Can We Stop Fat From Fueling Triple-Negative Breast Cancer?

Welcome to The Breastcancer.org Podcast, the podcast that brings you the latest information on breast cancer research, treatments, side effects, and survivorship issues through expert interviews, as well as personal stories from people affected by breast cancer. Here’s your host, Breastcancer.org Senior Editor Jamie DePolo.

Jamie DePolo: Hello, thanks for listening. When triple-negative breast cancer grows, the fat cells around it seem to shrink. A research team at the University of California, San Francisco, found that triple-negative breast cancer cells build molecular tunnels called gap junctions into nearby fat cells and use the fat cells’ energy for fuel. When the scientists blocked the gap junctions, the tumors stopped growing. The research was recently published in the journal Nature Communications.

I'm joined by Dr. Andrei Goga, who is professor of cell and tissue biology at UCSF, co-leader of the Breast Oncology Program in the UCSF Helen Diller Family Comprehensive Cancer Center, and senior author of the paper. He’s going to help us understand what these findings could mean for treating triple-negative disease. Dr. Goga, welcome to the podcast. 

Dr. Andrei Goga: Well, thank you very much for having me.

Jamie DePolo: So, just so everybody understands and also, so I understand, how did this research even come about? What made you think to look at this?

Dr. Andrei Goga: Yeah. Thanks for asking. There’s been quite a bit of interest in how things like diet affect cancer and it’s been known really for quite a while that there are quite a variety of different cancers, including breast cancer, but other cancers as well like ovarian, gastric, esophageal, etc., that are associated with adiposity, with patients having increased fat. You know, the risk is not super high, but it’s definitely been there.

So, that was one area that we were interested in.And then the other area is that breast cancer cells like other cancer cells, and in fact, all tissues within our body, rely on different nutrients to allow them to survive, and in the case of breast cancer to grow. 

And we and others at Baylor College of Medicine and other places had done research a number of years ago and found that the particularly aggressive subtype of breast cancer called triple-negative breast cancer, so it’s the subtype of breast cancer that doesn’t express the estrogen or the progesterone or the HER2. These are other subtypes of breast cancer. So, this is a subtype that’s really differentiated because it doesn’t have those things. But this is a subtype that tends to occur in younger patients, sometimes, and is particularly aggressive, often associated with metastases and since it doesn’t have those receptors, we don’t have very directed drugs. So, more often it comes to things like chemotherapy and in some cases, immunotherapy is used for those cancers.

But we observed that this type of breast cancer is particularly in need of using lipids. So, basically it uses fat as one of the energy sources that allows it to grow. And this is somewhat different from those other subtypes that, in particular the hormone-positive breast cancers, like the estrogen- and progesterone receptor-positive. They tend more often to actually need to synthesize new fat. So, there’s some difference there. And in this prior work again, both by us and other groups, found that if you could block the ability of these types of breast cancers, the triple-negative breast cancer, to break down, in particular these long chain fats, then you could really diminish their growth. So, we thought this was an important insight and might also, you know, again relate to the overall status of a patient whether they were overweight and so forth. 

So, the obvious next question was, well, where’s the fat coming from? So, the breast cancers develop within the breast area, which actually has a lot of adipocytes, which are fat cells, basically. And so, this led us to ask whether…in fact, we know that the tumor cells are next to the fat cells very frequently. And so this really led us to ask could they be directly stealing fats from the surrounding tissue? Is this an important way in which they might be supporting their growth? That was really the beginning of us investigating this.

Jamie DePolo: Oh, that’s very interesting. Now could you sort of explain to us a little bit how you did this study? Because if I'm remembering correctly, you looked at actual breast cancer cells from people who had been diagnosed and then I think mouse models as well. Is that right?

Dr. Andrei Goga: Yes. Yes. So, we used a variety of really complementary approaches, and I should say this is really a very collaborative study that involved, you know, pathologists and radiologists, and so forth.

