Some remarks on purported fraud concerned with αβ oligomer hypothesis in Alzheimer's research.

[NNadir]

Last edited Mon Jul 25, 2022, 09:36 AM - Edit history (1)

The news article in the prominent scientific journal, Science, to which I will refer in this post is this one: Blots on a Field Charles Pillar, Science 21, July, 2022.

The subtitle is here:

A neuroscience image sleuth finds signs of fabrication in scores of Alzheimer’s articles, threatening a reigning theory of the disease

Mr. Piller is an investigative journalist. Science is not generally - primarily - involved in "Investigative Journalism;" it is a well-respected prominent scientific journal published by the American Association for the Advancement of Science, the AAAS. I am a member of the AAAS by virtue of being a subscriber to Science; no evaluation of my credentials was required to become a member, but I am a working scientist, generally in a management role. One also gets a nice tee shirt every year when one renews one’s subscription to Science with the AAAS logo on it. My favorite is the one from a few years back that reads:

Facts are facts.

Somehow this tautology has become a controversial political statement.

I logged out of Science to ascertain that the full article is open sourced. Thus I will limit the number of excerpts of the full article; anyone who is seriously interested in the facts of the case can read it themselves. It is a compelling article and makes a very solid case that fraud may well have occurred. I certainly make a number of comments from time to time criticizing journalistic interpretations of science, making sweeping jokes that I apply to all journalists. My critique would not apply to Mr. Piller, from what I can tell. He is very fair and does a fine job engaging in the science underlying his piece; it does however, overall, read like investigative journalism. Of course, he has the benefit of producing his work in an organization led and supervised by scientists.

A post here at DU links has a more hyperbolic title than the Science article, Two decades of Alzheimer's research may be based on deliberate fraud that has cost millions of lives, however the author of the post was simply reproducing the title found in article at Daily Kos, to which the post referred, adding a tweet on the subject.

The Daily Kos article, which was written by a sometime Geologist who apparently now writes Syfy novels, and who has as well as a written popular science book, Mark Summer, a staffer at Daily Kos.

Two decades of Alzheimer's research may be based on deliberate fraud that has cost millions of lives.

Several other posts have appeared at DU with the same title. This stuff is getting a lot of air time.

Daily Kos is owned by a journalist, but I’ll make no further comment.

I commented on the Tweet in the DU post: The tweet may be a considerable overstatement.

I hadn't read the Daily Kos diary as of the time I read the post and responded to the "tweet" referring to the Daily Kos diary; I quickly read through it at this morning, an easy read with very little technical detail. Mr. Summer seems to have taken a look at the Science news article on which I will comment here, and has provided a link to the original Nature paper, whether he fully accessed and read either is not clear from what he wrote. Whether he did or not does not come through the innuendo clearly present, although he does make several brief statements which are clearly true, if superficial. The post is consistent with how I personally view as a Daily Kos level discussions of issues in science.

I have downloaded and looked through the Nature paper. It is here: Lesné, S., Koh, M., Kotilinek, L. et al. A specific amyloid-? protein assembly in the brain impairs memory. Nature 440, 352–357 (2006). I have highlighted the name of the scientist under suspicion who is discussed at length in the Science news feature. Another scientist discussed in the news article is Dr. Karen Ashe. As I understand it, she developed the transgenic mice.

The Nature article now comes with an "Expression of Concern" dated July 14, 2022:

14 July 2022 Editor’s Note: The editors of Nature have been alerted to concerns regarding some of the figures in this paper. Nature is investigating these concerns, and a further editorial response will follow as soon as possible. In the meantime, readers are advised to use caution when using results reported therein.

