Detection and quantification of Brucella abortus DNA in water buffaloes (bubalus bubalis) using droplet digital polymerase chain reaction

Abstract

Brucellosis represents a major public health concern worldwide. Human transmission is mainly due to the consumption of unpasteurized milk and dairy products of infected animals. The gold standard for the diagnosis of Brucella spp in ruminants is the bacterial isolation, but it is time-consuming. Polymerase Chain Reaction (PCR) is a quicker and more sensitive technique than bacterial culture. Droplet digital PCR (ddPCR) is a novel molecular assay showing high sensitivity in samples with low amount of DNA and lower susceptibility to amplification inhibitors. Present study aimed to develop a ddPCR protocol for the detection of Brucella abortus in buffalo tissue samples. The protocol was validated using proficiency test samples for Brucella spp by real time qPCR. Furthermore, 599 tissue samples were examined. Among reference materials, qPCR and ddPCR demonstrated same performance and were able to detect up to 225 CFU/mL. Among field samples, ddPCR showed higher sensitivity (100%), specificity and accuracy of 93.4% and 94.15%, respectively. ddPCR could be considered a promising technique to detect B. abortus in veterinary specimens, frequently characterized by low amount of bacteria, high diversity in matrices and species and poor storage conditions.

Introduction

Brucellosis is an infectious disease caused by different members of the genus Brucella. Brucellosis remains the most common zoonosis worldwide although human surveillance plans are issued in most countries. Currently, human brucellosis is reported in more than 170 countries, with approximately 500,000 new cases reported each year, nevertheless the actual number has probably a 10-to-25-fold higher rate, since in endemic areas, the disease is frequently misdiagnosed or underreported (Liu et al. 2023). Indeed, Laine et al. (2023) estimated, through the use of three different statistical models, that the annual global incidence of brucellosis is 2.1 million, definitely higher than the reported cases (Laine et al. 2023).

Therefore, the identification of new cases and the determination of the infection routes are the primary aims of human surveillance (3). Indeed, human transmission occurs mainly through the consumption of unpasteurized dairy products and direct contact with infected animals. In areas where the disease is endemic in livestock, brucellosis creates economic losses for the primary productions of dairy industries and the relative increase in management costs to control the disease and huge implications for animal health (World Organisation for Animal Health (WOAH), n.d.b; Dadar et al. 2021; Deka et al. 2018).

The disease has a wide range of hosts from several mammal species to marine mammals and amphibians. B. abortus, B. melitensis, B. suis are the main species responsible for human brucellosis (Yang et al. 2020; Dadar et al. 2021; Shakir 2021).

Control strategies for brucellosis in livestock have been issued in most countries, leading to the eradication of the disease in several geographic areas, nevertheless, Brucella spp. is still endemic in some European countries of the Mediterranean basin, north and east Africa, the Middle East, south and central Asia and Central and South America (Shakir 2021; Centers for Disease Prevention and Control (CDC) 2019), and Africa and Asia are the areas at highest risk (Laine et al. 2023).

Currently, the European Regulation establishes a compulsory eradication plan for Brucella abortus, B. melitensis and B. suis, using serological examinations for the detection of the antibodies in exposed animals with Rose Bengal Test (RBT), indirect ELISA, Complement Fixation Test (CFT) or competitive ELISA. However, several Gram-negative bacteria can induce false positive results, such as Yersinia enterocolitica 0:9, Escherichia coli 0157:H7, Salmonella N group and Vibrio cholerae O1 (Bonfini et al. 2018; Yang et al. 2020). These cross-reactions could create important diagnostic implications in brucellosis free farms for the suspension of the status, the culling of not-infected animals, movement and natural mating restrictions and limitations in the use of raw milk, as issued by international, national and regional regulations. Moreover, in case of falsely negative results, even worse implications could occur.