So, one of the approaches we took…we collaborated with Dr. John Shepherd and his group. He was at UCSF at the time and now he’s one of the leaders in breast imaging over at University of Hawaii. But they had developed a type of mammogram that’s slightly more complicated than the typical mammogram that allows you to…essentially it does imaging at different energy strengths. So, it can actually distinguish not only — this is called three component breast imaging — and so it allows you to detect different components of the breast, including how much fat is present in a given area. So, you can determine where the tumor is, but also kind of the extent of a lipid content, fat content around that tumor. 

And so, what we found that was surprising was that if you were to find where the breast tumor itself is, let’s say it’s small breast cancer, a centimeter or something like that, but then were to look in the area around the tumor, that the lipid content right around the tumor was much lower than further away, suggesting that the fat cells around there actually had less content.

So, just to remind people, a fat cell is essentially a specialized cell that has a giant kind of blob inside of it where the fat is stored. Kind of like a fat droplet if you were to drop it, you know, on a pan of water. This is the way it looks within the cell.

And they’re very dynamic, so under conditions where you lose weight, the size of the fat cells actually gets much smaller. And then let’s say you're under conditions where you're eating more and you're trying to store fat, you know, those cells can actually get bigger. So, the amount of — people had shown that the size of the adipocytes actually can be correlated to how much they’re breaking down fat. So, that was the first clue.

And then we were able again through this collaboration to get access to some patient samples from the same studies so we could directly actually measure the size of the adipocyte. So, this is a tumor section that shows where the cancer is and then we could look at the fat cells that are surrounding it and actually measure how much lipid is there. And so, that seemed to really correlate. And so the closer you are to the tumor the less lipid there are. If you get further away, then there’s actually more lipids and these adipocytes are much bigger. 

And then, that led us to do a variety of other things using other data sets and work from other groups that had been published that we could then access to actually look at both gene signatures. So, there are the types of genes that get turned on when a fat cell is induced to release its lipids or break down its lipids. And that’s a process called lipolysis. So, for example, if you take a certain medicine that causes the fat cells to start to break down the lipid, that’s called lipolysis. So, we found that the signature for the signal to break down the lipids was highest closest to the tumor and as you went further and further away, up to 4 centimeters away, that signature got much lower, and then normal breast tissue had an even lower signature. 

And then we collaborated with a group that had actually done very precise quantification of proteins within the breast using a process called mass spectrometry so they could look at particular proteins in different subtypes of breast cancer. Again, what was found is that some signatures, some proteins that are associated with this fat breakdown, was highest in the tissue that was right next to the tumor rather than the tumor itself.

So, that is what really led us to think that there’s an effect from the breast tissue overall. And then that led us to do more, what we’d say mechanistic studies. So, more detailed scientific analysis using both kinds of mouse models, as well as patient samples. 

So, one question we had is, if the lipids are being depleted when you're very close to the tumor, how is that happening? And so, there’s prior work to suggest that tumors can release factors, sort of like growth factors and things like that, or cause inflammation in the breast and some of that can lead the fat cells to release and break down the fat. But we wondered if there might be direct interaction. And so, this is where things are a little bit technical, but I think are important to understand.

So, many cells in our body make direct interactions with other cells. So they are bound very tightly together. So, it’s not like if you were to imagine, you know, you're eating cereal and whatever, it’s like alphabet cereal or something and everything is sort of floating around. Cells are typically not like that, but our tissues…our organs, you know, have a very clear structure and this direct cell-cell interactions are really important for example to allow us to have our form. 

And it’s known that fat cells can interact with one another through something called gap junctions. And so, these gap junctions are proteins that form a channel, they can form a channel between two cells, and they can transmit contents between one cell to the other. Now they’re pretty small so you can't actually get large fats to go directly through these, but you can send signals that can stimulate this breakdown of fats.

Jamie DePolo: So, then it would be smaller so it could get through that little…

Dr. Andrei Goga: Right. 

Jamie DePolo: …space. Okay. 

Dr. Andrei Goga: It’s about 1,000 kilodaltons or so. [Editor’s note: A kilodalton is 1,000 daltons. A dalton is 1/12 of the mass of a carbon atom.] So, it’s relatively small, so something like a nucleotide. Something like…in this case it’s something called cyclic AMP. So, it’s a small molecule that can stimulate this activation of the fats…for the breakdown of the fats.

