전사 🔗
WEBVTT
00:00:16.480 --> 00:00:20.719
Hello and welcome back to Better Biopharma, the official podcast of Bioprocess Online.
00:00:20.879 --> 00:00:24.719
I'm your host, Tyler Minachello, and on this episode, I'm happy to be joined by Dr.
00:00:24.800 --> 00:00:32.320
Susan Sharfstein, a professor of nanoscale science and engineering at the University at Albany's College of Nanotechnology Science and Engineering.
00:00:32.560 --> 00:00:35.439
Today we're going to be talking about her research in cell line development.
00:00:35.600 --> 00:00:37.439
Susan, thank you so much for being here.
00:00:38.000 --> 00:00:38.880
My pleasure.
00:00:39.280 --> 00:00:47.280
And so for our audience, would you mind uh introducing yourself, giving a bit of background and your research, um, specifically your research in cell line development?
00:00:47.920 --> 00:00:48.880
Certainly, Tyler.
00:00:49.119 --> 00:01:16.400
Um, I'm a biochemical engineer by training, and much of my research in the bioprocessing, biopharmaceutical space has focused on questions of the way I describe it, understanding the problems that people in industry would like to work on, but don't have the time to do, and working with companies to develop new technologies for advancing bioprocessing and biomanufacturing.
00:01:16.480 --> 00:01:22.480
So I work out with a lot of small businesses as well as with larger biopharmaceutical companies.
00:01:22.640 --> 00:01:38.000
And our work on cell line development actually started decades ago, where we received a collection of cell lines all making the same monoclonal antibody with different productivities.
00:01:38.159 --> 00:01:50.719
And what we were really doing there is trying to understand questions like what makes a good cell line, what happens when you increase copy number, how does that affect productivity, and things of that nature.
00:01:50.959 --> 00:02:02.560
And more recently, we've been interested in questions of what are ideal ways or better ways, at least, to isolate high productivity clones.
00:02:02.879 --> 00:02:19.599
That much of the industry has moved into automation, where the idea being that you just screen many more clones and therefore you're able to pull out a high-quality clone, but we're interested in, I would say, somewhat more elegant approaches.
00:02:21.520 --> 00:02:22.159
That's great.
00:02:22.319 --> 00:02:23.120
Thank you so much.
00:02:23.360 --> 00:02:24.400
Thank you so much, Susan.
00:02:24.560 --> 00:02:35.199
And um, would you mind telling us a little bit about the specifically the micro RNA and structurally interface interacting, structurally interacting RNA and your work in clonal cellline development?
00:02:35.840 --> 00:02:36.000
Absolutely.
00:02:36.319 --> 00:02:44.560
More succinctly, would you mind telling our audience a bit about your work on clonal cellline selection using microRNA and structurally interacting RNA?
00:02:44.960 --> 00:02:45.599
Absolutely.
00:02:45.759 --> 00:02:57.840
Uh so this is a collaboration between my lab and a company called SXRNA, where SXRNA stands for structurally interacting RNA.
00:02:58.159 --> 00:03:13.439
And currently, when we do cell line selection, that's typically done either using an antibiotic resistance gene or alternatively a metabolic selection gene like glutamine synthetase or dihydrofolate reductase.
00:03:13.599 --> 00:03:24.879
And so the gene that produces that protein is generally introduced into the cell on the same plasmid that the gene of interest.
00:03:25.120 --> 00:03:28.879
So, for example, your heavy chain or light chain if you're making a monocolor antibody.
00:03:29.039 --> 00:03:41.039
And the challenge of doing that is that you have to make a fairly significant amount of this selection protein, and that creates a burden on the cell.
00:03:41.199 --> 00:03:47.840
And also the selection protein and the gene of interest are not under the control of the same promoter.
00:03:47.919 --> 00:03:56.080
So there's not a direct correlation between the expression of the selection marker and your gene of interest.
00:03:56.240 --> 00:04:15.039
So the idea of our technology is that in the vector that has our gene of interest, we put generally in the three prime untranslated region, but it can also be in other places, uh, a microRNA that is co-transcribed with our gene of interest.
00:04:15.280 --> 00:04:24.560
That microRNA is then processed to it, it's cut out of the trans group to form a microRNA.
00:04:24.639 --> 00:04:50.160
And that microRNA, which is what we call structurally interacting RNA, will interact with an mRNA that has been transfected in separately, and it'll bind at the um 3' end, and it will facilitate the binding of an RNA binding protein that allows for translation.
00:04:50.480 --> 00:05:06.319
So what we anticipate is that the more we express our gene of interest, the more we will express that microRNA, and then the higher the concentration of that microRNA, the more it will bind to this mRNA and then translate the product of that mRA.
00:05:06.480 --> 00:05:12.639
And we envision that primarily that mRNA will be a fluorescent protein.
00:05:12.879 --> 00:05:17.279
So in the presence of the microRNA, we will produce a fluorescent protein.
