Production of granulomas in Mycoplasma bovis infection associated with meningitis-meningoencephalitis, endocarditis, and pneumonia in cattle

Abstract

Mycoplasma bovis, the most important primary pathogen in the family Mycoplasmataceae, causes pneumonia, arthritis, otitis media, and mastitis in cattle. Histopathologic pulmonary changes associated with M. bovis infection have been characterized as suppurative-to-caseonecrotic bronchopneumonia; infection in other organs has been reported in only a few studies that examined caseonecrotic endocarditis and suppurative meningitis. Granulomatous lesions associated with M. bovis infection have been reported only rarely. We studied the granulomatous inflammation associated with M. bovis infection in several organs of 21 Japanese Black cattle. M. bovis was detected by isolation and loop-mediated isothermal amplification methods; other bacteria were detected using culture on 5% blood sheep agar and a MALDI-TOF MS Biotyper. Tissues were examined by histopathology and by immunohistochemistry (IHC) using anti–M. bovis, anti-Iba1, anti-iNOS, and anti-CD204 antibodies. All 21 cases, which included 2 cases of meningitis-meningoencephalitis, 8 cases of endocarditis, and 11 cases of bronchopneumonia, had caseonecrotic granulomatous inflammation associated with M. bovis infection. The IHC for macrophages revealed a predominance of iNOS-labeled (M1) macrophages in the inner layer of the caseonecrotic granulomas associated with meningitis-meningoencephalitis, endocarditis, and bronchopneumonia in Japanese Black cattle naturally infected with M. bovis.

Mycoplasma bovis is a primary pathogenic bacterium in the family Mycoplasmataceae, class Mollicutes. M. bovis is unable to produce a eubacterial cell wall, ²³ and uses the cell membrane and membrane-associated proteins to adhere to a host cell. ⁴ An association between M. bovis and pneumonia, mastitis, otitis media, and polyarthritis has been reported in several studies.^1,3,21,25 The histopathologic changes of pneumonia associated with M. bovis infection have been characterized as suppurative-to-caseonecrotic bronchopneumonia, which is considered a typical lesion associated with M. bovis infection.^9,18 In contrast, reports are limited regarding histopathologic changes associated with M. bovis infection in other organs, especially the heart and brain. Endocarditis associated with M. bovis infection has been reported in only one study, ¹⁴ and only a few studies have reported progressive meningitis in M. bovis associated with otitis media.^3,17 In these studies, purulent meningitis consistent with M. bovis infection was observed in the cerebellar area; however, no definitive evidence of the M. bovis antigen was found in the lesion in one study. ¹⁷

Granulomatous inflammation is a distinct type of chronic inflammation in which cells of the monocyte-macrophage system predominate and take the form of macrophages, epithelioid macrophages (activated macrophages), and multinucleate giant cells. In granulomatous inflammation, the cells are dispersed as a sheet that is distributed at random (diffuse or lepromatous) within the parenchymal and connective tissues. In contrast, in a granuloma (nodular or tuberculoid), inflammatory cells are arranged in distinct masses or nodules. Granulomatous inflammation arises in response to endogenous or extrinsic antigens, or it can be idiopathic. ² The macrophages associated with granulomatous inflammation can be categorized as classically activated macrophages (M1), which display a pro-inflammatory host defense phenotype, or as alternatively activated macrophages (M2), which conduct tissue repair and phenotypic regulatory functions.^12,26 M1 macrophages are characterized by the expression of pro-inflammatory cytokines and by production of nitric oxide (NO) by nitric oxide synthetase (NOS). The expression of inducible NOS (iNOS), an NOS isoform, has been reported as a hallmark of M1 macrophages.^6,27 In contrast, M2 macrophages are phenotypically characterized by the expression of specific markers, mainly of macrophage scavenger receptors, such as CD204.^19,22