For full disclosure: I read the Science article, before reading anything on political blogs, this past Friday because I have been involved, peripherally, in supporting three different Alzheimer's projects for three different (gasp) pharmaceutical companies. I am not an expert in any sense of the word on the mechanism or physiology of Alzheimer's, although I make a serious effort to at least familiarize myself with the basic tenets of the science. As I am an old man, I also have some personal interest in dementia, since anyone my age should consider themselves as under a sword of Damocles in this respect. The work in which I am participating is involved with conducting analytical support, in two cases, measuring pharmacokinetic pathways, including in one case of monitoring certain biomarkers, in another, properties of a formulation. I cannot say much about this work; I am bound by CDA's, but none of the projects on which I am working is involved with Cassava Science, a company which comes up in the Science article. I have nevertheless also read the statement on this topic written by the Cassava CEO for general interest; their lead drug has followed the appropriate regulatory pathway and is about to enter Phase III clinical trials, although the company has lost over a billion dollars of valuation based on the issues discussed in this report. (Failures of Phase III clinical trials of Alzheimer's drugs are legion, but this does not rule out that anyone's phase III clinical trial, including that being conducted by Cassava, will not fail.) For the record and further full disclosure, one study in which I am involved will utilize the ?? generating transgenic mice - I have no idea whether these are the mice developed by Dr. Ashe - that have been engineered to produce the Amyloid beta precursor like protein 2 as well as the plaques resulting from their hydrolysis, however, the mode of action of the drug does not depend on the ?? oligomer subtype hypothesis intimately. Thus none of the projects on which I am working are at risk, at least in the absence of hysteria.

The CEO of Cassava Sciences is not a scientist. He's an investment type, and plainly confesses he doesn't know much about Western blots. His comments on the use of a "Citizens Petition" are well taken. I can recall no instance in my long career, in which a "Citizen's Petition" has been employed for clinical trials on an NCE (New Chemical Entity) for the treatment of human disease. I am aware of "Citizen's Petitions" involved with approved drugs, generally for patent cases as well as a number involved with the opioid crisis. I was involved with one filed in the latter category, and have been involved, indirectly, in producing data for the former case. I also fully credit the CEO of Cassava's note that money changed hands; that there was a financial position involved, something the Science news feature fully discloses in its generally and I think appropriately laudatory description of Dr.Matthew Schrag of Vanderbilt University. None of this of course has merit as to whether fraud has taken place or not, however.

I am in no position to state that the Cassava Sciences drug will fail either. The Citizens Petition filed by short selling scientists as described clearly, but anomalously in the Science news feature, has been rejected by the FDA.

I do have a scientific opinion on Western blots, although I have never run one or even actively interpreted one. I know how they work and what they are, and have occasionally looked at them in papers, without all that much sophistication. This said, I believe that overall the Western Blot technique should ultimately go the way of the dinosaur, dinosaurs being subject on which Mr. Summer apparently, from his bio, has some professional experience, although I strongly suspect his familiarity with Western Blots is not a sophisticated as even my very limited experience with them. Here's what I said about the subject in my comment in the DU post:

The claim was that a specific set of small oligomers derived from the the ?? protein were responsible for inducing the plaques. The Science article indicates that there is strong suspicion of manipulated images in a technique that, in my opinion, should rapidly become obsolete, Western Blot, kind of a cross between TLC, SEC, and electrophoresis. The technique utilized to raise the doubts involved computer processing of some images.

This said, Western blots are cheap and easy to run and do not require the very expensive equipment and sophisticated software that are required for the technologies I have implied might supplant them. They are, and have been, good scientific tools, particularly where cost is an issue, and, especially now, as the Science investigative journalism article shows, there are very sophisticated tools to determine or at least suggest whether one is cheating and manipulating images of their output.

About cost:

I recently screened for a potential purchase of a high end TIMS-TOF (Trapped Ion Mobility Spectroscopy Time of Flight) Mass Spec, which is said to offer unprecedented ion utilization efficiency along with an added degree of analytical orthogonality. A single instrument, which may have a run time as long as three minutes, or perhaps very much longer depending on the use chromatographic separations, thus placing limits on the number of samples that can be screened will run roughly at a million dollars, plus ancillary costs, not the least of which is reagents, obtaining samples and preparing them, as well as, most importantly, the cost and time involved in training staff to run the instrument and further, most importantly to interpret the results.

A very large screening project may require many such instruments, and many experiments utilizing them, and the potential for honest error is high, meaning that experiments with ambiguous results may require many repetitions. It is true that these instruments will produce a deeper level of data, and many instruments today are designed with software designed to be 21CFR Part 11 compliant, a kind of IT system that is designed to disallow or track data manipulation. However, not every lab can afford such luxuries, and in any case, there are certainly instances in which they are overkill.

