Knockout cell lines and their role in confirming antibody specificity

On-demand webinar

Summary:

Knock-out (KO) cells play an important role in demonstrating antibodies’ specificity to their targets and represent true negative controls when it comes to validation. But what should we be looking out for when it comes to using KOs in our research?

Join Will Howat, Director of Antibody Validation at abcam, as he explores the role of KOs in confirming antibody specificity and the challenges we may face as life scientists.

Key discussion points:

A basic understanding of knock-out cells

Some of the pitfalls when using knock-out cells

Discover how we utilize knock-out cells in our catalog:

We have over 2,800 KO-validated antibodies in our catalog, but why is knock-out validation considered the gold standard for antibody specificity?

The CRISPR revolution and its impact on the production of knock-out cell lines

The importance of using your knock-out in the relevant application

The benefit of purchasing abcam cell lines and lysates

Knockout Cell Lines and Their Role in Confirming Antibody Specificity

Video Transcript

00:00 - 00:13: Hello, welcome to Abcam’s webinar, “Knockout Cell Lines and Their Role in Confirming Antibody
00:13 - 00:17: Specificity”.
00:17 - 00:22: During this webinar, we will cover the following three objectives: gain a basic understanding
00:22 - 00:28: of knockout cells, learn about some of the pitfalls when using knockout cells, and discover
00:28 - 00:32: how Abcam utilizes knockout cells in their catalog.
00:32 - 00:33: Thank you and I hope you enjoy.
00:33 - 00:40: Hello, welcome to our webinar on knockout cells and their role in confirming antibody
00:40 - 00:41: specificity.
00:41 - 00:46: I’m Dr. Will Howat, and I’m the Director of Antibody Validation and Characterization
00:46 - 00:47: at Abcam.
00:47 - 00:52: So to kick off, let’s talk a little bit around what knockouts are.
00:52 - 00:58: Now, if you’re as old as I am, you’ll know that knockouts have existed for quite a while
00:58 - 01:00: and there are a variety of ways of doing it.
01:00 - 01:07: But more recently, the CRISPR-Cas9 technology came along, and that is what’s revolutionized
01:07 - 01:14: the knockout field going forward, and certainly now it’s the ubiquitous technology that’s
01:14 - 01:16: used for creating knockouts.
01:16 - 01:23: Now, it’s a very simple process that is involved here, where there’s a guide RNA guiding the
01:23 - 01:32: Cas9 enzyme to the area that you want to edit, and that enzyme cuts the DNA, and then through
01:32 - 01:39: a non-homologous end joining, you get a cut on the DNA, or you can actually insert DNA
01:39 - 01:40: if you wish to.
01:40 - 01:46: And in that process, you get an error-prone repair, which more likely than not will
01:46 - 01:53: create a stop codon or will create a mutation in the genome, which will mean that the protein
01:53 - 01:54: will no longer be expressed.
01:54 - 02:00: Now, when we start out on this technology, we will be using a haploid cell line, as shown
02:00 - 02:05: on the right; that’s advantageous in the fact that there is only a single copy of the gene.
02:05 - 02:09: It also means there’s only a single copy of the gene to knock out.
02:09 - 02:14: However, you know, as we’ve gone through this technology, it’s preferable to be using
02:14 - 02:19: a diploid cell line, and that is what we in Abcam use more frequently.
02:19 - 02:24: Those are the cells that are actively being used in cell culture, and a variety of different
02:24 - 02:31: backgrounds will be picked to then knock the relevant protein of interest out.
02:31 - 02:36: Now, in this case, it’s a little bit more complicated, and you’ve got two alleles to
02:36 - 02:41: knock out, and therefore you want to make sure you knock out both of those, but this
02:41 - 02:45: is the more likely scenario that we would do.
02:45 - 02:51: So to move forward, what we then look at is a set of knockouts.
02:51 - 02:58: All right, so these are examples of where you can have a knockout in different types
02:58 - 02:59: of applications.
02:59 - 03:04: I think it’s clear that you have to use the knockout in the way that you would be using
03:04 - 03:06: it in your own lab.
03:06 - 03:15: So if western blot is your final application, then that is where you and your own scientific
03:15 - 03:19: expertise will be where you want to have that knockout existing.
03:19 - 03:24: But if, like on the right-hand panel, you are more interested in an immunocytochemical
03:24 - 03:31: type analysis, then you will want to be able to use your knockout to demonstrate a knockdown
03:31 - 03:34: of that protein in the immunocytochemistry.
03:34 - 03:37: So these are two examples that we have in Abcam.
03:37 - 03:40: In the middle, you’ve got NF-kappaB.
03:40 - 03:46: As you can see here, you have a wild type, and that wild type is the parental cell.
03:46 - 03:51: You can call it parental, you can call it an isogenic control, a number of names.
03:51 - 03:53: We call it a wild type.
03:53 - 03:58: And in the middle lane here, you’ve got a knockout, and you can see in this Western Blot,
03:58 - 04:04: you have presence of the protein here within the wild type, and absence of the protein
04:04 - 04:09: present in the knockout using that anti-NF-kappaB antibody.
04:09 - 04:14: We would also recommend that you throw in a couple of additional controls.
04:14 - 04:23: In our case, we’re using two positive controls here, a HeLa cell line, and A431, representing
04:23 - 04:25: different levels of expression of that protein.
04:25 - 04:31: HeLa is very strong, A431 is much weaker, although we also recommend that you run a
04:31 - 04:36: loading control to help to demonstrate that you’re actually loading the correct amount
