Importance and Antimicrobial Resistance of Mycoplasma bovis in Clinical Respiratory Disease in Feedlot Calves

Simple Summary Bovine respiratory disease is a common health and economic problem that mainly affects calves raised in feedlots. Several viruses and bacteria may be involved, but Mycoplasma bovis can cause disease chronification and poor response to antimicrobial treatment. This study investigated the role of Mycoplasma bovis in cases of clinical respiratory disease unresponsive to treatment that affected feedlot calves in southeast Spain, and tested the in vitro susceptibility of a selection of isolates to the specific set of antimicrobials used for therapy in vivo. Mycoplasma bovis was found in 86.9% (20/23) of the calves, predominantly in the lungs (78.26%; 18/23) where it was involved in pulmonary lesions. Furthermore, the selected isolates were found to be resistant in vitro to most of the antimicrobials specifically used for treating the animals in vivo. These results highlight the implication of Mycoplasma bovis in the bovine respiratory disease affecting feedlot calves in Spain. Abstract Bovine respiratory disease (BRD) is an important viral and/or bacterial disease that mainly affects feedlot calves. The involvement of Mycoplasma bovis in BRD can lead to chronic pneumonia poorly responsive to antimicrobial treatment. Caseonecrotic bronchopneumonia is a pulmonary lesion typically associated with M. bovis. In Spain, M. bovis is widely distributed in the feedlots and circulating isolates are resistant to most antimicrobials in vitro. However, the role of this species in clinical respiratory disease of feedlot calves remains unknown. Furthermore, available data are relative to a fixed panel of antimicrobials commonly used to treat BRD, but not to the specific set of antimicrobials that have been used for treating each animal. This study examined 23 feedlot calves raised in southeast Spain (2016–2019) with clinical signs of respiratory disease unresponsive to treatment. The presence of M. bovis was investigated through bacteriology (culture and subsequent PCR), histopathology and immunohistochemistry. The pathogen was found in 86.9% (20/23) of the calves, mainly in the lungs (78.26%; 18/23). Immunohistochemistry revealed M. bovis antigens in 73.9% (17/23) of the calves in which caseonecrotic bronchopneumonia was the most frequent lesion (16/17). Minimum inhibitory concentration assays confirmed the resistance of a selection of 12 isolates to most of the antimicrobials specifically used for treating the animals in vivo. These results stress the importance of M. bovis in the BRD affecting feedlot calves in Spain.