00:05:17.439 --> 00:05:22.560
That fluorescent protein will then allow us to select for high productivity clones.
00:05:23.040 --> 00:05:29.199
After one to three days, uh the mRNA we've transpected in will go away.
00:05:29.279 --> 00:05:31.680
So you won't make any more of that fluorescent protein.
00:05:31.759 --> 00:05:53.920
So you have a transient selection marker, so you put no burden on the cells when you're actually in the productivity mode, and that we hope that the level of expression of that selection marker, particularly a fluorescent protein, will correlate well, ideally linearly, with the expression of your gene of interest.
00:05:54.879 --> 00:05:56.000
Thank you so much, Susan.
00:05:56.079 --> 00:06:00.079
That is um
Load full transcript
it's a very fascinating approach.
00:06:00.160 --> 00:06:08.160
And I'm curious if you could explain to me, because you know, obviously I'm not uh I'm not too deep in the world of cell line uh selection or engineering.
00:06:08.319 --> 00:06:12.560
How does this contradict or kind of go beyond traditional cell line selection?
00:06:12.959 --> 00:06:18.399
Uh so okay, as I mentioned, right, first of all, you don't have that selection marker that you're dragging along.
00:06:18.480 --> 00:06:21.279
But the other advantage is that it's much faster.
00:06:21.519 --> 00:06:39.680
So um typically when you do antibiotic selection, you put in your antibiotic or your uh methotrexate or whatever, and it takes somewhere between one and three, maybe four weeks, for you to be able to isolate your primary candidate clones.
00:06:40.000 --> 00:06:49.519
We envision, um, and we've done this um with negative selection, so slightly different technology, that you should be able to isolate candidate clones in about 10 days.
00:06:49.680 --> 00:06:57.279
And potentially, even if we're using this in conjunction with the transposon technology, that you could potentially even do it faster.
00:06:57.439 --> 00:06:59.839
So it will save you a lot of time.
00:07:00.079 --> 00:07:20.399
And of course, the challenge, right, that with the selection markers, like as you increase your selection stringency, hypothetically you get higher productivity clones, but there's no direct correlation between your selection strengthening and your um expression of your gene of interest.
00:07:20.720 --> 00:07:29.279
So we envision that um we should be able to say, ah, these are my highest fluorescing clones, these are my highest fluorescing cells, right?
00:07:29.439 --> 00:07:33.199
They are my best candidates for good productivity clones.
00:07:33.360 --> 00:07:40.879
So um better speed and better relationship between selection, your selection marker and your expression levels.
00:07:41.279 --> 00:07:42.720
Yeah, that makes sense.
00:07:42.800 --> 00:07:44.160
It makes intuitive sense too.
00:07:44.240 --> 00:07:52.480
I mean, I'm a visual kind of learner, so I think I much prefer to see visually what cells um or clones are or the highest producing, right?
00:07:52.560 --> 00:07:55.360
It like to me seems like an intuitive way to select.
00:07:55.519 --> 00:08:00.959
Um and I'm just kind of curious of if there what precipitated this research, right?
00:08:01.040 --> 00:08:06.079
Was there is this built on existing research or what inspired you to pursue this?
00:08:06.160 --> 00:08:08.800
Uh and I have some follow-ups, but okay.
00:08:08.879 --> 00:08:16.639
So the technology, the structurally interacting RNAs, um, is actually comes out of my colleague Scott Tennenbaum's lab.
00:08:16.879 --> 00:08:19.600
And he he's an RNA biologist.
00:08:19.759 --> 00:08:27.279
And so they had developed this technology and they were really interested in using it for things like um killing cancer cells
and things like that.
00:08:27.519 --> 00:08:45.120
And one of my colleagues in Nicole Borth in Austria invited him to come and speak at a workshop that um she was gonna that she was uh having, and I was gonna speak as well at a an invited conference.
00:08:45.440 --> 00:08:52.720
And so I get my right, it was a long time ago, but my recollection is like that I guess I introduced the two of them.
00:08:52.879 --> 00:09:03.919
And so the two of them started talking about his technology, and Nicole was very interested in microRNAs, and so they started discussing well, could we use this technology for cell line selection?
00:09:04.080 --> 00:09:06.720
So that was actually how the impetus for it was.
00:09:06.960 --> 00:09:13.120
And then uh so ultimately we had a multi-way collaboration between my lab and Scott's lab and Nicole's lab.
00:09:13.200 --> 00:09:22.320
And then more recently, um, as the technology has evolved, um, it's a little bit easier to work with people who are down the hall from you and then people on the other side of the ocean.
00:09:22.399 --> 00:09:32.480
Uh, but we did actually um publish a very nice paper out of all three of our labs, oh, about seven, eight years ago, something like that.
00:09:33.120 --> 00:09:34.080
That's awesome.
00:09:34.320 --> 00:09:35.919
And so you you mentioned Dr.