04:36 - 04:37: of protein.
04:37 - 04:42: And that was where we would be in a western blotting scenario.
04:42 - 04:44: But as I said, you can also use other methodologies.
04:44 - 04:51: This is the immunocytochemistry for a Ki67 recombinant RabMab, and in this case, where
04:51 - 04:56: you can see you’ve got a positive here, which is the wild type, and Ki67 expressing cells,
04:56 - 05:00: and the knockout on the bottom panel, Ki67 null cells.
05:00 - 05:05: So again, in this case, we haven’t got additional control cells in this particular panel, but
05:05 - 05:10: you can clearly see that the protein has been knocked out.
05:10 - 05:15: But it can apply to a number of different applications, not just western blot and immunocytochemistry;
05:15 - 05:16: it can apply to flow cytometry, and this is the example I want to show you for immunohistochemistry.
05:21 - 05:27: This knockout is generated by an older method, and it’s using TALEN constructs rather than
05:27 - 05:30: CRISPR-Cas9, but the principle is the same.
05:30 - 05:37: But in this circumstance, what you’re then doing is you’re using cells as a surrogate
05:37 - 05:38: for tissue.
05:38 - 05:43: So you’re taking those cells, you’re pelleting them; usually you’re putting them into some
05:43 - 05:48: kind of agarose plug to mimic the tissue, and you’re processing them as you would process
05:48 - 05:55: a piece of tissue into a paraffin block, and then you cut sections from them.
05:55 - 06:01: So although it’s not cell culture in the same sense as you would do with ICC, it’s a similar
06:01 - 06:06: process you would use for your formalin-fixed, paraffin-embedded tissue, and you can clearly
06:06 - 06:12: see here, this is a PD-L1 clone again, that you have a positivity in the parent, and you
06:12 - 06:19: have a clear knockout in the negative here, in the knockdown cell line, using immunohistochemistry.
06:19 - 06:23: So there’s a number of different ways, and again, recommending that whatever your final
06:23 - 06:28: assay is, you want to make sure that you’re using those knockout cells or those lysates
06:28 - 06:33: in the assay of interest for your cells.
06:33 - 06:38: And you know, in Abcam, we’ve been doing this for a little while, and we have to date
06:38 - 06:43: done, as I said, flow cytometry, immunohistochemistry, and western blot.
06:43 - 06:49: And at this point, we have more than 2,800 antibodies that have been knockout validated
06:49 - 06:54: in our online catalog, and that equates to quite a large number of tests.
06:54 - 06:58: So we’re talking around 7,000 western blot tests.
06:58 - 07:05: It’s our primary method of validation, but we also do immunohistochemistry, so 2,249
07:05 - 07:11: and 90 flow cytometry tests; very, very few are IHC knocked out at this stage, something
07:11 - 07:15: we would like to do more of in the future.
07:15 - 07:19: So now that we are getting to the point where we have quite a large number of different
07:19 - 07:25: targets that have been knocked out, we can start to see some similarities and some interesting
07:25 - 07:28: facets of having done this on a large scale.
07:28 - 07:32: So once you go to this large scale, and you have the number of antibodies that we have
07:32 - 07:38: in our catalog, you can start to see some things, one of which is when you take a number
07:38 - 07:45: of antibodies, such as Annexin A1 here, we have five different antibodies represented in this
07:45 - 07:46: particular panel.
07:46 - 07:52: And when we’ve knockout validated those in western blot using, as we said before, a wild
07:52 - 07:59: type, a knockout, HeLa in this case, and Hep G2, we’re able to demonstrate that four of
07:59 - 08:06: those antibodies, two of those are rabbit monoclonals, one is a mouse monoclonal, one
08:06 - 08:10: is a rabbit polyclonal antibody, are all knockout validated.
08:10 - 08:15: They’re all demonstrating the right band of interest and the right size.
08:15 - 08:21: We are able to demonstrate that the knockout is not present, and also we get the presence of
08:21 - 08:23: the HeLa control in there.
08:23 - 08:29: But another antibody that we have is a different rabbit polyclonal; we can show it does not
08:29 - 08:30: knockout.
08:30 - 08:35: So that helps to confirm for us that the knockout itself is working well, as well as our antibody
08:35 - 08:36: is not functioning.
08:36 - 08:42: And what I can say to you is that when we discover these products are no longer present,
08:42 - 08:45: they’re no longer working, we remove them from sale.
08:45 - 08:49: So the product is no longer available; you can’t buy it.
08:49 - 08:53: That applies in a number of circumstances, because we don’t want these antibodies to
08:53 - 08:55: be for sale for you to use.
08:56 - 09:01: And what is interesting is that in many cases, or in some cases, the highly cited antibodies
09:01 - 09:02: are not always the best.
09:02 - 09:12: So while the most citations for CD73 is for this antibody on the left here, which is AB81720
09:12 - 09:22: for CD73, actually, we find that that is not working accurately, and we remove this one
09:22 - 09:25: again from sale; you can no longer buy it.
09:25 - 09:31: Whereas the citation index in this one is only six citations; has been used by in publications
09:31 - 09:36: for; it’s actually the one that we would recommend going forward, where we can clearly see that
09:36 - 09:40: the knockout works here and the knockout is absent.
09:40 - 09:48: So again, we can use these different methodologies to try and make sure that we have confidence
09:48 - 09:52: in the antibodies that you’re buying from us.
09:52 - 09:53: But there are challenges.
