TNFα, TGFβ, Pseudomonas and Classically Activated Macrophages Conspire to Reprogram the Airways in Obliterative Bronchiolitis


Corresponding author: David C. Neujahr


A perfect storm of inflammatory signals and bacteria conspire to drive alveolar macrophages toward a phenotype capable of inducing small airway fibrosis in lung transplant recipients. See article by Borthwick et al on page 621.

The last 25 years of clinical lung transplantation have convincingly shown that acute cellular rejection predicts subsequent development of obliterative bronchiolitis (OB). Nevertheless it has been difficult to reconcile the fact that the histopathology of rejection involves invasion of the bronchioles with lymphocytes, while the pathology of OB is characterized by a paucity of cells, and replacement of the bronchiolar epithelium with fibroblasts. A growing body of work has developed the concept that these fibroblast-like cells arise from a process termed epithelial mesenchymal transformation (EMT) [1]. In EMT, cellular reprogramming causes an epithelial cell to convert into a mesenchymal cell such as a fibroblast. EMT is thought to be an essential component of organogenesis. In lungs for example, EMT is a normal process that exists to allow for branching of the airway during fetal development. Given the stakes for aberrant EMT in lung transplant, the field is in need of an explanation of how subtle inflammatory signals such as rejection get amplified and converted into airway remodeling.

Borthwick and colleagues report one potential mechanism in this issue of the journal: reprogramming of the bronchioles driven by classically activated macrophages [2]. To understand their important findings it is helpful to conceptualize the macrophage as the dominant resident cell in the alveolar space ready to respond to diverse signals. To the first approximation alveolar macrophages can be activated either classically (via lipopolysaccharide and INFγ), or alternatively (via IL4 and IL13). Classically and alternatively activated macrophages have very different effector functions. The former serve to augment antigen presentation and direct T cell activation in part through the elaboration of TNFa while the later augment responses against pathogens such as protozoa and fungi.

The authors [2] report on a cohort of lung transplant patients who subsequently developed the clinical correlate of OB. In the months just prior to a decline in lung function they found measurably increased TNFα in the lung lavages of these patients, hinting that a signal of classical macrophage activation was at play. They then undertook a series of complex cell biology experiments. Utilizing a macrophage cell line, THP1 and primary human bronchiolar epithelial cells they performed coculture experiments with either classically or alternatively activated macrophages. EMT was assessed by looking for the disappearance of E-cadherin and the expression of vimentin indicating that cellular reprogramming had occurred. When classically activated macrophages were cocultured with the epithelial cells nothing happened. However, when the classically activated macrophages were also exposed to the bacterium pseudomonas and the additional key ingredient TGFβ, epithelial features disappeared and mesenemchyhal features emerged. This effect was also dependent on the production of TNFα by the macrophages. At first glance this in vitro assay system with the requirement for so many disparate components may seem like a contrived way in which to probe pathogenesis. There are several important aspects of this work that deserve mention. The coculture experiments used primary human respiratory small airway epithelium—the precise cells implicated in terminal airway fibrosis. Given the growing progress on rodent models of OB [3, 4], these new in vitro findings could be further tested using animal models. Regarding the complexity of the experiments, it is not surprising that EMT required pseudomonas, macrophages, TNFα and TGFβ: there are multiple studies linking a risk of OB to both pseudomonas infection and elevations in TGFβ signaling [5, 6]. Further, biomarker studies in OB pathogenesis convincingly show an association between the chemokines CXCL9 and CXCL10 and airway fibrosis [7]. These chemokines are hallmark features of classically activated macrophages. Finally, the authors have taken the next step in adding a small pilot proof-of-concept study in TNFα blockade as a means to define a new therapeutic target. They treated five patients with the anti-TNFα agent Infliximab, a drug approved for the treatment of rheumatoid arthritis. Lung function improved modestly in four of the five patients.

Collectively the findings of these investigators offer a more refined mechanism into the pathogenesis of OB suggesting that the most common leukocyte found in the lung, the macrophage, may serve as a signal amplifier of otherwise subtle inflammatory cues. There are several important limitations to this study which will need to be further investigated. First, it is unclear how closely the in vitro effects of a transformed monocytic cell line correspond to the in vivo effects of a diverse population of lung macrophages. Understanding this will require better characterization of the types of lung macrophages found in human lung transplant patients across a range of conditions. Second, assuming the in vitro experiments mirror the in vivo events of OB, we still do not know if classically activated, TGFβ aided, pseudomonas-enhanced, TNFα producing macrophages are the dominant players in OB progression, or rather part of a supporting cast of cells promoting OB via diverse pathways. This later question becomes relevant to understanding if infliximab will be a breakthrough therapy for many, or a niche treatment in limited circumstances.


The author of this manuscript has no conflict of interests to disclose as described by the American Journal of Transplantation.