The ATF-4 protein level is dependent on PHD3 but not PHD2

FL/RL

p<0.01

Fig. 3.5: ATF-4 activity is increased in hypoxia and after DMOG treatment. HeLa cells were transiently transfected with an ATF-4 dependent firefly luciferase (FL) reporter gene plasmid (pATFx2-Luc). As control the renilla luciferase (RL) plasmid pRLSV40 was co-transfected. The cells were incubated in 20 % or 1 % O2 for 4 hrs, with or without treatment with 1 mM DMOG. Subsequently the cells were lysed and luciferase activities were determined. Shown are mean values of the FL/RL ratios (± SD) of three independent experiments.

As expected from the western blot analysis, only little ATF-4 transactivation activity was measured under normoxic conditions. Exposure to hypoxia, however, led to an enhanced FL activity. The inhibition of the PHD enzymatic activity by treatment of the cells with the 2-oxoglutarate analog DMOG also resulted in an increased FL activity even under normoxic conditions. These results suggest that the ATF-4 protein is not only more abundant after inhibition of PHD activity by hypoxia or DMOG, but is present in an active form in which it is able to induce target gene expression.

3.4 The ATF-4 protein level is dependent on PHD3 but not PHD2

To gain insight into the mechanism of hypoxia/DMOG-induced increase of ATF-4 protein levels, the kinetics of the hypoxic/DMOG stabilisation were determined. Cells were incubated under normoxic (20 % O2) or hypoxic conditions (1 % or 0.2 % O2), for 4 and 24 hrs (Figure 3.6). I experienced a lot of variation in the quality of the commercially available polyclonal anti-ATF-4 antibody. Therefore, I used cell extracts obtained from HeLa cells, which were treated with 300 nM thapsigargin for 4 hrs, as a positive control for the reliable detection of ATF-4 expression. Thapsigargin is a well described inducer of enhanced ATF-4 expression (Ord and Ord, 2003).

HIF-1

ATF-4 PHD3 PHD2

-actin

20% 1% 0.2% 20% 1% 0.2%

DMOG + - + - + - + - + - +

4 hrs 24 hrs

- T O2

Fig. 3.6: ATF-4 protein levels are elevated in hypoxia and after treatment with DMOG. HeLa cells were incubated for 4 or 24 hrs in 20 %, 1 %, or 0.2 % O2, with or without the addition of 1 mM DMOG. Subsequently cells were lysed. The protein levels of HIF-1, ATF-4, PHD2, PHD3 and -actin were analyzed by immunoblots. To obtain a positive control for the detection of ATF-4, HeLa cells were treated with 300 nM thapsigargin (T) for 4 hrs as indicated.

HIF-1 and ATF-4 were barely detectable under normoxic conditions, but after 4 hrs of exposure to hypoxia (1 % O2), both proteins showed strong western blot signals. These signals were even stronger after exposure of the cells to 0.2 % O2 for 4 hrs. In contrast to PHD2, PHD3 is barely detectable in normoxia. After treating the HeLa cells with DMOG, ATF-4 and HIF-1 were also detectable under normoxic conditions and even stronger signals were observed after DMOG treatment and the exposure to hypoxia.

In contrast to 4 hrs hypoxia, ATF-4 was not detectable after 24 hrs of hypoxia. This correlates with the increased PHD3 expression after 24 hrs of hypoxia. However, inhibition of PHD activity with DMOG resulted in an enhanced ATF-4 and HIF-1 abundance under hypoxic and normoxic conditions. This leads to the suggestion that ATF-4 stability is not only dependent on the interaction with PHD3 but also on the PHD hydroxylation activity.

To further study the functional impact of PHDs in ATF-4 expression, PHD2 or PHD3 expression was repressed via siRNA (Figure 3.7). The influence of PHD1 was not further investigated because it was not detectable at the protein level in HeLa cells in considerable amounts.

3. Results 56

20% O2 1% O2

HIF-1

ATF-4

PHD2 PHD3

-actin

2 3 2+3 2 3 2+

T

PHDsiRNA 3

Fig. 3.7: ATF-4 protein levels are elevated after silencing PHD3 expression. HeLa cells were transiently transfected with siRNAs against PHD2, PHD3 or PHD2 and PHD3 together. The cells were incubated in 20 % or 1 % O2 for 24 hrs and subsequently the cells were lysed. The protein levels of ATF-4, PHD2, PHD3, HIF-1

and -actin were analyzed by immunoblots and compared to the untreated cells (-). To obtain a positive control for the detection of ATF-4, HeLa cells were treated with 300 nM thapsigargin (T) for 4 hrs as indicated.

ATF-4 expression was, comparable to the results described above, barely detectable under normoxic conditions. In contrast to HIF-1, which is stabilized in normoxia after PHD2 down-regulation, silencing of PHD2 expression had no influence on ATF-4 protein levels.

Elevated ATF-4 protein levels were detected only after silencing PHD3 expression in normoxia and hypoxia. The normoxic induction of ATF-4 is therefore restricted to PHD3, in contrast to HIF-1proteinlevels, which are upregulated after PHD2 or PHD3 silencing.

To determine if the PHD3-mediated elevated ATF-4 protein levels are a result of an increased ATF-4 mRNA expression, the ATF-4 mRNA levels were measured. To this end, HeLa cells were transiently transfected with siRNA against PHD2, PHD3, a combination of both or a non-targeting control siRNA. Subsequently the cells were incubated in normoxia (20 % O2) or hypoxia (1 % O2) for 24 hrs and the mRNA levels of PHD2, PHD3 and ATF-4 were determined (Figure 3.8).

siRNA

Fig. 3.8: ATF-4 mRNA levels are not increased by hypoxia or after the knockdown of PHD2 or PHD3 expression. HeLa cells were transiently transfected with the indicated siRNAs. Subsequently cells were incubated for 24 hrs in 20 % or 1 % O2. The mRNA levels of ATF-4 (A), PHD2 (B), PHD3 (C) and L28 (housekeeping control RNA) were measured by quantitative RT-PCR. Shown are mean values ± SD of 3 independent experiments (ratios to the ribosomal protein L28 mRNA levels).

The mRNA levels were normalized in relation to the mRNA levels of the housekeeping gene L28, which encodes a ribosomal protein. No significant differences in the ATF-4 expression were observed after exposure of the HeLa cells to hypoxia compared to normoxia (see Figure 3.8 A). As expected, PHD3 expression was elevated after exposing the cells to hypoxia (Figure 3.8 C). The hypoxic induction of PHD2 mRNA was not that strong, which is in line with the literature (Cioffi et al., 2003) (Figure 3.8 B). For PHD2 and PHD3, a significant reduction of the mRNA levels was observed after transfection of the respective siRNA.

Neither down-regulation of PHD2 nor PHD3, however, affected ATF-4 mRNA levels. These results indicate, that the elevated ATF-4 protein level under hypoxia or PHD3 knockdown is

3. Results 58 not the result of an increased mRNA expression, but might be the consequence of a change in