High-resolution melting analysis for simultaneous detection and discrimination between wild-type and vaccine strains of feline calicivirus

Abstract

High-resolution melting (HRM) analysis, a post-polymerase chain reaction (PCR) application in a single closed tube, is the straightforward method for simultaneous detection, genotyping, and mutation scanning, enabling more significant dynamic detection and sequencing-free turnaround time. This study aimed to establish a combined reverse-transcription quantitative PCR and HRM (RT-qPCR-HRM) assay for diagnosing and genotyping feline calicivirus (FCV). This developed method was validated with constructed FCV plasmids, clinical swab samples from living cats, fresh-frozen lung tissues from necropsied cats, and four available FCV vaccines. We performed RT-qPCR to amplify a 99-base pair sequence, targeting a segment between open reading frame (ORF) 1 and ORF2. Subsequently, the HRM assay was promptly applied using Rotor-Gene Q® Software. The results significantly revealed simultaneous detection and genetic discrimination between commercially available FCV vaccine strains, wild-type Thai FCV strains, and VS-FCV strains within a single PCR reaction. There was no cross-reactivity with other feline common viruses, including feline herpesvirus-1, feline coronavirus, feline leukemia virus, feline immunodeficiency virus, and feline morbillivirus. The detection limit of the assay was 6.18 × 10¹ copies/µl. This study, therefore, is the first demonstration of the uses and benefits of the RT-qPCR-HRM assay for FCV detection and strain differentiation in naturally infected cats.

Keywords:

• High-resolution melting analysis
• feline calicivirus
• cat
• polymerase chain reaction

Introduction

Feline calicivirus (FCV) is a nonenveloped, single-stranded, positive-sense RNA virus that belongs to the genus Vesivirus, family Caliciviridae (Weiblen et al. 2016; Pereira et al. 2018; Vinje et al. 2019). FCV is a common causative agent of upper respiratory tract disease and stomatitis in cats (Hofmann-Lehmann et al. 2022). The virus was first discovered in America in 1957 (Fastier 1957) and subsequently spread to several countries in Europe and Asia (Bannasch and Foley 2005; Coyne et al. 2007; Hou et al. 2016; Afonso et al. 2017; Zhou et al. 2021). Formerly, most FCV-infected cats were considered self-limited and often presented mild upper respiratory symptoms and oral diseases (Gaskell et al. 2012). Other clinical symptoms, including abortion, acute jaundice, acute febrile lameness syndrome, chronic gingivostomatitis, severe pneumonia, and ulcerative dermatitis, have been reported in FCV-infected cats (Schorr-Evans et al. 2003; Hurley et al. 2004). Within the last decade, the emergence of a highly virulent systemic FCV (VS-FCV) strain has been the primary cause of the more severe, tentatively termed virulent systemic disease. VS-FCV infection results in a higher rate of mortality in FCV-vaccinated cats (Hurley et al. 2004; Reynolds et al. 2009; Battilani et al. 2013).

Based on the nature of the RNA virus, the lack of a proofreading mechanism of the viral polymerase gene plays a major role in the increased probability of RNA virus mutation and evolution rates (Coyne et al. 2012). In some virus families, including caliciviruses, intragenomic recombination may also influence rapid evolution (Posada et al. 2002). Intragenomic recombination can encourage increasing genetic diversity and dispose of the gathering of deleterious mutations (Worobey and Holmes 1999) that can be a major mechanism to contribute to the emergence and evolution of new variants (Zhou et al. 2021). Mostly, recombination events of caliciviruses, have been crossed-over point between the open reading frame (ORF) 1 and the ORF2, which is considered as an important region used for genotyping and vaccine-associated epitope (Bull et al. 2005; Coyne et al. 2006; Hansman et al. 2007; Symes et al. 2015). This phenomenon in FCV mutation often leads to a genetic recombination event in the ‘genomic hotspot’ between the 3′ terminus of ORF1 and 5′ terminus of ORF2 (Symes et al. 2015; Zhou et al. 2021) that encode the RNA-dependent RNA polymerase (RdRp) protein and leader of the capsid (LC) protein, the initial part of capsid protein (VP1), respectively. The RdRp and LC proteins are responsible for viral replication (Smertina et al. 2019) and promoting viral spread (Abente et al. 2013), respectively. Although there is still no reliable tool used for the FCV genotyping, this partial region between the RdRp and VP1 gene is commonly employed as a routine genotyping for Norovirus and Sapovirus (Tatusov et al. 2021). Moreover, there is a study that has revealed a significant association between the genetic diversity of FCV and the level of neutralizing titer, especially in recombinant strains (Zhou et al. 2021).