Mr. Summer takes time in his Daily Kos rant to complain about pricing of the only approved Alzheimer's drug, the controversial drug Aduhelm (Aducanumab). Single doses of the drug are given over periods several weeks, apparently at 21 day intervals. Thus the maximum number of doses in a year will be about 17. All of the development costs of the drug will need to be covered by very few doses compared to the doses of say, blood pressure drugs, or drug even antibiotics, which may involve hundreds, even thousands of doses, over a year. (Antibiotics are not huge money makers compared to chronic dose drugs; their economics as well as their widespread and entirely legal use and abuse will make this a problem for future generations, about whom many people my age apparently couldn't care less: All we seem to care about is our money and our convenience.)

I don't know if Mr. Summers was paid for his geological work on dinosaur fossils, or if his books are distributed for free. Writing a book, of course, is a risky enterprise if the goal is to make money, but hardly nowhere near as risky as developing an Alzheimer's drug, given the high rate of failure for many well known efforts as well as the legitimately questioned efficacy of Aducanumab. Writing a book or a blog post on Daily Kos doesn't require a one million dollar equipment investment, a computer from Best Buy or Staples or the Apple store will do, and not much training is required to use a computer.

People who develop drugs are in a morally difficult place; their products can do what was once magic, cure or at least treat the symptoms of disease. It is only human - I personally question the humanity of Republicans - to think that a sick person should be treated, and that cost should not be the only factor under consideration. Sometimes this exercises the noble altruism muscles of people who consider themselves altruistic because they bought an electric car from Elon Musk, among other things. I am proud of the projects on which I worked in my career that treated the ill, but regrettably perhaps, I do want to be paid for my work.

I do understand that there is considerable room for discussion of drug pricing, and egregious abuses are well known, but we should not isolate the cost of doing science from the potential rewards, some of which are financial, particularly if we are cherry picking only to emphasize abuses. Again, I want to be paid for what I do.

Perhaps I should apologize for that, but I won't.

One of the issues of the soundbite/tweet era is magnification of the trivial over greater good that requires critical thinking to recognize. (This is hardly limited to pharmaceuticals but also plays an important role in environmental decisions, most of which, left and right, are delusional.)

Now I'll briefly turn to the aptly named Mr. Piller's article to point to his description of the technology that can be used, from this time forward, to keep people who rely on Western Blots for their publications honest. I'll reproduce the text on the five steps that are used by PubPeer, the technology utilized by Dr. Schrag in his investigation. I'll produce the text from the news feature, as the paper is open sourced, the interested reader is invited and encouraged to look at the graphics attached. I won't reproduce them here.

Image in question Ashe uploaded this Western blot to PubPeer after Schrag said the version publishedin Nature showed cut marks suggesting improper tampering with bands portraying A?*56 and other proteins (black boxes added by Ashe). The figure shows levels of A?*56 (dashed red box) increasing in older mice as symptoms emerge. But Schrag’s analysis suggests this version of the image contains improperly duplicated bands...

The PubPeer link in this excerpt will take the reader to papers by Dr. Lesné that are under review for potential fraud.

Step One:

1 Spot the similarities. Some bands looked abnormally similar, an apparent manipulation that in some cases (not shown) could have made A?*56 appear more abundant than it was. One striking example (red box)ostensibly shows proteins saidto emerge later in the life span than A?*56.

Step Two:

2 Match contrast. Schrag matched the contrast level in the two sets of bands foran apples-to-apples comparison.

Step Three:

3 Colorize and align. Schrag turned backgrounds black to make the bands easier to see, then colorized them and precisely matched their size and orientation.

Step Four:

4 Merge He merged the sets of colorized bands. The areas of the imagethat are identical appear in yellow

Refer to the article to see the color.

5 Calculate similarity Schrag then calculated the correlation coefficient, showing the strength of the relationship between the merged bands. Identical images show a correlation of 1, and display as a straight 45° angle line. These bands show a 0.98 correlation, highly improbable to occur by chance.

