Any life scientist will readily sing you a song about expensive commercial antibodies which never work as described. Those more successful scientists will however instead tell you a tale of how an antibody which doesn’t really work for everyone else, performs beautifully in their own capable hands.

And this is exactly what Sholto David wants to tell you about. Hundreds of scientists in many labs have been using a totally wrong antibody simply because its target has a similar name to the gene they wanted to study. The really shocking part is that this wrong antibody miraculously produced exactly the “right” results in their hands. Well, maybe not miraculously. A triumph of the will, one could say.

Unlike in Sholto’s previously discovery of fake western blots which Thermo Fisher paraded in their antibody datasheets (see May 2026 Shorts), the fault here doesn’t lie with the vendors, which actually makes it much more serious. I contacted some of the corresponding authors mentioned in this article. Only one of them replied, and he didn’t sound worried. Which probably proves that their conclusions remain not affected?

Now, over to Sholto. I again meddled with some editing and bad jokes.

---

Mind over Antibody

by Sholto David

An almost unbelievable blunder has occurred in the field of senescence, hundreds of papers by dozens of labs have presented incorrect data by using the wrong antibody to analyse the critical tumor suppressor and cell cycle regulator usually referred to as “p16”. This mistake has been caused by researchers muddling two proteins with similar names but entirely different sequences and functions. High impact papers with hundreds of citations in Nature, Nature Medicine, Cancer Cell, and eLife have been caught in this mistake. First, a brief background:

p16-INK4a is a protein encoded by the CDKN2A gene. It acts as a tumour suppressor by halting the cell cycle. Accumulation of p16-INK4a is said to trigger a state of “cellular senescence”. p16-INK4a is widely studied in cancer biology and is considered a key biomarker of ageing. Because p16-INK4a is frequently studied alongside related proteins (especially p21 and p53) it is often abbreviated simply as “p16”.

Actin-related protein 2/3 complex subunit 5 is a protein encoded by the ARPC5 gene. It is a comparatively less well-understood component of the actin cytoskeleton. Various synonyms are circulating including ARPC5, ARC16, dJ127C7.3, and, fatefully, p16-ARC (this is an alternative name listed by UniProt). Importantly, this protein is not structurally or functionally related to p16-INK4a, and is not known to be involved in cell cycle control or ageing.

So in this story we have two unrelated proteins; a critical tumour suppressor that I will continue to refer to as p16-INK4a, and a comparatively obscure component of the actin cytoskeleton, which I will refer to as p16-ARC.

Antibodies are frequently used to investigate protein expression by western blot or immunostaining because of the specificity of their binding. Abcam is a major supplier of antibodies, and if you go to their website today and search for a “p16” specific antibody a list of potential choices is offered, importantly for this story a p16-ARC antibody is the first search result, the product code is “ab51243”. There is also another less frequently cited p16-ARC antibody “ab151303”. Similarly, if you head over to Santa Cruz (another supplier) you can find a p16-ARC antibody with the product code “sc-166760”. I think most people can guess where this is leading… How many researchers have muddled these two proteins and mistakenly ordered an antibody that binds p16-ARC when trying to investigate the expression of p16-INK4a?

I decided to check, searching for the “ab51243”, “ab151303”, and “sc-166760” product codes in Google Scholar I found over 400 unique research papers mentioning at least one of these p16-ARC antibodies (after removing duplicates, false positives, preprints, and dead links). Of these papers I could access the full text of 334, with the remaining 72 articles being paywalled. I reviewed each accessible paper to determine whether the antibody used in the paper was correctly intended for p16-ARC or incorrectly used to try and bind p16-INK4a

Here is the astonishing result: 95% of these papers have got it wrong. The vast majority of researchers who purchased either ab51243, ab151303, or sc-166760 have tried to use these antibodies to investigate p16-INK4a expression! Only seventeen used these p16-ARC antibodies correctly. It is truly an astonishing unforced error which might sound unbelievable at first. You are invited to check my working, so here is a spreadsheet of the identified papers.

