Validating antibodies with knock-out technologies

On-demand webinar

Summary:

Dr. Alejandra Solache, Vice President of New Product Development at Abcam, recently discussed how we validate our antibodies using knock-out (KO) technologies. Learn about the value of gene knock-out technologies in large-scale antibody screening by watching our webinar.

Why use knock-out validation:

Several studies have demonstrated the irreproducibility of life sciences reagents, which in the case of antibodies, frequently results from a lack of specificity to their targets of interest. This lack of specificity, related to issues such as cross-reactivity and off-target protein binding, has contributed to the so-called ‘reproducibility crisis’^1,2.

KO validation is one of the most accepted and trusted validation processes for antibody specificity^3,4. This robust technique can confirm the specificity of an antibody by testing it in a KO cell line, cell lysate, or tissue that does not express the target protein. A specific antibody should, therefore, produce no signal in the KO cell line but give a specific signal in the wild type sample. In this way, KO validation serves as a true negative control5.

Find out more about knock-out validation

Abcam’s knock-out validation initiative

To ensure that our antibodies are specific and provide reproducible results, we apply KO validation at a large scale. We validate our antibodies in multiple applications, including flow cytometry, western blot, and immunocytochemistry, using multiple antibodies per target. Antibodies that demonstrate specificity are given a special “KO validated” designation, whereas those exhibiting off-target binding are removed from our catalog. We have now confirmed the specificity of more than 4000 antibodies with KO validation for over 2000 protein targets.

Key learnings from Abcam’s KO validation initiative:

In general, recombinant monoclonal antibodies demonstrate superior specificity and consistency compared to traditional monoclonal and polyclonal antibodies.

A significant number of top-cited antibody clones have been observed to exhibit off-target effects.

We recommend that scientists choose antibodies that have been validated across multiple applications, ideally using KO technologies. Alternatively, you can validate an antibody yourself, using the appropriate KO cell line, KO cell lysate, or tissue.

Confirming the specificity of an antibody can take time and effort, including the generation of a KO cell line using gene-editing technologies (on average 6 months to generate KOs). To help you accelerate your research and avoid tedious CRISPR engineering experiments, we now provide one of the largest collections of CRISPR-Cas9 KO lysates and KO cell lines.

All our KO cell lines and lysates are reliable, ready-to-use, and enable you to confidently advance target identification and validation without first establishing a KO cell line. All our KO cell lines are validated by Sanger sequencing, and many include additional western blot data to confirm KO at the protein level.

Validating antibodies with knock-out

About Dr. Alejandra Solache

Dr. Alejandra Solache joined Abcam in December 2013 and manages Abcam’s Global New Product Development and validation teams in Cambridge, UK, Hangzhou, China as well as Eugene and Branford, CT in the USA. Alejandra has also been instrumental in developing Abcam’s innovation strategy and product pipeline.

Before joining Abcam, Alejandra led the Antibody and Assay Development teams as R&D Director at EMD-Millipore, where she held various positions for eight years. She trained as a life science researcher where she gained expertise in immunology, cell signaling, and epigenetics through post-doctoral fellowships at the University of California, San Francisco, and the Trudeau Institute, Saranac Lake, NY.