Even though FCV vaccination is widely conducted, according to the 2020 AAHA/AAFP feline vaccination guideline (Stone et al. 2020), it cannot cross-protect cats susceptible to other FCV field strains that have emerged (Hofmann-Lehmann et al. 2022), as evidence of VS-FCV outbreaks has been reported (Schorr-Evans et al. 2003; Hurley et al. 2004; Battilani et al. 2013). Regarding the rapid mutation of the RNA virus, FCV could escape vaccine-induced immunity, especially in cats vaccinated with the old FCV strain-derived vaccine. FCV-associated diseases in FCV-vaccinated cats have been reported, resulting in recommendations for further vaccination with other FCV strain-derived vaccines for broader protection (Lauritzen et al. 1997; Addie et al. 2008; Afonso et al. 2017; Smith et al. 2020; Zhou et al. 2021; Hofmann-Lehmann et al. 2022).

Although virus isolation is a gold-standard method, molecular techniques such as conventional reverse-transcription PCR (RT-PCR), nested RT-PCR, and real-time RT-PCR (RT-qPCR) have been developed, providing a faster and more reliable tool for virus detection. However, post-PCR analysis, such as genomic sequencing, is still required in the case of lineage classification and differentiation of viral strains, which is sometimes impractical and time-consuming. While nested RT-PCR offers increased specificity, it involves multiple steps, which can be considered time-consuming and may facilitate contamination; the major disadvantage of this technique is due to product carrying between two reactions (Akhoundi et al. 2017; De Silva et al. 2022). Over the last decade, the high-resolution melting (HRM) assay has been widely employed in several manners, such as for mutational scanning, repeat typing, genetic variant scanning, and genotype differentiation of the infected virus (Reed et al. 2007), in both human and veterinary medicine (Chua et al. 2015; Sun et al. 2019; Vaz et al. 2019). Additionally, the development of the HRM assay offers a cost-efficient, one-step closed-tube system that can reduce time, and contamination and enables high-throughput analysis without requiring any post-PCR processing. (Marin et al. 2016).

In addition, the HRM assay has become a new method that is simple, fast, inexpensive, and effective for initial scanning and monitoring genetic variants (Toi and Dwyer 2008; Tajiri-Utagawa et al. 2009; Marin et al. 2016). Therefore, this study aimed to establish an HRM assay for concurrent diagnosis and discrimination between the FCV-vaccine strain (FCV-Vac), virulent FCV (VS-FCV), and wild-type FCV-Thai strains (FCV-TH). The developed HRM assay was also validated with wild-type FCV-TH strains obtained from clinical samples of naturally infected cats.

Materials and methods

Animals and sample collection

Sixty-four samples comprised 22 nasal swabs (NSs), 42 oropharyngeal swabs (OSs), and two fresh lung tissue samples. The samples were obtained from 56 living cats (24 healthy cats and 32 cats with respiratory illness) and two cats who died from respiratory illness. All swabs were immediately immersed in 1% sterile phosphate-buffered saline and stored at −80 °C until used. All procedures were approved by the Chulalongkorn University Animal Care and Use Committee (No. 1631002). The cats’ owners were requested to sign a consent form before sampling.

RNA extraction and complementary DNA synthesis

The fresh-frozen tissues were aseptically homogenized using Tissue Rupture (Qiagen GmbH, Hilden, Germany) in 1% sterile phosphate-buffered saline and centrifuged at 6000 g for 1 min. Then, the supernatants were collected in a 1.5-ml microcentrifuge tube. Subsequently, viral RNA was extracted from 200 µl of tissue-derived supernatants or swab-derived solution using the QIAamp® cador^® Pathogen Mini Kit (Qiagen GmbH, Hilden, Germany) following the manufacturer’s instructions. The extracted RNA was qualified and quantified using a NanoDrop Lite Spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA). Then, the RT reaction was conducted using the Omniscript^® RT Kit (Qiagen GmbH, Hilden, Germany) following the manufacturer’s recommendations, and the constructed complementary DNA (cDNA) was stored at −20 °C until use (Chaiyasak et al. 2020).

