Expression of IL-8 is cell cycle regulated in eTat cells: induction of IL-8 transcription at S phase.
Using a cell line which constitutively expresses an active Tat protein, researchers recently reported that Tat activates transcription from the HIV long terminal repeat in a cell cycle-dependent manner (8
). This result prompted us to analyze the potential cell cycle regulation of cellular genes induced by Tat. We arrested the cell cycle of cultured eTat and control pCEP cells using Hu, which arrests cells in early S phase by inhibiting ribonucleotide diphosphate reductase (25
). Figure shows a typical fluorescence-activated cell-sorter (FACS) analysis conducted after an 18-h Hu block and subsequent release. After release, eTat and pCEP cells moved through S phase (represented by the 3 h postrelease sample [Fig. ]) and into the G2
/M phase by 6 h postrelease with a peak G2
/M population at 9 h postrelease. Cells then entered the G1
phase (12 h postrelease). The cell cycle profiles were similar in both cell lines.
FIG. 1 Cell cycle analysis of eTat and control pCEP cell lines following Hu block. Log-phase growing cells were blocked with Hu (2 mM) for 18 h and released by removing the inhibitor and adding fresh media. Samples were collected every 3 h, and the cells were (more ...)
To analyze cell cycle-dependent gene expression, pCEP and eTat cells were reversibly blocked with Hu for 18 h and then released. Cells were collected every 3 h for 9 h, and RNA was extracted. Fifteen micrograms of total RNA from each sample was then used in an RNase protection assay using the hCK-5 cytokine probe set (PharMingen). This allowed us to study the expression of a total of seven different cytokine and chemokine genes, including the genes for lymphotoxin, RANTES, IP-10, MIP-1β, MIP-1α, MCP-1, IL-8, and I-309, along with two housekeeping genes, the L32 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) genes, as controls (Fig. A). The results of this experiment demonstrated that the expression of one of the cytokine genes, IL-8, was regulated as the cells progressed through S phase. Strikingly, the IL-8 mRNA levels increased from 0 h (G1/S) to 3 h (S), then decreased at 6 h (G2/M) and further at 9 h (G2/M) in the eTat cells. In contrast, the levels of IL-8 mRNA were low and roughly constant in the pCEP cells. Densitometric analysis of the RNA levels (Fig. B) showed a 3.5- to 4-fold induction of IL-8 expression at 3 h in eTat cells. The level of RNA expression and the sensitivity of the RNase protection technique did not allow us to detect fluctuation in several of the other genes analyzed in this assay. We did reproducibly observe, however, that RANTES was expressed at higher levels in pCEP cells and that its transcription peaked at the G2/M phase (Agbottah et al., unpublished data). The levels of L32 and GAPDH mRNAs were constant in pCEP and eTat throughout the cell cycle.
FIG. 2 IL-8 mRNA is induced in a cell cycle-dependent manner in Tat-expressing cells. For the RNase protection assay, 15 μg of total RNA was obtained from 3 × 107 eTat and pCEP cells blocked with Hu. (A) Samples were collected at 0, 3, 6, and (more ...)
A similar experiment was conducted with Noco, a drug which reversibly blocks the cells at the G2
/M boundary (5
). In contrast to the results obtained as the cells transitioned through S phase, the level of IL-8 mRNA expression in G2
/M and G1
was not significantly above background (data not shown) (see below). These results demonstrate that IL-8 transcription is specifically induced during the S phase of the cell cycle in cells expressing the Tat protein.
IL-8 protein is overexpressed in eTat cells following Hu block and release.
We next investigated whether overexpression of IL-8 mRNA was reflected in levels of secreted IL-8 protein that were higher in the supernatant of eTat cells than in that of pCEP cells. Cell culture supernatants were collected for 9 h following Hu or Noco block and release and assayed by ELISA (Fig. A). The levels of IL-8 increased dramatically starting at 3 h postrelease (approximately 450 pg/ml) in the eTat supernatant, reaching as high as 828.0 ± 151.0 pg/ml at 9 h post-Hu release. In contrast, the levels of IL-8 were low and fairly constant in the supernatant obtained from pCEP cells (170 to 188 pg/ml) (Fig. A). To address the possibility that IL-8 expression might be due to clonal variation in cells, and not to Tat expression, we tested several other clones of pCEP and eTat cells for IL-8 expression (Fig. B). The results of this assay clearly demonstrate that IL-8 expression is induced in each Tat-expressing eTat cell clone during S phase. These results argue strongly that IL-8 induction is due to Tat and is not an artifact of clonal variation.
