A Murine Model of Mycobacterium abscessus Infection Mimics Pathology of Chronic Human Lung Disease

To the Editor:

People with chronic airway diseases are predisposed to nontuberculous mycobacterial lung disease (NTM-LD), caused predominantly by Mycobacterium avium complex and Mycobacterium abscessus (Mabsc) (1, 2), which increase morbidity and mortality (2, 3). A major hurdle to understanding pathogenesis and pathophysiology of Mabsc lung disease (Mabsc-LD) in people with chronic airway diseases is a lack of preclinical models that adequately reflect the features of Mabsc infection (4). Like humans, healthy mice clear Mabsc pulmonary infections relatively rapidly. Although immunosuppression can prolong Mabsc infection in mice (5, 6), the relevance of immunosuppressed models is limited, because most humans with Mabsc-LD have intact immunity. An alternative approach that is used with other opportunistic bacteria (7) is to embed Mabsc in agar or agarose beads that lodge in small airways and cause persistent infection (8, 9). Published reports have established that the Mabsc agar bead model can be used for testing antimicrobials during chronic infection, but they have not evaluated how well the pathology approximates human disease, which is essential if the agar bead model is to be used for understanding the pathophysiology of human Mabsc-LD.

Here, we demonstrate that the histopathology of lungs from the chronic phase of the murine Mabsc agar bead model closely mimics the lesions observed in surgically resected lungs from people chronically infected with Mabsc. We embedded Mabsc subspecies abscessus (ATCC 19977) (hereinafter abbreviated as Mabsc19977) in agar beads with a diameter of 100–200 μm (10) and inoculated female C57Bl/6J mice oropharyngeally with 2 × 10⁵ colony-forming units per mouse. Bead size was confirmed microscopically. Smooth morphotype Mabsc19977 was selected because initial human colonization is thought to be caused by smooth morphotype organisms (11). We opted for a mouse strain (C57Bl/6J) that demonstrates a high level of resistance to Mabsc when administered in suspension (12), because we hypothesized that the inflammatory response to impaction of Mabsc agar beads in small airways would replicate the damaged airway environment in people with chronic airway diseases who tend to have otherwise intact immune systems. To distinguish a pathology unique to the agar bead model, we compared this model with mice inoculated with Mabsc in suspension. At 1, 2, 3, and 7 weeks post inoculation murine lungs were recovered for histology and quantification of bacterial colony-forming units. Mice that were infected with Mabsc19977 agar beads displayed early bacterial growth and sustained infection over 7 weeks, a pattern reminiscent of Mabsc growth observed in immunodeficient mice that develop chronic infection (13). In contrast, mice that were infected with suspension bacteria experienced transient infection, with early and consistent bacterial killing, and eradication in most mice by 3 weeks (see Figure E1 in the data supplement).

We previously described the pathologic features of chronic Mabsc infection in people with underlying airway diseases, where granulomas were sometimes present (14) (Figures 1Ai and 1Aii). In the present study, the same lung pathologist (C.D.C.) performed blinded histopathologic comparisons of mouse lungs inoculated with Mabsc19977 agar beads. Semiquantitative scoring identified features shared between mice and humans (Figures 1Aiii–1Avi and E2 and Table 1). Nonnecrotizing granuloma formation was observed by 1 week post inoculation with Mabsc19977 agar beads and increased over time (Figures 1Aiii and 1Aiv and Table 1). We did not observe necrotizing granulomas over the duration of the study. Of note, in human NTM-LD, necrotizing granulomas were found in only a minority of Mabsc-infected specimens (29%) versus in 55% of M. avium–infected specimens (14). Mice that were inoculated with Mabsc19977 agar beads demonstrated early lung neutrophilia (at Week 1) and the presence of histiocytic aggregates (at Weeks 1 and 2), both of which resolved as granulomas and lymphocytic aggregates (precursors to lymphoid follicles) became the predominant pathologic findings later in infection (Figures 1Aiii and 1Aiv and E3 and Table 1). Immunofluorescence confirmed the presence of macrophages in mouse lung lesions with architecture congruent with human lung granulomas (Figures 1Av and 1Avi). In contrast, mice with transient Mabsc19977 infection neither exhibited neutrophilia at any assessed time point nor developed granulomas; however, lymphocytic aggregates were prominent during transient infection and persisted even after lung sterilization (Figures 1Bi–1Biii and Table 1). Mice that were inoculated with sterile beads did not develop granulomas or histiocytic aggregates, and they exhibited minimal lymphocytic pulmonary infiltration (Figures 1Biv–1Bvi and E4 and Table 1).