Primer design

Multiple full-length genomes of FCV strains available in the GenBank database, including classical, virulent, and commercial vaccine strains, were retrieved and aligned using BioEdit Sequencing Alignment Editor Version 7.2.5. The alignment included 14 sequences of both partial RdRp and full-length VP1 major capsid protein sequences of FCV-TH strains. Primers for the HRM assay were designed from the location between ORF1 and ORF2 of the FCV, where nucleotide sequences were significantly different between available FCV-Vac and wild-type FCV-TH sequences and those of other global FCV strains (Supplementary Figure S1). The nucleotide sequences of all primers used in this study are shown in Table 1.

Table 1. Nucleotide sequence, position, and amplicon size of primers used in this study.

Assay	Primer name	Sequence 5’ to 3’	Position^a	Amplicon size (bp)
Screening RT-PCR	ORF1 For	GTAAAAGAAATTTGAGACAATGTCT	1-25	414
	ORF1 Rev	GTGAGCTGTTCTTTGCACAT	414-394
Conventional RT-PCR	RdRp For	TATGGTGATGATGGWGTKTAYATGTT	4778-4805	696
	RdRp Rev	AATCCCACACAGTGCCAAAT	5473-5454
HRM assay	HRM For	GGYGTGGAGGCGCGGWC	5240-5256	99
	HRM Rev	CGTHAGCGCAGGTTGAGCACAT	5338-5317

^a Primer position was quoted to the sequence of FCV strain Urbana (GenBank accession number L40021).

Reference FCV strains for the HRM assay

To establish and validate the HRM assay, the six selected FCV strains were commercially synthesized as string DNA fragments (Invitrogen^TM GeneArt^TM Strings^TM DNA Fragments, Thermo Fisher Scientific^®, MA, USA). The fragments covered nucleotide sequences at the region between the 3′ terminus of the RdRp gene and the 5′ terminus of the VP1 major capsid protein gene, resulting in a 150-bp product length. The six string DNA fragments of FCV strains comprising four VS-FCV phenotype strains (UTCVM-H2 strain; accession no. AY560117 (Addie et al. 2008), SH/2014 strain; accession no. KT000003.1 (Guo et al. 2018), Jengo strain; accession no. DQ910793 (Battilani et al. 2013) and GX01-13 strain; accession no. KT970059 (He et al. 2016)), one commercial FCV-Vac strain (Nobivac^® 1-HCPCh, Intervet Inc, Ohama, NE, U.S.A), and one FCV-TH strain (FCV-TH/KP313, accession no. OM982631) were employed to determine the analytical sensitivity and efficacy of the HRM assay for discrimination between different FCV strains. Moreover, three commercially available vaccines including Felocell^® CVR (Ser No. 464968B, Zoetis, Lincoin, NE, U.S.A), Purevax^® RCPh (Lot No. E26368, Boehringer Ingelheim Animal Health France SCS, Saint Priest, France) and Feligen^® RCP (BN No. 8G8S, Virbac, Carros Cedex, France) were also employed in the HRM validation.

Establishment of the qPCR-HRM assay

To establish the qPCR-HRM assay, the qPCR was initially performed to amplify the 99-bp fragment product of FCV using Rotor-Gene^® Q (Qiagen GmbH, Hilden, Germany). Then, the HRM application was analyzed using the accompanying software (Rotor-Gene^® Q software 2.3.1.49). Template amplification was generated using the Type-it^® HRM^TM PCR kit (Qiagen GmbH, Hilden, Germany), in which EvaGreen^® was applied as an intercalating fluorescent dye. The reaction was conducted in a final volume of 25 µl containing 2× HRM PCR Master Mix, 0.5 µM of each designed primer (HRM For and HRM Rev; Table 1), and 2 µl of cDNA products from clinical samples. Instead of the cDNA, the string DNA fragments of the constructed FCV were used as a template for the positive control tube. The thermal cycling conditions were as follows: initial denaturation at 95 °C for 5 min, then 40 cycles of PCR at 95 °C for 10 s, 60 °C for 30 s, and 72 °C for 10 s. Then, the fragments were ramped from 65 °C to 95 °C in 0.1 °C increments by holding 2 s at each step. The data analysis was evaluated, and a report was generated by Rotor-Gene Q^® software 2.3.1.49 (Qiagen GmbH, Hilden, Germany).