In infection caused by Mycobacterium spp., macrophages are present in layers surrounding a central caseous granulomatous structure, which has 3 morphologic areas. The center is a cheese-like focus of cell necrosis, which is surrounded by a middle (inner) area containing macrophages, epithelioid macrophages, and multinucleate giant cells, and an outer area that consists of macrophages, lymphocytes, plasma cells, and a fibrous capsule. ² However, in M. bovis infections, granulomatous inflammation has been reported in only a few studies of bronchopneumonia. The granulomas are characterized by a caseonecrotic core surrounded by a rim of degenerate neutrophils, lymphocytes, and macrophages that is, in turn, surrounded by fibrous tissue. ¹¹

To our knowledge, M. bovis has not been reported in association with granulomatous inflammation in any organs other than the lung. The purpose of our study was therefore to use histologic and immunohistologic methods to examine the macrophages in the lesions of meningitis and/or meningoencephalitis, endocarditis, and bronchopneumonia associated with granulomatous inflammation in Japanese Black cattle following natural infection with M. bovis. Our overall goal was to obtain a better understanding of the host response to M. bovis infection.

Materials and methods

Sample collection

Our study was conducted from 2014 to 2021 and included 21 Japanese Black cattle. Thirteen poor prognosis cattle with various conditions caused by M. bovis infection, including otitis and pneumonia, were selected from 2015 to 2021 from local farms in south Kyushu; 2 otitis cases were 4 and 17 mo old; 11 pneumonia cases were 3–9 mo old. The selected cattle were euthanized, subjected to postmortem examination, and samples collected from lesions for bacterial detection and histologic study in the Laboratory of Anatomy, University of Miyazaki, Japan. The Animal Care and Use Committee of the University of Miyazaki approved the experimental procedures (2015-006). Hearts were collected for bacterial detection and histologic study from 8 cattle, 25–210 mo old, which were found to have endocarditis at slaughterhouses in south Kyushu from 2014 to 2019. ¹⁴

Bacterial detection and M. bovis identification

Hearts, lungs, and other organs with lesions were stamped onto 5% sheep blood agar (Nissui); the plates were incubated under microaerophilic conditions at 37°C for 24–48 h. Each bacterial colony on the 5% sheep blood agar plate was selected for identification using a MALDI-TOF MS (Biotyper; Bruker).

The same samples were subjected to parallel identification of M. bovis by incubating in Mycoplasma NK broth (Kanto Chemical) at 37°C for 24 h. The supernatant solution was subjected to a loop-mediated isothermal amplification method specific for M. bovis. ¹⁰ The supernatant solution was also spread on Mycoplasma NK agar (Kanto Chemical) and incubated under microaerophilic conditions at 37°C for 5–7 d to confirm the hallmark fried-egg colonies of M. bovis. ¹⁴

Histopathology and immunohistochemistry

Hearts, lungs, and other organs with lesions in the otitis cases were fixed in 10% neutral-buffered formalin, processed routinely, and 3-μm sections stained with H&E or used for immunohistochemistry (IHC). Sections were examined for the 3 layers characteristic of caseous granulomas associated with Mycobacterium spp. infection. ² IHC was performed using anti-rabbit polyclonal M. bovis antibody, ¹⁴ anti-rabbit polyclonal Iba1 antibody (Wako), anti-rabbit polyclonal iNOS antibody (Abcam), and anti-mouse monoclonal CD204 antibody (Cosmo Bio) for the samples that were found to have granulomatous inflammation with H&E staining. The sections were deparaffinized, dehydrated through an alcohol gradient, and then heated in an autoclave at 121°C for 15 min for antigen retrieval. The antigens for M. bovis antibody, Iba1 antibody, and iNOS antibody were retrieved in citrate buffer solution at pH 6.0, and the antigen for CD204 antibody was retrieved in Tris-EDTA, pH 9.0. Endogenous antigens and nonspecific antigens were blocked using 3% H₂O₂ in methanol and Blocking One solution (Nacalai Tesque), respectively. The primary antibodies were used at the following dilutions: 1:3,000 for anti-rabbit polyclonal M. bovis, 1:500 for anti-rabbit polyclonal Iba1, 1:200 for anti-rabbit polyclonal iNOS, and 1:1,000 for anti-mouse monoclonal CD204 antibody. Histofine MAX-PO (multi; Nichirei Biosciences) was used as a secondary antibody, and 3,3′-diaminobenzidine (DAB; MilliporeSigma) was used for visualization. Normal rabbit serum (Dako) and normal mouse IgG (sc-2025 serum; Santa Cruz Biotechnology) were used as negative controls for each antibody; tissue sections from a previous study ¹⁴ and from bovine brain were used as positive controls for M. bovis and for other antibodies, respectively. The Iba1-positive cells were considered macrophages, the iNOS-positive cells were considered M1 macrophages, and the CD204-positive cells were considered M2 macrophages.