00:09:36.000 --> 00:09:37.200
Scott Tinnenbaum and Dr.
00:09:37.279 --> 00:09:42.559
Nicole Barth in your interview with my colleague John O'Connell, your recent interview for bioprocess online.
00:09:42.960 --> 00:09:52.960
And in that interview, you also mentioned how biologists think differently about you think that you said that biologists think differently than bioprocessing experts and people.
00:09:53.120 --> 00:09:56.240
And I was hoping you could kind of explain that a little more for me.
00:09:56.320 --> 00:09:57.200
Um, how so?
00:10:07.120 --> 00:10:22.159
So one of the first things like that they they, I guess, Scott, and actually there was an earlier company that they envisioned, well, it was good enough to just like have a population of cells, and that would be fine for bioprocessing.
00:10:22.240 --> 00:10:44.480
And what they didn't really understand is that the FDA mandates clonality in biomanufacturing and mandates clonality to the degree that you know most companies are doing two levels of clonal isolation to ensure that they have a clonal cell and then you know they subclone again.
00:10:44.720 --> 00:10:56.720
So it and this is actually one of the reasons why uh flow cytometry is an ideal tool because as opposed to like the other option would be magnetic bead selection.
00:10:56.799 --> 00:11:02.080
But with magnetic bead selection, you're you're just pulling down positive cell lines.
00:11:02.159 --> 00:11:04.799
It's hard to get them individualized.
00:11:04.879 --> 00:11:22.159
Uh, whereas flow cytometry, you're you have one cell at a time and you can easily quantitate the level of fluorescence as opposed to if you were to do magnetic beads on a different cell surface marker where you're um you would just be like over threshold to pull down.
00:11:22.320 --> 00:11:31.600
So um, you know, it's a different perspective because the whole idea how that clonality is so important in biomanufacturing.
00:11:32.799 --> 00:11:33.200
Yeah.
00:11:33.360 --> 00:11:41.360
Um we actually have a by the time this airs, uh, there will be a bioprocess online live event that I have I will have done about cell line development.
00:11:41.440 --> 00:11:47.679
And um I have a couple questions for you about that, but all that to say, clonality is going to be a big talking point.
00:11:47.840 --> 00:11:59.759
Um and I remember when we spoke earlier, you were saying that the next kind of order of business is to see how you can get this microRNA selection to to hold up in a clonal in clonal selection.
00:11:59.919 --> 00:12:03.679
Uh again, like you said, regulators really look for that clonality.
00:12:03.840 --> 00:12:05.519
Um, so how's that going?
00:12:05.679 --> 00:12:07.039
How far away are you from that?
00:12:07.200 --> 00:12:10.080
Like kind of what what are the obstacles there?
00:12:10.159 --> 00:12:11.279
I'm just curious to hear.
00:12:11.679 --> 00:12:20.240
Um also some of the challenges that we have faced have been uh and the first one was well, where do you put the microRNA?
00:12:20.399 --> 00:12:21.840
Like, so where in the vector?
00:12:22.000 --> 00:12:30.960
And it turns out uh that you we get different results depending upon where in the vector we put the microRNA.
00:12:31.360 --> 00:12:32.960
So that was challenge number one.
00:12:33.279 --> 00:12:47.440
The other challenge that we have faced has been that the fluorescent RNA that or the RNH for the fluorescent protein, uh it has been a little bit leaky.
00:12:47.600 --> 00:12:54.240
So, I mean, it shouldn't fluoresce if the microRNA isn't present, and we are seeing some of that.
00:12:54.399 --> 00:13:03.120
So, this is part of the reason why we we are exploring other alternative RNA-based technologies that would have the same concept.
00:13:03.279 --> 00:13:17.440
I mean, there's the real key to the technology in my mind is the fact that the microRNA is co-transcribed with your gene of interest so that the microRNA level correlates with your gene of interest.
00:13:17.519 --> 00:13:24.000
Um, and that then the real thing we need to make sure is that we can demonstrate that correlation.
00:13:24.159 --> 00:13:28.879
So we've also been doing work where we have been quantitating the microRNA.
00:13:29.039 --> 00:13:52.879
So we've been looking here to say, okay, the cells that make a whole bunch of our gene of interest, can we isolate the microRNA and do quantitative real-time PCR on the microRNA so that we have an independent measurement besides the fluorescence of that is produced by the mRNA that should be turned on in the presence of the microRNA?
00:13:53.919 --> 00:13:56.320
It would work better
with a picture, I realize.
00:13:56.639 --> 00:13:57.519
No, yeah, it's fine.
00:13:58.159 --> 00:14:01.919
I think our audience is is is keeping up with you there.
00:14:02.559 --> 00:14:04.399
Um that's that's great.
00:14:04.480 --> 00:14:06.399
Thank you for breaking that down for me.