The qPCR-HRM reaction was carried out in triplicate in all clinical samples to ensure the intra-assay reproducibility; meanwhile, an FCV-Vac string DNA fragment and two clinical samples (FCV-TH/KP367 and FCV-TH/KP361, accession no. MZ542330) were incorporated into each run to verify the inter-assay reproducibility.

Analytical sensitivity and specificity of the qPCR-HRM assay

To validate the analytical sensitivity of the qPCR-HRM assay, tenfold serial dilutions of FCV-Vac string DNA fragments (Nobivac^® 1-HCPCh), with concentrations ranging from 6.18 × 10⁻¹ to 6.18 × 10⁸ copies/µl, were used as a reaction template. A cycle threshold (C_t) value lower than 35 was considered positive. To check the analytical specificity, the PCR-positive controls of feline herpesvirus-1 (FHV-1), feline coronavirus (FCoV), and feline morbillivirus (FeMV), that were obtained from naturally infected cats and were genetically characterized by sequencing, were employed. While commercially available vaccines of feline leukemia virus (FeLV; Leukocell^®2) and feline immunodeficiency virus (FIV; Fel-O-Vax-FIV^®) were utilized. All qPCR-HRM assays were run in triplicate with inter-assay and intra-assay validations.

Analytical accuracy and efficacy for strain discrimination

To ensure the accuracy and efficacy of FCV strain discrimination using the qPCR-HRM assay, two string DNA fragments (FCV-Vac and FCV-TH/KP313 strain) were used as templates, with altered concentration ratios from 1:4 to 4:1 (v/v). The template mixture was made following the qPCR-HRM assay as mentioned above. Then, the efficacy of the strain discrimination at each concentration was determined as representative of a mixed infection.

FCV detection by conventional RT-PCR

cDNA products deriving from extracted nucleic acid were also amplified for FCV detection using conventional RT-PCR. The specific primers (ORF1 For and ORF1 Rev) were designed for this study (Table 1). The 25-µl mixture of the RT-PCR reaction was composed of 3 µl of cDNA, 12.5 µl of GoTaq^® Green Master Mix (2×) (Promega, USA), 7.5 µl of nuclease-free water, and 1 µl each of forward and reverse primers (0.4 µM). The thermal cycling started with an initial denaturation (95 °C, 5 min), followed by 40 cycles of denaturation (95 °C, 30 s), primer annealing (59 °C, 30 s), and primer extension (72 °C, 1 min). Then, a final extension was run at 72 °C for 5 min. The positive control of FCV was the commercially synthesized string DNA fragment (Thermo Fisher Scientific^®, MA, USA) (Supplementary Table S1). The negative control of the reaction employed distilled water instead of a genomic template. The positive RT-PCR products (414 bp) were visualized using capillary electrophoresis (QIAxcel DNA Screening Kit, Qiaxcel^®, Qiagen GmbH, Germany). Then, the positive amplicons were purified using a commercial kit (NucleoSpin^® Extract II kit, Macherey Nagel, Germany) according to the manufacturer’s instructions. The purified PCR products were submitted for bidirectional Sanger sequencing (Macrogen©, Incheon, South Korea).

Data analysis and interpretation

Normalized and difference graph plots of the HRM assay were constructed to determine the sample deviation related to FCV strain typing generated on the Rotor-Gene^® Q software 2.3.1.49. Linear regression analysis was conducted to identify the correlation between the percentage of the C:G component and melting temperature (T_m) of each strain typing pattern. STATA statistical software version 16.1 (Stata Corp., TX, USA) was used for all statistical analyses.

Results

Establishment of the qPCR-HRM assay

Three string DNA fragments combined with one FCV-Vac strain (Nobivac^® 1-HCPCh; Intervet Inc, Ohama, NE, U.S.A), two VS-FCV phenotype strains (UTCVM-H2 and GX01-13), and eight cDNA products constructed from RNA-extracted clinical samples derived from FCV-infected Thai cats were employed to establish the qPCR-HRM assay. The eight clinical samples were selected based on their different nucleotide sequences to present different HRM patterns. The HRM analysis results revealed distinct HRM patterns that could be used to differentiate between the FCV-Vac, virulent, and wild-type Thai strains. Both normalized and difference melt plot curves of HRM analysis demonstrated eight different strain typing patterns: one typing pattern from the FCV-Vac strain and VS-FCV strain (UTCVM-H2), one typing pattern from one VS-FCV strain (GX01-13), and six typing patterns from eight FCV-TH strains (confidence value >95%) (Figure 1). Furthermore, the linear regression model revealed the correlation between the C:G percentage of the PCR product sequence in each strain typing pattern, and the T_m was clearly significant (R² = 0.9954, p < 0.001; Figure 2). We found that 1% C:G alteration could remarkably encourage a T_m modification of 0.25 °C, with p < 0.001 (Table 2).