The implications are not good, to put it mildly. And these are not just insignificant papers. As I mentioned previously there are papers with hundreds of citations in high impact journals claiming to probe for p16-INK4a with antibodies which (I repeat) do not bind p16-INK4a! Well, don’t take it from me, it’s right there on Abcam’s product page:

“This antibody (ab51243) is specific to the p16 ARC protein […] There is no observed cross reactivity to the CDKN2A/p16INK4a protein.”

Let’s look at some examples, and we may as well start with the error as published by Nature. This paper comes from the lab of Jan Karlseder, CSO and Senior Vice President at Salk Institute in La Jolla, USA, his colleague Reuben Shaw (director of NCI Cancer Center at Salk), and their collaborators at University of Heidelberg in Germany. The paper is focused on the cell cycle, of which p16-INK4a is a critical regulator. Here we’ll look at extended data Figure 1d:

Joe Nassour, Robert Radford, Adriana Correia, Javier Miralles Fusté, Brigitte Schoell, Anna Jauch, Reuben J. Shaw, Jan Karlseder Autophagic cell death restricts chromosomal instability during replicative crisis Nature (2019) doi: 10.1038/s41586-019-0885-0

The western blot is labelled ambiguously as “p16” (although the figure also presents p53 and p21 blotting alongside p16 which is a major hint), in the methods section the antibody is listed as “P16 (Abcam ab51243)”. The figure is discussed in the main text, quote:

“Human mammary epithelial cells (HMECs) escape from senescence through spontaneous silencing of P16INK4A and enter crisis at PD27 (Extended Data Fig. 1c, d).”

So there is simply no question that the listed antibody is wrong as they have named the protein as p16-INK4a!

Karlseder eventually replied to out inquiry with:

“I will discuss this with the lead author and get back to you. We have worked with the correct p16 antibodies for years, so I would be surprised if the wrong one was used here.”

Next, here’s a figure published in Nature Medicine, this one from a team that is just down the road from me, led by Simon Leedham at the Wellcome Trust Centre for Human Genetics and the John Radcliffe Hospital (University of Oxford, UK). In this case the authors specify that they stained p16-INK4a in human samples, see below. This is wrong as they also used the p16-ARC antibody ab51243.

Hayley Davis, Shazia Irshad, Mukesh Bansal, Hannah Rafferty, Tatjana Boitsova, Chiara Bardella, Emma Jaeger, Annabelle Lewis, Luke Freeman-Mills, Francesc C Giner, Pedro Rodenas-Cuadrado, Sreelakshmi Mallappa, Susan Clark, Huw Thomas, Rosemary Jeffery, Richard Poulsom, Manuel Rodriguez-Justo, Marco Novelli, Runjan Chetty, Andrew Silver, Owen J Sansom, Florian R Greten, Lai Mun Wang, James E East, Ian Tomlinson, Simon J Leedham Aberrant epithelial GREM1 expression initiates colonic tumorigenesis from cells outside the stem cell niche Nature Medicine (2015) doi: 10.1038/nm.3750

The authors, who include the Glasgow professor Owen Sansom (PubPeer record) and Michael Karin‘s former mentee Florian Greten (now eternally innocent professor at University of Frankfurt, Germany), mentioned another CDKN2A related antibody in this paper (PA1-30670 from ThermoFisher). They specified it as staining mouse p16-INK4a – this is also wrong though, PA1-30670 is supposed to bind p19-ARF which shares the same gene but is an entirely different protein because it is derived from a different reading frame (ARF stands for Alternative Reading Frame). So that’s awkward, they tried to stain p16-INK4a in humans and mice and failed both times by choosing the wrong antibody twice.

It would be unfair to pick on Oxford, so here’s a Cambridge team, led by emeritus professor Graham Burton, trying to western blot human placenta samples for p16-INK4a with the same wrong p16-ARC antibody (ab51243). Burton used to be founding chair of the Cambridge Reproduction initiative, allow us the pun to ask how reproducible his p16 result is:

Tereza Cindrova-Davies, Norah M.E. Fogarty, Carolyn J.P. Jones, John Kingdom, Graham J. Burton Evidence of oxidative stress-induced senescence in mature, post-mature and pathological human placentas Placenta (2018) doi: 10.1016/j.placenta.2018.06.307

Interestingly, the methods state that they also performed immunohistochemistry with this antibody although they only seem to show the results for p21 staining… I wonder why?