The ‘gold standard’ for the diagnosis of brucellosis is represented by bacterial isolation followed by species identification, albeit it is time-consuming and the sensitivity is relatively low, as Brucella often shows limited bacterial loads in field samples (Liu et al. 2023; European Union Reference Laboratory (EURL) for Brucellosis, 2021). Consequently, the development of novel diagnostic tools can represent a valid aid for disease control both in free or almost free countries (Bonfini et al. 2018), as well as in regions where the disease is endemic. Molecular examinations, primarily Polymerase Chain Reaction (PCR) and real-time quantitative PCR (qPCR), are rapid and reliable methods that have been introduced in several eradication plans, helping to obtain early detection, even in the presence of small amounts of targeted material (World Organisation for Animal Health (WOAH) n.d.a). However, to date, no innovative molecular techniques are available for the detection of Brucella in eradication plans in livestock.

A third generation PCR, droplet digital PCR (ddPCR), is a novel technique that is becoming widely used in the diagnosis of infectious diseases in matrices with low amounts of nucleic acids, such as Brucella spp., Yersinia pestis, Bacillus anthracis and Francisella tularensis (Du et al. 2022). ddPCR shows high fidelity in the absolute quantitation, without preparing standard curves, and lower susceptibility to amplification inhibitors than conventional PCR and qPCR, that could be responsible for false negative results (Li et al. 2018; Liu et al. 2023). Recently, it has been shown a higher diagnostic performance of ddPCR in the detection of Brucella nucleic acids in whole blood samples of human patients compared to qPCR. Indeed, ddPCR showed higher positivity rate of detection, thus identifying this method as a potential tool for the diagnosis and follow-up of brucellosis (Liu et al. 2023).

Herein we describe ddPCR assay for the detection and quantification of Brucella abortus, evaluating the performance of the test in both reference materials and field samples from water buffalo (Bubalus bubalis). ddPCR results were evaluated in comparison with those using qPCR and bacterial culture.

Materials and methods

Sample collection

During routine activity of Brucellosis eradication plan conducted by the official veterinary health authorities, 171 buffaloes from Caserta province, Campania Region (Italy) tested positive to Rose Bengal Test (RBT) and Complement Fixation Test (CFT) official serological tests, and were slaughtered according to the provisions of the national and regional regulations. The animals belonged to 38 different establishments, that showed a disease-free status for bovine tuberculosis and enzootic bovine leukosis, while a status of infected for brucellosis. At the slaughterhouse, retropharyngeal, submaxillary and mammary lymph nodes (n = 329), spleen (n = 79), genital organs (n = 112) (that included testis, seminal vesicles, foetal liquid and, uterus) and udder (n = 79), which were collected and singularly placed in sterile packages, hermetically closed with lead seals, refrigerated at 5 ± 3 °C and conferred to the Istituto Zooprofilattico Sperimentale del Mezzogiorno (IZSM) of Portici, Italy, for direct detection of Brucella spp. Surface cauterisation was carried out on each organ that was next incised using sterile scalpels and, from the inner parts of the organs, two specimens were collected using flocked swabs (Copan, Brescia, Italy). Next, one swab was processed for bacteriological analysis no more than 48 h after sampling, and the second one was stored at −20 °C for molecular analyses. Thus, a total of 599 swabs were randomly selected and underwent droplet digital PCR, in parallel with bacterial culture and real-time PCR.

Bacterial culture

Notably, Brucella spp. was determined according to the procedures described by the World Organisation for Animal Health (WOAH) (World Organisation for Animal Health (WOAH) n.d.a). In Biosafety Level 3 (BLS-3) labs, swabs were plated onto Brucella Agar (Oxoid, Rodano, Italy) and selective CITA Agar (Oxoid), both incubated at 37 ± 1 °C in 5-10% CO₂ atmosphere and daily checked up to 10 days, for the visualization of the development of the colonies. The colonies, presumed to be Brucella spp., were picked and plated onto Blood Agar (Oxoid) and Brain Heart Infusion (BHI) Agar to set up sub-colonies and incubated at 37 ± 1 °C in 5-10% CO₂ atmosphere for 48-72 h. The colonies underwent identification using biochemical macro-methods, with Gram stain, urease, oxidase and agglutination with hyper-immune anti-S-Brucella polyclonal serum (ThermoFisher Scientific, Waltham, MA, US), and miniaturised biochemical procedures, with VITEK 2 Compact system (bioMérieux, Lyon, France). Brucella spp. positive samples underwent species and Biovar identification, by the National Reference Centre for Brucellosis (NRCB), Istituto Zooprofilattico Sperimentale di Abruzzo e Molise (IZSAM) using Restriction Fragment Length Polymorphism (RFLP)-PCR (World Organisation for Animal Health (WOAH) n.d.a).