Figure 1. Feline calicivirus (FCV) strain differentiation by high-resolution melting (HRM) analysis. (A) HRM normalized the melt curve of eight unique patterns. (B) HRM difference graph of the eight FCV strains using the vaccine strain (nobivac^® 1-HCPCh) as a reference strain. HRM normalized and difference graphs could distinguish eight FCV-strain typing patterns from 11 strain templates.

Figure 2. Linear regression analysis of the relationship between the percentage of the C:G content of the PCR product and T_m value.

Table 2. Linear regression model showing the relationship between % C:G of PCR product sequence and melting temperature deriving from HRM assay.

Temp	Coef.	Std. Err.	t	P > | t |	[95% Conf. Interval]
CG	0.25	0.01	35.93	0.000	0.24 − 0.27
Cons	71.35	0.38	186.00	0.000	70.41 − 72.28

Efficacy of qPCR-HRM to differentiate between FCV-Vac, VS-FCV and FCV-TH strains

The differentiation efficiency of the established qPCR-HRM assay was further evaluated through the utilization of thirteen DNA templates composing of four string DNA fragments of VS-FCV phenotype strains (UTCVM-H2, SH/2014, Jengo and GX01-13 strain), four commercially available vaccines (Nobivac^® 1-HCPCh, Felocell® CVR, Purevax^® RCPh and Feligen® RCP) and five cDNA products that were synthesized from RNA-extracted clinical samples obtained from FCV-infected Thai cats. Of note, three of five cDNA templates were constructed from suspected VS-FCV illness cats (FCV-TH/KP361, FCV-TH/KP365, and FCV-TH/KP367), while the remaining two cDNA templates were chosen based on their distinct nucleotide sequences (FCV-TH/KP135 and FCV-TH/KP313), resulting in diverse HRM patterns (Supplementary Figure S1).

After HRM analysis, ten unique strain typing patterns were yielded and able to distinguish among the vaccine, virulent, and wild-type Thai strains of FCV. Ten typing patterns comprised two patterns associated with the FCV-Vac, four patterns specific to the VS-FCV strain, and four patterns related to the FCV-TH strain (confidence value >95%) (Figure 3). The results displayed a distinct discrimination pattern between the vaccine strains in both the normalized (Figure 3A) and differentiated (Figure 3B) melt plot curves. In addition, the unique melt plot curve patterns for each VS-FCV and FCV-TH strain were found to be in accordance with their individual sequences, as illustrated in Supplementary Figure S1.

Figure 3. Differentiation efficacy of qPCR-HRM between FCV-Vac, VS-FCV, and FCV-TH wild type (A) HRM normalized the melt curve of ten unique patterns. (B) HRM difference graph of the ten FCV strains using the vaccine strain (purevax^® RCPCh) as a reference strain. HRM normalized and difference graphs could distinguish between FCV-Vac, VS-FCV, and FCV-TH wild type.

Analytical sensitivity and specificity of qPCR-HRM

The analytical sensitivity of the qPCR-HRM assay using a tenfold serial dilution of the FCV-Vac string DNA fragment (Nobivac^® 1-HCPCh) was evaluated. The results showed that the lowest detection limit of the assay was at 6.18 × 10¹ copies/µl, with a C_t value lower than 35 (Figure 4a). All starting template concentrations revealed the same melt curve alignment in the normalized plot curve (Figure 4b). Additionally, the qPCR efficiency displayed a strong linear correlation between 10⁻¹ and 10⁸ copies/µl by the slope value of −3.418, correlation coefficient (R²) of 0.998, and reaction efficiency of 0.9613 (Figure 4c).

