Human alveolar macrophages predominately express combined classical M1 and M2 surface markers in steady state

1The healthy human alveoli are dominated by alveolar macrophages (AM). These cells play a critical role in regulating immune responses and maintaining homeostasis in the lung. Their plasticity to adapt to changes in their microenvironment is fundamental in retaining the lung health. Macrophages have been categorised as “polarised” towards either M1 (proinflammatory, mediating resistance to pathogens), or M2 mode (anti-inflammatory, promoting tissue remodelling). This categorisation is derived from murine models, and from in vitro polarisation of human macrophages using a combination of cytokines. However, whether AM polarisation in steady state fits the traditional M1/M2 dichotomy is still not well understood. We therefore investigated human AM polarisation in steady state, and ascertained the stability of the phenotype in two geographical locations and following in vivo exposure to bacterial and viral stimuli.

2Briefly, in the UK, healthy, non-smoking participants aged 18–50 years were recruited from an ongoing study of the Experimental human pneumococcal challenge (EHPC) model. A subset of 25 participants, 16 individuals who had become colonized (carriage +) with Streptococcus pneumoniae, and 9 who had not (carriage –) underwent research bronchoscopy at 4 to 7 weeks after the bacterial challenge. Alveolar macrophages were obtained from bronchoalveolar lavage (BAL), as described previously. In Queen Elizabeth Central Hospital (QECH), Malawi, BAL samples were obtained from selected asymptomatic adults (≥18 yrs), comprising ten HIV-1-uninfected healthy controls and ten age-matched ART-naïve HIV-1-infected individuals. We excluded participants with clinical evidence of active disease or recent history of severe respiratory illness.

3Flow cytometry-based immunophenotyping was used to characterise AM phenotype obtained from BAL fluid. The expressional levels of key surface markers were measured immediately after sample collection.

4Data from the two human cohorts from distinct geographical locations, UK (Additional file 1 Table S1) and Malawi (Additional file 1 Table S2), were collected independently and their comparison revealed a similar pattern of surface marker expression by AM. In both cohorts, the majority of AM were characterised as CD206hiCD86hi. M1-like phenotype (CD206loCD86hi) and M2-like phenotype (CD206hiCD86lo) subsets represented a small proportion (less than 1%) of the total AM population (Fig. 1a and c). In addition, the CD206hiCD86hi subset expressed greater levels of the M2 marker, CD163, compared to the M1 and M2 subpopulations (Fig. 1b and d). UK samples were further analysed for additional M1 markers (CD80, CD64) and an activation marker (HLADR). In these, the CD206hiCD86hi subset expressed the highest levels of CD80, CD64 and HLADR expression in comparison to the other AM subsets (Fig. 1e-g).

5We have previously shown that live intranasal pneumococcal challenge alters the alveolar environment by increasing the levels of pneumococcal-specific memory CD4+ Th17 cells in the human lung, and that HIV infection disrupts the alveolar cytokine microenvironment. We therefore investigated whether in vivo exposure to intranasal pneumococcal challenge or HIV infection polarise the AM phenotype. We compared experimentally pneumococcal-colonized individuals with non-colonized individuals, as well as, asymptomatic ART-naïve HIV-infected adults with age-matched HIV-uninfected healthy individuals. We found no difference in M1 or M2-like phenotype between individuals who became colonized following inoculation with S. pneumoniae compared with those who cleared the bacteria (Fig. 1a). Specifically, the expression of CD163 (Fig. 1b), CD80, CD64 and HLADR measured in the four AM subsets remain at the same levels between these groups (Fig. 1e-g).

6Interestingly, in the Malawian population, AM subsets from HIV-infected individuals followed the same pattern of subdivision described in healthy volunteers. No difference was found in the proportion of the investigated subsets between the ART-naïve HIV-infected adults and HIV-uninfected healthy individuals (Fig. 1c). However, the expression of CD163 was significantly reduced in the CD206loCD86hi subset in HIV-infected individuals (p = 0.0147) (Fig. 1d).

7In our study, we compared the polarisation of AM isolated from human volunteers from distinct geographical locations, UK and Malawi. We report that in healthy human pulmonary mucosa, AM adopt a hybrid phenotype, which shares features of both M1 (CD80, CD86, CD64) and M2 (CD206 and CD163) polarized macrophages. The same phenotype has been previously reported on the decidual macrophages (resident macrophages of the uterine lining), which exhibit characteristics of both pro-inflammatory and tolerogenic macrophages.

8We also found that nasal pneumococcal colonization, even though it alters CD4+ T cell responses in the alveoli, does not alter the expression of key M1 and M2 polarization markers on the surface of AM. This leads to the speculation that a nasal pneumococcal exposure episode is either incompetent to modulate AM polarization or its effect is transient and quickly irreversible due to macrophages high plasticity. By contrast, chronic HIV infection did affect the expression of CD163 on the M1-like AM subset (CD206loCD86hi), with a lower expression of CD163 observed in HIV-infected individuals than HIV-uninfected controls. This finding is intriguing and may be due to active HIV infection of this alveolar macrophage subset which has been shown to strongly repress CD163 expression on infected-macrophages. Furthermore, CD163 is shed during activation as soluble CD163 (sCD163) and this might in part explain the reduction of CD163 on the surface M1-like AM in HIV-infected individuals. However, even with in vivo exposure to pneumococci or HIV, the CD206hiCD86hi AM subset remained the major population, and this ascertains the stability of this phenotype.

9Our data provides strong evidence suggesting that not all macrophages fall into M1 and M2 subsets. The major AM population in steady state expresses a duo M1/M2 phenotype. This phenotype is present in individuals from two distinct geographical location and is stable even after exposure to stimuli known to alter the alveolar environment. This is consistent with recent call for a rethink of the M1/M2 macrophage paradigm. It is possible that the lack of stark polarisation is helpful in maintaining a healthy balance between immune tolerance and protective immunity in the alveolar space. However, whether other tissue macrophages, beyond decidual and alveolar macrophages, exhibit similar phenotypes in vivo warrants further investigation.

10The clear majority of alveolar macrophages combine M1 and M2 features in steady state, a phenotype that may allow brisk and adaptive responsiveness to multiple elements in the local milieu.