Here’s another example, this time from Immunity published by Cell Press, this team is from Stanford University School of Medicine, led by Jörg Goronzy and Staford’s former centre director Cornelia Weyand, both are now at Mayo Clinic in Rochester. Again, this paper also uses the ambiguous name “p16” but it is clear from the text that the authors intended to investigate p16-INK4a and not p16-ARC, for example, quote (highlight mine):

“To better define the aging process in naive CD4+ CD45RA+ T cells, we assessed the aging-associated cell-cycle inhibitors p16, p21, and p53″

Only p16-INK4a is an aging associated cell-cycle inhibitor related to p21 and p53, p16-ARC is certainly not. Now here’s Figure 5E and F:

Yinyin Li, Yi Shen, Philipp Hohensinner, Jihang Ju, Zhenke Wen, Stuart B. Goodman, Hui Zhang, Jörg J. Goronzy, Cornelia M. Weyand Deficient Activity of the Nuclease MRE11A Induces T Cell Aging and Promotes Arthritogenic Effector Functions in Patients with Rheumatoid Arthritis Immunity (2016) doi: 10.1016/j.immuni.2016.09.013

Note that CDKN2A is the gene for p16-INK4a making it clear (if it wasn’t already) which p16 the authors wished to investigate. We can also look at Figure 6H, and it is helpful to share the discussion of this figure, again linking the CDKN2A gene to the p16 protein (highlights mine):

“To examine a possible link between tissue invasiveness, inflammatory capability and cellular aging, we quantified tissue expression of CDKN2A, CDKN1A, and TP53 transcripts and measured p16 protein expression by dual-color immunohistochemistry.”

Here is a quote from the corresponding methods section (highlights mine):

“Slides were incubated with a primary antibody cocktail overnight at 2C to 8C. The cocktail contained both monoclonal mouse anti-human CD3 (Clone F7.2.38; 1:100, Dako), and anti-p16 ARC antibody (EP1551Y; 1:100, Abcam, ab51243).”

Interestingly, here the authors correctly describe the antibody as a p16-ARC antibody (presumably copied directly from the Abcam product page) but it is quite clear they had no interest in the actin related protein… the word “actin” does not even appear in this paper!

This isn’t the only paper from Rochester academics. And I was not going to write a blog without poking fun at Scientific Reports and their “excellent team”! So here is a study led by Irfan Rahman on a mouse model of COPD which they produce by forcing the mice to smoke “research grade” cigarettes. And then they blotted the wrong p16 (ab51243 again).

Kahkashan Rashid, Isaac K. Sundar, Janice Gerloff, Dongmei Li, Irfan Rahman Lung cellular senescence is independent of aging in a mouse model of COPD/emphysema Scientific Reports (2018) doi: 10.1038/s41598-018-27209-3

You might be interested to learn that University of Rochester’s professor Rehman is not only among the “top 0.03% of Research Scholars in the world“, but also owner of a disastrous PubPeer record, which, what a surprise, includes many duplicated western blots and microscopy images. Also on the topic of smoking, see for example Moodle et al 2004, Marwick et al 2006, Marwick et al 2010, Edirisinghe et al 2008, Arunachalam et al 2010 or Yogeswaran et al 2022, or if you are rather into curcumin, see Meha et al 2008. No prizes for guessing how Professor Rehman made the wrong p16 antibody produce the right results in his lab.

What’s better than Scientific Reports? MicroRNA + Scientific Reports. A legendary partnership. Here the Spanish authors from Madrid describe p16 as a “conventional cellular marker of senescence” and western blot with ab51243 again.

