Who: Allan Weissman
Position: Chief, Laboratory of Protein Dynamics and Signaling at the National Cancer Institute, 2003 to present
Background: MD, Albert Einstein College of Medicine, 1981; Postdoc fellow, National Institute of Child Health and Development, 1984 to 1989; Chief, Laboratory of Regulation of Protein Function, 2001 to 2003
While most research into cancer metastasis has looked at genetic or genomic changes, a team at the National Cancer Institute is studying a protein interaction in vivo that promotes the phenomenon.
Specifically, the researchers looked at gp78, a protein that tags specific proteins with ubiquitin, and found that silencing the protein increases the level of the KAI1 gene, which has been shown to interfere with metastasis.
Led by Allan Weissman, chief of the NCI’s protein dynamics and signaling lab, and Chand Khanna, head of the NCI’s pediatric tumor and metastasis section, the team found that in mouse models, high levels of gp78 correlated with lower levels of KAI1, and that silencing KAI1 in cancer cells encouraged metastasis and survival of the cancer cells.
Their research appears in the December issue of Nature Medicine,
This week, ProteoMonitor spoke with Weissman about the research. Below is an edited version of the conversation.
What’s the take-away message of this study?
I think the take-away message is that when you think about genes that are not only responsible for cancer but also for metastasis, you need to think about not only looking at the expression of the gene, the transcript, but also at the level of the protein, and that protein degradation is a critical regulator of many, many processes, including metastasis.
Wasn’t that known already?
Well, I think the idea that degradation is important in cancer is known. There are a lot of different oncogenes that are regulated by degradation … Here we have a gene that is specifically involved in metastasis that can be regulated by degradation, and that’s the KAI1.
What was your original hypothesis?
The original question we were asking was whether gp78 played a role in cancer because there had been some reports that it might.
We were interested in it because this protein was originally thought to play a role in cancer aggressiveness, but that was before we discovered it was a ubiquitin ligase, and then we became interested in whether its activity as a ubiquitin ligase was critical to its function in cancer or in metastasis.
And that’s really what we were interested in looking at.
Was this study built on other studies?
If you read the paper, we reference papers … where [researchers] had seen an association with a protein that they refer to as gp78 with aggressiveness of tumors.
Was that the reason you looked at this protein?
Right. But it isn’t entirely clear to us whether the protein they were looking at is actually the exact same protein that we were looking at. And that’s an issue of the antibodies that were being used.
Where is that issue now? Has it been determined one way or the other?
It’s still up in the air.
Had this inverse relationship between gp78 and KAI1 been suspected before?
This was a new finding, but I think it’s really important to understand that the KAI1 findings are limited in terms of looking at tumors. [It’s] limited to a relatively small number of tumors with [a] single array, and that that array that we looked at is not an outcome linked array. So we can’t establish in those tumors – I think this is really important not to overstate the importance – we can’t conclude in patients that gp78 plays a causal role in regulating KAI1 and in aggressiveness. We can only conclude that based on our animal studies.
The animal survival experiments were really our initial experiments looking at metastasis, and we don’t know whether KAI1 is the only factor that may be affected by gp78. We know that KAI1 can be affected by gp78 and seems to play a role in the sarcoma that we looked at.
But we don’t know whether gp78 is doing other things in the cell – it likely is doing other things in the cell – that may play a role in survival of tumor cells.
And there could be other proteins that play a role in regulating KAI1?
There could be. The more important issue, the take-home message, is that by targeting proteins for degradation you can affect metastasis. This is a ubiquitin ligase that specifically works by targeting proteins in the endoplasmic reticulum or from membranes.
This is an instance where a membrane-bound ubiquitin ligase, by targeting proteins for degradation, whether it’s KAI1 or KAI1 and other things, can play a role in metastatic progression.
Based on your study, do other proteins and protein interactions play a role in metastasis?
We don’t know. We are interested in the possibility that by preventing degradation of proteins from the endoplasmic reticulum, which is where gp78 largely works – now we haven’t proven that it targets KAI1 specifically in the endoplasmic reticulum, but that’s just a working hypothesis – we believe that by affecting that, that can affect cell stress.
When the ability of cells to target proteins for degradation from the endoplasmic reticulum is not adequate, there is something called ER stress. ER stress can cause something called an unfolded protein response, which is the cell’s response to this stress, and the primary idea behind the unfolded protein response is a compensatory mechanism by the cell to try and recover from the stress.
But when the cell’s ability to recover from the stress is inadequate, there is a component of the unfolded protein response that can lead to apoptosis. So we’re interested in not only whether gp78 may be playing a role by targeting KAI1, which is the conclusion of the paper, but whether some of its effect may be due to its causing ER stress because it is an important ubiquitin ligase involved in targeting proteins for degradation in the endoplasmic reticulum, and thereby can cause an unfolded protein response that might potentially lead to cell death.
And it may be, although we don’t know, that cells undergoing metastasis may be more susceptible to this type of stress than other cells. That’s a hypothesis but something that we’re very interested in.
Is that the follow-up to this study?
That might be one follow-up that we’re interested in.
And are you doing a follow-up to get a clearer relationship between gp78 and KAI1?
Yeah, we will be trying to look at additional types of tumors.
Your study is based on animal models. Are you interested in verifying your results in human patients?
We are interested, eventually, in the idea of effecting ER degradation in general, and ER stress and the effects of that on various tumors. And there are efforts underway at NCI to find compounds that affect ER degradation in general and maybe, as part of that maybe will affect KAI1.
But I think we want to be careful not to overstate the generalizability of the KAI1-gp78 relationship beyond what we’ve looked at, because the only setting where we have a very strong causal relationship is in the sarcoma cell line that we used for most of these studies.
You also looked very specifically at ubiquitylation.
We did look at ubiquitylation of KAI1.
Any plans to look at other types of post-translation modifications and what kind of role they can play in metastasis?
Right now, we’re mostly interested in ubiquitylation because it’s a tag for degradation and altering the levels of proteins. But I’m sure there are many laboratories where many people are studying phosphorylation as being critical for cancer metastasis.