Second line molecular diagnosis for bovine tuberculosis to improve diagnostic schemes

Surveillance of bovine tuberculosis (bTB) is partly based on the sanitary inspection of carcasses at the abattoir to detect bTB-like lesions which, in compliance with EU recommendations, are analysed by bacteriology and histopathology to disclose Mycobacterium bovis (or M. caprae) infection. Moreover, since 2012, a PCR method with similar sensitivity and specificity values of histopathology and bacteriology respectively is additionally employed in France, partially compensating for the weaknesses of classical diagnostic methods. We analysed a collection of bTB-like lesions from cattle presenting positive histological results albeit with negative PCR results. We present here the results of these samples, recovered from 292 animals culled between 2013 and 2016, analysed with a second line molecular diagnosis approach that consists in a combination of PCRs targeting the M. tuberculosis-M. avium complexes as well as the Mycobacterium genus and sequencing of hsp65 gene. These molecular analyses disclosed to identify the presence of non-tuberculous bacteria which could be responsible for most of these non-specific TB lesions: non tuberculous mycobacteria (24%) or Actinomycetales (56%) such as Rhodococcus equi (53%); 24% of the samples were negative. M. bovis -or any other MTBC members- was neither detected by molecular methods nor isolated in any of them at the end of the 3 months of culture. In conclusion, these results highlight the lack of specificity of histopathology and the usefulness of a first line PCR with a second line molecular diagnostic test to circumvent it. This diagnostic strategy makes it possible to reduce the number of suspect bTB cases raised at the abattoir or shortening their lock-up periods. By simplifying diagnostic schemes, the use of this tool could improve bTB surveillance and make eradication programs more efficient in the future.


Introduction
Bovine tuberculosis (bTB), mainly due to Mycobacterium bovis, is an important re-emergent zoonotic disease in Europe [1]. Although France has been officially bTB free (OTF) since 2001, the persistence of the disease in livestock and its occurrence in wildlife in some areas is of great concern [2,3]. The bTB control campaign in France, in compliance with the EU Directive 64/ 432, is mainly based on regular skin testing of animals and on detection of bTB-lesions during routine veterinary inspections at the abattoir. Indeed, inspection at the abattoir is a cost-effective method, especially in low prevalence areas or OTF countries [4]. In this particular context, the submission of all suspect lesions detected during meat inspection to the laboratory for histopathology and/or culture examination is necessary to increase the sensitivity of the surveillance system [5,6]. Still, only by implementing more epidemiologically adapted control measures could eradication be envisaged [7]. BTB diagnosis by bacteriology can take up to three months due to the slow growth of the Mycobacterium tuberculosis complex (MTBC) mycobacteria [8]. Alternative more rapid tools such as histopathology are employed to circumvent the bacteriology slowness drawback. However, although histopathology is a fast and sensitive method it lacks specificity [9]. Actually, in a recent study on the evaluation of sensitivity and specificity of confirmatory bTB diagnostic tests, it was shown that histopathology was less specific than bacteriology, albeit as sensitive as another rapid test, a MTBC PCR introduced in France as a first line method, in parallel with bacteriology, to detect bTB infected animals [10]. This PCR method thus compensates for the specificity deficiency of histopathology which has been until now the only recognised rapid test in accordance with the EU Directive 64/432. Furthermore, the bovine tuberculosis National Reference Laboratory (NRL) applies a second line molecular diagnosis method that enables the identification of mycobacterial species either on mycobacteria bacteriological isolates or directly on DNA extracted from animal samples. This method provides rapid information about bTB or any other mycobacterial infection.
A significant and increasing number of cattle samples were analysed at the NRL, after discordant histopathological positive-first line PCR negative results. In this study we summarise the results of 4 years analyses on such samples with our second line molecular diagnosis scheme, which made it possible to identify non-tuberculous bacterial agents giving rise to nonspecific bTB-like lesions, to avoid cumbersome -albeit official-diagnostic alternatives such as culture for confirming the bTB-free status of the herd, thus gaining diagnostic specificity and confidence for bTB status confirmation.

