Part 3: Development of CasMab

[00:00:00]Next, I would like to move on to the development of CASMAB. The idea behind this work is not something that came to me only recently. In fact, around the time when I became a principal investigator, I was already thinking that this was the direction I wanted to pursue. From the very beginning, I felt that we had to create something that pharmaceutical companies would take seriously. The reason is quite simple.

[00:00:21]What we can do in academia is in reality very limited. Ultimately, it is not academia that turns a discovery into a drug. That is the role of pharmaceutical companies. One might ask whether a venture company could do everything on its own. But that too is extremely difficult. Unless a pharmaceutical company becomes seriously involved, it is very hard to take a project all the way to an approved medicine. I had also worked in industry myself. So I understood this very well. And when I thought about it in that way, I realized that we would have to make an antibbody that left no room for criticism. Even if we worked very hard and succeeded in generating an antibbody against a cancer target, if that antibbody also reacted with normal cells, all that effort would come to nothing. This is a slide that I often show to remind myself of where we started. It is from a research plan I wrote 12 years ago and I included this idea in various grant applications. I submitted it to projects related to the Ministry of Education before AMED [music] was established and also to JST projects, but every one of them was rejected at the document review stage.

[00:01:21]And not only that, [music] the evaluations were quite poor. I still remember the frustration very clearly.

[00:01:27]When I look at this slide, those [music] feelings come back to me. The comments that were particularly severe were things like, "This is merely a desk theory," or, "Please submit [music] your application after you have shown actual examples." Those comments were written quite plainly, and I remember thinking that [music] research really is a harsh world. But precisely because of that frustration, I continued the work, determined that somehow we would make this [music] antibbody. Proteins expressed on cancer cells are almost always expressed on normal cells as well. In practice, there is no such thing as a protein that is simply absent from all normal cells. Even today, I sometimes see papers [music] making that kind of claim. And when I see young researchers working hard to find such proteins, I do want to encourage them.

[00:02:07]But in reality, proteins that are completely absent from normal cells are extremely rare.

[00:02:13]Most of them are expressed in normal tissues [music] as well. So why have antibbody drugs been successfully developed up to now? I think there are probably many reasons depending on the era. Regulatory standards were different and the way people [music] thought about acceptable toxicity was also different.

[00:02:28]For example, if an antibbody like Herpin were developed today and submitted for approval under current standards, I am not sure whether it would be approved.

[00:02:36]So our goal was this. Even if the target protein [music] is also present on normal cells and even if the amino acid sequence is exactly the same, could we generate an antibbody that reacts only with the cancer cell form of that protein? It was of course a rather reckless challenge. But I think many people working in antibbody [music] therapeutics are drawn to this kind of reckless challenge and without such challenges, I do not think this field can move forward. If such an antibbody could be obtained, then even if the target molecule were expressed [music] on normal cells, adverse effects against those normal cells might no longer be the central problem. In theory, that would mean that virtually [music] any membrane protein could become a potential target and it would open the possibility [music] of targeting many different types of cancer. I thought that this was exactly the sort of thing pharmaceutical companies were looking for. With that objective in mind, we began [music] the project. One of the first starting points came from clinical samples of glyobblastoma.

[00:03:30]These are data that we have published at the time. After proper ethical review and approval, [music] we prepared lysates from fresh clinical samples and performed western blotting. Using a commercially available antibbody, we found that keratin sulfate, a type of glycosaminoglycin or gag was present on modified proteins that were detected almost exclusively in glyobblasto. In contrast, these signals were not strongly detected in grade 2 or grade three glomomas.

[00:03:55]This was work I carried out in the year after I graduated from medical school and returned to research. Having gone through medical training meant that I could discuss clinical [music] matters more naturally with clinicians and I could also understand these clinical issues much more readily. This was knowledge that [music] I would have found difficult to acquire if I had stayed only within a pharmaceutical sciences environment. In that [music] sense, it was a very exciting period for me. We performed western blotting using samples provided by clinicians. Then we asked what would happen in cell lines.

[00:04:24]So we collected a number of glyobblastoma cell lines and examined them. In most of the cell lines, keratin sulfate was not detected. However, one cell line LN229 [music] did show keratin sulfate modification.

[00:04:36]We then asked what was special about this cell line. To investigate this, we performed comprehensive [music] profiling of glycosal transferase expression. What we found was that LN229 [music] was the only cell line in which all five glycosal transferases required for keratin sulfate synthesis were present.

[00:04:52]So we began to think that perhaps this cell line could be useful for something and we pursued the work further.

[00:04:58]This shows the result of that [music] profiling.

