Development of a loop-mediated isothermal amplification technique and comparison with quantitative real-time PCR for the rapid visual detection of canine neosporosis

Neospora caninum is an obligate intracellular protozoan that causes neosporosis in dogs and cattle worldwide. Although the structure and life-cycle of N. caninum are similar to those of Toxoplasma gondii, canids are the only definitive hosts of N. caninum, whereas felids are the only definitive host of T. gondii [1]. The infection has also been reported in horses, sheep, deer, and goats [2]. Neosporosis is a major cause of abortion and stillbirth in both dairy and beef cattle and is occasionally a problem in a livestock operation. Infected cattle may abort several times because of chronic neosporosis, and any offspring born alive are likely to be themselves infected and have a higher risk of future abortions [3]. Therefore, neosporosis is considered a veterinary health problem that has an economic impact on the livestock industry through losses in beef and milk production [4].

The transmission of N. caninum occurs from the ingestion of water or food contaminated with sporulated oocysts or tissue cysts (horizontal transmission), or transplacental infection (vertical transmission). Major modes of transmission in dogs include direct contact with cattle in rural areas enabling the ingestion of infected foetal material, the ingestion of infected intermediate hosts, e.g. rodents, birds, and other animals, and the direct ingestion of oocysts [5]. Because dogs are used on farms as herding or livestock guardians and can shed oocysts when they are infected, it is important for control programs to identify and remove these animals from breeding herds [1].

The diagnosis of canine neosporosis is based on combining appropriate clinical signs and serology such as immunofluorescent antibody tests and enzyme-linked immunosorbent assays (ELISAs) [6]. However, these assays may be limited by false-negative results in early or chronic cases of infection [7]. Moreover, standardised tests are not available, so it is not possible to compare results between different laboratories [2]. The definitive standard for identifying infection is the recognition of tissue cysts on histological examination, the discovery of faecal oocysts, or identification of the parasite by immunohistochemical staining [8]. Light microscopy alone is not enough to identify N. caninum oocysts in dog faeces as it is difficult to distinguish N. caninum oocysts from those of other canine coccidian parasites such as Hammondia heydorni, or from other contaminating oocysts such as H. hammondi and T. gondii [2]. Immunohistochemical examination, which would aid differentiation, is rarely performed [9]. Moreover, only a few reports have identified N. caninum oocysts in naturally-infected dogs [10–13]. Quantitative real-time PCR (qPCR) has also been developed to detect and quantify N. caninum DNA in mice [14], rats [15], cattle [16–19], horses [20] and dogs [7, 21]. Although this is highly sensitive and specific, in general, it can only be used in specialised laboratories. It is not applicable to field situations due to a variety of issues, including the high costs and the need for well-trained personnel for DNA isolation, preparation of reaction mixture, etc. [22].

During the last decade, loop-mediated isothermal amplification (LAMP) has been established as a simple, rapid, and cost-effective technique for the detection of pathogens. It offers high sensitivity and specificity even when little target DNA is available [23, 24]. Therefore, the present study aimed to develop LAMP for detecting N. caninum infection in dogs in the field. The technique was validated and evaluated using blood, and faecal samples from dogs and the prevalence of canine neosporosis in Thailand was investigated.

Dogs, especially farm dogs, are an important factor in the spread of N. caninum infection. They can become infected by ingesting tissues such as livers, lungs, and brains, as well as aborted foetuses and placental material from infected cattle [32]. Following infection, N. caninum oocysts are shed in their faeces which contaminate water, pasture, and the environment of cattle farms. Although there are few reports of dogs naturally shedding oocysts, 503,300 oocysts were detected in a dog that was fed with experimentally-infected calf tissues and 820,655 were found in a dog fed naturally-infected buffalo brain [33, 34]. Although infected dogs normally shed oocysts for only a short period, Gondim et al. [35] reported that they might re-shed oocysts after feeding on infected tissue 18–20 months after the primary infection, which might increase the transmission of the parasite to cattle but this remains in doubt.

To prevent significant losses in farm production, the rapid detection and control of N. caninum infection in dogs are required and have been an objective for many studies [2]. The implementation of qPCR assays has revealed their major advantages over conventional molecular detection techniques [7, 17, 21]. However, significant dominance in the use of LAMP assays with non-instrumented fluorescent detection is reported. In the present study, the lower limit of LAMP detection was equivalent to that of qPCR (Fig. 2), the per-test consumable costs were similar, but the equipment costs for LAMP carried out using a water bath, or heated block was considerably lower [36–39]. This potentially makes the LAMP technique more suitable for use outside conventional laboratory facilities or in small-scale diagnostic laboratories where established PCR-based methods are too slow, expensive, or complicated for routine use. It could also be beneficial to industry, where the rapid assessment of the parasite infection status in or near farms could be used to make decisions relating to the management of neosporosis. Moreover, the time to complete LAMP and qPCR is similar (within 1.5 h), which is much faster than that of nested or conventional PCR (approximately 2.5–4 h plus the time required for gel electrophoresis) [40].