Introduction
Bovine respiratory disease (BRD) is a major disease of feedlot cattle that affects the upper or lower respiratory tract, causing high mortality and carcasses of lower quality [1]. Typical clinical symptoms include fever, dyspnea, coughing, nasal or eye discharge, depression and decreased or no appetite [2]. BRD has a multifactorial etiology, including infectious agents, host predisposing factors and environmental stressors. The disease usually appears in calves after a stressful environmental event such as weaning, transportation, co-mingling or drastic diet or weather changes [3]. Infectious agents commonly associated with BRD are the viruses bovine viral diarrhea virus (BVDV), bovine herpesvirus type 1 (BHV-1), parainfluenza-3 (PI-3) virus, bovine coronavirus (BCOV) and bovine respiratory syncytial virus (BRSV), and the bacteria Pastereulla multocida, Mannheimia haemolytica, Histophilus somni, Trueperella pyogenes and Mycoplasma bovis [4][5][6]. Generally, viruses are considered to be primarily BRD initiators that then promote colonization by bacterial pathogens [7]. Among bacteria, M. bovis can also act as a primary pathogen and is recognized as an important cause of chronic pneumonia that seems poorly responsive to antimicrobial treatment [7,8].
Often, calves become infected with M. bovis by close contact with asymptomatic carriers, which are occasionally shedding the pathogen in colostrum, milk, nasal or genital secretions [9][10][11]. Some animals acquire the infection at the farm of origin and others become infected once arrived at the feedlot [12][13][14]. In northern Italy and northwestern Spain, the analyses of pneumonic lungs recovered from beef cattle with subclinical pneumonia revealed the presence of the pathogen in 25 (16/64) and 66% (33/50) of the animals [5,15]. In eastern and western France, M. bovis was isolated in 78.5 (106/135) and 52.1% (60/115) of the feedlot calves at the onset of BRD outbreaks, based on the analyses of broncho-alveolar lavage (BAL) and nasal swab samples, respectively [4,16].
Caseonecrotic bronchopneumonia with multiple foci of caseous necrosis is a pulmonary finding typically associated with M. bovis in naturally or experimentally infected calves [8,17,18]. Other authors consider the bronchiolitis as another distinctive lesion [19][20][21]. Furthermore, the pathogen may be involved in other lesions such as necrosis of the bronchiolar epithelium, bronchus-associated lymphoid tissue (BALT) hyperplasia, and bronchiolar fibrosis [19,20,22]. Additionally, M. bovis may be involved in lesions such as bronchopneumonia with foci of coagulative necrosis or abscesses, but usually in co-infection with other bacteria [12]. In such cases, the necrosis lesions originated from bronchioles or small bronchi are distinctive of the mycoplasma presence [22]. In addition to the histopathological analyses of pneumonic lesions, the presence of M. bovis has to be confirmed by bacteriological (culture and subsequent PCR) and/or immunohistochemical analyses [18,22].
Prevention and control of M. bovis pneumonia mainly rely on antimicrobial treatment as there are no efficient vaccines available [23]. However, in vitro antimicrobial resistance has been reported by many countries worldwide [24,25]. In Spain, recent studies have demonstrated the extended circulation of M. bovis in beef cattle herds and the circulation of isolates resistant to most antimicrobials in vitro [26][27][28]. For minimum inhibitory concentration (MIC) assays, those studies included some isolates from the lungs of young animals with clinical respiratory disease, but pulmonary lesions compatible with M. bovis were not evaluated. Hence, the role of M. bovis in the clinical respiratory disease of feedlots calves in Spain remains to be addressed. On the other hand, these studies tested the in vitro susceptibility of M. bovis isolates against a set of antimicrobials commonly used in the field, but did not consider the specific antimicrobials used for the treatment of each animal in vivo.
This study was conducted (i) to address the role of M. bovis in clinical respiratory disease unresponsive to antimicrobials in feedlot calves in Spain through bacteriological, histopathological and immunohistochemical techniques; and (ii) to determine the MIC values of the isolates recovered against the specific set of antimicrobials used for therapy in vivo.

Animal Sampling
All animal procedures met the conditions set out in the EU Directive 2010/63/EU for animal experimentation and had the authorization of the Ethics Committee on Animal Testing of the University of Murcia (Number: 307/2017).
In this study, 23 calves (50-350 kg), raised in 12 feedlots placed in the southeast of Spain, were sampled over a four-year period (2016-2019). The epidemiological background of the animals is summarized in Table 1. The animals showed clinical signs of respiratory disease and did not respond to antimicrobial treatment. Most of them (18/23) were euthanized with T-61 (MSD, Kenilworth, NJ, USA) by the feedlots' trained veterinary staff and the carcasses were submitted for necropsy. The necropsies and sampling procedures were conducted following the standard necropsy procedures for ruminants. In these animals, one nasal, auricular and conjunctival swab from the left and right side, one lung swab and one lung tissue specimen were collected (n = 5 per animal). The remaining five animals (5/23) were sacrificed at the slaughterhouse. In these calves, one lung swab and one lung tissue specimen were obtained (n = 2 per animal). In all cases, lung swabs and lung tissues were obtained from areas of cranioventral consolidation. In total, 100 samples were obtained. The sample collection was composed of auricular (n = 18), conjunctival (n = 18), nasal (n = 18) and lung (n = 23) swabs, as well as lung tissues (n = 23).
Spain VC-e Tilm, Flor, Marb France France VC-i Flor  1 Numbers in bold indicate the animals from which one isolate was selected for MIC determination. MIC values were calculated for antimicrobials used for the treatment in vivo. MIC values for amoxicillin and sulfadimidine were not determined, as mycoplasmas are intrinsically resistant to these antimicrobials [25]. A single isolate per animal was tested; if possible, the isolate obtained from the lung swab was used for MIC assays. When no isolate was obtained from the lungs, MIC was determined for the isolate obtained from the nasal swab as the second option, or from the auricular swab as the third option. Resistance breakpoints encompasses the intermediate breakpoints. 2 Different letter (a-l) indicates different feedlot. 3 Anatomical location as detected by culture and PCR except in lungs, where both culture and PCR and IHC were carried out.