Matilde Alique, Guillermo Bodega, Chiara Giannarelli, Julia Carracedo, Rafael Ramírez MicroRNA-126 regulates Hypoxia-Inducible Factor-1α which inhibited migration, proliferation, and angiogenesis in replicative endothelial senescence Scientific Reports (2019) doi: 10.1038/s41598-019-43689-3

Sorry, back to serious research in proper journals and this next paper is from Cancer Cell published by Stephen Baylin at The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University in USA. Once again we’re looking at a p16 western blot which was performed with the ab51243 antibody, the authors describe p16 here as a “known anti proliferative gene” (i.e. a tumour suppressor) meaning they must have intended to target p16-INK4a.

Xiangqian Kong, Jie Chen, Wenbing Xie, Stephen M. Brown, Yi Cai, Kaichun Wu, Daiming Fan, Yongzhan Nie, Srinivasan Yegnasubramanian, Rochelle L. Tiedemann, Yong Tao, Ray-Whay Chiu Yen, Michael J. Topper, Cynthia A. Zahnow, Hariharan Easwaran, Scott B. Rothbart, Limin Xia, Stephen B. Baylin Defining UHRF1 Domains that Support Maintenance of Human Colon Cancer DNA Methylation and Oncogenic Properties Cancer Cell (2019) doi: 10.1016/j.ccell.2019.03.003

In fact, Baylin and his former mentee Limin Xia are repeat offenders because they published another western blot using the same wrong antibody also in Cancer Cell two years prior! In this case they also discussed p16 in the context of senescence confirming their intended target.

Limin Xia, Wenjie Huang, Marina Bellani, Michael M. Seidman, Kaichun Wu, Daiming Fan, Yongzhan Nie, Yi Cai, Yang W. Zhang, Li-Rong Yu, Huili Li, Cynthia A. Zahnow, Wenbing Xie, Ray-Whay Chiu Yen, Feyruz V. Rassool, Stephen B. Baylin CHD4 Has Oncogenic Functions in Initiating and Maintaining Epigenetic Suppression of Multiple Tumor Suppressor Genes Cancer Cell (2017) doi: 10.1016/j.ccell.2017.04.005

Xia is now professor at Huazhong University of Science and Technology in Wuhan. And as for clues how she made this wrong antibody work: also Xia has her own PubPeer record of duplicated and cropped microscopy images, see for example Li et al 2018 or Huang et al 2015, the latter somehow corrected in 2022.

Here’s a recent paper in Science Advances from Jonathan Keller’s lab at the NIH’s National Cancer Institute in Bethesda, USA. From the abstract regarding the analysis of the results: “Mechanistically, p16 expression, senescence, and apoptosis were increased…” making it clear which p16 they were interested in. Here is Figure 5:

Shweta Singh, Kristbjorn O. Gudmundsson, Tanmoy Sarkar, Brad L. Jakubison, Holly M. Morris, Sandra Burkett, Karim Baktiar, Gary T. Pauly, Dina M. Sigano, Joel P. Schneider, Lino Tessarollo, Jonathan R. Keller Id1 promotes clonal hematopoiesis in mice with Tet2 loss of function Science Advances (2025) doi: 10.1126/sciadv.adr5867

I don’t think it’s quite worth trying to explain the whole figure but the method requires an antibody to stain cells as p16 positive or negative and the only p16 antibody provided in the methods is sc-166760, a p16-ARC antibody. In fact, Keller’s lab has been using this wrong antibody since at least 2018 (seven years earlier) to try to stain p16-INK4a, as he has another publication in the Cell family journal Stem Cell (Singh et al 2018). Keller has now graduated to NCI Scientist Emeritus.

How about a Russian-American-Spanish anti-ageing collaboration on the subject of senolytic drugs? If the Spanish succeed in reversing ageing they might reach their long-held ambition of 100% youth unemployment. Alas – they tried to blot p16-INK4a but used the wrong antibody (ab51243) and the experiments are doomed. I’m joking of course, senolytics isn’t real research anyway, it doesn’t matter what you blot.