09:53 - 10:00: So I want to be clear that when you do this in your own lab, using these reagents, they’re
10:00 - 10:08: not always as pretty as those ones I just showed you, because life is not that simple.
10:08 - 10:12: So there are things you’ve got to try and get right and one of them is represented here.
10:12 - 10:15: So this is a bit more of a complicated picture.
10:15 - 10:22: So on the left-hand panel, you’ve got a rabbit polyclonal in green here, and you’ve got a
10:23 - 10:25: mouse GAPDH in red as being the loading control.
10:25 - 10:30: On the right-hand panel, you’ve got a rabbit monoclonal and it’s got a single band just
10:30 - 10:33: there and GAPDH again as a loading control.
10:33 - 10:39: And what this is trying to represent is that CD63 is a protein that is secreted.
10:40 - 10:43: And so if you just go and look at CD63 on its own in a Western blot, you’re not going
10:43 - 10:44: to find very much of it.
10:45 - 10:49: But in this case, you have to use the cell treatment we’re recommending.
10:49 - 10:55: He is using Brefeldin A, which will keep the protein present within the cell before
10:55 - 10:59: we take that cell and lyze it in order to be able to create our lysate.
11:00 - 11:07: And what it then shows you is that on the left-hand side, with this rabbit polyclonal,
11:07 - 11:09: we have a lot of bands.
11:09 - 11:15: And those bands, whether they’re with Brefeldin A or not, it’s a very bandy pattern.
11:15 - 11:21: And more importantly, the knockout might be present.
11:21 - 11:28: There are far too many cross-reactive bands in this for this recommendation for this antibody
11:28 - 11:30: to be used.
11:31 - 11:36: And in contrast, on the rabbit monoclonal on the right-hand side, we see a much cleaner
11:36 - 11:37: wblot pattern.
11:37 - 11:42: We have CD63 sitting positive here, and the knockout is negative.
11:43 - 11:48: You only see that presence when the Brefeldin A is present, which is what we would expect
11:48 - 11:53: because CD63 should only be there when it’s being held within the cell.
11:54 - 11:57: And this is the product that we would take forward.
11:58 - 12:02: So again, we can use different cell treatments to help us to overcome some challenges.
12:03 - 12:07: And then the other thing that comes up quite frequently is that sometimes you still get
12:07 - 12:12: a band. So we’ve knocked out this particular protein.
12:13 - 12:15: But what we’ve left in is what we call a truncation.
12:15 - 12:19: So this is where it’s been an incomplete knockout;
12:19 - 12:22: we’re still getting production of that protein, but you can clearly see it on the western
12:22 - 12:26: blot because the band is a different size.
12:27 - 12:31: Now, what we recommend and what we try and do is we would generate another knockout, a
12:31 - 12:36: second knockout, which completely removes this protein.
12:36 - 12:38: We can demonstrate it here with knockout two.
12:39 - 12:41: Again, we’re confident that the antibody is working.
12:41 - 12:46: The antibody is clearly working in both cases and the knockout is removed in the second
12:46 - 12:52: case. But we just wanted to highlight that to you so that when you use them in your
12:52 - 12:57: circumstances, sometimes these things come up and it can be confusing what to do with
12:57 - 13:02: them. But it would represent in this case that the protein is still present.
13:02 - 13:07: So in some circumstances, if you’re using immunocytochemistry, you wouldn’t want to use
13:07 - 13:10: this knockout because it’s still there.
13:10 - 13:13: The protein is still present and knockout two is completely absent.
13:14 - 13:17: But if you’re in a western blot scenario, you can clearly see there’s a difference.
13:18 - 13:23: So again, these are the nuances that come about when you’re knocking out a variety of
13:23 - 13:26: antibodies and a variety of targets that we’re looking at.
13:28 - 13:34: OK, so coming to the end of my short overview webinar of knockouts, I want to
13:34 - 13:40: highlight a couple of things to you, one of which is that you can go to the Abcam website
13:40 - 13:41: and you can buy these knockout lysates.
13:41 - 13:45: It’s not just the antibodies that we are knocking out, which you can buy knowing that
13:45 - 13:50: they’re knockout validated, but you can buy the lysates themselves to do this in your
13:50 - 13:57: own lab. And you can also buy the cell lines so that you can use more functional tests
13:57 - 14:03: within your own lab, biologically relevant in your own lab to be able to do those as
14:03 - 14:09: well. And shameless plug here, I want to make sure that you’re also aware that, you
14:09 - 14:15: know, we want to help you in your expertise in using western blot.
14:15 - 14:21: So we collaborated with a company called LI-COR, and with them we produced a paper
14:21 - 14:28: recently, only last June, Journal of Biological Chemistry around the best ways of
14:28 - 14:29: validating antibodies for Western blot.
14:29 - 14:35: This is an image that comes from that paper and some key criteria that is good to use
14:35 - 14:38: when you’re using western blot in your own labs.
14:38 - 14:43: And I also want to highlight that within the Abcam website, you can see a number of
14:43 - 14:51: different webinars similar to this one around different parts of the validation process
14:51 - 14:52: in different applications.
14:52 - 14:57: So on the left-hand side, you’ve got western blotting by my colleague, Hannah Dreja, in the
14:57 - 15:02: middle, immunohistochemistry by Silvia Sbacchi, and on the right-hand side, Sara
15:02 - 15:06: Passos is giving an overview to flow cytometry.
15:06 - 15:13: So I hope you enjoyed this webinar and I hope that you will get a lot out of that.
15:13 - 15:14: So thank you very much for your time.