Anyway, so we looked in patient data from the cancer genome atlas and then other data stats including from some mouse models where we could generate breast cancers that look like triple-negative breast cancer, had the same sort of features. There’s at least 20 of these gap junction proteins. Not all of them are expressed in each cell, in fact, certain cells really prefer one type versus the other. 

So, we had identified one in particular called, the gene is GJB3, and then the protein is connexin 31, which we found to be the highest expressed within the triple-negative breast cancer subtype. It’s also expressed in the hormone receptor-positive, but just at a lower level. It’s most highly expressed in the triple-negative subtype. And that led us to ask a variety of different questions.

So, if you were to block these gap junctions what happens? Well, it turns out that this molecule that I mentioned called cyclic AMP is no longer transferred either between breast cancers, but also between breast cancers and the adipocytes. 

And so, we used some drugs that can inhibit gap junctions more broadly and we showed that actually that inhibited the growth of these aggressive breast cancer cells. And then, we also used genetic approaches.

So, your listeners may have heard of a technique called CRISPR, which was developed, you know, a number of years ago by researchers at UC Berkeley and also at Harvard that allows you to engineer cells. So, essentially to knock out or knock down genes. And so, we specifically knocked down this particular connexin 31. And the first thing we found that was interesting is we couldn’t completely knock it down if we tried. We never got cells that were viable that completely lost it. So, that told us that it has some important function, you know, in the growth of these cancers on its own. But when we were able to generate cells that had partial knockdown — about a third of the normal expression — they didn’t grow very well, and in particular when transplanted into the breast of a mouse. Right? So, they really weren’t able to grow very well. It took, you know, three or four times longer for them to grow. In some cases, they didn’t grow at all.

And then what we found was that the fat cells around these tumors, the ones that did grow, were much larger, suggesting that they weren’t actually able to do this process of causing breakdown of fats in the surrounding fat cells.

And then, I think one interesting experiment was that if we took one of these cancer models in which the connexin 31 was decreased and we supplemented by treating the mice with a drug that would cause breakdown of fatty acids sort of throughout the body, now we can stimulate their growth. So, this sort of showed that at least part of the way that this is working is by forcing the breakdown of the fatty acids. Does that make sense?

Jamie DePolo: Yeah. I do have a couple questions though.

Dr. Andrei Goga: Sure.

Jamie DePolo: So, what I'm wondering…you're talking about breast cancer cells are using these gap junctions to stimulate the fat cells to break down the fat and then they're kind of getting that out, a very simplistic explanation…

Dr. Andrei Goga: Right.

Jamie DePolo: And so, the fat cells around the breast cancer cells were smaller, but you also said that the gap junctions can only be so long, that they're small. So, I guess what I'm wondering, if the cancer cells deplete the fat cells around them, do they move closer to other fat cells? How does that…What’s going on there?

Dr. Andrei Goga: That’s a great question. The way we did these studies was to use kind of versions of the small molecules that can be visualized. So, there’s sort of dyes that can transfer from these and we showed that blocking these gap junctions prevented this transfer. So, that’s how we know that they can have a direct effect. But work from other labs in the past have shown that the entire fat tissue can actually respond to a signal to break down long chain fatty acids, not only within itself, but because the fat cells themselves have gap junctions with other fat cells, so the signal can actually extend over a larger space. So, it’s like the tumor cells talk to the fat cells. The fat cells talk to their neighbors and so on and so forth. 

Jamie DePolo: I see.

Dr. Andrei Goga: Eventually there’s a decrease in the signal, but it can extend much longer over a larger area than just the direct interaction between the tumor and the adipocytes. 

Jamie DePolo: So, it’s almost like the fat cells are playing a game of telephone with each other with this signal. It’s like it comes from the cancer cell, but then it gets passed on from fat cell to fat cell, sort of holistically. 