Video Transcript

00:00 - 00:19: Good afternoon, everyone. Welcome to our Biologics webinar. My name is Ese, and on behalf of
00:19 - 00:24: Oxford Global, I would like to thank you all for joining us today. The full recording of
00:24 - 00:29: the webinar will be made available to download on our secured website, and the details will
00:29 - 00:34: be sent to you after the webinar. Today, we will have Alejandra Solange, Vice President
00:34 - 00:40: of New Product Development from Abcam, presenting on the technologies that support Abcam in
00:40 - 00:45: the discovery, development, and validation of high-quality reagents. Please submit your
00:45 - 00:50: questions through the chat option on the bottom of your menu display. Without any more delays,
00:50 - 00:55: I would like to get started with Alejandra’s presentation. Alejandra, over to you.
00:55 - 01:01: I am very happy to be here and to discuss a topic that I believe is key for reproducibility
01:01 - 01:07: in scientific research. Today, we will discuss how antibody selection, and in particular,
01:07 - 01:13: how the use of recombinant antibody technologies and high standards in antibody characterization
01:13 - 01:19: in multiple applications, including knockout cell lines and other validation strategies,
01:19 - 01:21: support scientific reproducibility.
01:26 - 01:32: High-quality antibodies are a combined result of thoughtful, effective strategies for immunogen
01:32 - 01:39: design and protein expression, antibody development and production, as well as specificity
01:40 - 01:46: validation and strict batch testing. The choice of antibody format combined with high standards
01:46 - 01:53: in manufacturing has a great impact on specificity, antibody performance, and consistency.
01:53 - 02:00: This presentation will be divided into parts. We will start by discussing the advantages and
02:00 - 02:06: constraints of the available antibody formats and host species, as well as the benefits of
02:06 - 02:12: recombinant rabbit monoclonals and why, at Abcam, we have adopted them as the format of choice.
02:13 - 02:19: In the second part, I will discuss how Abcam’s validation is constantly evolving by adopting
02:19 - 02:24: new technologies to assess specificity. In particular, I will share with you one of the
02:24 - 02:30: most exciting initiatives that we have in Abcam, which is the incorporation of
02:30 - 02:33: knockout technologies in our antibody validation.
02:34 - 02:41: So, the choice of antibody format can significantly influence the consistency of our results,
02:41 - 02:45: and the performance of an antibody can be very closely linked to how they are manufactured.
02:47 - 02:53: Antibodies have been traditionally developed in two main formats, polyclonals or monoclonals.
02:53 - 02:58: The polyclonal antibody process generates a large variety of antibodies to different
02:59 - 03:05: epitopes on the antigen used for immunization. Although this can be advantageous for some
03:05 - 03:10: applications, this can also be a significant source of background and cross-reactivity.
03:10 - 03:17: The complexity of polyclonal antibodies is further complicated by their finite nature,
03:17 - 03:23: and once a batch has been depleted, you need to initiate another immunization,
03:23 - 03:26: and therefore the antibody performance may vary significantly.
03:26 - 03:32: In contrast, monoclonal antibodies are manufactured from hybridoma lines,
03:32 - 03:40: where the antibody-producing B-cells are isolated and immortalized by fusing them to
03:40 - 03:46: a myeloma or a cancer cell line, and the derived hybridoma-produced antibodies
03:47 - 03:52: recognize a single epitope in the target protein. Once a hybridoma clone has been
03:52 - 03:58: identified, it can continue to be produced by culturing the hybridoma cell line in tissue
03:58 - 04:03: culture, and if properly maintained, they should have no batch-to-batch variability.
04:03 - 04:10: However, in some instances, if hybridomas are poorly developed or maintained poorly,
04:10 - 04:16: it is possible that they will experience loss of yield or binding activity caused by sequence drift.
04:16 - 04:23: Polyclonal antibodies have been used for many years, and they became the reagent of choice when
04:23 - 04:28: access to monoclonal technologies was not widely available, and although there are some very good
04:28 - 04:35: and specific polyclonal antibodies, if you have the choice between a polyclonal or a monoclonal
04:35 - 04:42: antibody that has been validated
04:43 - 04:50: for the applications that you need for your experiments. Our recommendation
04:50 - 04:55: is to choose a monoclonal as, in general, it will provide a more reliable option.
04:58 - 05:03: An alternative to both polyclonal and traditionally produced monoclonals is to produce