The distribution of each antibody-labeled cell type was scored in the 3 areas of the granulomas (the caseonecrotic core, the inner layer, and the outer layer) viewed at the same light microscope magnification (400×, ocular field number 20 mm). The scores were as follows: ++++ = >80% positive cells; +++ = >50–80% positive cells; ++ = >20–50% positive cells; + = ≤20% positive cells; – = no positive cells.

The numbers of Iba1-labeled macrophages in 3 fields of each layer were averaged and used to represent all macrophages in each organ. The number of iNOS-labeled cells (iNOS⁺) divided by the average number of Iba1-labeled cells (Iba1⁺) (iNOS⁺/Iba1⁺) was considered to represent the M1 proportion. The number of CD204-labeled cells (CD204⁺) divided by the average number of Iba1⁺ cells (CD204⁺/Iba1⁺) was considered to represent the M2 proportion in either the inner or outer layers of the granulomas in each organ.

Data analyses

Normally distributed data were analyzed using the Kolmogorov–Smirnov test. The dependent t-test was then used to compare the proportions of M1 and M2 macrophages between the inner and outer layers of granulomatous inflammation in all cases and in each organ (except for the brain). The 95% CI of the difference (p ≤ 0.05) was considered to indicate statistical significance. All statistical analyses were performed with the computer programming language R (v.3.5.1; https://www.r-project.org/).

Results

Detection and isolation of M. bovis

Only M. bovis was detected and isolated from the brain of one animal with otitis; M. bovis was detected, along with other bacteria, in the other otitis case. Only M. bovis was isolated from the heart in the 8 endocarditis cases; M. bovis, or M. bovis with other bacteria, was isolated from the lungs in the 11 bronchopneumonia cases (Suppl. Table 1).

Gross and histologic findings

The 2 otitis media cases had drooping ears and brain lesions but no obvious lesions in other organs. One case had a white nodule (2.0 × 1.5 cm) on the right side of the pons, close to the cerebellar peduncle (Fig. 1A). The nodule was connected to the facial and vestibulocochlear nerves. The cut surface was white, with caseous necrosis and mineralization at the center. Caseonecrotic material covered the cerebellum in the other otitis case. Exudate was observed in the middle ear, housed in the temporal bone; the external ear canal had no remarkable lesions.

Figure 1.

Representative gross lesions in the brain, heart, and lung of cattle with Mycoplasma bovis infection. A. White 2.0 × 1.5 cm nodule (asterisk) on the right side of the pons. B. A caseous nodule (asterisk) on the endocardial surface of the left atrium. C. All lung lobes have diffusely distributed nodules. The cut surface of the lung has white-to-yellow caseous material in the lung parenchyma (asterisk). H = heart.

The 8 cattle with endocarditis had verrucous white-to-tan nodules of 5–12 cm diameter on the endocardial surface of the left atrium (Fig. 1B), without any obvious lesions in other organs.

The 11 cattle with pneumonia did not have grossly evident lesions in other organs. Eight of the 11 pneumonia cases had nodules in the cranioventral lung lobes; 3 cases had nodular lesions distributed in all lung lobes (Fig. 1C).