Figure 4. Analytical sensitivity of the high-resolution melting (HRM) assay. (A) The quantitative assessment in a tenfold dilution of feline calicivirus (FCV)-synthesis string DNA (6.18 × 10¹ and 6.18 × 10⁸ copies/µl) was evaluated using qPCR and analyzed by HRM software. (B) The normalized graph of HRM analysis revealed a normalization region between 80 and 89 °C with a confidence threshold of 95%. The cutoff C_t value of the positive sample was set at C_t < 35. (C) The standard curve for indicating the efficiency of the qPCR detection assay by the x-axis refers to the tenfold dilution of FCV-synthesis string DNA (ng/µl). In contrast, the y-axis refers to the corresponding C_t values. The assay is linear in the range of 6.18 × 10¹ to 6.18 × 10⁸ template copies/µl, with a determination coefficient (R²) of 0.998 and reaction efficiency of 96.13%.

The analytical specificity of the qPCR-HRM assay demonstrated positive results only for the FCV, from either the FCV-Vac string DNA fragment or a clinical sample. At the same time, other feline viral pathogens (FHV-1, FCoV, FeLV, FIV, and FeMV) could not be detected (Figure 5).

Figure 5. Analytical specificity of the high-resolution melting (HRM) assay. The graph shows positive results from the positive control (PTC: string DNA fragment from the feline calicivirus [FCV] vaccine strain with concentration at 6.2 × 10⁵ copies/µl) and FCV-positive clinical sample. The non-template control (NTC) and other feline virus templates, including FHV-1, FIV, FeLV, FCoV, and FeMV, showed amplification signals over positive cutoff values (C_t > 35).

Analytical efficacy for FCV strain discrimination

The developed qPCR-HRM assay successfully demonstrated the ability to distinguish the template mixture of double FCV strains employing two string DNA fragments (FCV-Vac and FCV-TH/KP313 strain) with an altered concentration ratio (1:4 to 4:1). Both normalized and different HRM graph plots exhibited strain typing patterns in the same manner, with the melt graph plot deviated along with a higher concentration template (Figure 6a-b). The melt curve peaks also corresponded to the T_m of each FCV strain. The T_m of the FCV-Vac and FCV-TH/KP313 templates was 84.8 °C and 86.4 °C, respectively, and appeared on the different plots as specific melting temperatures (Figure 6c). The pattern of the HRM difference graph displayed a strain-specific peak of either FCV-Vac or FCV-TH/KP313. At the same time, the mock mix infection showed graph variation depending on the mixture ratio (Figure 6b). Despite the similar concentration ratios (3:2 and 2:3), the HRM standard normalized melting curve exhibited different patterns for the FCV-Vac and wild-type strains (Figure 6a), and the HRM difference plot curves demonstrated their own specific patterns (Figure 6b).

Figure 6. Analytical efficacy of the high-resolution melting (HRM) assay for strain differentiation between the wild-type feline calicivirus-Thai (FCV-TH) strain and vaccine strain (FCV-Vac). The patterns of the HRM normalized graph (A) and HRM difference graph (B) demonstrate detectability patterns at each various concentration ratio. Melt curve analysis from the HRM assay (C) indicated the derivative melt peak patterns that seem likely to correspond to the divergence ratio of the strain mixture.

qPCR-HRM application on clinical samples

FCV detection in 64 clinical samples between conventional RT-PCR and qPCR-HRM methods demonstrated a higher sensitivity of the qPCR-HRM assay than the conventional RT-PCR assay (Table 3). The qPCR amplification was presented in 33 samples, revealing a C_t value less than 35. The amplicons contributed to the sufficient quality of signals for HRM analysis and could subsequently differentiate the strain typing patterns. Eight of 33 positive samples displayed different results between qPCR and RT-PCR. However, the detectable C_t values of those eight samples were between 33 and 35 due to the small number of templates.

Table 3. Result comparison between qPCR-HRM assay and conventional RT-PCR.

	qPCR-HRM assay	RT-PCR
Positive	33	25
Negative	31	39
	64	64

After HRM analysis of 33 FCV-positive samples, seven different strain typing patterns on the HRM melt plot curve were evident, based on the mean T_m shift with the altered C:G content. These seven strain typing patterns consisted of two FCV-Vac and five wild-type FCV-TH typing patterns, represented by the standard deviation of each group presenting between 0 °C and 0.099 °C (Table 4). Interestingly, nine wild-type FCV-TH strains showed melt curve patterns related to two vaccine strains that are commercially available in Thailand. Additionally, all nine cats were clinically healthy on the sampling date. Five cats had been vaccinated over three months before sampling, and the other four did not receive any vaccination against FCV. Moreover, the VS-FCV strain (FCV-TH/KP361) represented a mean T_m similar to that of the FCV-TH wild-type pattern two (Table 3) and demonstrated a unique HRM graph (Figure 4).

