F11R expression upon hypoxia is regulated by RNA editing

Abstract

F11R is a cell adhesion molecule found on the surface of human platelets. It plays a role in platelet aggregation, cell migration and cell proliferation. F11R is subjected to RNA editing, a post-transcriptional modification which affects RNA structure, stability, localization, translation and splicing. RNA editing in the 3'UTR of F11R and RNA levels are increased upon hypoxia. We therefore set to examine if RNA editing plays a role in the increase of F11R RNA seen upon hypoxic conditions. We show that ADAR1, but not ADAR2, takes part in the editing of F11R however editing alone is not sufficient for obtaining an elevation in RNA levels. In addition we show that hyper-edited mature mRNAs are retained in the nucleus and are associated with p54(nrb). We therefore conclude that hypoxia-induced edited RNAs of F11R are preferentially stabilized and accumulate in the nucleus preventing their export to the cytoplasm for translation. This mechanism may be used by additional proteins in the cell as part of the cell's effort to reduce metabolism upon hypoxic stress.

Figure 1. Protein expression of F11R upon hypoxia.

Western blot analysis was performed on protein extracts from LB cells 24h following DFO treatment or without treatment. Numbers under the F11R panel represent the relative quantification of the amount of protein. Actin was used as a loading control.

Figure 2. Knock-down of ADAR1.

LB cells were transfected with si-ADAR1 or with a control si molecule (siControl). Cells were either treated with DFO 16h post-transfection or untreated. RNA and proteins of all cells were extracted 40h post-transfection (24h post-DFO treatment). Error bars indicate standard deviation (±SD,**p≤ 0.01). (A) Relative quantification of mRNA. mRNA was quantified relatively to the siControl sample which was set as 1. Silenced cells showed low expression of both ADAR1 subunits both in normoxia and upon hypoxia. No effect of the silencing was seen on F11R mRNA which was increased upon hypoxia similarily to the control-transfected cells. (B) Percentage of average RNA editing in F11R upon ADAR1 silencing with or without DFO treatment. ADAR1-silenced cells showed a reduction in editing which was elevated upon DFO treatment however not to the levels seen in the control-transfected cells. p values refer to the difference from the siControl sample unless specified otherwise. (C) Western blot analysis. Numbers in between rows show the relative quantification of the amount of protein set at 1 in the siControl sample. Expression of both ADAR1 subunits was decreased upon silencing. F11R expression was not affected by ADAR1 silencing but was decreased upon DFO treatment. Actin was used as a loading control.

Figure 3. Overexpression of ADAR1-p150 and IFN treatment-α.

(A) Relative quantification of mRNA extracted from control cells (transfected with a control plasmid) and cells transfected with a plasmid overexpressing ADAR-p150. mRNA was quantified relatively to the control-transfected sample which was set as 1. Error bars indicate standard deviation (±SD,*p≤ 0.05). ADAR1-p150 levels were increased upon transfection however no increase in F11R expression was obtained upon p150 overexpression. (B) Relative quantification of mRNA extracted from control cells and cells treated with IFN-α. mRNA was quantified relatively to the control sample which was set as 1. ADAR1-p150 was increased upon treatment however no increase in F11R expression was obtained upon treatment. (C) Percentage of average RNA editing in F11R upon transfection with a p150-overexpressing plasmid. and upon IFN-α treatment. No significant changes were obtained in RNA editing in F11R.

Figure 4. Knock-down of ADAR2.

LB cells were transfected with si-ADAR2 or with a control si molecule (siControl). Cells were either treated with DFO 16h post-transfection or untreated. RNA and proteins of all cells were extracted 40h post-transfection (24h post-DFO treatment). (A) Relative quantification of mRNA. mRNA was quantified relatively to the siControl sample which was set as 1. Error bars indicate standard deviation (±SD, *p≤ 0.05, **p≤ 0.01) Silenced cells showed low expression of ADAR2. No effect of the silencing was seen on F11R mRNA which was induced upon hypoxia similarily to the control-transfected cells. (B) Percentage of average RNA editing in F11R upon ADAR2 silencing with or without DFO treatment. ADAR2-silenced cells did not show a change in editing. Editing in the control-transfected cellswas induced upon DFO treatment.

Figure 5. Percentage of average RNA editing in pre- and mature- F11R mRNA .

Pre- or mature- F11R mRNA from control and DFO-treated cells was amplified using specific primers. Significant average RNA editing was seen only in mature F11R mRNA which was further increased upon DFO treatment. Error bars indicate standard deviation (±SD, **p≤ 0.01) .

Figure 6. Treatment with α–amanitin.

LB cells were treated for 24h with α–amanitin with or without DFO treatment. (A) Relative quantification of mRNA. mRNA was quantified relatively to the non-treated sample which was set as 1. Error bars indicate standard deviation (±SD, *p≤ 0.05, **p≤ 0.01). ADARs and F11R expression was reduced upon α–amanitin treatment. Treatment with both α–amanitin and DFO showed an increase only in F11R mRNA when compared to the amount obtained with α–amanitin treatment only. (B) Relative quantification of mRNA following cell fractionation. mRNA was quantified relatively to the cytoplasmic control sample which was set as 1. Error bars indicate standard deviation (±SD, *p≤ 0.05, **p≤ 0.01). F11R mRNA was higher in the nuclear fractions. α–amanitin treatment reduced the amount of F11R transcripts and the addition of DFO induced them. (C) Percentage of average RNA editing in F11R. Cells treated with α–amanitin showed high levels of RNA editing in F11R. RNA extracted from the cytoplasm showed higher levels of editing when compared to those obtained in the RNA extracted from the nucleus. Error bars indicate standard deviation (±SD, **p≤ 0.01) (D) Percentage of RNA editing in F11R at additional sites. Treatment with α–amanitin revealed significant editing at seven additional sites. Upper panel: editing at these sites in the nuclear extracts following treatment. Lower panel: editing at these sites in the cytoplasmic extracts. Error bars indicate standard deviation (±SD). All the results obtained for the α–amanitin and the α-amanitin+DFO treatments have a p≤ 0.01. (E) Western blot analysis. Numbers in between rows show the relative quantification of the amount of protein set at 1 in the non-treated sample. No F11R protein was seen in the α–amanitin-treated cells. Low F11R expression was seen upon α-amanitin + DFO treatment. Actin was used as a loading control.

Figure 7. RNA immunoprecipitation with an anti-p54^nrb antibody.

(A) PCR amplification with F11R-specific primers on RNA extracted from a RIP experiment conducted with an anti-p54^nrb antibody on fractionated control cells and fractionated cells treated with DFO. PCR amplification was obtained in the RNA extracted from p54^nrb-precipitated RNA from nuclear extracts of DFO-treated cells. As a positive control PCR was performed on total RNA (B) Percentage of RNA average editing of F11R RNA that was bound to p54^nrb. High levels of editing were obtained in the RNA that was extracted from the p54^nrb-bound molecules.

Figure 8. A proposed model for the regulation of F11R expression following hypoxic stress.

Upon normoxia, F11R mRNA is transcribed and a fraction of the molecules are exported to the cytoplasm for translation. Upon hypoxia, editing of F11R is increased. The majority of the hyperedited molecules binds p54^nrb and is retained in the nucleus in large amounts. Translation of the small number of edited transcripts which escape nuclear retention, is attenuated by a yet to be determined mechanism. We propose that when oxygen levels return to normal, a large amount of F11R hyperedited molecules disengage from p54^nrb and thus are released to the cytoplasm for translation enabling protein levels to return to normal very quickly.