In a recent study published in EMBO reports, researchers observed that transposable elements (TEs) activation is impaired during infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).
Coronaviruses generally cause mild respiratory tract illnesses; however, SARS-CoV-2, the causal agent of coronavirus disease 2019 (COVID-19), could lead to a cytokine storm, resulting in acute respiratory distress syndrome (ARDS). SARS-CoV-2, unlike other viruses, is effective at evading early innate responses such as interferons (IFNs). TEs are abundant in mammalian genomes and contain numerous regulatory elements. Lately, TEs stimulated antiviral responses via cis or trans mechanisms.
The study and findings
In the present study, researchers examined the changes in the expression of TEs during SARS-CoV-2 infection in cellular models. First, they analyzed published datasets of SARS-CoV-2- or influenza A virus (IAV)-infected normal human bronchial epithelial (NHBE) cells. IAV infection, but not SARS-CoV-2, caused an increased expression of TE subfamilies across all TE families.
Similarly, IFN-β treatment also induced TE expression, as many TEs contain IFN-responsive sequences that are upregulated after the induction of IFNs. This led to a hypothesis that SARS-CoV-2 might elude the robust TE response. To this end, the data analysis was expanded to include more cell types and other viruses. Cells included cancer cell lines (Calu3 and A549) and induced pluripotent stem cell-derived alveolar epithelial cells type 2 (iAT2s), among others.
They noted that in SARS-CoV-2 infection, the strength of TE response was proportional to viral load. Further, the prediction accuracy of TE response improved when basal TE levels were incorporated into the model. A549 cells mounted a more robust TE response than NHBE cells, despite low viral loads in both.
This was because primary cells (NHBE) have a higher basal TE response than transformed cell lines (A549 or Calu3). Similarly, at higher viral loads, A549 cells expressing angiotensin-converting enzyme 2 (ACE2) had stronger TE responses than (primary) iAT2 cells. Therefore, the viral load was not the only factor for the upregulation of TEs, but the identity of the cells and the virus influenced TE expression.
Next, the researchers assessed the association between IFNs and TE expression during SARS-CoV-2 infection, given that IFN expression had been previously linked with TE expression. They found a significantly higher number of upregulated IFN genes in cancer cell lines than in primary cells one day post-infection (dpi) with SARS-CoV-2.
SARS-CoV-2-infected iAT2 cells had a milder TE response than NHBE cells at 1 dpi; however, the IFN response was weaker in these cells than in NHBE cells. Nevertheless, the TE response increased slightly by 4 dpi with a significant IFN response, suggesting that the TE response preceded the delayed IFN response. Moreover, they found that genes adjacent to upregulated TEs were generally more likely to be upregulated.
Notably, this effect was more substantial for IFN genes adjoining TEs, suggesting that adjacent TEs induce IFNs during SARS-CoV-2 infection. This meant that TEs induced during infection could act as cis-regulatory enhancers. Further, they investigated whether IFN induction during SARS-CoV-2 infection resulted from TEs acting in cis or trans. Approximately 60% of upregulated IFN-related genes were near upregulated TEs during IAV infection.
In contrast, 10% to 20% of IFN-related genes were nearby upregulated TEs during SARS-CoV-2 infection. Although TEs induced neighboring IFN genes during SARS-CoV-2 infection, IFN response did not occur due to TEs acting in cis. Overall, this meant that TE induction/upregulation in response to IAV preceded and contributed to the expression of IFN-associated genes. However, SARS-CoV-2 failed to upregulate TEs co-opted for immune activation.
Furthermore, the authors investigated the epigenetic signature of SARS-CoV-2-induced TEs before infection. Upregulated TEs in A549 cells were enriched for active histone marks, with some marked by H3K36me3 or a combination of H3K4me3, H3K27ac, H3K9ac, and H3K79me2. Short-interspersed nuclear elements (SINEs) were particularly enriched for active marks.
In contrast, long-interspersed nuclear elements (LINEs) were highly enriched for the repressive H3K9me3 and active H3K36me3 marks. Intriguingly, all SARS-CoV-2-induced TEs were depleted of the H3K27me3 mark, which was not the case for IAV-induced TEs. In contrast, IAV-induced TEs adjacent to IFN genes were highly enriched for H3K27me3. This suggested that failure to activate IFN by TEs during SARS-CoV-2 response was the result of not evacuating H3K27me3 from TEs near IFN genes.
The authors proposed that the high basal expression of TEs desensitizes the TE response induced by SARS-CoV-2. Accordingly, SARS-CoV-2-induced TE activation was associated with basal TE levels (with lower basal levels having stronger cellular responses). TE activation response in older cells (with high basal TE, thus, TE-desensitized) would be milder, leading to a less effective IFN response.
In summary, the study found that SARS-CoV-2, unlike IAV, does not upregulate TEs in normal lung epithelial cells. This meant that the weak TE activation resulted in a less effective IFN response that generally responds to TEs by a positive feedback loop.