Molecular examinations

Real-time quantitative PCR (qPCR)

In BLS-3 laboratory, swabs were first thermal inactivated at 99 °C for 10 min and then transferred in 2 mL tubes with the addition of 200 µm in diameter glass beads (NextAdvance, Troy, NY, USA). Samples were suspended in 1 mL sterile Phosphate-Buffer Saline (PBS) and incubated at room temperature for 2 h on magnetic stirrer. Next, samples were vortexed and mechanically homogenized using TissueLyser (Qiagen, Hilden, Germany) for 5 min at 30 Hz and subsequently clarified by brief centrifugation at 6800 × g (Eppendorf, Hamburg, Germany). Aliquots of 200 µL of supernatant were collected and nucleic acids extraction and purification were carried out with QIAsymphony automated system (Qiagen, Hilden, Germany) using with QIAsymphony DSP Virus/Pathogen Mini Kit (Qiagen). During the DNA extraction step an Internal Control specific for the PCR amplification kit used for the subsequent Brucella spp. detection (Qiagen) was added to all samples. Extracted DNA was eluted in 60 µL Elution Buffer and stored at −20 °C until use.

Brucella spp. detection by real-time PCR was conducted using the real-time thermal-cycler CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) according to the protocol of the European Union Reference Laboratory for Brucellosis (EURL) (EU Reference Laboratory for Brucellosis (EURL) 2019), targeting the Insertion Sequence (IS)711. Each reaction mixture contained QuantiFast Pathogen Master Mix at 1X final concentration (Qiagen), Internal Control Assay (Qiagen) at 1X final concentration, 0.5 µM of each primer, 0.2 µM of probe, 5 μL of DNA and 9.5 μL of nuclease-free water in a final volume of 25 μL. The sequence of primers and probes are reported in Table 1. The PCR thermal profile was as follows: initial template denaturation at 95 °C for 5 min, and 45 cycles including a template denaturation step at 95 °C for 15 s, followed by primer annealing/extension at 60 °C for 30 s. A negative control (purified PCR-grade water), a negative extraction control and a positive control (B. abortus genomic DNA) were included in all PCR assays. According to the guidelines of the EURL protocol, samples were considered as positive when showing a threshold cycle (Ct) value lower than 38 for the IS711 target (EU Reference Laboratory for Brucellosis (EURL) 2019).

Table 1.

Primer sets and probes for the detection of Brucella spp. by real-time PCR and B. abortus by droplet digital PCR.

Molecular assay	Sequence (5’-3’)
Real time PCR	IS421-For_5′-CGCTCGCGCGGTGGAT-3′
	IS511-Rev_5′-CTTGAAGCTTGCGGACAGTCACC-3′
	IS711-Probe_5′-FAM- ACGACCAAGCTGCATGCTGTTGTCGATG-3′- TAMRA
Droplet digital PCR	BRC-For-5′- GCGGCTTTTCTATCACGGTATTC- 3′
	BRC-Rev-5′- CATGCGCTATGATCTGGTTACG- 3′
	BRC-P-5′- FAM-CGCTCATGCTCGCCAGACTTCAATG- 3′- MGB

Droplet digital polymerase Chain Reaction (ddPCR)