Figure 1.

Lung histopathology during the chronic phase of infection in mice inoculated with Mabsc subspecies abscessus (ATCC 19977) (Mabsc-19977) agar beads resembles lung lesions in patients with M. abscessus lung disease. (A) Mabsc agar beads that lodge in small airways create granulomatous inflammation that is similar to chronic Mabsc lung disease in human subjects. (i and ii) Hematoxylin and eosin (H&E)–stained representative lung tissue from human Mabsc-LD. The patient was a 59-year-old woman with smoking-related lung disease, notable for bronchiectasis and peribronchial thickening, who underwent right middle lobectomy for treatment of chronic M. abscessus lung disease. (ii) High-power image of the area indicated by the square in (i). Arrow indicates a nonnecrotizing granuloma. (iii and vi) Histopathology of murine lungs at 7 weeks post oropharyngeal (o.p.) inoculation with Mabsc19977-embedded agar beads. (iii) H&E-stained mouse lung infected with Mabsc19977-embedded agar beads displaying granuloma formation (0.6×). (iv) High-power image of the area indicated by the square in (iii). Arrows indicate nonnecrotizing granulomas. (v) Immunofluorescence of mouse lung infected with Mabsc19977agar beads at 7 weeks displaying macrophages within the granulomatous lesions. Red (AF594) indicates CD68+ cells; blue indicates DAPI/nuclei; white indicates Lycopersicon esculentum tomato lectin-fluorescein isothiocyanate. (vi) High-power image of the area indicated by the square in (v). The images in (v) and (vi) are from the same part of the mouse lung shown in (iii) and (iv). Scale bars: A, i, 4 mm; ii, 200 μm; iii, 1.5 mm; iv and v, 200 μm; vi, 100 μm. (B) Representative lung pathology from mice inoculated with Mabsc19977 in suspension or sterile agar beads at 7 weeks post inoculation. (i–iii) Representative images of lung sections stained with H&E from mice that received an o.p. inoculation with Mabsc19977 in suspension, demonstrating that inflammatory pathology persists in mice infected with Mabsc19977 in suspension, but no granulomas form. (i) Full lung section. (ii) High-power image of section indicated by solid brown square in (i), showing congested parenchyma. (iii) High-power image of section indicated by dashed blue square in (i), showing patent airways. (iv–vi) Representative lung histology from mice that received an o.p. inoculation with 100–200 μm sterile agar beads. H&E-stained images are shown. (iv) Whole lung section. (v) High-power image of area indicated by dashed blue square in (iv). (vi) High-power image of area indicated by solid black square in (v). Scale bars: B, i, 1.5 mm; ii and iii, 200 μm; iv, 1.5 mm; v, 200 μm; vi, 100 μm.

Table 1.

Semiquantitative Scoring of Lung Histology

Histopathology	Not Infected	No. of Weeks Post Inoculation with:
		Mabsc19977 in Suspension				Mabsc19977 Agar Beads				Sterile Agar Beads
		1	2	3	7	1	2	3	7	1	2	3	7
Granulomas	0 (0, 0)	0 (0, 0)	0 (0, 0)	0 (0, 0)	0 (0, 0)	2.2* (0, 4)	2.75 (0, 8)	8.8† (5, 12)	6† (3, 10)	0 (0, 0)	0 (0, 0)	0 (0, 0)	0 (0, 0)
Lymphocytic aggregates	0 (0, 0)	9.4 (0, 18)	26.8 (14, 48)	21 (9, 37)	6.2 (4, 10)	0.2* (0, 1)	0† (0, 0)	11.4 (5, 19)	10.4 (5, 18)	0 (0, 0)	0 (0, 0)	0 (0, 0)	0.3 (0, 1)
Histiocytic aggregates	0/5	0/5	2/5	0/5	0/5	4/5†	5/5†	2/5	0/5	0/4	0/4	0/4	0/3
Neutrophils	0/5	0/5	0/5	0/5	0/5	4/5†	0/5	0/5	0/5	0/4	0/4	0/4	0/3