The histopathologic findings in all 3 organs revealed granulomas with 3 distinguishable areas: a caseonecrotic core with degenerate neutrophils, numerous large macrophages surrounding the caseonecrotic core in the middle or inner area, and an outer area consisting of macrophages admixed with lymphocytes, plasma cells, and fibrous tissue. In the 2 otitis cases, one case had caseonecrotic granulomatous meningoencephalitis, and the other case had caseonecrotic granulomatous meningitis. Caseonecrotic nodules with severe inflammatory cell infiltration were observed in the meninges in both cases; perivascular cuffing was observed in the brain parenchyma in only one case. Each nodule was composed of a caseonecrotic core with degenerate neutrophils, numerous epithelioid macrophages with occasional neutrophils, and fibrous tissue with macrophages and lymphocytes (Fig. 2A1). Cell necrosis with degenerate neutrophils was located in the center, with mineralization observed in one case. Epithelioid macrophages were present as a sheet and contained phagocytized neutrophils. These macrophages were admixed with neutrophils in the inner layer, whereas macrophages, lymphocytes, and plasma cells were located in the outer layer (Fig. 2A2). The facial and vestibulocochlear nerves had no remarkable changes. An accumulation of degenerate cells and neutrophils was observed in the middle ear, together with an infiltrate of lymphocytes in the lamina propria.

Figure 2.

Caseonecrotic granulomatous meningitis, endocarditis, and bronchopneumonia associated with Mycoplasma bovis infection. H&E. Asterisk in panels A1, B1, B2, C1, and C2 indicates the caseonecrotic core. Dotted frame in panels A1, B1, and C1 indicates the area at higher magnification in A2, B2, and C2. A1. Multifocal caseonecrotic granulomatous meningoencephalitis with 3 layers of granulomatous inflammation. A2. The core necrotic material on the left side consists of degenerate neutrophils, with neutrophil infiltration in the border. The macrophages in the inner layer, with abundant cytoplasm, surround the caseonecrotic core; an admixture of macrophages, lymphocytes, and plasma cells surrounds the outer layer on the right side. B1. Caseonecrotic granulomatous endocarditis on the endocardial surface of the left atrium. B2. The caseonecrotic core of the granulomas (lower) is surrounded by epithelioid macrophages. C1. Caseonecrotic granulomatous bronchopneumonia with epithelioid macrophages surrounding the caseous necrotic material. C2. Epithelioid macrophages surround the necrotic material, with some engulfed degenerate neutrophils (arrows).

In the 8 endocarditis cases, the endocardial surface of the left atrium had necrotic areas consisting of 3 layers of granulomatous inflammation (Fig. 2B1), with mineralization of the caseonecrotic core in 6 of 8 cases. The necrotic core was surrounded by an inner layer of activated macrophages with abundant cytoplasm. Macrophages, plasma cells, lymphocytes, and fibrous tissue were observed in the outer layer (Fig. 2B2). Neither myocardial necrosis nor myocarditis were observed. Fibrous tissue was deposited within the myocardium.

The caseonecrotic granulomatous bronchopneumonia in all of 11 pneumonia cases also had 3 layers of inflammation (Fig. 2C1). The large necrotic nodules were surrounded by several layers of epithelioid macrophages in the inner layer. These cells had abundant cytoplasm and resembled epithelial cells, and were easily differentiated from the macrophages in the outer area (Fig. 2C2).

Immunohistochemistry

M. bovis was very strongly labeled in the caseonecrotic core in all cases (Table 1). Two cases of meningitis-meningoencephalitis with strong labeling for M. bovis had more neutrophils in the necrotic border than in the caseous material (Fig. 3A). A few M. bovis–labeled cells were also observed in the inner layer in one case, but no M. bovis–labeled cells were observed in the outer layer in either case. All of the endocarditis cases also had very strong M. bovis labeling in the caseonecrotic material. Neutrophils, containing M. bovis–labeled antigen, were phagocytized by epithelioid macrophages (Fig. 3B). In the bronchopneumonia cases, M. bovis was also labeled in the caseonecrotic core, especially in the neutrophils and macrophages (Fig. 3C). M. bovis–labeled cells were observed in the inner layer in some cases (Table 1).

Table 1.

The immunohistochemical scores for the distribution of Mycoplasma bovis antigen, Iba1, iNOS, and CD204 labeling in the cells associated with granulomatous lesions in 21 Japanese Black cattle.