Knockout Cell Lines and Their Role in Confirming Antibody Specificity

Video Transcript

00:00 - 00:17: Hello and thank you everyone for joining Abcam’s live Q&A conversation for our “Knockout cell lines
00:17 - 00:24: and their role in confirming antibody specificity” webinar. Well I hope you all have had the time to
00:24 - 00:30: watch the on-demand video. So during today’s session I am joined by Will Howat who’s here
00:30 - 00:38: to answer your questions live. Actually Will, would you mind giving a brief introduction to
00:38 - 00:44: who you are as well please, is that okay? Okay so I am the Director of Antibody Validation and
00:44 - 00:52: Characterization in Cambridge which means that I lead the team in Cambridge which consists of a
00:52 - 00:59: group of staff and that includes Hanna who sits with us who leads the scientific quality
00:59 - 01:06: control team but also our team looking at immunocytochemistry, flow cytometry, immunohistochemistry
01:06 - 01:12: as well and some cell science teams and all of which impact into how we can validate those
01:12 - 01:18: antibodies that we produce in the most effective manner. So that’s my role in Abcam
01:19 - 01:24: and as I said I manage Hanna and Hanna you should introduce yourself because you will also
01:24 - 01:32: be participating in the Q&A on a very much on that technical level. Do you want to introduce
01:32 - 01:37: yourself and your team? Yeah of course I probably should have started with that so thank you Will.
01:37 - 01:43: Yes my name is Hanna Dreja and I lead the scientific quality control team and we work
01:43 - 01:50: very hard at validating our products predominantly using western blotting and this is where the
01:50 - 01:56: knockout samples for us become an amazing tool to use which you will hear more about today.
01:57 - 02:03: So to get us started we have received a number of questions through our inbox so anytime during
02:03 - 02:08: this session please submit your questions through the Q&A chat box at the bottom of your screen.
02:09 - 02:15: So if we are unable to answer all of your questions during the session we will follow up
02:15 - 02:28: with you via email okay and so I will go for yeah I will start the questions if you’re okay with that
02:28 - 02:40: Will. Absolutely yeah a little nervous but yeah. Yes I’m gonna see how this is a question from
02:40 - 02:48: well an anonymous attendee so just you know if you do not want your name being
02:48 - 02:54: shared or known by Abcam please click anonymous attendee. Unfortunately we will not be able to
02:54 - 02:59: directly contact you with any additional information or answers so if these are
02:59 - 03:05: any of the questions we will not get to today we will have difficulties contacting you.
03:05 - 03:14: So this listener asks how to actually confirm the specificity
03:15 - 03:20: of an antibody or antibodies by using knockout cell lines?
03:21 - 03:28: So I guess there’s a variety of different ways of doing that. The most common as
03:28 - 03:34: Hanna’s already pointed out is to take those cell lines and lyse them and then to run a western
03:34 - 03:40: blot and that’s primarily what we do the most of so then we would normally then run that western
03:40 - 03:49: blot with the cell line and our isogenic control alongside that and ideally when we can we would
03:50 - 03:57: add in a positive control and negative control cell lines as well to give us confidence on the
03:57 - 04:04: cell line itself and try and do that with as many of those antibodies as we can to confirm
04:04 - 04:12: the specificity of those antibodies. So that’s one way it’s obviously well known and it’s a good
04:12 - 04:21: method to use but in some cases where that antibody is really being used as a marker for
04:21 - 04:29: flow cytometry or immunocytochemistry and potentially where the target is not going to work
04:30 - 04:36: within a reducing SDS-page western blot then we would recommend doing another method. So
04:36 - 04:44: you can use flow cytometry in which case you’re looking for that peak to be disappearing back to
04:44 - 04:50: where your controls are or you can use immunocytochemistry and then we’re adding extra visual
04:50 - 04:56: information so clearly if you have a protein or a target that’s expressed in the nucleus
04:57 - 05:02: in your control it’s perfectly present and then your knockout you can determine that it’s absent
05:03 - 05:10: then that’s also giving you that peace of mind that you can visually see that there’s a difference
05:10 - 05:15: between those two controls and again any other cell lines you add in and I would also add
05:15 - 05:23: immunohistochemistry as well which is taking those cell lines to take them through a processing
05:23 - 05:28: method as you would do with tissues so you take a cell pellet you formalin fix the cell pellet
05:28 - 05:35: usually in combination with agarose and then process it like tissue and put it into a tissue
05:35 - 05:39: block and then cut sections from it again you’ve got the control and you’ve got the
05:39 - 05:44: knockout and you should be able to detect the protein and the lack of protein in those
05:44 - 05:50: two so there’s a variety of different ways you can use those cell lines in an application
05:50 - 05:57: sensitive manner. Wow thank you very much for that very comprehensive answer Will. I’ve got a
05:57 - 06:07: question from Alice and she wants to know if for any given Abcam antibody reference which methods
06:07 - 06:16: were used for validation of specificity so how do we determine the specificity of the Abcam’s
06:16 - 06:24: antibody in which method. That’s how I interpret that. So I would say the majority of the ones on
06:24 - 06:29: the Abcam website will be in Western blot and we will then put the image up which does the
06:29 - 06:37: side-by-side alignment of you know the knockout and the wild-type control
06:37 - 06:42. If it has been run with immunocytochemistry- and we tend to use Western blot first and then
06:42 - 06:49: move to immunohistochemistry -then you will then see a different panel and that again normally
06:49 - 06:56: is a four panel which will have the knockout and the wild type and you should again be able to
06:56 - 07:02: see for your own eyes that the knockout is verified so that’s how you normally would kind of drill
07:02 - 07:10: into the detail and using you know clicking through the data sheet online with the
07:10 - 07:16: images obviously in that top bar and it will say knockout validated in the top bar but you’ll