Ethical statement
BTB is a notifiable disease for which there are control and surveillance campaigns in France. Official methods for diagnosis of this disease are culture, PCR and histopathology. Therefore, all the samples included in this study are issued from animals analysed within an official context. No purpose killing of animals was performed for this study. All samplings were in complete agreement with national and European regulations. No ethical approval was necessary.

Sample collection
Samples included in our study presented macroscopic bTB-like lesions at routine abattoir inspection between 2013 and 2016. Inspection procedures for bovine carcasses implemented in France (DGAL/SDSPA/SDSSA/N2013-8123, https://info.agriculture.gouv.fr/gedei/site/boagri/instruction-N2013-8123) follow the regulation(EC) No 854/2004 (Annex 1, Section IV, Chapter I). Typically, lesions due to M. bovis have a centre of caseous necrosis, sometimes associated with calcification, surrounded by epithelioid cells, lymphocytes and neutrophils [6,11]. Samples had previously been submitted to first-line bTB diagnosis (bacteriology/PCR and histopathology) by authorised regional laboratories (RL) of the national surveillance network for bTB [10]. Briefly, histopathology was based on Hematoxylin-Eosine and Ziehl Neelsen staining. Bacterial culture is performed following the protocol established by the French NRL (NF U 47-104) for isolation of M. bovis. Two to 5 g of sampled tissues were crushed with a 4% sulfuric acid solution to decontaminate the tissue. After 10 min, the acid was neutralized by adding a 6% sodium hydroxide solution. After decontamination, the supernatant was seeded on two different solid media: Löwenstein-Jensen and Coletsos. All seeded media were incubated at 37˚C +/-3˚C for three months and exanimated every two weeks. Any isolated mycobacterial strain is submitted to the NRL for further characterisation. DNA from each sample was extracted by using the QIAamp DNA mini kit (Qiagen, Courtaboeuf, France) or by Magvet MV384 (Thermo Fisher scientific, Villebon-sur-Yvette, France) with a King Fisher KF96 automate, following the manufacturer's instructions and analysed with the LSI VetMAX Mycobacterium tuberculosis Complex Real-Time PCR Kit (Thermo Fisher scientific, Villebon-sur-Yvette, France). Samples submitted to the NRL for further molecular analyses were those that (i) presented a histopathological result suggesting tuberculosis, and (ii) showed a negative result with the first-line PCR.
Samples from 81 TB-free cattle without TB-like lesions following diagnostic slaughter and having presented a negative PCR result at the first-line bTB diagnosis were also included as a control population.

Confirmatory tests-second line molecular diagnosis
For further analyses at the NRL, original tissue, macerated tissue and extracted DNA were sent by the RLs. A first analysis on DNAs was done by real-time PCR targeting insertion sequences IS6110 and IS1081 for MTBC identification, IS1245 for Mycobacterium avium complex (MAC) identification, and the 65 kDa heat shock protein gene (hsp65) for Mycobacterium sp. detection (Table 1). Real-time PCR assays were performed in a final volume of 25 μl using the TaqMan Fast Universal PCR Master Mix (Roche Diagnostics, Meylan, France) at a 1X final concentration, with primers at 300 nM and probes at 250 nM. PCR cycling comprised of 2 min at 50˚C and 20 s at 95˚C, followed by 50 cycles of 2-step amplification of 3 s at 95˚C, and 30 s at 60˚C. If necessary (negative or doubtful results) a second analysis was done with a new DNA extraction with the High Pure PCR Template Preparation Kit (Roche Diagnostics, Meylan, France) from the original tissue and the macerated lesion. Quality of the DNA extraction and PCR inhibition was tested with DiaControlDNATM (Diagenode, Thermo Fisher, USAdiagnostics, Belgium). Table 1. Primers and probes oligonucleotides for real-time-PCR assays used in this study.

