.2%)
|
0.2271
|
8 (15.4%)
|
44 (84.6%)
|
0.1991b
|
|
Persian
|
17 (20.5%)
|
4 (23.5%)
|
13 (76.5%)
|
5 (29.4%)
|
12 (70.6%)
|
|
Maine Coon
|
10 (12%)
|
1 (10%)
|
9 (90%)
|
3 (30.0%)
|
7 (70.0%)
|
|
Birman
|
1 (1.2%)
|
0
|
1 (100%)
|
0
|
1 (100%)
|
|
Siamese
|
3 (3.6%)
|
0
|
2 (100%)
|
0
|
3 (100%)
|
DSH: Domestic shorthair;
statistical significance: *p<0.05; **p<0.01
aComparing young adult and kitten together with mature adult and senior together
bComparing DSH with all the breeds combined
No cats were FIV positive and 5.5% (4/73) were FeLV positive. Out of the 80 cats tested for FCoV, 52 (65.0%) were positive and 28 (35.0%) were negative. One cat was positive for feline associated cyclovirus and FeSCV at the same time, and only one cat tested positive for FeLV, FCoV and FeSCV at first sampling. This patient was retested after four months and remained positive for FeLV and FCoV but tested negative for FeSCV. Because a large number of animals showed different results for FCoV and FeSCV between the first and second sampling, the statistical analysis was carried out with a total number of samples (134), instead of the number of cats (80). Among FeSCV positive samples, 58.8% (10/17) were also FCoV positive (Table 2). Out of those, 50% (5/10) showed clinical signs of FIP, which was statistically relevant (p-value = 0.006861). The majority of the circovirus positive samples (6/8 - 75.0%) were FCoV positive, but it was not statistically significant. It was not possible to associate risk factors with retroviruses since only 4 cats were positive for FeLV and none for FIV.
Table 2: Association between FeSCV and FCoV
|
FCoV
|
Number of samples
|
Circoviruses
|
FeSCV
|
|
Positive
|
Negative
|
p-value
|
Positive
|
Negative
|
p-value
|
|
FCoV +
|
with FIP
|
8
|
1 (12.5%)
|
7 (87.5%)
|
0.1618
|
5 (62.5%)*
|
3 (37.5%)
|
0.3983
|
|
w/o FIP
|
58
|
5 (8.6%)
|
53 (91.4%)
|
5 (8.6%)
|
53 (91.4%)
|
|
FCoV-
|
68
|
2 (2.9%)
|
66 (97.1%)
|
7 (10.3%)
|
61 (89.7%)
|
|
Total
|
134
|
8 (6.0%)
|
126 (94.0%)
|
|
17 (12.7%)
|
117 (87.3%)
|
|
*p-value = 0.006861, comparing FCoV positive samples with and without sign of FIP
This study shows the occurrence of FeSCV in feline stool outside of Asia, a previously unknown fact. In Japan, FeSCV was identified in 65.0% (13/20) of fecal samples from domestic cats by PCR with specific FeSCV primers. Among the positives, 76.9% (10/13) were diarrheic and 23.1% (3/13) were healthy [7]. Similar to our study, the circoviruses hemi-nested PCR screening had less sensitivity, so, it is essential to perform the specific PCR for FeSCV. In China, Hao and colleagues [9] found FeSCV in 20.0% (2/10) of samples from diarrheic cats with the FeSCV specific primer, similar to the occurrence in our study (19.3%), but unfortunately, we did not have information about stool consistency.
Due to small sample sizes and scarcity of studies, there is still not enough evidence to determine FeSCV pathogenicity. Both studies suggest it is possibly related to enteric diseases in domestic cats, especially in co-infection with enteric viruses, such as FCoV, Feline Panleukopenia Virus, Feline Astrovirus, Feline Kobuvirus and Feline Bocavirus [8, 9]. In our study there was no statistical association with co-infection with FeSCV and FCoV, despite 58.8% (10/17) of FeSCV positive samples being also positive for FCoV, with half of them having FIP signs. This indicates that animals with FCoV, especially those with signs of FIP, may be more susceptible to FeSCV infection (p-value = 0.006861). Unfortunately, it was not possible to verify the correlation with retrovirus infection, which is highly related to decreased cat immunity, as there were no FIV and FeLV animals in the sample in significant numbers.
FeSCV-positive animals from the same cattery were diagnosed in the same sampling. In cattery São Paulo 7, all the positive results (4/10) were on the second sampling (one year after the first). During the interval of sampling, 4 new cats were introduced into the cattery, and it is possible that one of these animals introduced the virus to the group. However, there were no samples of these cats. In this cattery, one animal (sample 38) was positive for FeSCV and feline associated cyclovirus at the same time. Felines from the catteries São Paulo 1 (3/10), São Paulo 4 (1/9) and Rio de Janeiro (1/19) were positive in the first collection, only one of them was still positive on the second sampling. This cat was FCoV positive in both collections but didn’t show FIP signs. This demonstrates that it is possible to recover from FeSCV infection. Jundiaí cattery had 4/5 FeSCV positive cats. These animals had been sampled only once, they were FCoV positive and showed signs of FIP. This study as well as Takano and colleagues [7] sampled cats from catteries. Enclosed spaces with multiple animals, such as catteries, may favour dissemination of viruses, as the animals are in closer contact with each other.
Although two novel cycloviruses were found, the sequenced fragments were small (341-500 nt), therefore, it would be necessary to sequence the entire virus genome to determine the species precisely. Other Cressdnaviricota species have already been described in domestic felines, such as canine circovirus (species Circovirus canine) [6]. In wild felids, a new species of Circovirus was identified in stool samples of Lynx rufus (Bobcat). The study was carried out in Mexico and obtained 13.6% (3/22) of positive samples [2]. Susceptibility to circoviruses in phylogenetic close species has already been observed by the detection of PCV type 3 in 10.0% (7/70) of serum samples from free-ranging wild boars in Brazil [16]. Further investigation of Cressdnaviricota viruses in domestic cats and wild felids, such as epidemiological data on viral distribution and development of clinical symptoms and associate the co-infection with other feline viruses, and analysis of the entire genome is required for phylogenetic classification and for a better understanding of the risks to animal and human health.