LAMP has successfully been used in the identification of parasitic infections in dogs, including giardiasis, leishmaniasis, and babesiosis [36–38]. To the best of our knowledge, this is the first LAMP technique to be developed for the diagnosis of neosporosis. It offers high specificity because it uses six primers to recognise six to eight regions on the target DNA for amplification [40]. Although the assay was not tested on parasites phylogenetically related to N. caninum such as Hammondia and Sarcocystis because of the absence of these organisms in our sample repository, we designed LAMP primers based on the unique, repetitive Neospora-specific Nc5 sequence which has not been detected in other taxa [7, 23]. We also checked the primer specificity against the GenBank database to ensure that non-specific binding was eliminated. Our results proved that the LAMP primers are highly specific and sensitive for the detection of N. caninum DNA (Figs. 1a, 2a). These results also demonstrate that qPCR and LAMP can be used to accurately detect target DNA over a wide range of concentrations with a detection limit equivalent to one genome. With regard to sensitivity level achieved by LAMP in this study, similar results were observed by Ghalmi et al. [7] who reported the same sensitivity of the equivalent of one genome by qPCR systems, while Han et al. [21] demonstrated a lower sensitivity of 38 molecules per microliter by using multiplex qPCR.

The detection of LAMP results is usually determined by observing turbidity with the naked eye or adding intercalating dyes such as SYBR green [36] or propidium iodide [39] to the reaction tube after amplification. However, these methods require a degree of skill or UV light to assess the result. With the use of HNB, LAMP products can be readily detected by visual inspection, thereby reducing the reaction time and enabling the assay to be used for field testing. Additionally, HNB minimises the risk of obtaining false positives through aerosol contamination because it is added before amplification [41]. For these reasons, the LAMP technique developed here appears to be a promising alternative to qPCR for the detection of N. caninum.

In Thailand, the seroprevalence of bovine neosporosis in dairy cattle farms ranges from 3 to 47% depending on the type of serological test and cut-off level used, and the areas surveyed [32, 42–44]. Few studies have reported the presence of N. caninum in dogs, although 1.2% of dogs near farms in Nakhon Pathom Province were shown to be N. caninum-positive by competitive ELISA [42]. In another study carried out in Western Thailand, 7% of dogs were found to be positive for the same test, and 4.4% by an indirect fluorescent antibody test [32]. Although serological antibody testing is the most common method used for dogs that have the past or recent contact with the parasite, it is not correlated with the shedding of oocysts [45]. Moreover, several studies have shown that dogs can shed oocysts without demonstrating seroconversion [11, 33, 45].

Despite the importance of dogs in the epidemiology of the disease, only a few studies have reported the prevalence of N. caninum from faecal samples or by using molecular techniques [46]; ours is the first such study in Thailand. Therefore, the prevalence and incidence of N. caninum oocyst shedding by dogs remain unclear. We detected N. caninum DNA in 3.8% of blood samples by LAMP and qPCR which was lower than that reported from a molecular survey of dogs in Algeria (13%) [47]. However, we observed a perfect agreement between the LAMP and qPCR results for the analysis of blood samples. It is possible that the marked difference in prevalence between these two studies reflects the geographical region because the seroprevalence in Algerian dogs was as high as 22% in stray dogs and 44% in farm dogs [47].