Definition of abbreviation: Mabsc19977 = Mabsc subspecies abscessus (ATCC 19977).

Lungs from mice inoculated with Mabsc in suspension, Mabsc agar beads, or sterile beads were recovered for scoring of histopathology. Data for granuloma and lymphocytic aggregates are presented as average number (lowest number, highest number) for each cohort; Data for histiocytic aggregates and neutrophils are presented as number of mice exhibiting specific cellular feature/total mice in cohort.

^*P < 0.05 for Mabsc19977 in suspension versus Mabsc19977-embedded agar beads.

^† P < 0.005 for Mabsc19977 in suspension versus Mabsc19977-embedded agar beads.

Granulomatous inflammation, the hallmark feature of NTM-LD, was observed both in the present study and in previously published Mabsc19977 agar bead models (8, 9); however, key differences are notable between the three studies (see Table E1). Despite using a similar inoculum and bead size as those used by Riva and colleagues (8), we observed faster bacterial clearance and dissolution of the agar beads. Riva and colleagues used male C57Bl/6N mice, whereas we used female C57Bl/6J mice, and both sex (12) and genetics (15) have been shown to influence immune phenotypes. Agar bead degradation required the presence of bacteria, as sterile beads persisted in airways at 7 weeks (Figure E4); thus, the speed of both bacterial clearance and bead degradation likely reflects the magnitude and/or type of inflammatory response. Bead size influences both the location of the infection and the host response. In a study using Pseudomonas aeruginosa, beads with a diameter smaller than 80 μm were deposited in alveoli and respiratory zone airways, producing inflammatory responses that were different from those produced by beads with a diameter of 100–200 μm that obstruct small conducting airways (16). The number of bacteria in inoculating beads likely also alters host response. Yang and colleagues used beads with a diameter of 50–150 μm (vs. 100–200 μm, as used in the present study and by Riva and colleagues), which contained 10-fold less Mabsc19977 than our inoculum, and they observed infection that lasted 3–4 months. However, in contrast to our findings and those reported by Riva and colleagues, the “granuloma-like” pathology in the model by Yang and colleagues completely resolved with eradication of infection (9).

The precise mechanism(s) by which the bead model results in prolonged infection and pathology is not completely understood. Some postulate that the beads physically protect bacteria from clearance by immune cells (7, 9), which may, in part, explain the initial growth of the bacteria in the first week of infection (Figure E1). However, we observed non–bead-associated Mabsc in lung tissue at 6 days post inoculation (see Figure E5), indicating that protection from host cells may not fully explain prolonged infection. Furthermore, coadministration of P. aeruginosa mixed with beads (rather than embedded within the bead) also generated prolonged infection (17), suggesting that the deposition of bacteria in small airways promotes chronic infection. Indeed, in lung diseases such as bronchiectasis, chronic bacterial infections persist in the small airways, not in the alveoli, which supports the idea that infection location in the lung is a key contributor to susceptibility to opportunistic pathogens in otherwise healthy humans or mice.

In conclusion, our study establishes that inoculation of wild-type mice with Mabsc embedded agar beads not only generates a prolonged infection but also produces lung histopathology that can be observed in Mabsc-LD in people with chronic airway disease. Thus, in addition to being used for testing antimicrobial therapies, this model may also have utility for studying the still poorly defined cellular and molecular processes that predispose humans to NTM-LD. A better understanding of factors that contribute to the development of Mabsc-LD could help identify which people are most at risk and provide strategies for enhancing antimycobacterial treatments.