Dr. Andrei Goga: Absolutely. Exactly. So, we think this cyclic AMP can be transferred not just from the tumor to the adipocytes, but more extensively. 

Jamie DePolo: So, was your research just looking at early-stage, like stage I through III breast cancer cells? Because I'm wondering about metastatic disease. Like maybe metastatic disease in bone — and I am not a doctor or researcher — but I don’t know that there are that many fat cells in bone. So, I'm wondering how does this work with metastatic disease?

Dr. Andrei Goga: This is fantastic. Yeah. Great question. This is actually one of the areas that our group is most interested in because most patients who die from breast cancer, it’s really because of metastatic disease and bone metastases, or metastases, you know, into the bone marrow, within the bone, is really a very common site and perhaps one of the most common sites. And it turns out that in young people there is about 30% to 50% of the bone marrow is fat and in older patients it’s as much as 70%. So, there’s definitely a large component of adipocytes. We haven’t looked at this. 

You know, this is actually an area that we’re very interested in looking at next. We’ve submitted some grants and so forth. I think this could be very important. We know that sometimes breast cancer cells have spread, you know, even at the time when we first see a patient in clinic we think the cancer is localized, but then, you know, unfortunately despite our best efforts, you know, months, years, or even decades later they may recur and often it’s in the bone. 

So, could these interactions between the tumor cells…could the fat within the bone actually be supporting some of these tumor cells perhaps even for long periods of time? So that is something we don’t know, but we’re excited to pursue. And it may be that the same process is going on, and it involves this connexin 31. It may be that it’s a different gap junction or maybe a different process altogether. But I think it’s something worth investigating. 

And then, more broadly related to that, our lab and many others are also interested in metabolism in general. I think there’s, you know, in the popular literature people have talked about how tumors can use sugar, as an example, to support the growth, but I think it’s actually more than that. I think, you know, cancer cells are pretty opportunistic, so depending on the cancer genes that are stimulating the cells they can be reprogrammed in a way to use different energy sources. 

So, our work has suggested or shown that fatty acids are an important energy source, but it may be that in other locations if, for example, let’s say in the lung, there is also metastases to the lung and typically there isn’t a lot of fat in the lung. It’s more, you know, alveoli and so forth, but there are other nutrients that could be present there. So it’s a very large capillary bed, so there could be blood flow, but there’s also lots of oxygen. There may be other ways that breast cancer cells survive and are able to be successful at growth in different tissues. And I think understanding that process will allow us to do a better job. 

Jamie DePolo: Yeah. That’s fascinating because it also could potentially hint at, you know, why does one person get metastasis to the bone, somebody else gets metastasis to the lung, like why isn’t it always the same? That’s very interesting to me if we could kind of figure out what the cells are feeding on. 

Dr. Andrei Goga: Right. And you know, as a medical oncologist, we treat patients when they have localized disease and that’s largely very effective with, you know, combination surgery and chemotherapy and so forth. And the chemotherapy is largely to prevent this recurrence, right? To try to eliminate the cells, if there are any, that have already sort of spread beyond the local breast. But you know, most therapies are designed and are sort of developed to shrink the local tumor and really what we would I think need to prevent recurrence is to also target these pathways that might allow the early metastatic cells that have, let’s say, already spread, that really are too few for us to detect with existing technologies. 

But let’s say we knew that they needed to be supported by fat metabolism, perhaps giving some treatment for a period of time to try to really prevent the subsequent distant metastases. Those are not easy studies to do, right? Because sometimes they don’t recur for many years. So, historically, it’s been challenging to get the interest of pharma because, you know, they have a limited shelf life for their particular medications and if we knew the medication worked and prevented recurrence, then that would be very helpful and I'm sure they would be very interested, but to wait for a study that might take five years to read out is obviously a challenge from a…like their business perspective. But nevertheless, I think understanding this is super important because ultimately, you know, we want to cure patients and we rarely say that, you know, once metastases has occurred.