Organ	ID	M. bovis			Iba1			iNOS			CD204
		C	I	O	C	I	O	C	I	O	C	I	O
Brain	1	++++	+	–	–	++++	++	+	+++	++	–	+	++
	2	++	–	–	+	++++	+++	+	++++	++	–	++	+++
Heart	1	+++	–	–	–	+++	++	+	++++	++	–	++	+++
	2	++++	+	–	–	+++	++	+	+++	++	–	+++	+++
	3	++++	+	+	+	+++	+++	+	++++	++	–	+	++++
	4	++++	+	–	+	+++	+++	+	+++	+	+	++	++++
	5	+++	+	+	+	+++	++	+	+++	++	+	++	+++
	6	+++	–	–	–	+++	++	–	++	+++	–	+	+++
	7	+++	–	–	–	+++	++	–	+++	++	–	++	++++
	8	++++	+	–	–	+++	++	+	+++	+	–	+	++
Lung	1	++++	+	–	–	+++	++	–	++++	++	–	+	+++
	2	++++	–	–	–	+++	+++	–	+++	++	–	+	++
	3	++++	++	–	–	++++	+++	+	+++	++	–	+	++
	4	++++	–	–	–	+++	++	–	+++	+	–	++	++++
	5	++++	++	–	+	++++	+++	+	+++	++	+	++	+++
	6	+++	–	–	–	++++	+++	–	+++	++	–	++	++++
	7	+++	–	–	–	+++	++	–	++++	++	–	–	–
	8	++++	+	–	+	++++	++	+	+++	+	+	+	++
	9	++++	+	+	+	++++	++	+	+++	++	+	+	+++
	10	+++	–	–	+	++	++	+	++++	++	+	++	+++
	11	+++	+	–	–	+	++	–	++++	++	–	+	+

C = caseonecrotic core; I = inner layer; O = outer layer. The immunohistochemical scores were scored in each area: ++++ = >80% positive cells; +++ = >50–80% positive cells; ++ = >20–50% positive cells; + = ≤20% positive cells; – = no positive cells observed.

Figure 3.

Immunohistochemistry of Mycoplasma bovis antigen distribution, including the caseonecrotic core of granulomatous inflammation. A. Meningitis. B. Endocarditis. C. Bronchopneumonia.

In all cases, strong Iba1 labeling was observed in macrophages surrounding the caseonecrotic core (Table 1). A few Iba1-labeled macrophages were observed within the caseonecrotic core in one case of meningitis. Numerous Iba1-labeled macrophages were observed in the inner and outer layers in both cases of meningitis-meningoencephalitis (Fig. 4A1). Only a few iNOS-labeled (M1) macrophages were observed in the caseonecrotic core; numerous iNOS-positive (M1) cells were observed in the inner layers (Fig. 4A2). No CD204-labeled (M2) macrophages were observed in the caseonecrotic core. Fewer CD204-labeled (M2) macrophages were observed in the inner layer than in the outer layer (Fig. 4A3).

Figure 4.

Immunoreactivity representative of granulomatous inflammation in cases of meningitis, endocarditis, and bronchopneumonia associated with Mycoplasma bovis infection. Asterisk indicates caseonecrotic core in all panels. A1–A3. Meningitis case. A1. Iba1-labeled macrophages surround the caseonecrotic core. A2. iNOS-labeled M1 macrophages are located in the inner layer of the granuloma. A3. CD204-labeled M2 macrophages are scattered in the outer layer of the granuloma. B1–B3. Endocarditis case. B1. Iba1-labeled macrophages surround the caseonecrotic core. B2. iNOS-labeled M1 macrophages. B3. CD204-labeled M2 macrophages. C1–C3. Bronchopneumonia case. C1. Iba1-labeled macrophages surround the necrotic center. C2. iNOS-labeled epithelioid macrophages. C3. CD204-labeled M2 are scattered in the outer layer of the granuloma.