00:14:06.639 --> 00:14:17.679
And as it relates to your research, but also in general, how do you think, because you know, you're obviously an academic, which we don't have many on the show, but it's always a joy, a treat for me to have some academics on.
00:14:17.840 --> 00:14:27.840
Um I'm curious from your point of view, how do you think cell selection in an academic or research setting differs from that of, you know, in an industrial application?
00:14:29.519 --> 00:14:35.440
So um, you know, academics typically do non-clonal cell selection.
00:14:35.519 --> 00:14:44.799
I mean, you know, that well, not to say they never do, uh, you know, because it's really the question, like if you're making a cell line, well, what's the question that you're going to answer?
00:14:44.960 --> 00:15:02.000
You know, the reality is part of the reason that you don't have a lot of academics on your um podcast is because there aren't very many academics, particularly in the US, working in biomanufacturing um in part, um, you know, because there's not a lot of funding for it, uh, which is frustrating.
00:15:02.159 --> 00:15:32.720
Uh but you know the but I would say, you know, like my collaborator, you know, um Scott Tenenbaum, you know, in looking at this SXRNA technology, he has been working um with uh a um biopharmaceutical tool manufacturing company, and they are looking at putting his technology into a kit for academics because it might be a good deal easier.
00:15:32.799 --> 00:15:46.559
And if you're not, if all you care about is getting good productivity and you're not concerned about clonality, stability, that it may work very nicely for people who want to do cell line selection for non-biomanufacturing, you know.
00:15:46.639 --> 00:15:53.840
So if you want to put in, you want to engineer cells to make something or to have different properties, so it may have a lot of utility there.
00:15:55.279 --> 00:15:56.399
Yeah, that's great.
00:15:56.559 --> 00:16:17.039
And typically something like what you're working on, right, with the the micro RNA or structurally interacting RNA uh for selection, how long do you advancements like that or technologies like that typically take to translate into kind of biomanufacturing that sector, right?
00:16:17.200 --> 00:16:21.840
Like from research, you know, what are the kind of because it's not like a therapy you're developing, right?
00:16:21.919 --> 00:16:33.840
You're developing a tool that that hopefully one day people in that are developing therapies can can leverage in in their biomanufacturing and to accelerate the whole the whole thing top to bottom.
00:16:34.080 --> 00:16:41.679
So what's what is the typical timeline for something like that where you're developing a tool that biomanufacturers can use as opposed to like an FDA-proof therapy?
00:16:41.919 --> 00:16:47.600
Yeah, so we are hoping, you know, in the one to three years kind of time frame, if things work.
00:16:47.919 --> 00:16:58.320
I mean, that would be what we, you know, and I think the company's exit strategy, and you know, I so on, I do not hold equity in the company.
00:16:58.480 --> 00:17:00.000
Um, I don't get royalty.
00:17:00.080 --> 00:17:11.759
So just to make it clear that I have no financial interest in the company, um, you know, their vision for an exit strategy is absolutely to license to somebody like Milliport Sigma or Thermo Fisher.
00:17:11.920 --> 00:17:26.400
And so really it's about, you know, figuring the finding the right partner, licensing the technology, and then they would, you know, and then the tool manufacturer, you know, would be the one who would be promoting the technology.
00:17:26.480 --> 00:17:36.000
And you know, you definitely see, you know, what there are technologies, you know, like that have become incredibly widespread.
00:17:36.079 --> 00:17:49.599
You know, I remember um hearing the developer of Wave of the Wave Bag at an AI ACS meeting decades ago, and I was like, uh I wasn't excited.
00:17:49.680 --> 00:17:50.799
Let me let me say that.
00:17:50.960 --> 00:17:57.759
Um but that's an academic perspective because it's like, well, you know, it I mean, it was it was a blood bag on a rocket cable, right?
00:17:57.920 --> 00:18:06.079
And now you go to every single biotech company, biopharmaceutical company, and they have wave bags.
00:18:06.400 --> 00:18:11.279
And what you're seeing, right, is you know the adoption of single-use technology.
00:18:11.359 --> 00:18:27.279
So it really wasn't so much the you know, the rocking wave bag that turned out to be the phenomenal technology, it was the whole introduction of single-use technologies that has actually really revolutionized the industry.
00:18:27.759 --> 00:18:28.400
Yeah.
00:18:28.960 --> 00:18:30.480
It just goes to show you, right?
00:18:30.559 --> 00:18:35.680
It's the um, it's something that might be overlooked in the moment that could change the uh the paradigm in the future.
00:18:36.000 --> 00:18:36.480
Absolutely, yeah.
00:18:36.640 --> 00:18:37.599
It's a good reminder.
00:18:37.920 --> 00:18:38.720
That's awesome.
00:18:38.880 --> 00:18:42.880
And I love how you're just kind of involved in this project for the love of the game, right?
00:18:42.960 --> 00:18:45.519
You you clarified you don't have any financial stake in the compass.