Mycoplasma Cultures, DNA Extraction and PCR
Swabs were put into a sterile tube with Aimes agar transport medium (Deltalab ® , Barcelona, Spain) and preserved at 4 • C for culture and molecular analysis. Lung tissues were put into a sterile container and fixed in 10% neutral-buffered formalin for histopathology and immunohistochemistry (IHC).
For mycoplasma isolation, swabs were incubated at 37 • C with 5% CO2 for 24 h in 2 mL of SP4 medium [29], modified as previously described [28,30]. Cultures were purified through a membrane filter of 0.45 µm (LLG-Labware, UK) and incubated for 48 h before plating 5 µL onto SP4 agar. The plates were incubated at 37 • C with 5% CO2 and checked daily under a light microscope for the presence of mycoplasma colonies.
DNA was extracted from 200 µL of broth culture [31] and the presence of M. bovis was confirmed by PCR [32]. By picking single colonies, PCR-positive cultures were three times cloned and the species of the final isolate was examined again by PCR.

Histopathology and IHC
The formalin-fixed lung tissues were embedded in paraffin wax and cross-sectioned 4-5 µm thick with a microtome for histopathology and IHC.
For histopathology, sections were stained with hematoxylin and eosin (H-E) and examined under a light microscope. On sections with necrotic foci Gram-stain was also used.
Histological sections were analyzed to assess the presence or absence of the following changes: bronchiolar necrosis, intrabronchiolar and alveolar neutrophils, bronchiolar and alveolar fibrosis, bronchiolar and alveolar syncytial cells, foci of coagulative necrosis, abscesses, foci of caseous necrosis, alveolar and septal thrombosis, fibrinous pleuritis, pleural fibrosis, BALT hyperplasia and alveolar fibrin exudation. The presence or absence of Gram-positive and Gram-negative bacteria and areas of mineralization was analyzed in the necrotic foci.
The detection of M. bovis antigen was carried out by IHC on paraffin-embedded sections as previously described [17], using a rabbit polyclonal antibody (Ref. PA295) raised against whole cell antigen of M. bovis, diluted 1:500, and the avidin biotinylated enzyme complex (ABC) method (Vector Laboratories, Burlingame, CA, USA). Substitution of the primary antibodies with mouse non-immune serum, and lung tissue from the control calves, served as negative controls. Then, the presence or absence of M. bovis antigen was assessed by examining the lung sections.

MIC Assays
MIC values were calculated only for the antimicrobials used for the treatment in vivo and with recognized antimycoplasmic effect. MIC assays were carried out when two conditions were met: (i) the treatment received by the animal in vivo was provided by the feedlot's veterinary staff and (ii) at least one M. bovis isolate was obtained from the animal. In those cases, isolates obtained from lung swabs were used for MIC assays. If no isolate was obtained from the lung swab, the assay was carried out with an isolate obtained from the nasal swab or the auricular swab.
Stationary-phase cultures of M. bovis isolates were used for MIC assays. The reference strain PG45 was used as a control. The microbroth dilution method was carried out in 96-well microtiter plates. Mycoplasma cultures and MIC assays were carried out as already described [28], and following previous recommendations [33]. All the assays were repeated twice. If the results of the repeated tests differed in only one dilution, the higher MIC value was used. If the MIC value differed in more than one dilution, a third repetition was carried out and the final MIC value was the mode of the three values. MIC values were interpreted by considering breakpoints for M. bovis proposed by other authors or analyzing whether mutations related to antimicrobial resistance had been described for those values [28,34]. For gentamicin and lincomycin, MIC values were interpreted according to breakpoints proposed for other mycoplasma species [33,35].