Jamie DePolo: Right. Well, and it also kind of reminds me…I talked to Dr. Andrea DeMichele at UPenn, and she has a study, the CLEVER trial, and they’re looking at identifying and sort of neutralizing dormant cancer cells that are left behind after treatments like chemotherapy and surgery. And I guess what I was thinking is, okay, you know, you have chemotherapy, and most chemotherapies target fast growing cells, like cancer cells, but you know, also your hair and the lining of your gut, which is why you have the side effects. But if the cancer cells are dormant, they’re not really growing quickly so the chemotherapy may not kill them. Like say, they’re hiding out in the bone marrow. So, it seems like your approach and maybe her approach like together would be amazing because if you could find these cells, figure out how they were staying alive even though they weren’t super active and get rid of them, that would be amazing. 

Dr. Andrei Goga: Yeah. That would be fantastic, especially if the treatment is, you know, is non-toxic or has low toxicity for patients. I mean, it’s kind of what is already done for hormone receptor-positive breast cancer. We know that things like tamoxifen and aromatase inhibitors, other sorts of anti-hormone therapies decrease the risk of recurrent disease. But presumably some of the small amounts of disease are actually killed off, but that’s typically five to 10 years of anti-hormone therapy and it may be that combining it with something else, like maybe some sort of metabolic inhibitor at least for a period of time if it’s, you know, safe and otherwise effective, could further decrease that risk and that’s definitely an area of interest. 

Jamie DePolo: Yeah. Now, I believe I read that there are medicines that can block these gap junctions, as you talked about. And people are looking at using them in other cancers, but not breast cancer. Am I understanding that correctly?

Dr. Andrei Goga: Yeah. So, as far as I know, a drug was developed that seems to be selective for a different connexin called connexin 43 or cx43. So, this is a connexin that’s been shown to be important for the interaction between neurons. So, brain cells and their surrounding glial cells or astrocytes. And I think it was initially developed or tested for migraines. It was safe from what I understand, and I wasn’t involved in any of this development so my knowledge of the specific trials is somewhat limited, but my sense was it didn’t really work for migraines, but there’s a lot of interest to try to develop it for other potential interactions. 

So, a couple things that are relevant here. One, is that brain cancers called glioblastoma also interact with their surrounding tissues and my understanding is there’s a trial to explore this in combination with more standard chemotherapy for brain cancers using these gap junction inhibitors. There was also a really nice study from Joan Massagué’s lab a number of years ago where they showed the same cx43 could allow breast cancers that had metastasized to the brain to interact with these same glial cells. They didn’t look at metabolism, but they looked at some other inflammation and some other things. So, it’s possible that those could have a role, although we don’t think it’s in the same way that this particular connexin that we identified. 

So, this idea that you could inhibit gap junctions and perhaps they would have utility in treating breast cancer or treating breast cancer metastases, I think is out there. You know, unfortunately we don’t have a very selective inhibitor for the gap junction that we identified, which is the cx31, but I think there’s a proof of concept out there that these other drugs have been developed for a different gap junction. So, it seems possible. There’s been some attempts to use small peptides. So, these are tiny fragments of proteins, which could be more selective. You know, typically those are not great drugs per se, but at least would provide sort of a proof of concept that something like that could be used and made more selective for the gap junctions that we’re interested in. 

Jamie DePolo: Is that something that your lab would do? Would be to work on drug development like that? Or are you more into studying how the cancer cells get their fuel?

Dr. Andrei Goga: I mean, I think the next steps…you know, we’re interested in many levels. So, I think one is really understanding, as we talked about, is this something that’s unique to the primary breast tumor? Or you know, do similar interactions occur in metastases, in particular in bone metastases, but perhaps in other tissues? We identified cx31 primarily or refocused on it, primarily because it was the most highly expressed. It was the one that was at highest levels from other data sets, but one could take a more systematic approach where you try to examine all the ones that are present to see which one is truly the most important. So, I think that would be an important area. 