00:18:47.200 --> 00:18:47.920
Love to see that.
00:18:48.240 --> 00:18:50.160
I like working with companies.
00:18:50.319 --> 00:18:58.160
Um, you know, I've worked with big and small companies um because it really makes sure that your research is relevant.
00:18:58.319 --> 00:18:58.480
Okay.
00:18:58.880 --> 00:19:07.279
Uh and it's you know, the the graduate student
working on this project uh is really, really entrepreneurial.
00:19:07.440 --> 00:19:13.759
So I think, you know, I told you like the alternative mRNA technology that we can't, or the microRNA technology we can't talk about.
00:19:13.920 --> 00:19:25.519
Um if that works out, I expect that he will start a company based on that technology, uh, which a I I I think could be really cool.
00:19:25.599 --> 00:19:31.920
I I've had um I had one other student spin out a technology, not for biomanufacturing, um, out of my lab.
00:19:32.079 --> 00:19:34.000
Um so it's it's a lot of fun.
00:19:34.160 --> 00:19:38.480
Um I think I got for that technology a couple of hundred dollars in royalty.
00:19:38.559 --> 00:19:40.400
So I'm like, I don't quit my day job on that.
00:19:40.559 --> 00:19:47.359
But uh but it is, I mean, you know, because it's like if you can see these things come to fruition, it's really, really fun and exciting.
00:19:47.680 --> 00:19:48.480
Yeah, absolutely.
00:19:48.559 --> 00:19:51.839
And I can imagine it's very satisfying as well, right?
00:19:51.920 --> 00:19:54.559
Like to see your your work have a tangible impact.
00:19:54.720 --> 00:20:00.079
Um in some indirect way, um, touch patience, right?
00:20:00.160 --> 00:20:00.640
Down the line.
00:20:00.880 --> 00:20:01.519
That's terrific.
00:20:01.599 --> 00:20:02.559
Thank you, Susan.
00:20:02.720 --> 00:20:12.799
So, like I said before, um, I'm getting ready for a bioprocess online live event around cell cell line development selection and um specifically next generation cell line development.
00:20:12.880 --> 00:20:15.279
I will link that in the description of this episode.
00:20:15.440 --> 00:20:20.559
But by the time this comes out, that will be in the past and available for people to watch.
00:20:20.720 --> 00:20:22.160
So I wanted to ask you a couple of questions.
00:20:22.240 --> 00:20:28.799
I'm planning to ask that panel, and um, and then we'll see how it how it compares to their answers and once once they're both available.
00:20:29.039 --> 00:20:40.079
But to start, um I want to hear from you on what technologies or advancements you think are changing cell line development in our in the present day for the better.
00:20:40.240 --> 00:20:43.200
Um, what what research going on are you excited by?
00:20:43.279 --> 00:20:46.400
Uh I know you attend industry conferences pretty pretty regularly.
00:20:46.559 --> 00:20:55.440
Is there any any particular team or or lab that company doing work that you're just you think it's the next big thing you're excited about?
00:20:56.240 --> 00:21:07.119
So definitely transposons, uh, you know, because transposons uh are a much easier way to get DNA into the cells and less randomly.
00:21:07.200 --> 00:21:13.519
And the biggest challenge with the old-style cell line development is you know that the random integration.
00:21:13.680 --> 00:21:31.680
And one of the things that you know we working with SXRNA envision is that we would ultimately
partner with a transposon company uh to use to introduce our uh vectors into cell lines.
00:21:31.839 --> 00:21:42.880
And the other thing is uh site-directed integration, and that's a lot harder, but it has huge potential.
00:21:43.039 --> 00:21:58.799
I was at a conference and I remember hearing uh folks from Pfizer and say that so they have a site uh you know, site directed integration, and that they screen a dozen clones.
00:22:00.160 --> 00:22:08.720
Uh I was like, you know, not 120, not 1200, but a dozen quilts, maybe 20, right?
00:22:08.960 --> 00:22:13.359
And so that really does move things a lot.
00:22:13.519 --> 00:22:34.960
The challenge will be as you get to more challenging modalities where you have to vary potentially the copy number of your of one chain versus another, um, then you have to do more engineering if you have site-directed integration than if you just put everything in on one vector.
00:22:35.119 --> 00:22:37.680
Um, but I think that that that has potential.
00:22:37.839 --> 00:22:41.519
I and so I I think that we will see more of those.
00:22:41.599 --> 00:23:14.960
You know, the flip side is that the molecules that we are making are becoming more and more challenging as we move to buy and tri-specifics and also moving towards um antibiotic drug conjugates because it's not then it's not just enough to get the molecule made, but you have to have the correct conjugation and make sure that you have the conjugation levels um where they need to be, because otherwise you aren't controlling the amount of the drug that you're delivering.
00:23:16.160 --> 00:23:16.799
Yeah.
00:23:17.119 --> 00:23:18.000
That's great.