07:16 - 07:19: have to click into the images to find that out I think that’s accurate is it not Hanna?
07:21 - 07:27: Yes so but please see there our website we at Abcam we try to be as transparent as possible so
07:27 - 07:34: we provide that information if we’ve got it it’s there and obviously there’s a lot of additional
07:34 - 07:41: validation data that we have on these data sheets that are not necessarily by using knockout samples
07:41 - 07:47: but yes there’s a lot of information there that one can use for to reference our
07:47 - 07:56: products. Thank you Will and this is a really interesting question for you Will
07:56 - 08:05: and it’s from Sajid Dhul and I do apologize if I mispronounce your name. What experiments can
08:05 - 08:09: we do through knockout rather than antibody validation?
08:12 - 08:18: A knockout or knockdown? It says knockdown actually yeah.
08:21 - 08:26: Let’s do that as two questions as in the knockout part of that and the knockdown because they are
08:26 - 08:32: different and particularly where I’ve got a higher background than immunohistochemistry
08:33 - 08:39: quite a lot doing that and knockdown. So if we say what experiments can we do through knockout
08:41 - 08:48: rather than antibody validation? I mean I guess the answer is that we can get a better clarity
08:49 - 08:54: with the knockout than we can through other methods that we do. With antibody validation
08:54 - 08:58: normally we would add in you know we draw through positive and negative cell lines and
08:58 - 09:06: the difficulty there is of course that we’re using various tools, databases etc and to try
09:06 - 09:13: and understand whether those cell lines are truly positive or truly negative and obviously we also
09:13 - 09:19: need to be able to add in cell treatments to sometimes upregulate or for that matter down
09:19 - 09:28: regulate that protein of interest and you know that gives a you know it’s a biological system so
09:28 - 09:33: sometimes that not necessarily give you the data you want and that’s where the knockout comes in.
09:33 - 09:39: So particularly on that knockout question then by eliminating to having confidence that we have
09:39 - 09:47: taken that gene away by doing PCR amplification that area understanding that that gene is gone
09:47 - 09:56: therefore we are hoping that the protein is also no longer present we can then eliminate that
09:56 - 10:01: and test the antibody against that protein and know that it’s absent. So that’s where it would
10:02 - 10:07: certainly add weight of course there are things that can go wrong there but it certainly adds
10:07 - 10:15: weight to the antibody validation process. So from a knockdown perspective something I used to use
10:15 - 10:21: quite a lot and particularly on the visual side of things so immunocytochemistry or
10:21 - 10:28: immunohistochemistry assays it’s been a transient knockdown using siRNAs which are still very
10:28 - 10:36: popular and relatively easy to do then actually they can have some added advantages. I mean there’s
10:36 - 10:42: not a true knockout so you don’t get zero but in some ways that can actually be positive because
10:42 - 10:46: you have some that are knocked down and some that are not knocked down you have a
10:46 - 10:51: gradation of scale you have got within that an inherent positive control so your antibody is
10:51 - 10:55: staining the things that haven’t been knocked down and it’s not staining the things that have
10:55 - 11:01: been knocked down. So providing you’re confident and I would usually recommend that when you’ve
11:01 - 11:07: done those kind of knockdown experiments visually for ICC or IHC that you add in something like a
11:07 - 11:12: Western blot to help to validate that it has actually gone away otherwise you know you could
11:12 - 11:16: just be looking at a cell line that’s growing funny and it’s not expressing the protein the
11:16 - 11:21: right way so you give yourself the additional confidence that you’re controlling is right
11:22 - 11:27: and then you can add in your antibody into that see those relative changes from that knockdown
11:28 - 11:32: and be confident again that is the accurate result and it gives you confidence that your
11:32 - 11:37: antibody is working. So that’s how I would approach knockdowns rather than knockouts.
11:39 - 11:43: And we actually have a follow-on question but I think you partly answered that already
11:44 - 11:54: Will and by the way thank you for that answer. So Sajid Dhul he wants to understand this difference
11:54 - 12:01: between knockout, knockdown and gene silencing and you did mention the siRNA for the knockdowns
12:01 - 12:06: and is there anything else you think you would like actually you would like to expand on that?
12:07 - 12:14: I mean that’s how I’ve done it in the past mainly because I’m not young anymore so therefore that
12:14 - 12:21: was prior to CRISPR being such a technique that is well known and being used in so many different
12:21 - 12:27: areas but you know transfections with lipofectamine or whatever it’s fairly common.
12:28 - 12:35: So that kind of knockdown methodology has got its place it’s relatively easy to do
12:36 - 12:44: it’s relatively cheap as long as we are comfortable with understanding and interpreting
12:44 - 12:50: the data that comes out of it. So for example the western blot because you’re lysing all those cells
12:50 - 12:57: and there’s not a 100% knockdown efficiency you’re going to get a band and you know if you’ve got
12:57 - 13:02: your loading control accurately lined up right so that you know you’re loading the same amount
13:02 - 13:09: of lysate so you’ve got your positive your non-knockdown and your knockdown aligned with it
13:09 - 13:14: you’re going to see a band but you’re actually basing on the fact the band is less and providing
13:14 - 13:19: your loading controls you’ve got to you absolutely have to have that loading control in there
13:19 - 13:24: otherwise your data isn’t valid and the same applies when you use your immunocytochemistry
13:24 - 13:29: methodologies you know that you’re going to see some that are weak you might not have any that
13:29 - 13:34: are completely obliterated in terms of the protein you might only knock them down a little bit
13:35 - 13:38: but you’re able to then distinguish that so it’s really about the quality of your
13:39 - 13:44: methodology the reproducibility of your methodology as well so that you could repeat
13:44 - 13:50: that experiment make sure you’re getting the same result because it’s not a true knockout so you’re