Uxía Nogueira-Recalde, Irene Lorenzo-Gómez, Francisco J. Blanco, María I. Loza, Diego Grassi, Valery Shirinsky, Ivan Shirinsky, Martin Lotz, Paul D. Robbins, Eduardo Domínguez, Beatriz Caramés Fibrates as drugs with senolytic and autophagic activity for osteoarthritis therapy eBioMedicine (2019) doi: 10.1016/j.ebiom.2019.06.049

Here’s another US-based lab blotting the wrong p16. The lead author Victor Thannickal is now Chair of the John W. Deming Department of Medicine at Tulane University. Again these authors used the ab51243 antibody.

Yan Y. Sanders, Hui Liu, Xiangyu Zhang, Louise Hecker, Karen Bernard, Leena Desai, Gang Liu, Victor J. Thannickal Histone Modifications in Senescence-Associated Resistance to Apoptosis by Oxidative Stress Redox Biology (2013) doi: 10.1016/j.redox.2012.11.004

Also Professor Thannikal has a PubPeer record of questionable science, with duplicated western blot bands and microscopy images, see for example Jiang et al 2021, Ghatalk et al 2017, Huang et al 2011 or Hecker et al 2009.

The following is a charming study that seems to involve bullying mice to death; title “Social stress shortens lifespan in mice”. The authors, mostly at the University of Minnesota, claimed to measure a senescence signal in mouse organs, partly by p16-INK4a expression… but they’ve used the wrong antibody (ab51243).

Maria Razzoli, Kewir Nyuyki-Dufe, Allison Gurney, Connor Erickson, Jacob McCallum, Nicholas Spielman, Marta Marzullo, Jessica Patricelli, Morito Kurata, Emily A. Pope, Chadi Touma, Rupert Palme, David A. Largaespada, David B. Allison, Alessandro Bartolomucci Social stress shortens lifespan in mice Aging Cell (2018) doi: 10.1111/acel.12778

The coauthor and Minnesota professor David Largaespada is a former collaborator of the infamous Catherine Verfaillie, and coauthor on her now-retracted Nature paper Jiang et al 2002, which fraudulently claimed that bone marrow cells were pluripotent.

Another coauthor, former Indiana University dean David Allison (now at University of Alabama at Birmingham), is actually a leading champion of reproducibility, celebrated by Nature:

We will see how hard it will be to fix this mistake in Allisson’s own paper.

A natural question that follows from these observations: If p16-ARC has no relationship to the cell cycle or ageing, what exactly should we see when it is accidentally investigated in these papers? Data collated by the Human Protein Atlas for p16-ARC (ARPC5) shows comparatively high levels of transcription and (where measured) protein expression across commonly used cell lines. p16-ARC is known to be a component of the actin cytoskeleton, so logically (in most circumstances) it should track β-actin expression.

In other words, as far as western blots go, it should look like a loading control!

In fact several papers appear to (accidentally) stumble into this finding. Here’s an example; researchers from Fudan University in China analysed the response of senescence related genes in cells exposed to increasing doses of radiation:

Jiangtao Bai, Yuyang Wang, Jianping Wang, Jianglong Zhai, Feilong He, Guoying Zhu Irradiation-induced senescence of bone marrow mesenchymal stem cells aggravates osteogenic differentiation dysfunction via paracrine signaling American journal of physiology. Cell physiology (2020) doi: 10.1152/ajpcell.00520.2019

Quote from this paper related to the above figure (highlight mine):

“Furthermore, results of the mRNA and protein levels of senescence-related genes showed that irradiation caused upregulation of p53 and p21 from 2 Gy to 10 Gy, whereas p16 had no obvious change at any dose of gamma rays (Fig. 4, E–G)”

The explanation, of course, is that they were not blotting p16-INK4a, but p16-ARC, which will likely track β-actin expression (as it does here). Another paper with a similar experiment made the same observation, this time from Third Military Medical University in Chongqing:

Lili Chen, Qian Ran, Yang Xiang, Lixin Xiang, Li Chen, Fengjie Li, Jiang Wu, Chun Wu, Zhongjun Li Co-Activation of PKC-δ by CRIF1 Modulates Oxidative Stress in Bone Marrow Multipotent Mesenchymal Stromal Cells after Irradiation by Phosphorylating NRF2 Ser40 Theranostics (2017) doi: 10.7150/thno.17853

Quote from the authors regarding this figure (highlight mine):

“Several proteins related to cell senescence were detected by western blot after irradiation at 48 and 72 h (Figure 1F). The results showed that protein levels of RB, p53, and p21 significantly increased, confirming that cell senescence was induced by irradiation, while no remarkable change of p16 was observed.”