FIG. 3 IL-8 is secreted from Tat-expressing cells in a cell cycle-dependent manner. (A) ELISA of IL-8 from supernatant of eTat or control cells. Control pCEP and eTat cells were blocked with Hu (2 mM) or Noco (50 ng/ml) for 15 to 18 h, washed, and released by (more ...)
Consistent with the RNA analysis, after a Noco (G2/M) block, IL-8 protein levels did not significantly change in either cell line following release into G1 (Fig. A). Taken together, these results confirm the mRNA data showing that IL-8 production is an S-phase phenomenon that is independent of the blocking agent used and is not due to cell cycle arrest per se.
To demonstrate that the induction of IL-8 was specific for the cell cycle and not due to non-cell cycle-related activity of Hu, we also used a thymidine block to synchronize the cells at the G1/S border. Following release, we collected the supernatant and analyzed IL-8 secretion by ELISA. The cells were also analyzed by FACS to assess the cell cycle progression. As observed after the Hu block, we detected a significant increase of IL-8 protein in the supernatant of eTat cells after release as the cells progressed through S phase, whereas the levels of IL-8 were constant and low in pCEP (data not shown).
Induction of NF-κB binding to the IL-8 promoter at the S phase of the cell cycle in eTat cells.
The IL-8 promoter was previously shown to be regulated by NF-κB. We conducted a series of EMSAs to measure NF-κB binding activity during the cell cycle. pCEP and eTat cells were blocked with Hu, released, and collected at 0, 3, and 6 h postrelease. Cells were subsequently fractionated into nuclear and cytoplasmic extracts. We examined the levels of nuclear NF-κB by mobility shift assay with a probe corresponding to the IL-8wt sequence (Fig. A). Three hours following release from the Hu block, an increase in NF-κB binding in eTat but not in pCEP cells was noted (Fig. A, lanes 6 and 9). NF-κB binding decreased at 6 h postrelease (S to G2/M transition) (lanes 7 and 10).
FIG. 4 NF-κB is induced during S phase in eTat cell nuclear extracts. (A) EMSA using 5 μg of nuclear extracts obtained from control pCEP and eTat cells after Hu block (2 mM, 15 h) and release. Lane 1, IL-8wt probe alone; lane 2, eTat untreated (more ...)
To determine the specificity of the DNA binding activity, competitor oligonucleotides were synthesized with base substitutions in the NF-κB (IL-8mκB) or neighboring C/EBP (IL-8mC/EBP) site (Fig. B). The results of competition experiments demonstrate that the gel shift complex is due to NF-κB binding. An excess of unlabeled wild-type oligonucleotide strongly diminished the signal (Fig. B, lane 3). In contrast, IL-8mκB had little effect (lane 4). The IL-8mC/EBP oligonucleotide, which possesses only the NF-κB binding sequence, competed for binding as efficiently as the wild type (lane 5). We also used oligonucleotides containing the NF-κB binding site from the HIV promoter as competitors. The results presented in Fig. C demonstrate that the HIV NF-κB oligonucleotides, but not the HIV TATA or initiator oligonucleotides, compete for binding of protein to the IL-8 NF-κB site. We also analyzed the gel shift activity of transcription factor Sp1 in the eTat and pCEP cell extracts. No significant change in Sp1 binding activity was observed in the 0-, 3-, 6-, and 9-h samples from eTat and pCEP cells (data not shown).
NF-κB complexes bound to the IL-8 probe are more stable in eTat cell extracts.