13:50 - 13:56: not taking it away it’s just knocked down a bit so it’s understanding how that applies into your
13:56 - 14:01: method and being confident that you’ve got that data and it’s publishable I’ve published
14:01 - 14:08: data that’s knockdown data and you provide you’ve got all those controls in there as I said I when
14:08 - 14:12: I was doing IHC then aligning that up with a western blot that shows it also goes down
14:13 - 14:20: is perfectly adequate for that knockdown methodology. Absolutely and obviously there
14:20 - 14:25: are a range of genes that we can’t knock out because they are essential for the survival
14:25 - 14:33: of the cells so in those cases going for a transient knockdown may be the answer if this
14:33 - 14:42: is if you’re interested in that in that particular target so yes thank you. So I’ve got a question
14:42 - 14:50: from Jeff and he asked that how we mentioned the validation we start with western
14:50 - 14:58: blot. Is there a reason for doing that? So for example Jeff needs an antibody for IHC and
14:59 - 15:04: he understands now that the validation that fails on western blot could still be
15:04 - 15:11: successful in IHC so does the western provide other important information to my IHC antibody
15:11 - 15:19: validation. I completely agree with you Jeff and the reason we will go western blot first is
15:19 - 15:26: it’s mainly for ease of use and my, Hanna’s, team have have got many skills they are able to run that in a
15:26 - 15:33: good very quick fashion as you know with pelleting out cells for IHC there’s additional processes
15:33 - 15:37: involved in that. It’s more difficult than just taking those cells and lysing them for example
15:38 - 15:43: so that’s one of the it’s more for convenience than we do it that way but you are right that
15:44 - 15:53: antibodies that do not work in a western blot can work perfectly well in IHC
15:53 - 16:02: I’ve seen many um so if I would recommend ultimately that if you are mainly wanting
16:02 - 16:08: an antibody for a particular application you use the application of interest so if that’s IHC
16:08 - 16:14: then you go down that route of validating it in IHC first and IHC particularly it doesn’t
16:14 - 16:20: really line up with anything else and nor does it line up with frozen IHC or immunocytochemistry
16:21 - 16:27: being the method that it is you have to have done it in that method before anybody is
16:27 - 16:32: going to accept it in that kind of knockout way providing you can again coming back to Hanna’s
16:32 - 16:38: point sometimes it’s not possible but then and then also lining up your other positive, negative
16:38 - 16:46: controls ultimately that as you know with IHC then once you get into tissue a validation of the cell
16:46 - 16:53: line is one thing entirely and it tells you your antibody’s binding specifically to that protein
16:53 - 16:59: but it doesn’t really tell you what other things it could randomly pick up because formalin
16:59 - 17:04: fixation changes things so you know you do still need to do that tissue validation with the right
17:04 - 17:10: positive negative controls um you know and ensure that it’s still getting that sensitivity
17:10 - 17:18: that you require in your assay. Great yeah thank you very much and I have a question again from
17:18 - 17:28: Sajid Dhul can you explain briefly the knockout procedure I know that you’re not um
17:29 - 17:35: you haven’t spent your entire career doing CRISPR- I haven’t no I haven’t spent any of my career
17:36 - 17:42: to be perfectly honest Hanna! Would you be able to go into more detail around
17:42 - 17:49: how that works? Yes so the idea is and I think actually you you did include it
17:49 - 17:57: really nicely in the webinar that people you if we we want to edit the gene and there are
17:57 - 18:03: different ways of doing it and one of the approaches we have done here at Abcam is using
18:03 - 18:08: this CRISPR-Cas9 system that has been around for the last probably four or five years
18:09 - 18:16: the first step is that you have a guide RNA and this guide RNA hybridizes with your gene of
18:16 - 18:22: interest so you would have to design that so it hits the part of the genome that you want to
18:22 - 18:28: modify or the region then you have the Cas9 protein which I think is a bacterial protein
18:28 - 18:38: and it will basically join the guide RNA and cleave the DNA what happens then is that the cell
18:38 - 18:46: this is the general approach is that the cell will realize it’s got a broken strand of DNA
18:46 - 18:53: and it’s going to try to repair it and during that repair mechanism something will most likely happen
18:53 - 19:00: so you will end up with a few nucleotides missing maybe a chunk missing and could only be one could
19:00 - 19:08: be two it could be three but obviously if you get one two four five what will ultimately happen is
19:08 - 19:16: that you will have caused a frameshift so when the cell then starts making RNA it will incorporate
19:16 - 19:24: that mutation and ultimately it cannot then be translated because there are the ribosomes will
19:24 - 19:32: hit a stop codon at one point which because you have this frameshift and basically the protein
19:32 - 19:38: will not continue to be expressed and also what happens is something called nonsense mediated RNA
19:38 - 19:45: degradation and there is a sensing mechanism within the cell realizing that this RNA is a bit
19:45 - 19:51: weird because it doesn’t translate to the very end so the cell themselves will actually degrade
19:51 - 19:59: the RNA as well so in the end you will have a you have incorporated a slightly random location
20:00 - 20:05: of mutation well you know roughly where it’s going to be very close to the RNA the guide RNA
20:05 - 20:14: and ultimately then that mutation causes an RNA that is not translatable basically and therefore
20:14 - 20:22: you will not have expression of your full protein so that is CRISPR-Cas9 in a nutshell
20:22 - 20:29: and there are different versions of it different other Cases that are not number nine and they
20:29 - 20:36: have slightly different actions but that’s the principle of it there are other ways of
20:37 - 20:46: mutating your genome of course but at the moment this is the approach that Abcam has taken
20:46 - 20:52: and but yes there’s a lot of really interesting literature of alternatives out there so
20:52 - 20:59: I hope that helps a little bit but please if I could clarify anything just put the
20:59 - 21:07: question back. Hanna, I understand that we can also use Cas9 to add in things so for production of cell
