At A Glance
Name: Florence Cabon
Position: Team leader, French National Center for Scientific Research (CNRS)
Background: Principal investigator, CNRS — 1995-1997; Assistant professor, French Institute of Health and Medical Research (INSERM) — 1991-1995; PhD, molecular biology, Pierre & Marie Curie University — 1990; MS, biochemistry, Rene Diderot University — 1983
At CNRS, Florence Cabon is working in the lab of Annick Harel Bellan to investigate angiogenesis and its role in cancer. Recently, she took time to talk with RNAi News about her research and how it has extended into the RNAi field.
How did you get started working with RNA interference?
We have been working on angiogenesis for a long time now, and about three years ago we arrived at a situation where we showed that thrombospondin was able to inhibit angiogenesis and tumor growth, but we found resistance in tumors that over-expressed VEGF.
We wanted to demonstrate that, and just at this time [Thomas] Tuschl published his paper in Nature [showing that short pieces of double-stranded RNA could trigger RNAi in mammalian cells]. We rapidly demonstrated that VEGF siRNA were efficient in vitro, then we decided we wanted to try that in vivo.
We did very simple things at the beginning; we first implanted tumors in mice that were constitutively expressing luciferase, and we tried different routes of administration to inhibit luciferase in these tumors. We used very small amounts of siRNA because we were not rich — we are an academic lab and we were not able to afford 50 milligrams per kilogram. So we just used three micrograms per mouse and, to be honest, we were surprised to observe about 50-percent inhibition of luciferase in the tumors after a single injection of siRNA.
We were very encouraged by that, and we did the same experiments [later on] but using siRNA targeting VEGF, not luciferase. We could really demonstrate that we could inhibit VEGF in the tumors, and this resulted in an inhibition of tumor growth, as we expected. In the same paper [where these data were presented], we also published that we could combine siRNA against VEGF with thrombospondin to have some synergistic effects.
We published that in Cancer Research about one-and-a-half years ago, and a group [led] by [Mark] Duxbury [at Harvard Medical School] also published similar approaches using intravenous injections of low amounts of siRNA in vivo.
We have repeated [our experiments] with different target genes in different tumor models and, for example, we have very nice inhibition of prostate-tumor growth. The message is that we have not only gotten results with one gene in one tumor — we have that in different tumor models with different genes.
Also, in an experiment in a paper that has been submitted [for publication], we demonstrate that we can inhibit the expression of target genes in normal organs, such as testes or liver, with very small amounts: 125 micrograms per kilogram. We have inhibition in normal tissues and in tumors.
Are these siRNAs modified?
No. Not at all. They’re synthetic siRNAs, not even purified by HPLC.
The very surprising thing for me is that we do not even vectorize them. They’re just diluted in saline and injected as is — and it works.
These are injected directly into the organs?
No. This was really puzzling for me because at first, when I started these experiments, I thought that direct injection into the tumor would be my control, because when you inject antisense oligonucleotides it works. But it did not work [with siRNAs].
I think the main reason for this is that the pressure inside of a tumor is very high and most of the injected material just goes out. I guess this is the reason because otherwise I don’t see why it doesn’t work.
We tried different routes of administration — we injected our siRNAs intraperitoneally. Then, we injected the siRNA by intravenous routes or subcutaneously. The three routes worked perfectly. Routinely, we use IP because we inject our mice every day, and it’s much more simple to do that then to inject by IV.
How long were the knockdown effects seen? What was the duration?
We did not [look] at that in vivo — we inject every day. We have many question we want to answer. First we injected three micrograms, and I know this is provocative but I’m not even sure that we need to inject three micrograms every day. Maybe once every other day or once every three days is enough, I don’t know.
We have to try that, but the problem is that when you do in vivo experiments it’s tough to do — you cannot use as many animals as you want for ethical reasons. But this is something we are just trying to [investigate] now.
Can you talk about what you’re working on in your lab now?
Currently, we are working on prostate cancer, and we’re trying to design new siRNAs against different targets to inhibit prostate cancer growth in mice. The idea is also to combine these siRNAs with other therapies such as anti-angiogenics and also chemotherapy.
Is the idea to find new targets for small molecules, or do you see RNAi as being a therapeutic itself?
Both, actually. Here [my colleague] Annick Harel Bellan is developing an siRNA platform that will be very useful to identify new target genes, first in vitro, but the idea is go after that in vivo and then into the clinic. We want to try to develop some clinical trials with siRNAs, as well.
We have a lot of work to do before then, especially in mice, because we need to do toxicity studies and so on. But the interesting point is that if you compare our technology to Alnylam [Pharmaceuticals’] technology [highlighted in a recent Nature paper], we use 400 times less siRNA than they do [and] we do not add anything to the siRNAs such as cholesterol or lipids. So I believe that we really minimize the toxic possibilities or side-effects you could have because of the siRNA or because of the vectorization.
Do you anticipate that there’s going to be a need to modify siRNAs if they are to be used as therapeutics in humans?
Honestly, I think it’s very interesting to stabilize the siRNA for some target genes, but maybe it’s not necessary for all [targets]. You have to take care that if you modify something you will increase, probably, the toxic properties of the siRNA or induce some immunological reactions. I think if you can stay as close as possible to the normal siRNA, you minimize that.
My opinion is that the ideal situation is to use as little as possible of non-modified and non-vectorized siRNA each time this is possible. But I do think that in some situations, for some genes expressed at low levels in poorly vascularized tissues for example, vectorization and/or stabilization may be necessary.
It may be we were especially lucky in the choice of our first targets. Gene dosage is very important for VEGF and deletion of just one allele in mice is embryonic lethal. We also targeted successfully transcription factors. In that case, this inhibition modifies the expression of many genes downstream and likely amplifies the signal. Therefore partial inhibition of these genes in vivo in probably enough to observe a physiological response. But we also have examples, for example with GFP siRNA, [where] it works very nicely without any vectorization.
I think siRNAs injected as we do really goes into the tissues, we have the proof of that because the target gene is inhibited. But we cannot measure the proportion of siRNA that penetrates into the tissue because the amount of siRNAs we use is so small that you just cannot find them in the tissue after that. There is something that is really puzzling to me: If you inject, for example, peptides derived from thrombospondin into man, the half-life of these peptides is very short — it’s about 15 minutes. But they’re really efficient in inhibiting tumor growth and angiogenesis. So the idea that you have to stabilize siRNA because they have to stay for a long time in circulation is not so obvious to me.
I think it’s very important to have siRNA that go as soon as possible into the target tissue, penetrate there, and stay there. But in the circulation, I don’t really see the reason why people want to have some siRNA stable in serum. If they just need to stay there for one minute, what’s the interest in stabilizing them for one hour?
What about off-target effects?
It’s an important question. But sometimes, especially when you inhibit a transcription factor or a growth factor, it is very difficult to discriminate between off-target effects and normal, expected down target effects.
I think that when you use very small amounts of siRNA, you minimize off-target effects. And, independent of the cost of such experiments, this is the best reason to use as little siRNA as possible. Of course we verify that some unrelated genes are not affected by the introduction of a given siRNA. But when your siRNA stops the growth of a cell for instance, to find such genes may give a hard time to researchers.
Finally, what about collaborations? Are you working with anybody from the industry?
With siRNAs, we are essentially working with people here in France, but not through main collaborations.
Are these academic collaborations?
Yes. Mainly academic collaborations, for the moment. We have some collaborations with biotech, also, to try different things, but for the moment we have no results we want to publish.
This work with the biotech is in RNAi?
I don’t want to say because I don’t know what I’m allowed to say.