Droplet digital PCR, that was performed using QX200 Droplet Digital PCR Systems (Bio-Rad Laboratories, Hercules, CA, USA) in 22 μL volume, containing 5 μL template (< 100 ng/μL), along with ddPCR Supermix for Probes (Bio-Rad Laboratories) at 1X final concentration, 0.9 µM of each primer, 0.25 µM of probe and nuclease-free water to reach the final volume. The primers and probe used to set up the ddPCR specific for B. abortus (Table 1), belonged to the real time PCR protocol of Dal et al. (2019) (Dal et al. 2019) targeting the IS711. Each working session was conducted in the presence of a negative control (nuclease-free water), a negative extraction control, and a positive control, NCTC_DNA 10093 Brucella abortus (Star Ecotronics, Milan, Italy). Each sample was partitioned into approximately 20,000 nanoliter-sized droplets with AutoDG automated droplet generator (Bio-Rad Laboratories) using QXDx AutoDG oil for Probes. Next, the plate was sealed with pierceable foil heat-sealed (Bio-Rad Laboratories) at 180 °C using PX1 PCR plate sealer (Bio-Rad Laboratories) and PCR amplification was carried out in T100 Thermal Cycler (Bio-Rad Laboratories). The thermal profile was composed of initial denaturation at 95 °C for 10 min, followed by 40 cycles of 94 °C for 30 s and 60 °C for 1 min, 1 cycle at 98 °C for 10 min of denaturation, annealing and extension, respectively. Finally, the plate was loaded into QX200 Droplet Reader (Bio-Rad Laboratories) that automatically read the droplets in the wells. QuantaSoft software was used to count the fluorescent positive and negative droplets to calculate target DNA concentration. ddPCR results were converted into copies/µL by multiplying the concentration obtained by the total volume of the reaction mixture (22 µL) and then divided by template volume. Samples were considered as positive when at least three positive droplets containing the target DNA were detected, while samples were considered as negative when no positive droplet or less than three droplets were revealed, as per manufacturer’s guidelines (https://www.bio-rad.com/webroot/web/pdf/lsr/literature/Bulletin_6407.pdf).

Brucella abortus ddPCR performance

A total of 20 spleen homogenates belonging to a proficiency test provided by the NRCB for the diagnosis of Brucella spp. were used to evaluate the performance of the assay. Samples were blindly administrated to the operators, that underwent bacterial examination, q-PCR and finally ddPCR. Droplet digital PCR was validated using samples in triplicates. The obtained results were compared to bacterial examination and real-time PCR to verify the agreement with the expected results provided by the NRCB.

In BSL-3 laboratory, it was determined the lower limit of detection (LOD) of the test using Brucella abortus serovar 1, which was previously isolated and characterized by the national reference centre for brucellosis. In particular, a pure Brucella abortus culture was obtained on Brucella agar (BA) and a colony was collected and suspended in 3 mL sterile 0.9% NaCl to evaluate bacterial concentration with DensiCHEK Plus (BioMérieux). A set of standards consisting of a blank, 0.0 McF, 0.5 McF, 2.0 McF, and 3.0 McF were first used and then, through the measurement of the turbidity at 580 nm wavelength, the absorption of light determined the McF value. Next, using a sterile 10 µL inoculation loop, another colony was collected and suspended in 100 µL sodium chloride, thus obtaining the stock solution (CM), that was used to carry out a 10-fold serial dilution up to reach 10⁻¹⁰ final dilution (C1-C10), by adding the 100 µL of CM and suspended in 900 µL sodium chloride into 2 mL tubes. Moreover, a negative control (C) composed exclusively of sodium chloride was included. Next, using sterile flocked swabs was introduced into each tube and plated onto Brucella agar plates in duplicates (BA1 and BA2), incubated at 37 °C and daily inspected for 72 h, Using sterile swabs, each dilution was plated onto Brucella agar in duplicates (BA1 and BA2), and incubated at 37 °C. A first colony count was carried out at 48 h (CFU/mL 48 h) and 72 h (CFU/mL 72 h). Furthermore, aliquots of 500 µL from each dilution tube were collected, thermally inactivated at 99 °C for 10 min and submitted to nucleic acids extraction to carry out in parallel RT-PCR and dd-PCR.