Table 4. Mean melting temperature (T_m) and standard deviation (SD) within the group of each feline calicivirus (FCV) strain typing pattern detected from clinical samples.

Strain typing pattern	Number of samples	Mean (Tm^oC)	SD
FCV-TH wild-type pattern 1^#	3	84.33	± 0.023
FCV-Vac related pattern 1^a	6	84.74	± 0.018
FCV-Vac related pattern 2^b	3	85.01	± 0.069
FCV-TH wild-type pattern 2*	3	85.46	± 0.094
FCV-TH wild-type pattern 3	5	85.77	± 0.074
FCV-TH wild-type pattern 4	12	86.11	± 0.068
FCV-TH wild-type pattern 5	1	86.40	0

^a FCV-Vac (Nobivac^® 1-HCPCh; Intervet Inc, Ohama, NE, U.S.A).

^b FCV-Vac (Felocell CVR^®; Zoetis, Lincoin, NE, U.S.A).

* FCV-TH/KP361 (VS-FCV).

^# FCV-TH/KP367 (VS-FCV).

Discussion

FCV has been a common endemic respiratory pathogen in cats, even though vaccination is regularly applied. The emergence of vaccine-resistant strains due to long-term use of a modified live vaccine is still a concern (Afonso et al. 2017; Smith et al. 2020). Regarding the high genetic diversity of the FCV, identifying genetic markers for FCV strain differentiation, especially between virulent and nonvirulent FCV strains, is still challenging (Brunet et al. 2019). Although conventional methods, such as genome sequencing, are still necessary for strain characterization and identification, they are time-consuming and require post-PCR procedures and analysis.

Although there is no specific region for FCV genotyping, the nucleotide region between ORF1 and ORF2 has been recognized as a hot-spot recombination area of many Caliciviridae viruses, including human norovirus and the FCV (Katayama et al. 2002; Tajiri-Utagawa et al. 2009; Symes et al. 2015). This region is also used for Norwalk-like virus genotyping (Katayama et al. 2002). Hence, our study developed an HRM-based qPCR with designed primer pairs that flank the junction region of ORF1 and ORF2 to differentiate the FCV genotype. According to the sequence alignment between our 14 wild-type FCV-TH and FCV-Vac strains, this region is suitable and could be a potential region for FCV differentiation. This study also displayed a distinct melt peak pattern that can distinguish between the wild-type FCV-TH, FCV- Vac and the VS-FCV found in the Thai cat population. However, in-silico analysis based on multiple FCV alignments obtained from various FCV strains and implementing our developed HRM-qPCR assay with other FCV genotypes might be more beneficial to confirm this speculation.

The HRM assay has been recognized as the simplest method for genotyping and mutation scanning that can identify a single-base change in short fragments (Reed and Wittwer 2004; Tamburro and Ripabelli 2017). In this study, our designated primer targeted to amplify a 99-bp segment revealed the capability to differentiate between FCV-Vac, wild-type FCV-TH, and VS-FCV; both global and detected in Thai cats. As a result, the distinctive melt plot curve patterns of each strain also correlated with their respective sequences (see Supplementary Figure S1). According to the product datasheets from four vaccine companies available in Thailand, there are three strains offered. These include the F9 strain found in Nobivac^® 1-HCPCh, Felocell^® CVR, and Feligen^® RCP, as well as FCV431 and G1 strains present in Purevax^® RCPCh. The results of this study demonstrate the efficiency of differentiating between vaccine strains by revealing distinct melt curve differences in both the HRM normalized curve and HRM difference graph. Furthermore, our self-designed primers could also distinguish between two FCV vaccines containing single-nucleotide polymorphisms at T5269C (Felocell CVR^®) and T5270G (Nobivac^® 1-HCPCh) that are commercially available in Thailand. The sequence data that present the correlation between HRM results and amplicon sequences are revealed in the Supplementary Figure S1.

The qPCR-HRM assay established in this study demonstrated the efficiency of strain discrimination that could be used in cats co-infected with different FCV strains or cats suffering from respiratory illnesses that were recently vaccinated with an FCV vaccine by showing double peaks or strain-specific peaks on melt peak analysis. Moreover, with co-infection, the peaks tended to deviate from the template with higher concentration. Although the HRM application in an in-vitro mixed infection model in this study could not be used to present the recombination evidence that seems possible to emerge in long-term co-infection, this HRM assay can distinguish a new recombinant strain if the recombination event occurs at the amplification region. Additionally, the authors suggest that the divergence of the HRM graph pattern is an advantage in the surveillance of co-infection episodes, which can lead to recombination afterward.