Detection of M. bovis in Different Anatomical Sites
In this study, M. bovis was detected in 86.9% (20/23) of the calves and 53% (53/100) of the analyzed samples (Table 1). Among the different anatomical locations studied, the pathogen was most commonly found in the lungs (78.26%; 18/23). In these samples, IHC immunolabeled 73.9% (17/23), and culture and subsequent PCR detected 65.2% (15/23). Most animals diagnosed as M. bovis PCR-positive were also positive by IHC, with only one exception (animal nº 8). Other PCR-positive samples were nasal (11/18), conjunctival (6/18) and auricular swabs (4/18). Generally, animals that carried M. bovis in these anatomical areas also carried the pathogen in the lungs. Only two exceptions were found, the animals nº 12 and nº 14, which were identified as auricular, and conjunctival and nasal carriers, respectively (Table 1).
The field isolates tested were recovered from lung (n = 9), nasal (n = 2), or auricular (n = 1) swabs. Individual MIC values for each isolate are shown in Table 1. MIC values were >128 µg/mL for tulathromycin, tilmicosin and lincomycin; 32 µg/mL for enrofloxacin; ≥8 µg/mL for oxytetracycline; and ≥4 µg/mL for florfenicol and gentamicin. MIC values were > 64 µg/mL for three of the four isolates tested against marbofloxacin and 0.5 µg/mL for the remaining isolate. These values reflected the low susceptibility of the isolates to the antimicrobials tested, with only two exceptions. One was the isolate from animal nº 20, recovered from the lung and susceptible to gentamycin (MIC = 4 µg/mL). Lesions typically attributed to M. bovis such as caseonecrotic bronchopneumonia or bronchiolitis and M. bovis antigen were observed in this animal ( Table 2). The other exception was the isolate from animal nº 14, recovered from the nasal swab, and with a low MIC value for marbofloxacin (0.5 µg/mL) ( Table 1). No lesions typically attributed to M. bovis, nor M. bovis antigen, were observed in this animal. Instead, lesions compatible with other bacteria such as fibrinous and suppurative bronchopneumonia were found (Table 2). Similar findings were observed in the animals nº 8 and nº 12, from which resistant isolates had been recovered from the lung and the auricular canal, respectively (Tables 1 and 2). From the remaining eight animals, multiresistant isolates were recovered from the lung (n = 7) or the nasal (n = 1) swab. All of them were IHC-positive and presented at least one of the two lesions characteristic of M. bovis (Tables 1 and 2).