Statistical analyses

Accuracy, repeatability and reliability of ddPCR were evaluated using Choen’s kappa (k), comparing the obtained results of the proficiency test through droplet digital PCR and the expected outcomes. Furthermore, using field samples, McNemar’s non-parametric statistical test for paired data was used to evaluate whether the difference between qPCR and ddPCR data distribution was statistically significant. The test assumes, as null hypothesis, that there is no difference between the two tests while, as an alternative hypothesis, it is stated there is a statistically significant difference between the two tests. The significance level was set at p ≤ 0.05. Moreover, the coefficient of variation (CV) was calculated. This parameter measures the average variation of a phenomenon in relation to its arithmetic mean and is a useful for comparing the relative variability of a phenomenon under different circumstances. Thereby, in order to evaluate the intra and inter assay average variability, the coefficient of variation (CV) was calculated for each C (10-fold dilution) and the average of the previously obtained coefficients of variation was subsequently calculated.

All statistical analyses were carried out using RStudio version 4.1 software.

Results

Accuracy, reliability and repeatability of ddPCR were assessed using the proficiency test for the detection of Brucella spp., composed of 14 positive and 6 negative spleen homogenates examined in triplicates. The obtained results were in accordance with the outcome of the bacterial culture and q-PCR, and the replicates showed repeatable results. Thus, a Choen’s kappa equal to 1 (k = 1) was obtained. Furthermore, in order to evaluate the LOD of ddPCR, a pure colony of Brucella abortus serovar 1 was used. Bacterial concentration was assessed through a turbidity test that determined 0.5 McF value, corresponding approximatively to 5x10⁸ CFU/mL. Thereby, a 10-fold dilution was performed. Thus, starting from 10⁻¹ dilution (corresponding to 5x10⁷), all the dilutions were plated on the two solid agars and q-PCR and ddPCR were carried out. Good efficiency and linearity for q-PCR and ddPCR assays were demonstrated on the serially diluted colony. For each dilution, q-PCR showed a reduction of three threshold cycles (Ct) and 10-fold reduction of the copy number by ddPCR. Furthermore, both the techniques were able to detect Brucella DNA up to 10⁻⁶, that determined at bacterial culture, a mean of 225 CFU/mL. Thus, this value was considered as the LOD of both the assays. At higher dilutions, the nucleic acids were not detectable, as RT-PCR and ddPCR showed 40.8 Ct value and 0.06 copies/µL (1 event/reaction) respectively, that, according to the relative protocols, are intended as negative (Table 2).

Table 2.

Results of bacteria count on serial dilution of B. abortus serovar 1 pure colony related to real time PCR and droplet digital PCR and intra and inter assay coefficient of variation.

		Bacteriological examination (CFU/mL)		Real time PCR			Droplet digital PCR
Dilution	(CFU /mL)	48 h	72h	Result	Ct mean	CV	Result	copies/µL mean	CV
C (blank)	Ø	Ø	Ø	N	N.A		N	N.A.	N.A.
C1 (10−1)	5x107	Massive	Massive	P	17,00	0,00	P	9046,67	0,11
C2 (10−2)	5x106	Massive	Massive	P	22,00	0,03	P	399,00	0,04
C3 (10−3)	5x105	Massive	Massive	P	25,00	0,00	P	34,67	0,03
C4 (10−4)	5x104	Massive	Massive	P	28,00	0,00	P	4,73	0,18
C5 (10−5)	5x103	Massive	Massive	P	31,00	0,00	P	0,48	0,61
C6 (10−6)	5x102	Ø	225	P	35,30	0,02	P	0,13	0,09
C7 (10−7)	5x101	Ø	22	N	N.A.	N.A.	N	0,00	0,00
C8 (10−8)	5x100	Ø	2	N	N.A.	N.A.	N	0,00	0,00
C9 (10−9)	5 × 10−1	Ø	Ø	N	N.A.	N.A.	N	0,00	0,00
C10 (10−10)	5 × 10−2	Ø	Ø	N	N.A.	N.A.	N	0,00	0,00
Intra-assay CV					0,91			3,00
Inter-assay CV					0,004			0,11

C (blank): control; C1-10: 10-fold serial dilution of B. abortus colony. N.A.: Not applicable. Ct: cycle threshold; CV: Coefficient of Variation.

