Discussion
This is the first study to examine the clonogenic potential of human ASM cells. The main objective of this study was to determine whether different lineages of airway mesenchymal cells can be derived from early passage cultures of human ASM cells by single-cell cloning. The next aim was to investigate their morphological, phenotypic and functional characteristics. Critically, our studies demonstrate that the majority of single cells present within early passage human ASM cell cultures have the potential to create expanded cell populations. The clonal populations exhibited dramatic differences in their doubling times ranging from 1.1 to 18.3 days in the 58% of clones which proliferated sufficiently to be quantified. Furthermore morphological and functional heterogeneity were also seen to exist within these clonal populations. These findings confirm the existence of a heterogeneous group of cells within these cultures which may have the potential to develop diverse phenotypes (hypercontractile, contractile or synthetic). Our observation that the majority of cells isolated from early passage cultures of human ASM are capable of expansion from a single isolated cell indicates the presence of large numbers of cells with significant proliferative potential in this cell population.
Functional data from our studies demonstrate varied proliferative response by different human ASM clones, both in the absence of mitogens and also following stimulation with PDGF. In our study these clones were classified into groups based on their doubling time in the absence of the mitogens. Interestingly, the fast-growing clones did not only have higher basal proliferation rates but also exhibited a higher proliferative response following PDGF stimulation suggesting these cells have an intrinsic pro-proliferative phenotype.
Based on previously published studies we hypothesized that the majority of the cells within the fast-growing clonal population would be of a "synthetic" phenotype and the slow-growing clonal populations would comprise more of the "contractile" forms. Hence we attempted to look at the phenotype characteristics of these two different clonal populations by immunocytochemistry. We used a range of previously published markers to identify human ASM cells (the classical contractile form) and human airway fibroblasts (the classical synthetic phenotype). We have previously reported the overlap between ASM and fibroblasts and from this study chose to use α-smooth muscle actin, FSP, cathepsin K, thromboxane synthase and α8 integrin. Contrary to expectation, the fast growing clonal population had, if anything, slightly increased levels of α smooth muscle actin and the slow growing populations had increased levels of FSP. It is important to recognise that both the standard culturing conditions used and the specific conditions required for the different experiments each have the capacity to promote different phenotypes. For example culturing with serum (as was done through the clonal expansion phase of the study) would be expected to promote the synthetic phenotype whereas the serum deprivation performed prior to immunofluorescent and second messenger studies should promote an increased proportion of cells exhibiting contractile characteristics. These experimental conditions are largely unavoidable i.e. attempting to derive clonal populations from single cells is unlikely to be successful. Another consideration is how the stiff plastic substrate on which all cells were cultured and on which experiments were performed itself promotes the enrichment of certain phenotypes. Whilst it is widely recognized that extracellular matrix composition alters ASM phenotype and signaling capabilities, it is increasingly being recognised that the physical properties of the substrate alone can also affect phenotype i.e. static vs dynamic, 2D vs 3D, differential stiffness as recently reviewed. Whether the expansion of clonal populations in a more physiological environment would result in different outcomes to those reported here is unknown.
The coexistence of cell populations with diverse proliferative responses is key in understanding the relationship between ASM cell heterogeneity and airway remodeling. It is important to remember that the cells described in this manuscript represent a population which has been effectively enriched for proliferative properties with non-proliferative cells being lost following the original isolation of the ASM cells and early passaging. However, that there remains such cellular diversity in phenotype from a culture of cells which have already been through two passages is still perhaps surprising. So what could be the origin of the observed populations? And why would a clonally-derived population of cells exhibit diverse morphological differences in identical culture conditions? Possible sources of these highly proliferative populations are mesenchymal stem cells or progenitor cells such as fibrocytes either from peripheral blood or from within the tissue. To date it is unclear whether mesenchymal stem cell populations exist within airway smooth muscle cell bundles and are able to contribute to the increased smooth muscle mass observed in asthma. A little more is understood about lung fibrocytes which are fibroblast-like progenitor cells first detected outside the circulation in human bronchial mucosa in patients with chronic allergic asthma. Intriguingly, fibrocytes exposed to TGFβ have been observed to become more smooth muscle-like in phenotype. In addition, levels of fibrocytes have been correlated with the yearly decline in lung function in patients with asthma presenting with chronic airflow obstruction. Recently the same group reported an increase in the accumulation of fibrocytes in bronchial walls of patients with the same clinical phenotype. This is in agreement with the study by Saunders et al., where elevated number of fibrocytes were observed in the ASM bundles of individuals with asthma compared with controls, however there was no correlation between this and lung function. Another possible source of myofibroblasts into the airways is via epithelial-mesenchymal transitions, however emerging evidence from lineage-tracing studies raises the possibility that the latter may just be a consequence of in vitro culture and not occur in vivo.
It is however important to note that the cultures used for the experiments we describe here were derived from nonasthmatic human airways and hence would be unlikely to contain fibrocyte-derived populations if such cells only traffic to the lung following airway inflammation.
It was interesting to note that the proliferative potential of each clonal population did not appear to influence either (pro-relaxant) cyclic AMP-mediated or (pro-contractile) IP3-mediated signaling, especially when considered with the observation that cells with an increased proliferative rate exhibited increased α-smooth muscle actin and decreased FSP expression. These data suggest that it is perhaps not feasible to dichotomize the ASM populations into "good" (anti-proliferative, anti-contractile, pro-relaxant) and "bad" (being the opposite) as might have been assumed from previous studies where a proliferative, synthetic phenotype is associated with reduced contractile protein. However, as evidenced by the diverse morphology observed within clonal populations (which we did not address quantitatively), there are subpopulations of cells within the clonal populations which are quite different. So whilst we observed differences between the clonal populations, it would be revealing to understand how subpopulations within these further differ, particularly in terms of GPCR density or signaling capacity. Seminal studies in this field from Halayko and colleagues using canine tracheal smooth muscle cells both characterized subpopulations which arose after prolonged serum deprivation. Here one sixth of the cells were observed to exhibit a contractile phenotype as characterized by elongated morphology, alignment into bundles, increased expression of smooth muscle α-actin, smooth muscle myosin heavy chain and SM22 and, importantly, expression of muscarinic M3 receptors which are usually lost from smooth muscle cells through culturing. Thus whether subpopulations within the clonal populations exhibited diverse GPCR expression remains to be explored.