05:04 - 05:10: antibodies through a recombinant platform. A recombinant antibody is generated by using the
05:10 - 05:17: antibody coding genes from a B cell producing the antibody and cloning those genes into a vector
05:17 - 05:26: and expressing them in a mammalian cell. This can be done from the beginning, as
05:27 - 05:35: what you do for a polyclonal, immunizing a rabbit, for example, and isolating circulating B cells and
05:35 - 05:43: then cloning the antibody-producing cells, or utilizing an existing hybridoma where you can
05:43 - 05:50: convert the recombinant antibodies by synthesizing a clone using the hybridoma DNA sequence and
05:51 - 05:59: expressing it in a mammalian cell line as well. Alternatively, the recombinant monoclonals can
05:59 - 06:06: be developed in a totally in vitro approach and phage display has become incredibly useful for
06:06 - 06:12: this application. Phage display offers the opportunity to make antibodies against difficult
06:12 - 06:19: targets such as toxins that were previously difficult to achieve using in vivo technologies.
06:19 - 06:23: Abcam’s Axiomics high throughput phage display technology
06:26 - 06:34: uses a high diversity synthetic library which is designed to allow efficient affinity maturation
06:34 - 06:40: and rapid cloning to give a full recombinant antibody. Antibody discovery using this technique
06:40 - 06:47: has benefits across a wide array of applications and can provide a suitable discovery approach for
06:47 - 06:54: certain challenging targets. So recombinant antibodies have multiple advantages over
06:54 - 07:00: polyclonal and standard hybridoma produced antibodies, such as improved consistency and
07:00 - 07:07: reproducibility. Once sequenced, the recombinant antibodies are not subject to loss of activity
07:07 - 07:13: caused by sequence drift or loss of yield, both of which can affect hybridomas if they are poorly
07:13 - 07:19: maintained. The engineered nature of the recombinant antibodies makes it easier to
07:19 - 07:26: improve both antibody specificity and sensitivity if required. And once an antibody has been cloned,
07:26 - 07:31: recombinant antibody production provides a scalable, fully in vitro system to
07:31 - 07:38: consistently produce antibodies that are identical from batch to batch, which eliminates the variation
07:38 - 07:45: in results between batches that is inherent to antibodies like polyclonals. In terms of
07:45 - 07:50: antibody selection for your own experiments, a recombinant antibody is the most robust format
07:50 - 07:56: available and we recommend them if you have a choice over standard hybridoma monoclonals and
07:56 - 08:04: certainly over polyclonals. Another critical consideration when developing and selecting
08:04 - 08:11: antibodies is the species where the antibody has been generated. Multiple host species have
08:11 - 08:17: been utilized over the years to develop antibodies, particularly when the monoclonal technologies were
08:17 - 08:24: not fully developed. Polyclonal antibodies were the choice and species such as goat, sheep, chicken,
08:24 - 08:31: and particularly rabbits were widely used. Until early 90s, when rabbit monoclonal technology
08:31 - 08:43: became available with the discovery of a rabbit myeloma cell line, known as 240E,
08:45 - 08:53: these most monoclonal antibodies were typically produced in mice, yet their antibody-producing
08:53 - 09:00: cells are unable to generate every kind of antibody required as we discussed. In contrast,
09:00 - 09:07: rabbits have a unique immune system compared with other species such as mouse, rat, and chicken.
09:08 - 09:13: Rabbits have a large B-cell repertoire that can generate a diverse range of antibodies.
09:14 - 09:21: The rabbit’s immune system also optimizes affinity by using two different mechanisms
09:21 - 09:27: of affinity maturation. One is gene conversion and the second one is somatic hypermutation.
09:27 - 09:33: This means that rabbit antibodies are better at distinguishing subtle differences such as
09:33 - 09:39: epitope variations, post-translational modifications, and conformational changes due to
09:39 - 09:47: larger and more diverse B-cell repertoire. Rabbits also produce high-affinity antibodies,
09:47 - 09:53: even against targets that are not immunogenic in mice, having a significantly higher affinity for
09:54 - 09:58: a target antigen on average than most mouse monoclonals.
10:00 - 10:05: And to overcome the limitations of the mouse model at Abcam, we combined the unique properties
10:05 - 10:13: of the rabbit immune system and antibody technology, monoclonal antibody technology,
10:13 - 10:19: to generate rabbit monoclonals. When compared to other common antibody platforms,
10:19 - 10:23: rabbit monoclonal antibodies have the highest affinity and specificity.
10:25 - 10:31: But going even further, we have combined the high affinity and specificity of rabbit monoclonals