Among 599 field samples, a difference in the detection of Brucella DNA was revealed between q-PCR and ddPCR. Indeed, out of 534 q-PCR negative samples, 35 tested positive by ddPCR method, obtaining a total of 100 positive cases, compared to 65 of q-PCR. For each matrix, the details of the results obtained by q-PCR, ddPCR and bacteriological culture are reported in Table 3. No sample tested positive by real time and negative by ddPCR. Sensitivity, specificity and accuracy of the test were evaluated, showing a sensitivity of 100% and specificity of 93.4%, as well as an accuracy of 94.15%. Furthermore, McNemar non-parametric statistical test for paired data showed these results were statistically significant (p < 0.05), thus demonstrating the difference in the performance of the tests were not a random effect (Figure 1).

Table 3.

Results of real time PCR, droplet digital PCR and bacteriological culture among field samples.

		q-PCR		ddPCR		Bacteriological culture
		Positive		Positive		Positive
Matrix	Total	n	%	n	%	n	%
Lymph nodes	329	36	10.94	53	16.11	45	13.67
Spleen	79	2	2.53	8	10.12	6	7.59
Genital Organs	112	25	22.32	35	31.25	32	25,57
Udder	79	2	2.53	4	5.06	4	5.06
Total	599	65	10.85	100	16.69	87	14.52

Figure 1.

Output obtained by statistical analysis software for the assessment of the performance of droplet digital PCR compared to real time PCR.

Discussions

Currently, qPCR is the most widely used molecular technique to detect and characterize microorganisms and gene targets in a broad type of samples, as it is a versatile, easy to perform and cheap tool. Nevertheless, this assay shows some limits, mostly identified in the need of constructing a standard curve to obtain an absolute quantification of the investigated target (Boulter et al. 2016), as well as the influence that natural inhibitors including inorganic and organic substances, such as phenolic components, heavy metals, haemoglobin, immunoglobulin G (IgG), proteinases, and fats of certain matrices, that may induce to misleading results (Schrader et al. 2012; Kojabad et al. 2021; Wang et al. 2022) and reduce the sensitivity of PCR. Droplet digital PCR is a more recent third generation polymerase chain reaction (Baettig et al. 2023) that is acquiring growing application in the early detection of cancer genes, biological agents used for bioterrorism purposes, microorganisms in the environment and in food supply chain (Li et al. 2018; Chen et al. 2021; Du et al. 2022; Baettig et al. 2023). Furthermore, growing attention has risen for the diagnosis of human infectious diseases in detecting low abundance of nucleic acids (Li et al. 2018). This novel assay represents an evolution of the well-known and consolidated real-time PCR technique. It combines TaqMan-based PCR technology to an innovative element of the ‘droplet’ system (Chen et al. 2021) through the creation of an emulsion phase in the reaction mixture, conceived to have in a single droplet, a unique nucleic acid molecule. The main advantages of this method are the absolute and direct quantification of the target gene, thereby, technical replications are deemed unnecessary (Baettig et al. 2023), a reduced susceptibility to the interference of inhibitors, and high sensitivity in the detection of low concentrated nucleic acids (Li et al. 2018; Chen et al. 2021). Accordingly, some studies have evaluated ddPCR LOD for some bacteria, showing that this technique is able to detect 50-100 CFU/mL (Abram et al. 2020; Chen et al. 2021), even though other authors observed a LOD of 10° CFU/mL and 10¹ CFU/ml, for Lacticaseibacillus casei and Salmonella enterica, respectively (Kim Choi et al. 2023; Kim, Yang et al. 2023). In this context, the present study aimed to evaluate the use of ddPCR in biological matrices to detect Brucella abortus in livestock, through the validation of the protocol using proficiency test samples and assessing the LOD, in term of detectable number of bacteria.