[00:05:01]LN229 clustered within the region corresponding to grade 4 glyobblasto and it was clearly separated from grades 1 to three. At that time in our laboratory we could profile around 180 glycan related genes [music] and glycosal transferases in a single day. These days we do very little work on glycans or glyosol transferases but at that time we were using these technologies extensively for this kind of analysis.

[00:05:23]We then examined the molecule I mentioned earlier ptolanin. [music] We came to the point where we could say with reasonable confidence that keratin sulfate was attached to ptolanin on the cell membrane. So we decided to generate an antibbody. What we could do first was to immunize mice and obtain antibodies.

[00:05:39]Here for the potlanin antibodies [music] shown in this slide, I have written non-casmab. The conventional pottolanin antibodies [music] used around the world are basically of this type. The two panels at the top show flowcytometry data. If the peak shifts, it means the antibody is reacting. When a conventional potlanin antibody is applied to glyobblastoma cells, it reacts very strongly. But at the same time it also reacts very strongly with lymphatic endothelial cells and with kidney derived cells such as 293 cells.

[00:06:08]If someone from a pharmaceutical company sees data like this they will usually say this is going to be difficult.

[00:06:14]In contrast let us look at LPAB 2 which is a castmab against potlanin. This is an antibbody from the time after I became a PI when I was still doing almost everything myself from immunization through [music] screening.

[00:06:25]With this antibbody there is almost no reaction with normal cells. That said, nowadays pharmaceutical companies are extremely strict even about the phrase almost no reaction. If they say isn't there a slight reaction here, it becomes very difficult to move forward. But in a sense, I take some [music] pride in the fact that we ourselves have helped make this field more stringent. This was the first CASMAB we reported in 2014, which is now about 11 years ago. The next thing people inevitably said was, "But you have only shown this in cultured cells, haven't you?" Then they asked us to show it by imuninohistochemistry.

[00:06:58]In this way, the hurdles kept getting higher.

[00:07:01]In the case of potlanin, conventional antibodies stain not only the cancer cell regions but also lymphatic endothelial [music] cells. Since potlanin is also a marker of lymphatic vessels, staining of lymphatic vessels is of course expected.

[00:07:14]In addition, cancer associated fibroblasts or calves are also stained.

[00:07:18]There are of course researchers who are interested in [music] targeting those cells, but what we were aiming for was the cancer cells themselves. If you look at the CASMAB staining image on the left, only the cancer cells are stained while the surrounding tissue is unstained. You can see that there is no brown staining around the [music] surrounding regions. In other words, only the cancer cells are being stained.

[00:07:37]Alvola epithelial cells are also a concern because they express podlanin.

[00:07:41][music] With the non-casmab antibody shown on the right, more than 90% of type 1 alvola epithelial cells are stained.

[00:07:48]However, with the kasmab against pottolanin, the alvola epithelial cells [music] were not stained at all. So, we were also able to demonstrate cancer specificity by imuninohistochemistry.

[00:07:59]The next question is whether it has anti-cancer activity. And here again, people quite naturally said it probably [music] won't have any effect, will it?

[00:08:06]This experiment was carried out by our collaborators and the result was quite striking. When the antibbody was administered, the tumors hardly grew. In particular, in the lung cancer model, tumors usually continue to grow rapidly with a [music] control antibbody. But with this antibbody, tumor growth could be suppressed. However, after about a month, the tumors sometimes began to grow again. This point connects to what I will discuss later regarding therapeutic modalities. It suggests that we will probably need a range of modalities such as ADCs and CARTT cells.

[00:08:35]Even so, in terms of ADCC and CDC activity, this antibody is at least not inferior to the non-casmab antibodies that we had examined up to that point.

[00:08:43]Whenever we show data like this, we are inevitably asked, what about heterogeneity? But even when that question is asked, it is not easy to answer it completely. In the end, the only thing we can do is continue to accumulate data of this kind and present the evidence step by step. We have also examined this antibody [music] in the context of carti cells. Here the blue line represents the non-casmab antibody, the NZID1 series that I mentioned earlier. The red line represents the casmab, the [music] clone we are discussing here. If you look at the period beyond about day 40, the anti-tumor effect is not very different between the two when they are used as carti cells. This tells us one important thing. Converting the casmab into a carti format does not impair its anti-tumor activity. However, the crucial point here is that NZI1 is not a casmab. Therefore, if it were administered systemically as a carti therapy, it could very well [music] damage alvola epithelial cells and cause serious problems. Naturally, a therapeutic product of that kind would not be approved. By contrast, the CASMAB does not react with alvolar [music] epithelial cells. So, we believe that this may finally have the potential to become a genuine therapeutic product. I believe development is still continuing, but once the project has been handed over to collaborators, there are aspects that are no longer under our control. As a basic researcher, that can be rather difficult. It would be a little unsatisfying to end the story here. So, I will now [music] move on to the development of a CASMAB against her, too.