10:31 - 10:39: with our rabbit recombinant technologies through NGS, hybridoma, or direct B-cell cloning process.
10:39 - 10:46: And the result of these features is a recombinant rabbit monoclonal antibody that is specific and
10:46 - 10:54: very sensitive to the target or molecule of interest, therefore providing a significant
10:54 - 11:03: advantage over rabbit monoclonals and mouse monoclonals. Our recombinant rabbit monoclonal
11:03 - 11:09: technology was recently used to develop a panel of antibodies to a promising immunotherapy and
11:09 - 11:17: diagnostic target known as B-cell maturation antigen, or BCMA. BCMA regulation has been
11:17 - 11:25: identified in B-cell malignancies, malignancies such as multiple myeloma. Antibodies that bind
11:25 - 11:34: both mouse and human versions of BCMA are needed to support fundamental research into the biology
11:34 - 11:41: of BCMA in parallel to efforts being carried out to advance protein therapeutics. The desired
11:41 - 11:49: antibody that we wanted to generate against BCMA had to bind both mouse and human versions of BCMA.
11:50 - 11:56: And as the rabbit immune system offers options for both human and mouse reactivity, the RabMAb
11:56 - 12:03: technology was an obvious choice to produce a series of high affinity, sub-nanomolar affinity,
12:03 - 12:13: anti-BCMA antibodies. Using the Octet system from ForteBio, we were able to identify recombinant
12:13 - 12:22: rabbit monoclonals that bind with high affinity to human and mouse BCMA. And as specified and
12:22 - 12:31: requested, they do not recognize any other protein targets. These recombinant rabbit monoclonals
12:31 - 12:37: are also suitable for multiple applications, including immunocytochemistry, flow cytometry,
12:37 - 12:44: western blotting, and immunoprecipitation. Furthermore, antibodies displaying antagonism
12:44 - 12:56: for BCMA ligand binding were also characterized. So in addition to being able to generate specific
12:56 - 13:01: immune oncology antibodies for particular application requirements, our expertise
13:01 - 13:07: in rabbit monoclonal technology also offers a significant advantage when generating antibody
13:07 - 13:14: pairs that can work together in the same assay. The challenge here is to find two antibodies
13:14 - 13:19: against the same target that recognize different epitopes and work well together.
13:20 - 13:25: Developing these antibody pairs in mice is difficult due to the limited immune repertoire
13:25 - 13:31: of the mouse, and rabbit monoclonals provide a much better alternative. As you can see here,
13:31 - 13:41: when we generated these antibody pairs utilizing rabbit polyclonal pair against human IL-1
13:41 - 13:48: receptor alpha, the sensitivity wasn’t that great. When we actually generated this
13:49 - 13:57: assay utilizing a rabbit monoclonal pair, we identified a tenfold increase in sensitivity.
13:58 - 14:05: So this basically highlights another great use of this, and we are utilizing all of these rabbit
14:06 - 14:14: recombinant antibodies also for the generation of our Simple Stabilizer and also multiplex assays.
14:14 - 14:25: So, as we discussed before, generating antibodies with high specificity and consistency requires a
14:25 - 14:30: multifaceted approach, and we have embedded the highest quality standards in every step,
14:31 - 14:37: as well as utilizing the most reliable antibody development process. Because of this, we are
14:37 - 14:44: moving primarily to recombinant antibody generation for all of our new antibodies.
14:45 - 14:51: However, regardless of the antibody format, antibodies require in-depth antibody characterization
14:51 - 14:59: in order to ensure that they bind the target of interest in multiple applications
14:59 - 15:03: and that they do not cross-react with proteins of high homology, for example.
15:04 - 15:07: At Abcam, we are also working towards providing the highest level
15:07 - 15:13: of validation and striving for 100% batch-to-batch consistency.
15:17 - 15:23: So, at Abcam, we have incorporated a range of assays and applications to ensure that we can
15:23 - 15:29: assess specificity and selectivity during early development at the clone screening stage,
15:30 - 15:37: as well as during the final characterization of the antibody. During the antibody screening process,
15:38 - 15:46: assays such as ELISA and label-free assays are essential as they are high throughput and provide
15:47 - 15:53: a lot of information on the binding and affinity of the antibody for the target.
15:53 - 15:59: We aim to provide validation in as many applications as possible, including western
15:59 - 16:06: blotting, ICC, IHC, flow cytometry, IP, ChIP. And we tailor these applications depending on
16:06 - 16:12: the target of interest. For example, if it is a phosphorylated target, we will try,