00:23:18.240 --> 00:23:24.480
Um sorry, you said transposons as opposed to what was the traditional kind of route of transfection.
00:23:24.799 --> 00:23:27.839
Oh, well, oh random integration, right?
00:23:27.920 --> 00:23:39.519
Um, where you know you will get a what we what we've observed in general is when you do random integration, you do get most of the DNA into a single site, into a single locus.
00:23:39.839 --> 00:23:42.480
Uh, but it could be anywhere.
00:23:42.720 --> 00:23:58.880
What we found when we screened all of these clones that we got a number of years ago is that the high productivity ones that we saw, they were all located in the telomeric region of the Chinese hamster ovary chromosome one.
00:23:59.200 --> 00:24:13.920
Uh that's probably because we had pre-selected clones that probably you get much more random integration, but the good ones tend to be um near in the telomeric region of chromosome one.
00:24:14.160 --> 00:24:20.079
But even then, you know, you will get um many copies, some of them won't be linearized the same way.
00:24:20.240 --> 00:24:32.960
They will be um you know you will get sometimes you will get silencing, and you know, what particularly when you do a lot of gene amplification, then you can get have like hundreds of copies, but you don't know how many of them are active.
00:24:33.119 --> 00:24:43.519
Whereas you know, with the transposon technology, they all wind up in specific loci and um at a generally lower copy number.
00:24:43.759 --> 00:24:48.559
Um, and that so that tends to give you more reproducibility.
00:24:49.680 --> 00:24:50.319
Thank you.
00:24:50.480 --> 00:25:02.480
I appreciate it, and thank you to the uh the cell line development scientists out there bearing with me in my naive questions, but uh that inner biology student in me, um curiosity starts coming up.
00:25:02.640 --> 00:25:04.880
Um thank you for explaining that, Susan.
00:25:05.039 --> 00:25:10.079
And a couple more questions here that I'm going to present to our panel at that event.
00:25:10.240 --> 00:25:12.000
I'm curious to hear from you.
00:25:12.319 --> 00:25:19.680
What do you think most teams overlook when it comes to cell line selection or cell line development?
00:25:20.000 --> 00:25:27.119
And maybe you'll have an interesting opinion or take on this, being that you're a little, you know, a degree away from industry, you know, you're more in research.
00:25:27.200 --> 00:25:30.799
But I'm just curious because you've attended a lot of these same conferences I have.
00:25:31.519 --> 00:25:44.720
So I think one of the bigger challenges, and this is not so much for innovators, um, but for biosimilar manufacturers, it's about the product quality attributes.
00:25:45.039 --> 00:25:57.599
Because we've had we've just done work where we've looked at multiple clones, all the same transfaction method, uh, same selection, uh, and same product, of course.
00:25:57.920 --> 00:26:07.839
And when you look at the glycosylation patterns on the antibodies, you find that they differ dramatically.
00:26:08.079 --> 00:26:21.839
And that's a big deal because when you're making a biosimilar, you need to, you don't have to be identical, of course, because it's a biosimilar, but you need to relatively closely match the innovator glycosylation pattern.
00:26:22.000 --> 00:26:34.000
And the challenge, and you know, to be fair to the biosimilar manufacturers, is we have no idea why one is better than, like, you know, what gives us that variation.
00:26:34.319 --> 00:26:42.640
We have looked at expression of the um glycosylation enzymes, and that doesn't seem to make a difference.
00:26:42.799 --> 00:26:57.200
So I we think that a lot of it has to do with things like precursor availability, uh, and that may have to do with underlying aspects of cell metabolism that occur when you do selection, and we don't really understand.
00:26:58.880 --> 00:26:59.200
Okay.
00:27:00.160 --> 00:27:00.720
Got it.
00:27:00.880 --> 00:27:01.599
Thank you.
00:27:01.920 --> 00:27:03.680
Thank you for breaking that down.
00:27:04.240 --> 00:27:10.880
Um in a kind of follow-up to that, what do you think are the most common mistakes in cell line selection or development?
00:27:11.599 --> 00:27:17.920
Um, I mean, I think one of the biggest mistakes is focusing solely on
productivity.
00:27:18.160 --> 00:27:32.799
Um that not the title's not important, but you know, um, if you don't get stability, product quality, uh, and um good growth, uh then it's challenging, right?
00:27:33.039 --> 00:27:44.640
So uh uh and again, you know, I I do I get I do some expert witness consulting, so I get to actually see like sometimes how biosimilar manufacturers go about things.
00:27:44.799 --> 00:28:07.119
And a lot of times what happens is they'll pick a clone and then they'll discover that their clone doesn't meet their critical quality attributes, and then they have to go back and redo their process to improve their critical quality attributes, where they might not have had to do that had they looked at other clones.
00:28:09.839 --> 00:28:10.160
Okay.