Discussion
In this study, M. bovis was found in 86.9% of the feedlot calves (20/23) with clinical respiratory disease unresponsive to antimicrobial treatments. This value was comparatively higher than those reported in western (78.5%; 106/135) and eastern French calves (52.1%; 60/115) at the onset of BRD [4,16], and reinforces the hypothesis of a previous study sustaining that the infection may have become endemic in Spanish feedlot herds [28]. In that study, M. bovis was identified in 40.9% (84/205) of the beef cattle analyzed. More specifically, the pathogen was mainly detected in feedlot calves (81/183) and to a lesser extent in pasture-raised animals (3/22) housed in 26 different farms from five Spanish regions [28].
The M. bovis diagnosis in lung samples differed slightly depending on the technique used, with 73.9% (17/23) of positives by IHC and 65.2% (15/23) by culture and subsequent PCR. This difference may be because the diagnosis based on culture and PCR is limited by the viability of the mycoplasmas present in the sample. Only one animal (nº 8), showing suppurative and fibrinous bronchopneumonia, was found to be PCR-positive and IHC-negative. A possible explanation is that M. bovis was present and viable in the lung but in low concentration, so IHC failed to detect it. This hypothesis agrees with other authors observations who reported that IHC failed to detect small numbers of intralesional M. bovis organisms [5]. All in all, the combined use of the two methods augmented the sensitivity of the diagnosis.
Given the multifactorial character of BRD, it is also common to find diverse patterns of pneumonic lesions in a single animal [21]. Our study was no exception as most of the calves (82.6%; 19/23) had a mixed pulmonary pattern, the quadruple combination of caseonecrotic, suppurative and fibrinous bronchopneumonia and interstitial pneumonia being the most frequent (30.4%; 7/23). In this context, viruses generally act as primary pathogens that, damaging the innate immunity and the epithelial surface of the airways, enable the participation of opportunistic bacteria. Among them, M. bovis can act as a primary pathogen [7]. However, this role is still controversial in some scientific communities and countries, mainly because this mycoplasma species is often found in asymptomatic carriers. Notably, caseonecrotic bronchopneumonia was the only pattern found in two animals (nº 20 and nº 22) and no Gram-positive nor Gram-negative bacteria were detected in the foci of necrosis. Although M. bovis was likely the primary cause, other bacteria, such as M. haemolytica, could have initiated the foci of necrosis and been removed by the antimicrobial therapy, as previously proposed [12]. Nevertheless, this finding still argues in favor of the role of M. bovis in respiratory disease. On the other hand, interstitial pneumonia and multinucleated syncytial cells were the only findings in one animal (nº 9). In effect, some viruses, such as BRSV and BHV-1, can induce life-threatening disease without bacterial superinfection [7,39]. Fibrinous bronchopneumonia was the only pattern found in another animal (nº 12), although no bacteria were observed in the foci of necrosis. They could have been eliminated by the antimicrobial treatment administered to the animal.
MIC values showed the low in vitro susceptibility of the M. bovis isolates (n = 12) to most of the antimicrobials received in vivo by the calves. This could be the result of resistance acquired after treatment [24,25]. Another possible explanation is that M. bovis isolates involved in the BRD episode were already resistant before any antimicrobial treatment, as recently observed in France [16]. Indeed, multiresistant strains currently circulate in that country, as it occurs in Spain [26][27][28]. Some authors propose that therapy in vivo may fail not because of the involvement of a resistant strain but because of the limited drug distribution into the caseous foci where M. bovis bacteria are most numerous [12]. This might explain the case of animal nº 20, with caseonecrotic bronchopneumonia and bronchiolitis, but a low MIC value for gentamicin (4 µg/mL). Nine of the 12 animals from which one resistant isolate was recovered had M. bovis antigen in lung lesions. The remaining three strains were recovered from animals without M. bovis antigen nor with lesions normally attributed to M. bovis. Given the multifactorial etiology of BRD, several variables may contribute to the clinical evolution of these animals, but still, the involvement of multiresistant M. bovis strains is likely contributing to the maintenance of the disease.

Conclusions
M. bovis plays a significant role in cases of clinical respiratory disease unresponsive to antimicrobial treatment that affects feedlots calves in Spain. The combined use of culture, PCR and IHC increases the sensitivity of M. bovis diagnosis in lung samples. Caseonecrotic bronchopneumonia is the morphological pattern most frequently observed in animals infected with M. bovis, and patterns indicative of other bacteria species and viruses can be concurrently detected. In some cases, M. bovis could have acted as the primary pathogen. M. bovis isolates recovered from animals with clinical respiratory disease are resistant in vitro to most of the antimicrobials specifically used for therapy in vivo.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.