16:12 - 16:20: we will do a stimulation of the cells to increase the phosphorylation of that particular target.
16:21 - 16:32: And if it is, for example, an anti-histone modified protein, we will test in peptide
16:32 - 16:40: arrays, for example, and ChIP or ChIP-seq. Abcam’s validation is constantly evolving
16:40 - 16:47: and by adopting new technologies to confirm specificity. And we now have access to new tools,
16:47 - 16:53: new tools such as mass spectrometry or high throughput arrays, as well as we have the
16:53 - 16:59: ability to knock down or knock out genes and alter protein expression in cells and animals,
16:59 - 17:05: which allow us to better assess reagent specificity and to be more precise than ever before.
17:06 - 17:12: Although there is not one single solution to antibody validation, and each method has pros
17:12 - 17:21: and cons, we always try to use a combination of these methodologies. And we have embarked on
17:22 - 17:31: large-scale antibody validation using knockout models, as in the context of antibody specificity,
17:31 - 17:35: knockout models provide an excellent standard for antibody validation as they represent
17:35 - 17:43: a true negative control for most proteins expressed. Of course, using knockout cell lines
17:43 - 17:48: as negative control won’t always be appropriate as essential genes cannot be knocked out.
17:49 - 17:56: However, this can be utilized as a model for a lot of the proteins of interest. The way we do
17:56 - 18:01: these assays is to have a comparison of the wild-type cell line and the knockout cell line,
18:01 - 18:07: and we also have additional control cell lines in which the target of interest is endogenously
18:07 - 18:13: expressed. We select multiple antibodies to the target and we do the experiment in several
18:13 - 18:18: applications. Here I’m showing an example of a western blot application when we are testing
18:19 - 18:26: different antibodies, where we have tested antibodies to PKC-alpha. And in this case,
18:26 - 18:33: what you see is that the band of interest, the band representing PKC-alpha, goes away in the
18:33 - 18:41: knockout. So, this highlights that the antibody is specific for the target of interest. You can also
18:41 - 18:49: have the other side of the coin, and this is when you see off-target binding of the antibody.
18:50 - 18:56: And in this particular case, you will see the off-target binding by the band of interest
18:57 - 19:07: doesn’t disappear in the knockout sample. Whenever we see these cases, we will remove the product
19:07 - 19:12: from the catalog and we will contact the scientists that have purchased this antibody in the past
19:12 - 19:15: to let them know that this antibody was not specific.
19:19 - 19:24: We started this validation using mostly western blot, but we have expanded it to several other
19:24 - 19:31: relevant applications that are pertinent for the target of interest and the area of research where
19:31 - 19:37: you will be utilizing these antibodies. In this example, we can show that we have tested this
19:37 - 19:51: antibody in ICC and flow cytometry. And so far, we have tested over 2,000 antibodies that have
19:51 - 19:59: been validated to recognize the target protein through knockout models. And we continue to
19:59 - 20:06: validate a growing number of antibodies. We have learned a lot through this endeavor,
20:06 - 20:14: and we have found that recombinant antibodies were found to be much more specific than other
20:15 - 20:19: monoclonal antibodies, and particularly over polyclonal antibodies.
20:21 - 20:28: However, other interesting learnings are that top-sided antibodies are not always the best
20:28 - 20:39: antibody, as we have found some top-sided antibodies that were cross-reacting with
20:39 - 20:46: proteins in the knockout samples. And this is just to highlight that we need to do our homework when
20:46 - 20:52: looking and searching for antibodies. As a top-sided antibody won’t be necessarily the best,
20:52 - 21:02: we will need to find evidence of specificity wherever, whenever possible. And the best
21:02 - 21:07: is to search for antibodies that have been validated in multiple applications
21:08 - 21:12: and confirm specificity by multiple strategies.
21:13 - 21:18: So, being specialists at making antibodies has been core to our mission for over 20 years.
21:18 - 21:24: We are also actively working to meet the industry needs for the development of exceptional,
21:24 - 21:31: high-performing IHC antibodies to key diagnostic and therapeutic targets. An example of an antibody
21:32 - 21:39: that is supporting in vitro diagnostics is our anti-pan-TK, or antigen, antibody.
21:39 - 21:44: Roche has recently announced the launch of the first in vitro diagnostic
21:46 - 21:55: for detecting variations in TRK receptors by IHC, which utilizes our rabbit monoclonal anti-pan-TRK.
