This page was created as part of an undergraduate assignment at Davidson College.



Paper to be Reviewed:

Identification and functional characterization of a novel binding site
on TNF-(alpha) promoter

Xiaoren Tang, Matthew J. Fenton, and Salomon Amar

Proceedings of the National Academy of Sciences, Volume 100, Number 7, April 1, 2003 4096-4101


Tumor necrosis factor (TNF)-(alpha) is an endogenous pleiotropic cytokine that is stimulated in the process of inflammation. With respect to the inflammatory system, TNF-(alpha) can be both helpful and/ or detrimental, therefore it is important to be able to regulate its production through the TNF-(alpha) promoter.

There are 2 basic levels at which the TNF-(alpha) promoter can be regulated: transcriptional and posttranscriptional. This paper focuses on regulation at the transcriptional level. One of the transcriptional factors that bind within the TNF-(alpha) promoter region is lipopolysaccharide (LPS)-induced TNF-(alpha) factor (LITAF). LPS induces the production and secretion of TNF-(alpha) by stimulating monocytes and macrophages. The functional role of LPS (in vivo) is unclear. While LPS is known to up-regulate DNA-binding activity of inducible transcription factors, such as NF-(kappa)B, it is also known to down-regulate the DNA-binding activity of some constitutive transcription factors. One of the inducible transcription factors that LPS effects is NF-(kappa)B, which may up-regulate TNF-(alpha) transcription in response to LPS in some cancerous human cells, but the role of NF-(kappa)B is ambiguous. Another transcription factor that might have a clearer role in the activation of hTNF-(alpha) transcription is hLITAF, which mediates hTNF-V transcription. The main goal of Tang et al. was to determine the mechanism of hLITAF and hTNF-(alpha) interaction.
In order to accomplish this goal, Tang et al. decided to go after the binding site of the TNF-(alpha) promoter—first they identified the site via DNase I footprinting, then they characterized it via different shift assays (Electrophoretic Mobility Shift Assay, EMSA and Enzyme-Linked Immunosorbent Assay, ELISA).

The data obtained from footprinting revealed a protected region on hTNF-(alpha) DNA by the protein hLITAF from nucleotides –515 to –511, a CTCCC motif (Figure 2a, b). However, the footprinting was not run with whole hLITAF; the protein was broken into restriction fragments, bound to a fusion protein, then mixed with the hTNF-(alpha) DNA. Therefore, Tang et al. were able to determine the specific sequence of hLITAF that bound to hTNF-(alpha) as well. From the data, one can see that the only regions in amino acid sequence 1 to 228 were capable of protecting the binding site of hTNF-(alpha). Therefore, the binding interaction between hTNF-(alpha) DNA and hLITAF occurs on nucleotides –515 to –511 and somewhere in the amino acid sequence 1 to 228, respectively (Figure 2a).

The EMSA (or band shift assay) was used to further establish the protein-DNA interaction. In this assay, Tang et al. tested smaller regions of fragment 1-228 of hLITAF in order to be able to focus on a more precise region of hLITAF that is binding to the hTNF-(alpha) DNA. Shifted DNA bands indicate binding: there were 2 detected, in regions 152 to 228 and 1 to 228 (Figure 3a). Therefore, the binding region must be within the amino acid sequence 152 to 228. A similar, but even more specific EMSA was run next: the data revealed binding in regions 152 to 228 and 1 to 228 again, but also in the region 152 to 228 with a deletion from 180 to 195 (Figure 3b). Therefore, the binding region must be within the amino acid sequence 165 to 180.

The next step for Tang et al. was to make sure that the hLITAF sequence was doing what they thought it was doing. Therefore, based on the amino acid sequence of hLITAF, they constructed 3 peptides, inserted them into THP-1 cells, and tested for TNF-(alpha) secretion. The 3 peptides were cleverly named peptides A, B, and C. Secretion of TNF-(alpha) was tested at several different concentrations of the 3 peptides. Peptides A and C did not have a significant effect on TNF-(alpha) secretion, but peptide B induced the secretion of significantly more TNF-(alpha) than a control peptide did (Figure 4). This corresponds well with the previous results of the paper, because peptide B is found in amino acid sequence 165 to 180 of hLITAF (Figure 1a).

The final goal of the paper was to determine if the region -515 to -511 in the hTNF-(alpha) promoter was indeed responsible for hLITAF binding. Tang et al. cloned 4 different constructs, 2 of which had the CTCCC sequence of -515 to -511 deleted. They transfected the constructs into THP-1 cells and stimulated them with either LPS (for a positive control) or peptides A, B, or C. Then they detected TNF-(alpha) promoter activity with a luciferase assay. Only TNF-(alpha) promoters with the CTCCC sequence in tact showed above control activity; among those, the only cells that showed close to wild-type activity were those stimulated with LPS or peptide B (Figure 5a, b).

From these results, Tang et al. conclude that the binding sequence of hLITAF to the hTNF-(alpha) promoter is CTCCC (in the -515 to -511 region of the DNA). They also suggest that the hLITAF sequence that binds to hTNF-(alpha) is between amino acids 165 and 180. These studies were all done in vitro, but Tang et al. hypothesize that the DNA binding region of hLITAF (165-180) also functions in vivo to regulate hTNF expression.




First off, this paper deserves credit for using a functional approach to clarify the details of this mechanism. The most convincing results Tang et al. had was the EMSA in which they determined that the most likely binding sequence on hLITAF is amino acids 165 to 180 (Figure3a, b). There was one problem I had with the figure: lane 7 of figure 3b has a blot where it was expected, however, there seems to be another prominent (albeit significantly less prominent) band right below it. Tang et al. did not explain that shifted DNA band.

Another point of contention that I had was that while it is compelling that the location of peptide B corresponds to the amino acid sequence 165 to 180, I am not thoroughly convinced that this region of hLITAF is definitely the binding site. While there is much evidence to support that the binding region is within amino acids 152 and 228, and it is not in the region 180 to 195, Tang et al. failed to test two critical regions: 152 to 165 and 195 to 228. I would be much more convinced if they had experimented with those two regions in addition to the other regions they tested.

Overall, the science was very clean and well-controlled. Each experiment had both positive and negative controls, which made their results quite convincing to me.

Future Areas of Research


Tang et al. themselves indicated that they would like to see how their theory works in the human body, therefore it would be interesting to find a way to conduct an experiment with the same goals in vivo. I would also conduct another EMSA including the previously mentioned fragments in the amino acid sequence 152 to 228 that were neglected in this study.

The next phase of this study should aim to further elucidate the mechanisms of hLITAF/ hTNF-(alpha) interactions. One possible area of research could be to figure out what else is going on in the hTNF-(alpha) promoter when it stimulates hLITAF activity. Tang et al. have suggested that other transcription factors such as NF-(kappa)B might be involved.

The main goal of my future research (if I were to do it) would be to determine other regions of the hTNF-(alpha) promoter that are induced by LPS. I would perform the same procedures that Tang et al. did, except I would use hNF-(kappa)B instead of hLITAF. I would expect to find similar results to Tang et al., except I would expect the binding region not to be within the nucleotide sequence –515 to –511.

Suzy's Home Page

Molecular Biology Main Page

Biology Home Page

Davidson College

Please send your questions, comments, or concerns to Suzy Rizi