21:07 - 21:14: lines for example add in a bacterial resistance gene or some such to ensure that we can kind of
21:14 - 21:20: select out the ones that have been cut versus the ones that haven’t been cut, is that true?
21:21 - 21:30: Yes I believe that the guide RNA will pull the protein of interest to the site
21:31 - 21:40: and then you can incorporate other genomic pieces that will then go into that area so yes
21:40 - 21:46: we can use that to focus the mutation and integration of other pieces of DNA so
21:46 - 21:53: for example as you said resistant plasmids or other exchanges of DNA can be
21:56 - 22:03: done by using this system as well so yes it’s a very versatile tool and I think we’ve just
22:03 - 22:11: seen the beginning of what we can do with it so yes thank you. I have another question
22:11 - 22:17: and I have to admit I’ve been hesitating asking Will and myself that because I don’t think
22:17 - 22:22: neither of us will have the answer but I’m going to be brave. “What is the best approach
22:23 - 22:35: to evaluate the number of off-side CRISPR-Cas9 editing?” I wonder that I read that question as
22:35 - 22:41: well and I’ve been thinking about it and I wonder whether the origin of that question is
22:43 - 22:49: around the proof of where we are editing so normally most people will do some kind of
22:49 - 22:56: basic amplification or Sanger sequence to understand the genome area and ensure that that
22:56 - 23:06: cut has taken place or that the nonsense is there and I wonder whether the answer to that is
23:07 - 23:17: some form of RNA-seq or sequencing paradigm in order to try and understand more about where
23:18 - 23:26: off-side CRISPR interactions have been. That logically is the only way I could see that we
23:26 - 23:36: could do that is to do some kind of sequencing- larger scale sequencing to see what off-side
23:37 - 23:44: damage has been done in that process. I don’t know if I directly answered the question and
23:44 - 23:53: certainly if the person there wants to follow that up with an answer or for that matter
23:54 - 23:59: the question then please do and if we can’t answer it now we’ll try and get back to you.
24:00 - 24:09: Yes I think next generation sequencing is probably the way of exploring what has gone on in the
24:09 - 24:15: genome but yeah please let us know if you need any further explanation or details and also if you
24:18 - 24:22: could provide your name and we could get back to you directly if you’re interested.
24:22 - 24:26: Also I’m going to make a call out we may very well have some experts sitting here
24:27 - 24:33: typing questions or listening in and you may have the answer so maybe we have an expert.
24:33 - 24:40: So yeah please inform us if anything we say you want to add anything so yes thank you.
24:43 - 24:55: “So which procedure is more reliable is it qPCR or western blot and is it always reproducible
24:55 - 25:01: in gene and on gene and protein level?” and this is from Shai Jedul again.
25:05 - 25:14: I think that’s one for you Hanna. I will try and so qPCR this is a quantitative PCR
25:14 - 25:21: determining what is probably relevant here is the RNA level so what is being expressed
25:22 - 25:33: and now it’s interesting one would expect/hope that the RNA level is completely associated with
25:33 - 25:40: the protein levels and even if we go away from necessarily knockout we know for a fact that
25:42 - 25:43: RNAs have different
25:43 - 25:51: stabilities so even some cells expressing low levels of an RNA may actually have quite a
25:51 - 25:57: lot of proteins in there because it’s not only the quantity of the RNA is there how long is
25:57 - 26:03: the RNA stable for how long is the protein stable for so there is already a lot of discrepancy and
26:04 - 26:11: so it’s not a one-to-one ratio. The qPCR question around in the
26:14 - 26:18: genetically modified cell is another challenge as well because
26:19 - 26:27: there has been evidence and there’s a paper by Smith et al from a Nature publication from
26:27 - 26:36: November showing that even if you genetically cause a mutation in the DNA it doesn’t always
26:36 - 26:43: and clearly remove all the RNA because it can be that the cells themselves try to overcome this by
26:43 - 26:50: either using a new start site and or doing some exon skipping so the question is which part of
26:50 - 26:58: the RNA do you need to look at to be able to really understand whether the protein will be gone
26:58 - 27:06: so I think RNA is possibly a good screening method if you really are critical and only
27:06 - 27:12: look at things that really have very low levels of RNA following your CRISPR-Cas9
27:12 - 27:22: and CRISPR-Cas9 clonage however I do if we are interested in proteins I would always look at
27:22 - 27:31: proteins so ultimately once you’ve made your mutations please explore what is in the cell
27:31 - 27:37: on the protein level and that is the safest bet to know whether your protein is fully gone or not.
27:38 - 27:46: So I guess the follow-up to that Hanna is when we’re doing this internally obviously one of the
27:46 - 27:55: starting points for the cell line is that we know it has got the protein present so how do you
27:55 - 28:02: do we do that internally to determine that the protein is present do we use protein and or do
28:02 - 28:10: we use RNA or do we use a mix of both? That’s a great question now it’s we are very fortunate
28:10 - 28:17: Will and I so at Abcam we have a lot of data for a lot of proteins so for example if we
28:17 - 28:23: wanted to particularly target well we want to target a particular protein of interest most
28:23 - 28:30: likely we have a lot of validation data already for that target meaning that we probably have
28:30 - 28:37: not only Western blot data we have IHC data, ICC data, flow data which already gives us a
28:39 - 28:45: good set of data to predict whether protein is expressed in the cell certain cell line or not.
28:46 - 28:52: Obviously there are cases where there is a limited amount of data possibly on our or other
28:52 - 29:00: other companies’ websites but obviously the go-to for all you scientists is obviously
29:00 - 29:06: the literature what is out there what evidence is there of protein expression and that is
29:08 - 29:17: really useful for us we always we always establish that the protein is indeed expressed
29:17 - 29:22: in the cell lines if we choose if we want to make a new cell line. There are obviously
29:23 - 29:29: challenges here there are targets which may not be that well known so therefore this