In 3 of 8 endocarditis cases, a few Iba1-labeled macrophages were observed in the caseonecrotic core; strong Iba1 labeling was observed in the cells of the inner and outer layers. The cytoplasmic membrane of some epithelioid macrophages had slight Iba1 staining in some cases (Fig. 4B1). Some iNOS-labeled (M1) macrophages were observed in the caseonecrotic core, especially at the border of the core, whereas numerous iNOS-positive (M1) cells were observed in the inner and outer layers (Fig. 4B2). CD204-labeled macrophages were scattered in the caseonecrotic core. Mild-to-moderate labeling was observed in the inner area, whereas numerous CD204-labeled (M2) macrophages were present in the outer layer in all cases (Fig. 4B3; Table 1).

In the bronchopneumonia cases, very few Iba1-labeled macrophages were observed in the caseonecrotic core, whereas they were numerous in the inner and outer layers (Fig. 4C1). Five of 11 bronchopneumonia cases had a few iNOS-labeled (M1) macrophages in the caseonecrotic core (Table 1), especially at the border. Strong iNOS labeling was observed in the (M1) macrophages in the inner layer of the granulomas (Fig. 4C2). A few CD204-labeled (M2) macrophages were also observed in the caseonecrotic core in 4 cases. Numerous CD204-labeled (M2) macrophages were observed in both the inner and outer layers, with greater numbers in the outer layer than in the inner layer (Fig. 4C3; Table 1).

The nominal M1 proportion (the number of iNOS⁺/Iba1⁺ labeled cells) was significantly higher in the inner layer than in the outer layer in all organs (p = 0.007; Table 2). The M1 proportion did not differ between the inner and outer layers of heart lesions; the M1 proportion in the pneumonia cases was significantly higher in the inner layer than in the outer layer (p = 0.008). The nominal M2 proportion (CD204⁺/Iba1⁺ labeled cells) was significantly lower in the inner layer than in the outer layer in all cases (p = 0.000; Table 2). The M2 proportion was significantly lower in the inner layer than in the outer layer in both the endocarditis and the pneumonia cases (p = 0.020, p = 0.006, respectively).

Table 2.

The proportion of immunoreactivity representative of M1 and M2 macrophages in caseonecrotic granulomatous inflammation associated with Mycoplasma bovis infection.

Proportion of macrophages/organ	No. of cases	Inner layer	Outer layer	p
M1 (iNOS+/Iba1+)
All organs	21	0.672 ± 0.050	0.509 ± 0.050	0.007
Brain	2	0.607 ± 0.093	0.649 ± 0.146	ND
Heart	8	0.627 ± 0.047	0.517 ± 0.097	0.224
Lung	11	0.717 ± 0.088	0.478 ± 0.055	0.008
M2 (CD204+/Iba1+)
All organs	21	0.313 ± 0.051	0.565 ± 0.056	0.000
Brain	2	0.188 ± 0.137	0.415 ± 0.058	ND
Heart	8	0.451 ± 0.100	0.720 ± 0.068	0.020
Lung	11	0.235 ± 0.049	0.471 ± 0.082	0.006

Results are mean ± SE; p ≤ 0.05 is considered significant. ND = not done.

Discussion

The histopathologic changes that we observed in the cases of meningitis-meningoencephalitis, endocarditis, and bronchopneumonia in the 21 cattle in our study included 3 areas: a caseonecrotic core surrounded by an inner layer of epithelioid macrophages and an outer layer of macrophages, lymphocytes, and fibrous tissue. These histologic characteristics were similar to histologic findings of granulomas in Mycobacterium spp. infection. ² The epithelioid macrophages surrounding the caseonecrotic core had Iba1, iNOS, and CD204 labeling consistent with granuloma formation. In addition, M. bovis was detected consistently in the lesions by M. bovis culture and IHC, confirming granuloma formation caused by M. bovis. The formation of granulomas has not been reported previously as a characteristic of the lesions associated with M. bovis infection, to our knowledge.