Targeted genes or sequences
Primers-Probe name Sequence 5'-3' Identification of non-tuberculous mycobacteria species was done by sequencing using primers targeting the 65 kDa heat shock protein gene (hsp65) [12] or the β subunit of bacterial RNA polymerase sequence (rpoB) [13]. The obtained sequences were compared to the Gen-Bank/EMBL/DDBJ databases using the BLAST program.

Results
This retrospective analysis includes all cattle cases sent to the NRL which had a negative result for MTBC PCR but with a bTB suggestive histopathology. Between 2013 and 2016, we analysed samples of 292 cattle from 278 herds ( Table 2) Table). For the other nine samples, the results suggest co-infection as different pathogens were identified by bacteriology and PCR on tissue samples.
Of the 292 cases submitted to the NRL, 24% were NTM, 56% were Actinomycetales and 24% were negative based on sequencing (Table 2). In 11 cases, we identified co-infection, either between two NTM (M. avium avium and M. nonchromogenicum) or between a NTM (M. aichiense, M. bourgelatii, M. kansasii, M. nonchromogenicum or MAC) and an Actinomycetale (Rhodococcus equi). Identification of the bacteria species by sequencing highlighted that 95% of the Actinomycetales were R. equi. Among the NTM, MAC represents 26% of the cases, while the other half were various mycobacteria species (Table 3). These NTM and Actinomycetales were found in various samples, most frequently in retropharyngeal LN (31%), followed Of the 240 LN from control animals without bTB-like lesions and a negative first-line PCR results, 96% remained completely negative to the second line diagnosis (Table 4). Only two NTM were identified in mediastinal LN (M. lentiflavum and M. gordonae) and 8 actinobacteria from various genus: Corynebacterium sp., Streptomyces sp., Arthobacter sp. and Nakamurella sp. These results confirm that bacteria identified in nonspecific bTB-like lesions are not ubiquitous and accordingly that they are the real causative agents of them.