29:30 - 29:38: makes it quite difficult and so again then having to go back to the RNA level
29:38 - 29:47: can be useful and we can certainly say that a cell line that has zero RNA in it will most
29:47 - 29:53: likely not have any proteins in it. It’s a cell line that has heaps of RNA in it
29:53 - 29:57: it’s more likely to have more proteins in it than the one that doesn’t have any
29:57 - 30:02: so that we use that as a guideline as well to determine which cell line to go for.
30:03 - 30:11: So it’s a combination of data sets and hopefully one cross validates the other two
30:11 - 30:16: because it’s you know it’s still a piece of work to pick the cell line and then make that knockout
30:17 - 30:20: and if it doesn’t have the protein then obviously it’s not going to knock out.
30:21 - 30:26: And there may be there may be data that is antibody independent out there so
30:26 - 30:31: if there’s mass spec data demonstrating that actually you know what that protein was
30:31 - 30:38: has been picked up in a cell line that gives you that confidence as well which is antibody
30:38 - 30:42: independent you know the protein is there you’re not reliant on an antibody so yeah
30:42 - 30:52: that’s another that’s a very useful tool as well. I have a question coming in from Anastasia
30:52 - 30:59: and so thank you again Anastasia for joining us in our meet the experts talks.
31:02 - 31:10: She’s asking about how to precisely evaluate off-site editing of CRISPR-Cas. I would like to
31:10 - 31:19: ask what solutions you can propose from Abcam and so she’s looking into what kind of equipment and
31:19 - 31:26: reagents to use for this kind of evaluation assay and this is fantastic I’ve got her contact details
31:26 - 31:33: and as I mentioned we actually not surprisingly we actually have a well we have a group of experts
31:33 - 31:40: within Abcam but neither Will nor I are that familiar with the pieces of equipment you use
31:40 - 31:47: so we will get back to you and give you a much better answer to help you with your research
31:47 - 31:51: Anastasia I hope that’s okay. Is there anything else you would like to add to that Will?
31:52 - 31:56: No I think that’s good I can think of a couple of people who’ve got that more of that background
31:56 - 32:05: particularly young who’ve been doing CRISPR-Cas for quite some time that could probably help to
32:06 - 32:14: do that and whether there’s I mean to be honest if we’re recommending going down a DNA level there
32:14 - 32:20: may not actually be that many products in Abcam that will help you with that but we will try and
32:20 - 32:29: highlight those that can. Yeah no that’s that’s great. I’ve got one other question from
32:30 - 32:39: from William “Can the choice of an antibody based on knockout material be different
32:40 - 32:42: in a different application?”
32:45 - 32:55: I think I partly answered that before with the Western blot versus IHC problem but yes it can
32:55 - 33:02: but this also came up in in webinar during reproducible science from Peter McPherson
33:02 - 33:09: where he’d been using a number of antibodies from a number of different vendors for
33:09 - 33:17: CRF72 I believe and what he demonstrated was really interesting and he talked about it was
33:18 - 33:22: that there were some antibodies that worked really well in western blot
33:23 - 33:29: but if you took those forward in ICC which was his final application he wanted to use
33:29 - 33:36: then you could show that they were okay for ICC but weren’t the best and actually the best
33:36 - 33:44: antibody which was from another manufacturer than Abcam the best antibody for the
33:44 - 33:51: immunocytochemistry actually gave up fairly on the western blot it was a bit okay but
33:51 - 33:58: you wouldn’t have picked it out of the bunch so I think it’s not that it wasn’t it wasn’t valid
33:58 - 34:03: because it was valid it was just that if you were going to run a western blot you would want
34:03 - 34:08: antibody A but if you’re going to run an immunocytochemistry experiment you would want antibody B
34:08 - 34:15: and the same actually applied in different flavors for IP as well so again it’s coming back to that point
34:15 - 34:20: that you want to be validating ultimately in the in the one that you’re going to be using
34:21 - 34:27: and your experiment is all designed around immunocytochemistry it’s whatever that might be
34:27 - 34:32: whether it’s cell growth assays with an antibody staining or you know biomarker discovery whatever
34:32 - 34:39: you want to be mainly focusing on that and as your application of interest and aligning the
34:39 - 34:44: western blot if it works then great it’s great data to put alongside it when you’re writing
34:44 - 34:48: your publication but if it doesn’t work there might be other reasons why it doesn’t work
34:48 - 34:52: and as I said the target just might not like reducing gels.
34:53 - 35:03: Yeah well thank you so much for that for that for that answer so Will I think you have exhausted
35:03 - 35:12: our listeners you have managed to answer all their questions and unless we have anything else
35:12 - 35:20: coming through I would suggest well I would thank you well thank you to Will for this great great
35:21 - 35:29: conversation and I would like to thank you all for taking your time and listening in to Abcam’s
35:29 - 35:37: meet the experts week.

About the presenter:

Will Howat leads Abcam’s global imaging and immunohistochemistry team as Director of Antibody Validation. Prior to joining Abcam in 2018, he worked at AstraZeneca as Team Leader of its Molecular Pathology Group, where his team developed biomarker assays for preclinical, phase 1 and phase 2 clinical trials. Before that he was at the Cancer Research UK’s Cambridge Institute, where he set up and ran the Histopathology/ISH core facility for nine years. He has also worked at the Wellcome Sanger Institute on the Atlas of Protein Expression and was responsible for R&D in within the Immunohistochemistry group.

Will has a BSc (Hons) in Immunology & Pharmacology from the University of Strathclyde and a PhD in Pathology from the University of Southampton. His publications span key journals such as The Lancet, Nature, Science, Nature Genetics, American Journal of Pathology, and the Journal of Pathology.

Knockout cell lines and their role in confirming antibody specificity

On-demand webinar

Summary:

Key discussion points:

Video Transcript

Video Transcript

About the presenter:

You may be interested in...