21:57 - 22:02: And this assay provides an important new tool to help us better understand the role of
22:02 - 22:08: TRK protein expression, particularly fusions in cancer, and to identify cancer patients who may be suitable for several targeted cancer treatments.
22:10 - 22:27: The rabbit monoclonal anti-pan-TRK antibody used in this assay was developed using recombinant technology
22:27 - 22:33: and validated in multiple applications for specificity. These traits make the antibody
22:33 - 22:39: ideal for use in companion diagnostics, and they allow clinicians to generate consistent
22:39 - 22:43: and reproducible results between every diagnostic test they run.
22:47 - 22:52: In summary, we are very proud to be generating high-quality recombinant antibodies and validating
22:52 - 22:59: using the highest standards. In order to provide you with antibodies that have excellent sensitivity
22:59 - 23:05: with the highest degree of consistency, we have engineered recombinant versions of our rabbit
23:05 - 23:13: monoclonal antibodies, and we have now over 18,000 recombinant monoclonal antibodies,
23:14 - 23:23: over 1,800 recombinant matched antibody pairs, SimpleStep ELISAs, and multiplex assays.
23:24 - 23:30: And a lot of these antibodies have been validated in knockout models.
23:30 - 23:37: As I mentioned before, these reagents are now utilized in diagnostic tests and supporting
23:37 - 23:43: translational research through our partnerships across pharma and diagnostics. And this really,
23:44 - 23:49: this really increases the bar in our responsibility not only to the scientific
23:49 - 23:54: community in terms of reproducibility, but by providing the highest-quality reagents,
23:54 - 23:57: we are also contributing to improving health outcomes.
24:01 - 24:06: Now, I would like to thank ForteBio for inviting us to present here,
24:06 - 24:09: and I would be very happy to take any questions.
24:14 - 24:18: Thanks, Alejandra, for this wonderful presentation. We really enjoyed it.
24:18 - 24:22: Now, I would like to open the floor for discussion. Please remember to send us
24:22 - 24:27: your questions using the chat feature of the platform. How many antibodies have
24:27 - 24:31: you tested that haven’t fulfilled the criteria described in your talk?
24:32 - 24:42: Yeah, there have been around 500 antibodies that have not passed our specificity criteria in the
24:42 - 24:47: knockout models. And this is something that I’m really proud to say that we have removed them
24:47 - 24:54: all from the catalog and contacted the scientists that have purchased those antibodies in the past.
24:55 - 24:59: Great. Thank you very much. The next question we received is the following.
24:59 - 25:03: To rule out antibody cross-reactivity as part of antibody validation,
25:03 - 25:09: are you mapping the actual linear or conformational epitope by, for example,
25:09 - 25:12: customized peptide arrays as an advanced quality control?
25:15 - 25:23: Yes, absolutely. We are just utilizing a proprietary peptide array, particularly for
25:23 - 25:32: histone modified targets. And in those cases, we have an array with all of the potential
25:32 - 25:44: modifications in the histone targets. And we test all of our antibodies against modified histones
25:44 - 25:49: in this target. And that’s just one example. Great. Thank you very much for this.
25:49 - 25:55: Why haven’t you tested all of your antibodies for specificity trait target using knockout cell lines?
25:57 - 26:06: So, as I mentioned, some of these knockout cell lines cannot be used for all targets, since
26:09 - 26:17: any genes that are essential will not be, we won’t be able to knockout as the cell will not survive.
26:17 - 26:26: And therefore, this is one area that we have, that we are utilizing different types of
26:26 - 26:34: technologies to test, as well as we are going through where we have available knockout cell
26:34 - 26:41: lines, as well as where they match with the antibodies that we have.
26:41 - 26:46: Thank you very much for your questions. And once again, Alejandra, thank you very much for this
26:46 - 26:52: excellent presentation. The 2020 Proteins and Antibodies UK Congress will be taking place in
26:52 - 26:58: April next year in London, UK. If you would like further information on the Congress, or would like
26:58 - 27:03: a copy of the conference program, please contact the marketing department or visit the event’s
27:03 - 27:08: official website. On behalf of Oxford Global, I would like to thank everybody for joining us today.
27:08 - 27:13: I would like to say a big thank you once again to Alejandra, who has put a lot of time into
27:13 - 27:19: making sure that this webinar was as interesting and valuable as possible. All that remains to be
27:19 - 27:24: said is have a great day, and we look forward to welcoming you at one of our future events.