Well, once again, the explanation is rather simple, the expression of p16-ARC should probably track β-actin (and it does!).

Perhaps more interesting is that so many of these papers have managed to produce the “expected” behaviour of p16-INK4a in response to various experimental manipulations despite using the wrong antibody. Let’s look at another example, in eLife, the authors affiliated in China and Australia measured what they believed to be p16-INK4a expression by immunofluorescence and western blot, quote:

“p16 expression decreased after siRNA-mediated knockdown, which decreased the proportion of NP cells that demonstrated a senescent phenotype. By contrast, p16 overexpression caused a marked opposite effect…”

Hui Che, Jie Li, You Li, Cheng Ma, Huan Liu, Jingyi Qin, Jianghui Dong, Zhen Zhang, Cory J Xian, Dengshun Miao, Liping Wang, Yongxin Ren p16 deficiency attenuates intervertebral disc degeneration by adjusting oxidative stress and nucleus pulposus cell cycle eLife (2020) doi: 10.7554/elife.52570

The authors only provided one p16 antibody product code (ab51243), along with specified dilutions for western blotting and immunofluorescence.

What are we to make of cases like this where the wrong antibody was used but the authors still manage to rustle up interpretable results? One explanation is that the data has been manipulated to conform to the researcher’s expectations or desires… i.e. fraud. And some of these are the most obvious frauds, at least four have been retracted, here’s one in the style of Elisabeth Bik’s famous tadpole papermill. It wasn’t noted at the time, but they used the wrong antibody for p16-INK4a here (ab151303 is a lesser cited p16-ARC antibody).

Zhijia Yang, Guoqing Li, Chao Ding, Wencong Sun, Ji Zhang Long non-coding RNA HULC exerts oncogenic activity on papillary thyroid cancer in vitro and in vivo Artificial Cells Nanomedicine and Biotechnology (2020) doi: 10.1080/21691401.2019.1703730

Another potential explanation for these impossible results is that the experiments actually were performed with a p16-INK4a specific antibody, but at the moment of writing the paper, a p16-ARC antibody code was retrieved from Abcam or Santa Cruz’s website and inserted into the methods section erroneously. This second explanation might seem more charitable in some sense, but I’m not sure it is… are we to believe that antibody codes in papers are selected at random by a lab tech based on the first search result? For papers to be meaningful the methods have to describe what was actually done and the results have to show what was observed.

In the end there is likely a mixture of different explanations ranging from outright fabrication, selective reporting, writing errors, and some teams blindly publishing contradictory findings without further questioning or curiosity. There is something unsettling about the idea that hundreds of papers can be published using completely the wrong antibody without anyone noticing, as if the research results are only loosely connected with biological reality. How many antibodies could you swap for loading controls until someone actually takes notice of what the results are telling them?

Anyway, I’m not going to write about all the papers here, but I will try to post them all to PubPeer eventually. There are plenty more interesting examples, including one paper where some Chinese researchers blasted some cells into space (Mao et al 2024), then stained them with the wrong p16 antibody, shame, perhaps they’ll have to repeat the rocket launch? Or what about this experiment on microplastics and mouse testicle ageing (Wu et al 2023), already cited over 100 times I expect it will be extensively laundered into the scientific literature by very serious reviews, but how competent are the researchers who can’t even order the right antibody?