And then, thinking about ways to go after them. I think gap junctions could be one. So, doing small molecule screens. You know, we have a screening center, it’s called Small Molecule Development Center, that we could screen for drugs. But I think even more broadly, blocking fatty acid oxidation, blocking the breakdown of fatty acids or the ability of the tumor cells to use them, I think would be an important area. There are no approved drugs yet, although some, you know, made it reasonably far in clinical development for other indications like for heart failure. 

So, perhaps repurposing some of those drugs or going back and redeveloping them in a way to make them more selective and more effective would be an area to go after. And that’s certainly an area of interest for our group.

So, I’ll give you an example. So, this notion that different tissues can use different energy sources is actually quite an important concept. So, in the heart, for example, you know, oftentimes cardiomyocytes, the heart cells, will use lipids and the thought is that if you could block their ability to use lipids then they would instead use sugar as an energy source, which is perhaps a more efficient way. And so, there have been some attempts to block the metabolism of fatty acids in the heart to improve things like patients who have heart failure, congestive heart failure where the muscle that’s present isn’t pumping as well as it should and so if you could switch the energy source. 

So, there was a trial of a drug that made it to phase III. It seemed like some of the endpoints they were looking at, which were how far patients could walk, you know, after taking the drug seemed to look positive, but a few patients developed liver toxicity from blocking the ability to use the fatty acids. And so, I think it was like two or three patients out of 300 had some liver toxicity. Of course, the utility in that context is very different, right? It was a drug to give daily, you know, forever, or for a long period of time. Whereas chemotherapy or these sorts of approaches typically are much shorter. You know, chemotherapy is typically given in cycles for short periods of time. So, there may be some opportunities to either revisit some of those drugs or make better drugs like those, that if tailored to cancer therapy, you know, might be better tolerated and could be given in a way that’s actually effective. 

Jamie DePolo: Sure. That all makes sense. I just have one last question. You mentioned that hormone receptor-positive breast cancer cells – it sounded like they were making these gap junctions, too, but the protein that stimulated that or the gene that stimulated that was expressed at a lower level. Did I understand that right?

Dr. Andrei Goga: That’s correct.

Jamie DePolo: So, do you think potentially…obviously right now the focus is on triple-negative as you said, but as the years go by and say this proves successful, do you think it could then be applied to these other subtypes of breast cancer?

Dr. Andrei Goga: Yeah, absolutely. I think more research is really needed. One observation has been that in the hormone-positive subtype of breast cancer more often those tumor cells seem to require desynthesis of fatty acids and so you could imagine that they could be synthesizing fatty acids, but they could also be taking in fatty acids as well. And so, there have been some studies, mostly preclinical studies, so in, you know, experimental models, to show that if you block synthesis of fatty acids that could be effective in those hormone receptor-positives. But again, we do see some expression of these gap junctions, and it may be that other kinds of gap junctions are present in those cells. Overall, the decrease in lipids seem to be present both in the triple-negative and in the receptor-positive. So, it may be that there’s a parallel mechanism or maybe even similar mechanism, but they’re just doing something different with the lipids. Instead of immediately burning them to generate energy they may be using them in different ways as precursors or something like that.

Jamie DePolo: Interesting. Dr. Goga, thank you so much. This is so fascinating. I learned a lot. I hope everybody else did too. And we will eagerly await the next steps in your research to find out where this is going. Thank you so much.

Dr. Andrei Goga: Yeah. It was great. Thank you so much. Can I just give a shout out that, you know, lots of people supported our research, including the NIH and I think NIH is so important for the type of research we do, but also the Department of Defense Breast Cancer Program and the BCRF. And really some, you know, a variety of different folks who directly donated funding to UCSF and the Breast Program, including the Bechtle family, and the Atwater family, and the Subramanian Breast Cancer Support Fund. So, thank you so much. 

Thank you for listening to The Breastcancer.org Podcast. Please subscribe on Apple Podcasts, Google Podcasts on Android, Spotify, or wherever you listen to podcasts. To share your thoughts about this or any episode, email us at podcast@breastcancer.org or leave feedback on the podcast episode landing page on our website. And remember, you can find out a lot more information about breast cancer at Breastcancer.org and you can connect with thousands of people affected by breast cancer by joining our online community.