The gross and histologic appearance of the suppurative-to-caseonecrotic lesion associated with M. bovis infection has been reported in bronchopneumonia, and this is considered a lesion typically associated with M. bovis infection. ⁵ However, granulomatous changes associated with M. bovis have been reported in only 2 studies^11,15; 1 on North American bison (Bison bison), and the other on calves. The first report ¹¹ described central areas of necrosis that were occasionally mineralized and that were surrounded by a rim of degenerate neutrophils, lymphocytes, plasma cells, and macrophages, and then outer layers of fibrous connective tissue. The other report ¹⁵ described well-delineated, round-to-oval foci of caseous necrosis surrounded by granulation tissue, but without epithelioid granulomatous formation. The granulomas in our meningitis-meningoencephalitis, endocarditis, and bronchopneumonia cases associated with M. bovis infection all had the 3 layers of the granulomatous reaction demonstrated histologically in granulomas associated with Mycobacterium spp. infection. ² Our IHC results revealed M. bovis–labeled cells and cell debris in a caseonecrotic core that was surrounded by Iba1-labeled macrophages, thereby confirming these lesions as granulomas.

The iNOS labeling suggestive of M1 macrophages was stronger in the inner than in the outer layer; the CD204 labeling suggestive of M2 macrophages was stronger in the outer than in the inner layer. The M1 proportion was also significantly higher in the inner layer than in the outer layer of the granulomatous reaction in all organs. However, a comparison of each organ revealed no significant difference in the M1 proportion between the inner and outer layers in the heart; the granulomatous reaction was significantly stronger in the inner layer than in the outer layer in the lung. The M2 proportion was significantly higher in the outer than the inner layer of the granulomatous lesions in all organs. This may suggest that M1 macrophages predominate in the granulomatous reaction associated with M. bovis infection. M. bovis infection in Japanese Black cattle progresses rapidly to the chronic phase, and granulomas develop as a result of accumulation of macrophages with engulfed necrotic M. bovis–containing neutrophils. We speculate that the host response might differentiate macrophages into the predominant epithelioid M1 phenotype to limit the infection.

We found that the proportion of M1 macrophages was significantly higher in the inner layer of the granulomatous lesions, and the proportion of M2 macrophages was higher in the outer layer of the granulomatous lesions. This is consistent with a study that reported noticeable iNOS immunoreactivity at the cellular (middle) zone surrounding the caseonecrotic center in the early stage of infection of an experimental bovine tuberculosis study. ²⁰ The proportion of M1/M2 polarization in M. bovis granulomas might be similar to the granuloma formation in bovine tuberculosis.^8,19 Other researchers have reported prolonged survival of M. bovis in neutrophils. ¹⁶ Thus, we hypothesize that the development of M1-polarized granulomas in our study might reflect a response by neutrophils and macrophages to M. bovis invasion of host tissues, with the inflammatory cells degenerating and becoming necrotic, thereby increasing the amount of caseous material containing M. bovis. The M1 macrophages would proliferate and produce granulomas in response to persistent M. bovis infection, whereas M2 macrophages might develop in the later healing process, as demonstrated in our study by the admixing of M1 and M2 polarizing macrophages in the outer layer of the granulomas.

Granulomatous inflammation associated with Mycobacterium spp. infection has been described as occurring in 4 stages, from stage I (small, unencapsulated granulomas without necrosis) through stage IV (large, thickly encapsulated, necrotic granulomas).^20,24 A study reported a positive association between iNOS immunoreactivity and the early stages of granulomas (stages I–II) and a decrease in the late stages (stages III–IV). ¹⁹ Another study reported that M1 polarizing macrophages predominated in mycobacterial granulomas, whereas M2 polarizing macrophages predominated in the diffuse granulomatous form. ⁷ However, all of the caseonecrotic granulomatous cases in our study were multifocal forms that expressed the iNOS antibody. Our results are consistent with an in vitro study that reported increases in apoptotic neutrophils after incubation with M. bovis and activation of the type 1 T-cell response associated with the M1 reaction. ¹³

Our detection of other bacteria in some cases raised the possibility that the observed granulomas were not caused by M. bovis infection. However, the histologic findings when other bacteria were observed were similar to those when only M. bovis infection was present, suggesting that all of these granulomas are associated with M. bovis infection. Nevertheless, other bacterial or viral pathogens might provide a synergistic microenvironment for M. bovis infection.