2.6 Data Visualisation
3.1.6 Discussion
HaCaT cells express all three isoforms of IFI16, cGAS and STING at constant levels throughout an 8 hour time course (Fig 3.1.1A). This is distinct from other cells which have been used to study the immune function of IFI16, such as human foreskin fibroblasts, vascular endothelial cells and monocytes. These cells express low basal levels of IFI16 unless stimulated with IFN or DNA (Iqbal et al., 2016; Orzalli et al., 2015; Unterholzner et al., 2010). This suggests that expression of these receptors may be regulated differently in keratinocytes. In contrast to these other cells, keratinocytes constitute the outer most layer of skin and are constantly exposed to pathogens and environmental insults. Keratinocytes could require more consistent PRR expression to provide defence in skin against the near ubiquitous presence of pathogens in the environment.
Skin acts as a mechanical barrier to pathogen invasion and while skin possesses a wealth of nonspecific defence mechanisms (e.g. secretion of anti-microbial peptides), it is not impervious to infection. As a result, skin cells must possess the capacity to communicate infection to localised populations of immune cells and the wider immune system in order to maintain homeostasis (Reviewed by Pasparakis et al., 2014).We
show that HaCaT cells produce IFN-β, ISG56, chemokines; CCL5 and CXCL10, and
pro-inflammatory cytokine IL-6 when stimulated with DNA, showing that DNA sensors are active within these cells and that keratinocytes can communicate the presence of infection to cells of the immune system. Other cytokines, chemokines and anti-microbial peptides are also likely expressed in response to DNA stimulation and viral infection within these cells but have yet to be tested.
The ability of IFI16 to recognise pathogen DNA has been observed in a wide variety of cell types (Table 1.3). However the discovery that cGAS knockout cell lines and mice are unable to respond to DNA stimulation or HSV-1 and HIV-1 infections has obscured the role of IFI16 in innate immunity (Gao et al., 2013; Li et al., 2013). As the studies in (Table 1.3) rely on IFI16 or p204 depletion by siRNA it is difficult to ascertain whether IFI16 has essential or redundant functions in initiating IFN-b
production in response to hCMV infection in primary fibroblasts. However, it should be noted that the experiment the authors rely on for this conclusion examined IFN-β
production using a CRISPR/Cas9 pool of partially depleted cells without selection, thus it is comparable to other depletion approaches such as siRNA. Therefore, the use of complete IFI16 knockout cell lines has afforded us conclusive insights into the role of IFI16 in innate immunity in keratinocytes (Fig 3.1.4).
We found that HaCaT cells lacking IFI16 are compromised in their ability to produce IFN-β, anti-viral cytokine or chemokine mRNAs in response to DNA, suggesting that IFI16 is required for complete activation of DNA sensing pathways in keratinocytes (Fig 3.1.5-7). We also show that HaCaT cells without IFI16 are compromised in their ability to produce CCL5 in response to a variety of different forms and concentrations of DNA or when faced with HSV1 infection (Fig 3.1.8-9). cGAS has been observed to be essential for DNA sensing in every cell type examined thus far (Gao et al., 2013a; Li et al., 2013b). Other experiments performed within our group have confirmed that cGAS is also essential for DNA sensing in keratinocytes (Almine et al., 2017). Therefore, these results indicate that neither IFI16 or cGAS is redundant for the immune response to DNA in keratinocytes and that there is co-operation between both receptors for functional DNA sensing in this cell type.
Since its discovery as a DNA sensor, IFI16 has been proposed to initiate IFN production via STING, albeit through an unknown mechanism (Unterholzner et al., 2010). In contrast, cGAS produces cGAMP which binds to directly to the cyclic dinucleotide cleft of STING, triggering its activation (Gao et al., 2013c; Sun et al., 2013a; Wu et al., 2013b). STING activation is measured by STING trafficking from the ER into activated punctuate structures at ERGIC. This results in the recruitment of TBK1 to STING, which undergoes auto-phosphorylation before phosphorylating STING and the transcription factor IRF3.
The STING pathway is shown to be dysfunctional in IFI16(-/-) HaCaT cells in (Fig 3.1.10 and 3.1.11). In Fig 3.1.10 we demonstrate that STING trafficking in response
severely reduced in the IFI16(-/-) HaCaT cells. While in Fig 3.1.11 we confirm the downstream STING signalling pathway is not activated with DNA stimulation in the IFI16 knockout cell line. Cells lacking IFI16 show no activation of TBK1 or IRF3 in response to DNA while these responses persist with RNA stimulation (Fig 3.1.11).
These results clearly show that IFI16 acts on the level of STING or is involved upstream in the STING signalling pathway. As is unlikely that IFI16 is able to produce cyclic di-nucleotides itself, one mechanism by which IFI16 could potentially function upstream of STING is by promoting cGAS activity thereby enabling cGAMP production.