00:28:10.480 --> 00:28:23.039
So at the you're saying that they select they're too focused on productivity, and that in turn can change whether it's the glycoform or just some some kind of CQA about the product.
00:28:23.359 --> 00:28:24.640
Oh, interesting.
00:28:24.960 --> 00:28:29.119
Have you seen that uh play out with companies before?
00:28:29.200 --> 00:28:30.960
It sounds like you you've seen that first thing.
00:28:31.200 --> 00:28:31.359
Right.
00:28:31.440 --> 00:28:36.480
So, I mean, you know, again, I you know, I can't obviously talk about things in great detail that I do in expert witness consulting.
00:28:36.559 --> 00:28:55.519
Um, but I definitely, you know, get to see, you know, what how companies do their cell line development, and then I do, you know, because I mean one of the things you find, you know, when you do patent litigation is that much of patent litigation is um has to do with cell culture processes.
00:28:55.680 --> 00:29:04.240
And so you'll see that, you know, a biosimilar manufacturer added compound X to get their critical quality attribute to be something.
00:29:04.319 --> 00:29:13.920
And then the problem is that the concentration of compound X that they've added um infringes on the patent that the innovator has.
00:29:14.079 --> 00:29:22.640
Um, so now they're in this situation where that the only way they can make their product and meet the CQAs um infringes on the innovator patent.
00:29:22.720 --> 00:29:32.400
So that means that they're gonna have there will be increased cost to them in launching that product uh when they as they settle the litigation.
00:29:34.160 --> 00:29:34.960
Got you.
00:29:35.279 --> 00:29:36.640
Got you, thank you.
00:29:36.960 --> 00:29:39.839
Um this this question just popped into my head.
00:29:39.920 --> 00:29:49.359
I want to ask you while while we're talking for my event that I'm getting ready for, which I hope uh I could send you a link after this, and hopefully you could attend.
00:29:49.519 --> 00:30:07.680
But for that for that event, what's something you think I should ask them, or something you should you think I should infuse into the conversation or bring forward during that discussion in terms of what do you think people should be talking about more when it comes to cell line development?
00:30:11.920 --> 00:30:13.359
Throw your curveball here.
00:30:13.759 --> 00:30:32.720
So well, okay, so the other thing, you know, that that is significant are way okay, so host host cell line development is a big one, you know, particularly you know that most of the innovator companies have their own host cell lines.
00:30:32.960 --> 00:30:41.759
And so the role in designing your host cell line for a process, you know, I think is increasingly significant.
00:30:42.079 --> 00:30:49.119
You may be aware that uh I don't know, two years ago, something like that, two, three years ago, time blurs when you get old.
00:30:49.200 --> 00:31:01.359
Um, Amgen announced that they had, for the first time in 20 years, redesigned their host cell line for improved productivity.
00:31:01.440 --> 00:31:08.720
And they've got, well, improved productivity and to function in the context of process intensification.
00:31:08.880 --> 00:31:15.359
And by their VP of research, you know, was at a conference and said basically, they're getting 25 grams per liter.
00:31:15.519 --> 00:31:30.000
Now, that's probably not for everything, but um thinking about how your host cell lines um could improve your cell line development, I think that's significant, particularly for innovators that often have better control over that.
00:31:30.079 --> 00:31:39.920
I think you know, for many biosimilar manufacturers, they're working with um CDMOs and so on and so forth, you know, to develop their cell line.
00:31:40.000 --> 00:31:41.680
So they're in a little bit different position.
00:31:41.839 --> 00:31:45.680
The other thing is the whole question of cell line editing.
00:31:45.759 --> 00:31:53.119
And you know, this has been somewhat of a fraught topic because of the IP around CRISPR, right?
00:31:53.359 --> 00:32:09.599
But there are now cell line, you know, cell editing technologies out there that don't have fall into the CRISPR, you know, Berkeley, Broad Institute conundrum.
00:32:09.680 --> 00:32:21.519
Um, and so I think that there are some real possibilities there for thinking about how you might relatively straightforwardly engineer cell lines um for better growth or productivity or things like that.
00:32:21.680 --> 00:32:26.640
That was much harder before we could do cell line editing.
00:32:28.720 --> 00:32:29.440
That's great.
00:32:29.680 --> 00:32:30.559
Thank you.
00:32:30.799 --> 00:32:33.680
Um, and just two last questions here.
00:32:34.000 --> 00:32:44.559
One, looking at your scope of research, the work you've done in your career, how do you hope long term that your research impacts cell line development more broadly?
00:32:44.880 --> 00:32:49.440
And a follow-up to that, is there anybody you you mentioned um Dr.
00:32:49.599 --> 00:32:56.720
Tenenbaum before, but are there others that you're working with who you want to shout out or that you're working with in conjunction
with the in in the industry?
00:32:57.359 --> 00:32:58.400
You want to shout out?
00:32:58.720 --> 00:32:59.200
Yeah, okay.
