The researchers demonstrate that different basal transcription factors drive expression of the histone gene cluster, lending new insight into the regulation of metazoan transcription.

"This study surprised us on 2 levels; one was the preponderance of TRF2 dependent promoters; the other was the differential usage of TRF2 versus TBP within a gene cluster generally thought to be coordinately regulated. Just goes to show that dogma shifts constantly in this field of transcriptional control, explains Dr. Tjian.

In eukaryotic cells, gene transcription is initiated when the RNA polymerase II machinery recognizes and binds to specific core promoter sequences in the gene. While some genes contain a TATA box core promoter element that is recognized by TBP (the TATA-box binding polypeptide), the majority of core promoters fall into various TATA-less categories. A family of TBP-related factors (called TRFs) have been identified, but their core promoter recognition functions have not yet been elucidated.

In this paper, Dr. Tjian and colleagues identify novel TRF2 target promoters, effectively distinguishing between three classes of genes: TBP-dependent ones, TRF2-dependent ones and a small class genes that utilize both TBP and TRF2. They show that TRF2 is used as an alternative core promoter recognition factor to drive transcription of the TATA-less Histone H1 gene, while the other core Histone genes (H2A, H2B, H3 and H4) are dependent upon TBP. Furthermore, depletion of TRF2 in Drosophila cells resulted in reduced ribosomal gene transcription, abnormal cell changes and chromosomal defects.

cshl/

"These proteins enter the cell nucleus to turn on transcription of the EPO gene," Yeh says. They found that SENP1 controls EPO production by regulating one particular HIF protein, HIF1a. "When there isn't any SENP1, HIF1a is very unstable," he says. "It is not detectable in the embryo, compared to an embryo that has the SNEP1 gene."

It was already known that SUMO plays a role in the hypoxia process, Yeh adds. "We know that when you lower oxygen, HIF1a enters the cell's nucleus, and is quickly modified by SUMO."

But they discovered that there was one more step before HIF1a becomes active, producing EPO proteins to make more blood cells, and other proteins like VEGF that build more blood vessels to seek new sources of oxygen. They found that SENP1 needs to snip SUMO from SUMO-modified HIF1a before HIF1a can be active in transcription.

But that still didn't explain why HIF1a was missing in the nucleus of cells without SENP1. That led them to another, surprising finding - that if SENP1 does not clip off SUMO from SUMO-modified HIF1a when it is inside the nucleus, that SUMO then acts like ubiquitin, targeting destruction of HIF1a.

"This is the first example that SUMOylation of a protein can lead to its destruction," Yeh says. "That goes against the dogma we all believed in: that SUMO can change the location of a cell, but not degrade it. SUMO can do everything under the sun, including what ubiquitin can do. This vastly increases the functions of SUMOylation."

All this makes sense as far as cancer is concerned, Yeh says. HIF1a expression plays a role in many cancers and to date SENP1 has also found to be over-produced in prostate cancer. "This tells us that SENP1 is potentially involved in the overall regulation of tumorigenesis."

If true, Yeh says, that suggests it could become an Achilles heel for cancer. "These findings imply that you could inhibit SENP1 in tumors and let SUMO target HIF1a for destruction," Yeh says. "If tumors can't grow, these cancers could not continue to build a blood supply and grow and thrive."

mdanderson/