We next analyzed the composition of the NF-κB complex bound to biotinylated oligonucleotides containing either the wild-type or mutant IL-8 NF-κB binding site. Nuclear extracts obtained from pCEP or eTat cells, blocked with Hu and released, were incubated with the probes. Following pull-down, the complexes were subjected to Western blot analysis, using antibodies specific to the p65 NF-κB subunit and to CBP. As seen in Fig. A and B, and consistent with the gel shift analysis, bound p65 was most abundant in nuclear extracts prepared from the eTat cells at 3 h after release from Hu (Fig. A, lane 2, versus Fig. B, lane 2). Similar to the RNA induction curve and gel shift binding, NF-κB binding was transient. The level of NF-κB binding activity returned to baseline by 6 h after release. The lack of p65 binding in pCEP nuclear extracts was not due to the absence of p65 in the nucleus of these cells, as demonstrated by straight Western blotting (Fig. C). To demonstrate the specificity of NF-κB binding, parallel assays were run with an oligonucleotide containing a mutant IL-8 NF-κB binding site. Consistent with the results presented in Fig. A, a significant level of NF-κB p65 was observed 3 h after release when the IL-8wt probe was used (Fig. D). In contrast, no binding of NF-κB p65 was observed when the IL-8 NF-κB mutant probe was used (Fig. D).
FIG. 5 The NF-κB species p65 binds strongly to the IL-8 promoter sequence in Tat-containing cells. eTat and pCEP cells were blocked with Hu (2 mM) and released. (A, B, and D) Nuclear extracts taken at different points were incubated with either the IL-8wt (more ...)
It has recently been reported that the association of the coactivator CBP with p65 is important for the induction of transcription activity (26
). To determine whether CBP was present in the IL-8 NF-κB binding complex, the 3-h nuclear extract from eTat and pCEP was incubated with the biotinylated oligonucleotide, centrifuged, washed, and analyzed by Western blotting. The results presented in Fig. E demonstrate that CBP was present in the complex of proteins bound to the IL-8 NF-κB site. Consistent with the results of the p65 Western blot, the level of CBP bound to the IL-8 NF-κB probe was significantly higher in the NF-κB pull-down from eTat extracts. The CBP input Western blot (Fig. F) did not reveal a significant difference between CBP levels in eTat and pCEP cells.
MG-132 treatment abolishes NF-κB binding to the IL-8 promoter sequence as well as IL-8 protein synthesis.
Since we demonstrated that the IL-8 promoter sequence is regulated by NF-κB through the cell cycle, it was of interest to determine what effect the proteasome inhibitor MG-132 would have on IL-8 expression. MG-132 inhibits proteasome function and thus inhibits IκBα degradation, interfering with NF-κB activation and translocation to the nucleus. eTat cells were treated with MG-132 immediately after release from the Hu block, and samples were harvested as described above. An IL-8 ELISA was conducted on the culture supernatants. The results of this experiment demonstrate that treatment with MG-132, which inhibits NF-κB translocation to the nucleus, significantly reduced the levels of IL-8 protein (Fig. A). These observations are consistent with the enhancement of IL-8 expression by a transient activation of NF-κB during S phase.
FIG. 6 A proteasome inhibitor blocks NF-κB signaling and IL-8 production during S phase. (A) ELISA of IL-8 from supernatant of eTat cells. eTat cells were blocked with Hu (2 mM) and released. MG-132 (50 μM) was added at the time of release. The (more ...)
eTat cells were treated with MG-132 immediately after release, and time course nuclear extracts were also prepared as described above. As seen on the gel shift (Fig. B), MG-132 treatment abolished NF-κB binding on the IL-8 NF-κB promoter sequence. This observation explains the inhibition of IL-8 production (Fig. A).
Finally, the cytoplasmic fractions from control and MG-132-treated cells were run on a sodium dodecyl sulfate gel and analyzed by Western blotting using an antibody which recognizes IκBα (data not shown). In the MG-132-treated cells, a slower-migrating species of IκBα, corresponding to the phosphorylated form, was detected at 3 h postrelease (data not shown). This result suggests that NF-κB induction occurs at 3 h post-Hu release (S phase) in eTat cells by phosphorylation of IκBα and its subsequent degradation by the proteasome.