00:32:59.279 --> 00:33:12.799
All right, so I would have to say, and this is not what you're expecting, but um my biggest impact is the students I train who will go out to work for um the biotech and biopharmaceutical companies.
00:33:12.960 --> 00:33:27.920
And you know, more recently, my not that I don't still do biomanufacturing research, but um my research has taken other directions, uh, in part because it's easier to get funding on things that are sexier than biomanufacturing, which is unfortunate.
00:33:28.079 --> 00:33:43.680
It is one of the uh real weaknesses, I think, in the research enterprise in this country, uh, that, you know, it's it's much in Europe, it's much, much easier to train students to do biomanufacturing.
00:33:43.839 --> 00:33:48.480
Uh and so I think that that's but I I have you know a good group of students out there.
00:33:48.720 --> 00:33:51.920
And I so I think that that's a big impact.
00:33:52.160 --> 00:34:09.039
The other thing, you know, that I think is an impact is to think biologically, to really sort of like to not think about cells as a black box, but to really dig in to the cell physiology.
00:34:09.199 --> 00:34:21.519
And that's certainly work that I have done that um my you know, my academic colleagues, you know, Mike Bietenbaugh, Calvin Lee, uh Sarah Harcombe to some extent, Ann Robinson.
00:34:21.599 --> 00:34:31.360
Um, but you know, there aren't very many of us, uh, and that that's unfortunate uh because there really should be more to do this.
00:34:31.599 --> 00:34:35.519
And then uh there was one more question.
00:34:36.639 --> 00:34:42.559
Oh, um I think that's I guess I asked if you're working with anybody directly in the industry that you'll use the shutter.
00:34:43.360 --> 00:34:43.840
Yeah, right.
00:34:44.000 --> 00:34:50.960
So we're working with we're working with one other small business, and this is not on the cell line development side.
00:34:51.199 --> 00:35:07.119
This is actually on the um antibody engineering side, and they are doing they're using artificial intelligence to identify molecules that have a good potential for high levels of expression.
00:35:07.280 --> 00:35:17.519
And because that's one of the real challenges, particularly with um you know uh biospecifics and things like that, is that the levels of expression are so low.
00:35:17.760 --> 00:35:19.840
And you know, we know a lot.
00:35:19.920 --> 00:35:33.199
Uh we we can you know we can look at a um protein sequence, and we have these days with AI, you know, some reasonable insight into um its folding and how well it will fold.
00:35:33.280 --> 00:35:39.280
Uh, and that's necessary, but it's not sufficient to get good expression.
00:35:39.440 --> 00:35:42.159
And so I'm really excited about this.
00:35:42.320 --> 00:35:45.920
Uh, that you know, we do we we finished our uh phase one grant.
00:35:46.079 --> 00:35:48.480
They've put in their phase two grant.
00:35:48.559 --> 00:35:57.440
Uh and so I might you know I'm fingers crossed that we will get the phase two award uh and that we'll get to do some really cool, exciting things with them.
00:35:57.920 --> 00:35:58.639
It is exciting.
00:35:58.719 --> 00:36:00.079
I'm crossing my fingers for you.
00:36:00.159 --> 00:36:01.519
I hope that happens.
00:36:01.760 --> 00:36:02.960
Um and Dr.
00:36:03.119 --> 00:36:04.719
Sharstein, last question here.
00:36:04.880 --> 00:36:11.039
It's a question I ask every guest on the show, and that is big or small, in your opinion, how do you think we can better biopharma?
00:36:11.840 --> 00:36:12.800
Ah, okay.
00:36:13.119 --> 00:36:24.320
So in my mind, the biggest thing we need to do is we need, and I'm wearing my my Nimble clothes um because it's cold in my office.
00:36:24.480 --> 00:36:28.239
Um John Erickson at Nimble.
00:36:28.559 --> 00:36:38.079
I mean, his number one goal is to bring down the cost for biologics so we can make them available to the entire world.
00:36:38.159 --> 00:36:48.960
Uh, because these are amazing life-saving drugs that are only available to 20% of the world's population, something like that, right?
00:36:49.119 --> 00:36:51.360
And and that's and that's not easy.
00:36:51.440 --> 00:36:58.239
I mean, that's you know, gonna have to require really, really thinking outside the box.
00:36:59.519 --> 00:37:00.159
Yeah.
00:37:00.639 --> 00:37:01.280
Absolutely.
00:37:01.440 --> 00:37:05.679
I think that's a great, great response and a great way to end this interview.
00:37:05.760 --> 00:37:07.119
Thank you so much for being here, Dr.
00:37:07.199 --> 00:37:09.599
Sharpstein, for our audience at Better Biopharma.
00:37:09.679 --> 00:37:18.320
I'm Tyler Minachello, and for much more on cell line development and every aspect of bioprocessing, head on over to Bioprocess Online to see everything we have to offer.
00:37:18.480 --> 00:37:18.880
Thank you so much.