The Human Respiratory Microbiome: Implications and Impact

Alicia B. Mitchell, BMedSci (Hons); Allan R. Glanville, MBBS, MD, FRACP


Semin Respir Crit Care Med. 2018;39(2):199-212. 

In This Article

Virome Studies

There are limited data in the area of the virome due to the inherent limitations in isolating and defining this part of the microbiome. In the lungs, the bacterial biomass is reduced compared with other compartments such as the intestine and the oral cavity. The amount of viral genomic material is even lower. Isolating viral nucleic acids without the use of a conserved sequence such as 16S in bacteria or ITS region for fungi and effectively depleting both human and bacterial genomic material present many challenges and may account for the lag in our knowledge of the viral component of the microbiome.

The first studies in the area of the lung virome were in patients with CF, paralleling the early studies of the bacterial component of the microbiome. Bacteriophage communities in a small cohort of individuals with CF were compared with healthy controls, to demonstrate that there appears to be a core set of bacteriophages found in the healthy lung and an additional core set of phages found in individuals with CF which correspond to concurrent pathogenic bacterial species. In CF, the lung bacteriophage community appears to be driven by airway pathology and the persistence of certain bacteria.[113] The same group conducted a small case series evaluating spatial heterogeneity within CF lungs. A large range of DNA viruses and bacteriophages was observed within the lungs of individuals with end-stage lung disease, with clear heterogeneity in the presence of viruses within different lobes of the lungs.[38]

Since then, a handful of studies has largely characterized the virome in individuals with acute viral infections. In this case, there is an overwhelming abundance of the causative organism which is often concurrently detected using PCR methods. However, a range of other viral species including torque teno viruses, herpes viruses, and community-acquired respiratory viruses such as picornavirus and adenovirus has been characterized in these samples.[114,115] Furthermore, a study characterizing the respiratory virome in febrile and afebrile children has provided evidence to show that a resident lung virome is likely established in healthy children within the first 2 years of life which may also impact the development of the immune system.[116]

The human respiratory virome has been investigated in lung transplantation. Young et al have characterized the DNA virus profile in the oropharyngeal washes and BAL samples post-transplant.[117] A range of DNA viruses was detected using sequencing, including herpes viruses, human papillomavirus, and bacteriophages. Anelloviruses, including torque teno virus, dominated these samples accounting for over 68% of reads. These anelloviruses were further quantified using real-time qPCR. Transplant patients were shown to have significantly greater viral loads in both the oropharyngeal wash and BAL samples compared with healthy controls.[117] A further study explored the role of torque teno virus in lung transplant recipients in the perioperative period. Torque teno virus was shown to increase in the immediate postoperative period; however, the magnitude of viral load increase was shown to be associated with the development of primary graft dysfunction.[118] Therefore, monitoring changes in the virome may, in the future, assist in determining risk profiles for graft injury and adequate levels of immunosuppression to guide therapeutic actions post-transplantation (Figure 3).

Figure 3.

Dynamics of the human respiratory virome after lung transplantation. This concept diagram demonstrates a potential temporal relationship between resident viral species within the lung and transient species. Specific events lead to blooms of resident species such as the ex-vivo stage of lung procurement and heightened immune suppression. It is likely but not proven that acute community acquired respiratory virus (CARV) infection may suppress resident species temporarily, while the impact of the development of chronic lung allograft dysfunction (CLAD) may depend on phenotype and also on therapies employed.

There is some evidence to suggest that viral infections are associated with the development of BOS. Garantziotis et al showed a relationship between infection with influenza and subsequent decline in lung function associated with early graft dysfunction and/or BOS.[119]

A further study has indicated that the 1-year incidence of BOS is significantly increased (25%) in patients who have had a respiratory virus detected by PCR compared with virus-negative patients (9%). However, detection of a respiratory virus in patients who already had diagnosed BOS had no impact on progression.[120] This was further evidenced by a retrospective 5-year cohort study on lung transplant patients, where detection of a lower respiratory tract viral infection was significantly correlated with the development of BOS and BOS-related mortality.[121] A further study that compared patients with symptomatic viral infection with asymptomatic controls demonstrated PCR-detected virus in the symptomatic group which was associated with both acute rejection events and progression to BOS.[122]

The herpes group viruses have been studied extensively as a quasi-resident species within the lungs of transplant recipients. Cytomegalovirus (CMV) and EBV as well as human herpesvirus type 6 (HHV-6), in particular, have received attention but data regarding their relationship to allograft dysfunction are somewhat confounded by the efficacy of prophylactic use of specific antiviral therapies.[123] Some relationships are clear. An EBV naïve recipient of an EBV-positive lung allograft has the highest risk of developing post-transplant lymphoproliferative disease (PTLD) (50% in some series) which carries a high mortality rate. This has been somewhat ameliorated of recent times by the selective use of anti-CD20 monoclonal antibody therapy to target EBV-driven B cell clonal expansion as an adjunct to reduction in immune suppression which, of itself, was commonly associated with the development of acute cellular rejection and graft dysfunction. This relationship between the risk of PTLD and EBV viral load has led some units to pursue a policy of peripheral load monitoring in blood using a quantitative EBV PCR with the intention of reducing immune suppression pre-emptively if EBV load increases to circumvent the risk of PTLD.[124] Conversely, an absent EBV load raises the question of acute rejection risk. The oncogenic potential of EBV, Kaposi's sarcoma herpesvirus (KSHV), and human T-lymphotropic virus type 1 (HTLV-1) has been confirmed by high-throughput RNA sequencing data from 50 common lymphoma cell culture models from the Cancer Cell Line Encyclopedia project.[125] The role of HHV-6 in the pathogenesis of organizing pneumonia, a not uncommon finding on transbronchial biopsy deserves further study.[123] Repeated EBV DNA detection in blood, possibly reflecting chronic EBV replication, has been associated with the development of BOS.[126]

CMV can cause fulminant pneumonia and death, which since the development of effective treatment modalities[127] is now rare, except for drug resistant strains, an ever-present risk of the use of widespread prophylaxis.[128] A largely intracellular virus, CMV, can be detected in BAL fluid and analysis of paired samples suggests detection of CMV DNA in BAL fluid reflects virus replication in the lung rather than oropharyngeal contamination.[129] Whether BAL load monitoring assists in the determination of a therapeutic decision is debatable versus a qualitative assessment. However, peripheral load monitoring in blood using a quantitative PCR is the gold standard for initiation of therapy versus prophylaxis. CMV naïve recipients of a positive donor graft hold the highest risk of CMV pneumonia; therefore, there exists some evidence for ongoing prophylaxis.[130]

To date, there are very limited studies characterizing the human virome, and none in the areas of COPD or IPF. Further research may help to elucidate the changes in the resident virome between different disease states, and how these impact progression. Early studies in lung transplantation show distinct differences in the respiratory virome composition and diversity between lung transplant recipients and healthy controls. However, the detection of causative viral organisms in symptomatic individuals has been associated with negative transplant outcomes. The presence of certain viral species and their viral load may help to guide treatment strategies and modifications to the respiratory virome may improve outcomes for patients in the future (Figure 3). In the majority of lung virome studies published, sequencing has focused on the DNA virus community due to the inherent limitations of working with RNA viruses which are much more prone to degradation. Little is known about the role of RNA viruses in both healthy individuals and in disease states, including after transplantation. Further studies, particularly those with a prospective longitudinal design, are needed to describe the large collection of viruses detected by NGS which remain unnamed and uncharacterized and to determine their role within the greater microbiome in both health and disease.