I did wonder how widespread this problem is. Well, speculation is irresponsible and unscientific so let’s do that. If we take this question to mean something like: “Of those papers that probe for p16-INK4a, what percentage mistakenly used a p16-ARC antibody?” I would guess likely less than 10% but maybe more than 5%. If that sounds like a lot… Well, I already have a list of over 300 incorrect papers so in order to prove me wrong you would need to find at least 6000 papers which correctly used a p16-INK4a antibody. Based on my searches for valid p16-INK4a antibody product codes I think that might be difficult! And this remains a highly active problem, in fact 2025 was a record year for wrong p16 antibody papers. As long as the first search result for “p16” on the Abcam website is a p16-ARC antibody I think this will continue.

For sure this error does not impact as many papers as related incidents where non-specific antibodies have been distributed by manufacturers, Entrop et al convincingly argued in 2024 that >1400 papers have used a non-specific antibody for BAX sold by Santa Cruz (sc-7480). But the case presented here has a different quality, more like an own goal, given that these p16-ARC antibodies were never claimed to be p16-INK4a antibodies in the first place.

So here is a quick post mortem on what went wrong here: On the side of the researchers who made this error, obviously the naming collision is unfortunate, and these proteins have similar names precisely because they are about the same molecular weight, thus making the confusion harder to spot by western blot because they will run to the same size on a gel.

On the side of not making thoroughly clownish mistakes this could have been so easily been avoided by running simple well-established controls. ThermoFisher has a very clean example for what this might look like on one of their p16-INK4a antibody pages. Well, maybe don’t trust a ThermoFisher blot (see May 2026 Shorts as to why), but in principle they show a sensible verification that would have immediately caught this problem, the key control here is the PC-3 cell line (presumably the lung carcinoma cell line) which has a homozygous CDKN2A deletion and therefore cannot express p16-INK4a. Not to be confused with the prostate cancer cell line with the same name, but I digress, another naming confusion for another day. Similar experiments with knockouts or siRNA are possible.

However, it seems as if most researchers purchased these p16-ARC antibodies and must have immediately implemented them in their research without further testing or controls. The Nature paper cited above includes a statement about antibody validation in the reporting summary: “Antibodies were validated by the size of the band and by targeting the corresponding product by siRNA.” – it would be interesting to see the data!

I’m quite sure there are more soundalike/sizealike proteins to find. Amusingly the senescence related protein usually referred to as “p21” also has an imposter p21-ARC protein! In these two papers by Chinese researchers based in Chengdu the team have managed to simultaneously blot for the wrong p16 (sc-166760) and the wrong p21 (sc-166630).

Peng Zhang, Qian Wang, Lulingxiao Nie, Rui Zhu, Xinyi Zhou, Pengfei Zhao, Ning Ji, Xing Liang, Yi Ding, Quan Yuan, Qi Wang Hyperglycemia-induced inflamm-aging accelerates gingival senescence via NLRC4 phosphorylation Journal of Biological Chemistry (2019) doi: 10.1074/jbc.ra119.010648

The second paper was published in the previously well regarded Journal of Biological Chemistry:

Lulingxiao Nie, PengFei Zhao, Ziqi Yue, Peng Zhang, Ning Ji, Qianming Chen, Qi Wang Diabetes induces macrophage dysfunction through cytoplasmic dsDNA/AIM2 associated pyroptosis Journal of Leukocyte Biology (2021) doi: 10.1002/jlb.3ma0321-745r

This p21 error not only effects Chinese labs, but even the highly respectable former rector of the University of Messina, one Salvatore Cuzzocrea, who is surely familiar to readers.

So, here it is:

Francesco Molinari, Nicla Tranchida, Francesca Inferrera, Roberta Fusco, Caterina Faggio, Federica Impellitteri, Salvatore Cuzzocrea, Marika Cordaro, Rosanna Di Paola Biocide mixture (CMIT/MIT) induces neurotoxicity through the upregulation of the MAPKs signaling pathways Journal of Neurophysiology (2025) doi: 10.1152/jn.00104.2025

And I could also talk about p38 and direct you to Han et al 2023… but I must stop writing now, or this could go on forever.

I considered whether I should email Abcam and ask them to change the name of the ab51243 antibody on their website to reduce confusion, but I won’t be doing that. It feels too much like asking Abcam to put the bowling lane bumpers up. I will instead continue to watch the carnage, perhaps it has only just started!

---