Currently, the ‘gold standard’ for the diagnosis of brucellosis in livestock is still considered the bacterial culture, as reported by international and national regulations, which permit to isolate the bacterium. On the other hand, molecular analyses are considered as additional tools for the detection of Brucella targets. Thereby, whenever possible, bacteriological culture shall be performed (World Organisation for Animal Health (WOAH) n.d.a), albeit it is time-consuming, needing over 3 days for bacterial grow, the risk of laboratory-acquired infection and the low isolation rate, as Brucella often have limited loads within the samples (Bonfini et al. 2018; Liu et al. 2023). Consequently, the use of molecular examinations is increasing for the rapid and sensitive diagnosis. Herein, we have compared results obtained by real time PCR and ddPCR both in reference material and in field samples.

Notably, ddPCR showed high reliability and accuracy (Choen’s kappa = 1), correctly identifying positive and negative samples, as well as good repeatability. Moreover, by the pure colony 10-fold serial dilution of B. abortus, RT-PCR and ddPCR showed a comparable performance, as they both correctly detected the nucleic acids at 10⁻⁶ final dilution, corresponding to a Ct value of 36.6 and 0.2 copies/µL (3 events), respectively. At this concentration, bacteriological culture gave rise to a mean of 225 CFU/mL that was set as the lowest limit of detection of this assay as, at the following dilution (10⁻⁷), 22 UCF/mL was revealed and a Ct value of 40.8 and 0.06 copies/µL (1 event) were detected. Therefore, results shall be considered as negative. Interestingly, same LOD values were revealed by G¨urbilek and colleagues who have recently developed an antigen capture-ELISA for the detection of Brucella spp. in milk samples. They revealed that this assay was able to detect up to 50-100 bacteria per well, equivalent to 2 × 10³ CFU/mL (Gürbilek et al. 2023). The evaluation of the coefficient of variation (CV) showed that ddPCR demonstrated an average coefficient of variation value of approximately 10%, while RT-PCR recorded an average of 0.4%. It should be noted that the coefficient of variation calculated within each group is higher for ddPCR because the variability in the groups is probably more affected by the dilution effect in the absolute quantification of ddPCR, compared to the RT-PCR Ct value.