It has previously been shown that N. caninum DNA can be amplified from the blood of infected aborted and pregnant cows [17, 48], bulls [49], mice [50] and sheep [51, 52]. Ghalmi et al. [47] also showed that circulating parasites or parasite DNA could be detected in the blood of naturally infected dogs. In the present study, we confirmed that the parasite DNA could be detected in the blood of dogs, indicating that either an acute infection resulted in the circulation of tachyzoites in the blood or that a latent infection released bradyzoites which differentiated into circulating tachyzoites that later formed a new cyst [17, 53]. Information concerning the duration of parasitemia in infected dogs is limited. When toxoplasmosis in cats was assessed by PCR, a cat that was positive in blood would likely have shed oocysts within the previous 35 days [54, 55]. As N. caninum and T. gondii are very closely related parasites, although faecal examination under the microscope was not performed in the present study, dogs with positive results were also likely to have shed oocysts during the period. However, care must be taken in making generalisations about N. caninum based on the biology of T. gondii because neosporosis and toxoplasmosis are biologically distinct diseases [2]. Molecular assays are useful for early detection of N. caninum in asymptomatic animals [7, 17, 47, 52] because blood seems to provide the transport medium for tachyzoites between body tissues [48]. However, establishing an epidemiological relevance or cause-effect relationship between abortion and N. caninum is difficult because it was not known that detected DNA derived from multiplying tachyzoites in blood or free floating DNA, and finding the presence of the parasite DNA does not mean that N. caninum caused the abortion [1]. Unfortunately, from 15 positive blood samples, only two faecal samples were collected from the same dogs (see Additional file 2: Table S1; Sample no. 40 and 99) but the sample no. 40 showed ambiguous results between blood and faeces. Therefore, if detection of N. caninum DNA in blood corresponds to the detection of DNA in faeces remains unclear. Epidemiological or experimental studies using molecular techniques versus immunological techniques in dog blood with variable aspects should be performed in detail to understand the implication for farm management better.

Neospora caninum oocysts are highly resistant to disinfectants and can remain infectious for many months in the environment; however, they are killed by temperatures above 100 °C [56]. Identifying oocysts in environmental samples is undertaken by dairy farms that have experienced an outbreak to prevent neosporosis-associated abortions [1]. Several methods are available to diagnose neosporosis as a cause of abortion. The preferred samples in cases of abortion include one or more aborted foetuses submitted with the placenta and sera from the dam [57]. The positive results obtained in the present study from the placental tissue of an aborted calf prove the applicability of LAMP to many sample types and suggest that it could be used as a definite diagnosis to confirm neosporosis.

In the present study, the prevalence of detected DNA in dog faeces was shown to be 7.8 and 4.3% by LAMP and qPCR, respectively, similar to the previous reports in Ethiopia [10] and Costa Rica [58] but somewhat lower compared to China [59], and higher compared to Iran [11]. The discrepancy of LAMP and qPCR perhaps related the very low levels of DNA in the samples, which may have been beyond the detection limit of qPCR; LAMP is typically one to two orders of magnitude higher in sensitivity than qPCR [60]. Alternatively, the difference in sensitivity may be related to the presence of inhibitors, such as bile salts and plant-based polysaccharides that are known to be found in faecal samples [61, 62].

Theoretically, Bst polymerase used in LAMP has a higher tolerance of inhibitors than Taq polymerase used in qPCR [63, 64]. This suggests that the sensitivity of LAMP is slightly higher than that of qPCR, which is useful for the direct application of field faecal samples. In this study, most dogs that tested positive for N. caninum infection were from farms that allowed them to eat foetal tissues, placentas, or calf carcases. These findings are consistent with reports of high neosporosis prevalence in farm dogs which is linked with the existence of infective materials [47, 58, 65]. Hence, good management practices are important to reduce the risk of transmission from cattle to dogs and vice versa, especially in highly endemic areas. However, the positive coprological LAMP detections may also be due to the possible presence of sexual parasite forms in the infected enteric epithelial cells and included in the samples that DNA was extracted from [66]. An analysis to determine LAMP detection sensitivity, timing, consistency and the stage of N. caninum detected compared with traditional methods (microscopy and bioassay) should be explored in future studies.

Despite obvious advantages of the LAMP technique used in this study, future issues remain to be addressed and improved. These include further validation with other tissues such as brain, heart, liver, placenta, and body fluids, as well as samples from the field to prove that this assay can be used to detect N. caninum infection in cattle rapidly. Secondly, the application of faecal sample preparation and DNA extraction by heat and physical disruption would greatly reduce the need for instruments and the overall cost, making the test simpler to perform in remote areas [62, 67]. Finally, the LAMP assay could be used to monitor amplification in real-time, which could be achieved with a turbidimeter [40, 68]. However, the high amplification efficiency should be taken into account, which could increase the occurrence of carry-over contamination compared with conventional PCR [69–71]. Hence, all precautions to prevent cross-contamination should be carefully observed.

We have developed the first known LAMP technique to detect N. caninum with high specificity and sensitivity. The immediate visual detection of positive results and the use of basic heating devices make this method simple for use in the field. This LAMP assay has been validated in dog samples and has the potential to achieve the rapid detection of N. caninum. Examination of both blood and faecal samples would probably have increased the number of positive dogs whose faeces may contain oocysts. Further development, modification, validation, and standardisation will provide an asset for diagnosis, epidemiology, control, and prevention strategies for neosporosis.