Our developed HRM assay had a detection limit at 6.18 × 10¹ copies/µl (Ct values lower than 35) and was delicate, with a 0.25 °C T_m shift at 1% C:G modification. Furthermore, our developed HRM analysis could distinguish the melt graph pattern with a T_m difference of 0.2 °C. The quality of the DNA template is also necessary for the optimal potential of HRM analysis (Toi and Dwyer 2008). According to the results of detection and strain typing of the samples obtained from clinical samples, we found eight FCV-positive clinical samples that were positively detected between C_t values of 33 and 35, presenting a poor signal to segregate HRM melt curve plots. Hence, the results should be cautiously interpreted when late amplification occurs (C_t between 30 and 35) because it is usually due to a low starting template concentration and/or a high degree of sample degradation (mService QT 2015).

After employing HRM analysis on 33 FCV-positive samples, they demonstrated seven melt curve patterns, two of which are FCV-Vac-related patterns. Interestingly, nine wild-type FCV-TH strains from this study presented melt curve patterns related to two commercial vaccine strains (Felocell CVR^® and Nobivac^® 1-HCPCh) that are widely used in Thailand. Moreover, a previous publication mentioned retaining the pathogenic potentiality of live vaccines by improper administration, such as accidentally dropping them into the environment or on the fur, and ingesting them during grooming (Pedersen and Hawkins 1995; Radford et al. 1997; Hofmann-Lehmann et al. 2022). Thus, this finding suggests the potential presence of the FCV-Vac strain within the Thai cat population, and its capability to induce disease should be monitored in the future.

On the contrary, identification of the FCV RNA in the cats could not indicate the infectivity of the virus; it was possibly the remnant of environmentally contaminant FCV RNA. In this study, the five melt curve patterns of wild-type FCV-TH were significantly different from the vaccine strain pattern on both the normalized melt curve graph and the difference graph. This study is the first report demonstrating the application of qPCR-HRM analysis to differentiate between FCV-Vac and FCV-TH strains.

In conclusion, in this study, we have developed the qPCR-HRM assay, which is a one-step closed-tube system for the detection of FCV. This streamlined protocol offers several advantages over other molecular techniques, such as RT-PCR and nested RT-PCR, as it significantly reduces both time consumption and the risk of contamination associated with multi-step protocols. Moreover, the established qPCR-HRM assay is helpful for simultaneous detection and strain differentiation between the vaccine and wild-type FCV-TH strains. This assay can also distinguish between the virulent FCV-TH (FCV-TH/KP361) and another wild-type nonvirulent FCV-TH strain. This developed simultaneous diagnostic method is rapid, sensitive, specific, and cost-effective for FCV differentiation within a closed single-tube system.

Ethical approval

All procedures in this study were approved by the Chulalongkorn University Animal Care and Use Committee (No. 1631002). The cats’ owners were requested to sign a consent form before sampling.

Acknowledgements

K.P. received grant funding from the Ph.D. program at Chiang Mai University. C.P. received a grant from the Ratchadapisek Somphot Fund for Postdoctoral Fellowship at Chulalongkorn University.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors guarantee that the data supporting the findings in this study are available within this article and its supplementary material. The raw data that encourage the finding of this study are available from the corresponding author, upon reasonable request. Nine sequences of the FCV-TH strains (FCV-TH/KP80, FCV-TH/KP82, FCV-TH/KP100, FCV-TH/KP103, FCV-TH/KP105, FCV-TH/KP181, FCV-TH/KP276, FCV-TH/KP313, and FCV-TH/KP361) obtained from this study have been deposited in NCBI GenBank under accession numbers OM982622, OM982623, OM982625, OM982626, OM982627, OM982629, OM982630, OM982631, and MZ542330, respectively.

Additional information

Funding

This study was supported by the 90th Anniversary of the Chulalongkorn University Fund (Ratchadaphiseksomphot Endowment Fund). ST was partly supported by the National Research Council of Thailand (NRCT) through R. Thanawongnuwech NRCT Senior Scholar 2022 #N42A650553.