Discussion
We studied bTB suspect cases resulting from abattoir inspection which presented nonspecific histopathology bTB-like lesions and a negative first line MTBC PCR and focused on the identification of the bacteria responsible for them. The number of this type of nonspecific suspicions increased since 2013 as a result of awareness campaigns organised for abattoir agents, one of a series of measures introduced to reinforce the national bTB control campaign in 2011 [14]. None of these cases were real bTB infections, but other bacteria, i.e. NTM or Actinomycetales, have been identified in the lesions. Indeed, bTB-like lesions could also be caused by other granuloma forming organisms such as NTM or Nocardia species [15]. Moreover, mycobacteria and some Actinomycetales (of the genus Nocardia, Rhodococcus or Corynebacterium) shared the same tinctorial properties and thus are identified as acid resistant bacilli by Ziehl Neelsen staining [16]. Many environmental mycobacteria may interfere with the bTB surveillance program at post-mortem inspections at the abattoir [17]. NTM have recently been detected in lymph nodes of clinically healthy Swiss cattle, emphasizing the need of more specific diagnostic tools [18]. A study in Northern Ireland tried to identify mycobacteria in lymph nodes of cattle belonging to herds with previous evidence of bTB. The identified bacteria species were almost the same as in our study but with different proportions: a majority of M. nonchromogenicum, few MAC, few M. kansasii and only one R. equi [19]. NTM as well as R. equi, have been recognised in lymph node infection of domestic and wild animals (swine and wild boar (Sus scrofa), roe deer (Capreolus capreolus) or red deer (Cervus elaphus)), which could lead to a possible misdiagnosis of M. bovis [20][21][22]. Indeed, real tuberculosis lesions in cattle are commonly found in retropharyngeal LN (29.4%), mediastinal LN (28.2%) and tracheobronchial LN (18%) [6], i.e. the same locations as the non-tuberculous agents in our study.
Our results strongly suggest the link between R. equi and the presence of nonspecific bTBlike lesion as this species was not identified in the LN of cattle without any bTB-like lesions. Rhodococcus equi and non-tuberculous mycobacteria (especially MAC) are facultative intracellular pathogens surviving inside macrophages and inducing granulomatous inflammation [17,23]. Rhodococcus equi (formerly Corynebacterium equi) is a coccobacillus bacterium commonly found in soil which is pathogenic for domesticated animals such as horses, pigs and cattle [24,25]. Even if its pathogenicity is low in cattle, it may occasionally cause lymph node granulomas, which are detected at abattoir post-mortem examination [26]. The interference caused by R. equi in the monitoring of bTB was already acknowledged 35 years ago [27]. This bacterium has the capacity to modify the phagocytic vacuole of host macrophages and present similarities on cellular responses attributed to resemblances in cell wall composition and antigenic structure with bTB agents [9,25]. Granulomas caused by R. equi are most frequently observed in retropharyngeal, bronchial and mediastinal lymph nodes [25] and are really difficult to differentiate from M. bovis granulomas, even if the presence of a heavy infiltration of neutrophils and/or extensive sheets of macrophages could presumably allow distinction [26]. Non-tuberculous mycobacteria and MAC are ubiquitous in the environment and particularly found in wet soil, water or plants [20,22]. Infection of animals in our study probably occurred by the oral route through ingestion of food or water contaminated by these environmental organisms [25]. Some of the identified NTM species are recognised as leading to misdiagnosis of bovine tuberculosis, especially MAC [17] and Mycobacterium nonchromogenicum which is known to interfere with ante-mortem diagnosis of bTB [17,28,29]. The role of the other identified species as cattle pathogens is unclear. Mycobacterium bourgelatii, closely related to M. intermedium and described for the first time in 2013, was isolated from cattle lymph nodes [30]. M. intermedium is classified as a 'pathogen', together with M. gordonae and M. kansasii, in a recent phylogenetic analysis [31]. Thus, their role in nonspecific bTB diagnosis cannot be ruled out [17]. Mycobacterium aichiense closely related to M. gilvum [32], M. petroleophilum closely related to M. aurum [33], M. pyrenivorans and M. thermoresistibile are all rapidly growing mycobacteria found in the environment, potentially opportunistic in humans and included in the same phylogenetic group [31]. Mycobacterium genavense, slow growing mycobacteria, is responsible for infection both in birds and humans and has also been isolated from the environment [34]. Mycobacterium shimoidei, slow growing mycobacteria, is an opportunistic pathogen of humans but few pulmonary cases have been reported worldwide [35].
In our study, the majority of the samples presented a histological profile with an encapsulation of the granuloma and the presence of giant Langhans cells, sometimes in association with others cells (lymphocytes, neutrophils or epithelioid cells). The presence of epithelioid macrophages and Langhans cells is not pathognomonic as these cells are seen in immunologic granulomas [36], especially in tuberculous ones [37]. The histopathological results of our cases were quite the same for lesions due to NTM or Actinomycetales, indicating a clear lack of specificity in histopathological diagnosis. However, the use of immunohistochemistry, in complement of special staining, could have increased the specificity of this diagnostic test by demonstrating M. bovis antigen immuno-localisation [36]. In a previous study, the sensitivity and specificity of confirmatory tests (bacteriology, histopathology and PCR) were estimated under French field conditions [10]. Histopathology was found to be as sensitive as PCR but less specific than bacteriology or PCR, which means that this test cannot be used alone as a confirmatory test. Our results confirm and explain histopathology's lack of specificity. Besides, second line PCR showed its usefulness as it clearly improves bTB diagnosis by disclosing false MTBC-infected cases. The second line molecular tests used at the NRL have shown an excellent negative predictive value and have quickly reduced the number of bTB suspicions through the identification of other non-tuberculous bacteria; however, as the proposed molecular diagnosis scheme is not a EU officially recognised strategy, animal movement restrictions in the 278 incriminated herds were maintained during at least 3 months awaiting for declaration of a negative M. bovis culture. In conclusion, second line molecular tests used at the NRL could confidently be added as an official test in order to accelerate the diagnosis process and improve bTB surveillance in Europe and to render control and eradication programs more efficient in the future.
Supporting information S1