Conversely, among field tissue samples, a different performance was observed. ddPCR was able to detect 35 additional positive samples compared to qPCR, thus obtaining a positivity rate of 16.7% compared to 10.8% of qPCR, obtaining high sensitivity (100%) and specificity (93.4%) values. Furthermore, statistical analysis confirmed the difference between the results of the two assays was statistically significant (p < 0.05). This disproportion was also observed in whole blood human samples. Indeed, among SAT positive patients, ddPCR detected more positive cases than qPCR, and, more interestingly, among suspected sero-negative samples, ddPCR detected seven further cases (19/33) in comparison with qPCR (12/33) (Liu et al. 2023). Altogether, these results underline that, although ddPCR and qPCR show similar results under controlled laboratory conditions, among field samples ddPCR demonstrates higher suitability for the detection of Brucella DNA. Indeed, in experimental conditions, with artificial infection or direct exposure to the pathogen, optimal circumstances are obtained but, on the other hand, the conditions of field samples are extremely different, characterized by the possible presence of inhibitors (Schrader et al. 2012; Kojabad et al. 2021; Wang et al. 2022) or not always adequate amount of Brucella (Liu et al. 2023), that may create results difficult to interpret or false-negative results (Schrader et al. 2012). Indeed, a study evaluated the inhibitory effect of sodium dodecyl sulfate, heparin and EDTA, some of the main PCR inhibitors, on both q-PCR and ddPCR using serial dilution of the substances. The authors concluded that ddPCR is more tolerant to inhibitors than q-PCR, due to the partitioning of the reaction in the microreactions inside a droplet. On the other hand, q-PCR reactions are not partitioned, so that the amplification is strictly dependent on the presence and concentration of the inhibitors. Thereby, an increasing number of amplification cycles are needed to obtain a fluorescent signal (Dingle et al. 2013). Same results were obtained by Rački and colleagues (Rački et al. 2014) that evaluated the effects of inhibitors among seeds, plants, soil and wastewater, using serially diluted pectin, dextran sulfate, tannic acid and humic acid. One of the method that is widely applied to evaluate amplification inhibition, is the use of internal controls (ICs), such as beta actin or beta-2-microglobulin. In particular, it is encouraged to include ICs in molecular assays for veterinary specimens, that easily contain PCR inhibitors (Yan et al. 2020). In order to minimize their effect, dilution of the template is frequently carried out, diluting sample inhibitors, too (Schrader et al. 2012). Nevertheless, if on the one hand this technique reduces the effect of the inhibitors, on the other hand it leads to the dilution of the investigated target. Therefore, if the target has rather low starting load, as typical field conditions, the risk of not detecting the nucleic acid rises exponentially. In our case, the scarce presence of bacterial target in the sample, which frequently characterizes field samples (Liu et al. 2023), as well as the presence of possible tissue inhibitors, such as fat, immunoglobulins, collagen, etc. that can be found in tissue samples, especially of veterinary origin (Schrader et al. 2012) may have influenced the sensitivity of ddPCR compared to qPCR. Despite a sampling method based on flocked swabs and not tissue homogenization, it is plausible that inhibitors may also have been collected. Indeed, flocked swabs are constructed with open-fiber structure onto the nylon-tipped surfaces that confer, through an electrostatic mechanism, high performance in collection and improved release efficiency (Viviano et al. 2018; Freddi et al. 2021; Kumarajith et al. 2024).

In conclusion, ddPCR has proven to be a promising technique to detect Brucella abortus in veterinary specimens. Nevertheless, long processing time is required and it is still more expensive than qPCR, thus, it may have limited applications in routine analyses. Thereby, RT-PCR remains a more cost-effective technique (Baettig et al. 2023). However, it could be a useful tool to investigate and clarify doubtful cases or to test samples that show inhibitors or low bacterial loads. Furthermore, it could find its application in the early detection of Brucella in suspected farms located in non-free areas, pending the results of bacteriological analyses, in order to contain the spread of the disease. Considering the high sensitivity of the method, it could contribute to reveal false negative results by serological tests. Finally, a possible future application could be the detection of the bacterium in non-invasive matrices where Brucella abortus is highly diluted, such as bulk milk.

Funding Statement

This work was supported by the Ministero della Salute with the Ricerca Corrente ‘Metodologie molecolari innovative nella diagnosi della brucellosi bufalina’, RC IZSME 13/22 under grant CUP C75E22000500001.

Authors contribution

This study was conceived and designed by G.F., E.D.C., M.N. and S.R.; G.F., L.C., E.S., C.d.M and G.B. drafted the manuscript; M.O. and R.P. conducted statistical analyses; G.B., A.P., A.C., F.D.F. and G.P. conducted molecular examinations; O.V., conducted bacteriological examinations; M.T., S.R. and E.D.C. final revision and approval of the manuscript. All the authors have read and approved the final version of the manuscript.

Disclosure statement

The authors report there are no competing interests to declare.

Ethical approval

All samples were collected by the Veterinary Local Health Authorities during mandatory routine activities for national and regional Brucellosis Eradication Plans. Moreover, the IZSM is the Official Laboratory designed by the Italian Ministry of Health, thus, according to national regulation and internal policy, ethical approval was deemed unnecessary.

Data availability statement

All data supporting the results of the paper are included within the manuscript.