In this case a high correlation is a bad thing, not a good thing. Similarity algorithms play a large role in many areas of science, they can be displayed graphically and are in the article, to which, again, it is useful to refer, showing similar (undesirable if fraud is being investigated) and dissimilar (desirable to disprove fraud) images.

It's a very good technique, I think, and makes a good case that Dr. Lesné may be "cooking" (the scientific) "books."

But what does this mean and how does it compare what does Mr. Summers' headline implies?

Well first of all, in my view, the statement that this fraud "cost millions of lives" is a extremely dubious statement. The involvement of ?? plaques in Alzheimer's disease has been known since the early 20th century. It's what Dr. Alzheimer, for whom the disease is named, discovered. Dr. Alzheimer lived in a time where immunology was poorly understood; indeed proteomics itself was a practically non-existent science. The question is whether the plaques are a cause of the disease or whether they are a symptom of the disease. The first hypothesis, that they are a cause, was around probably when Dr. Lesné was in diapers. His highly cited paper - Google Scholar indicates 3,265 citations - was offered as evidence that the first hypothesis was correct, but if, as the Science News Feature suggests is fraudulent, this does not prove that the hypothesis is incorrect, only that it is not supported by this particular finding.

Amyloid-beta precursor protein, accession number P05067 · A4_HUMAN in the Uniprot protein database has many important cell functions, described at the accession number link. In the section on diseases in which it is involved it says this about Alzheimer's Disease 1:

The disease is caused by variants affecting the gene represented in this entry

Description

A familial early-onset form of Alzheimer disease. It can be associated with cerebral amyloid angiopathy. Alzheimer disease is a neurodegenerative disorder characterized by progressive dementia, loss of cognitive abilities, and deposition of fibrillar amyloid proteins as intraneuronal neurofibrillary tangles, extracellular amyloid plaques and vascular amyloid deposits. The major constituents of these plaques are neurotoxic amyloid-beta protein 40 and amyloid-beta protein 42, that are produced by the proteolysis of the transmembrane APP protein. The cytotoxic C-terminal fragments (CTFs) and the caspase-cleaved products, such as C31, are also implicated in neuronal death.

I'd like to emphasize the 1. As is the case with the generic term, "cancer" it now seems possible that Alzheimer's is not just one disease. The alternate inflammatory hypothesis may account based on similarities for a relational dementia between Alzheimer's and CTE, the disease most often found in football players as the result of multiple concussions. CTE may or may not account for Georgia Senate candidate Hershel Walker; I'm not qualified to say, but I can suggest as much anyway. The inflammatory pathway results in some grousing by its supporters as noted in Mr. Piller's investigative journalism piece:

Scientists who advance other potential Alzheimer’s causes, such as immune dysfunction or inflammation, complain they have been sidelined by the “amyloid mafia.” Forsayeth says the amyloid hypothesis became “the scientific equivalent of the Ptolemaic model of the Solar System,” in which the Sun and planets rotate around Earth.

The second sentence may be a bit of an overstatement by Dr. Forsayeth. It may Ptolemaic for one "AD" and not for another if in fact, "Alzheimer's disease" is more than one disease. Let's not "jump the gun."

As for Mr. Summers, he's making a wildly unjustified assumption with his headline. He's assuming that if the ?? process had not been supported by the suspect work, lives would have been saved. In his contempt for the pharmaceutical industry, he assumes, with zero justification that if money had been spent differently, a cure would exist. The polite word for this is "bullshit." Americans, somewhat arrogantly believ based on the Manhattan Project and the Apollo project, that every problem can be solved if only enough money is thrown at it. This is not true. I personally regard the money thrown at so called "renewable energy," a policy I very strongly am inclined to believe that Mr. Summers supports, as all having been wasted. Trillions of dollars thrown at so called "renewable energy" has done nothing, zero, zilch, nada, zip, to address climate change as we can see because the planet is burning.

There is no evidence, zero, zilch, nada, zip, that spending money differently on Alzheimer's research other than the way money was spent would have "saved lives." It may have done so, or it may have not done so. I personally hope that the immune/inflammatory hypothesis will work; I have great respect for, and rather like, the people I know working on it. Full disclosure, they help finance my paycheck.

It is possible that if the same money that was thrown at the “amyloid mafia” was thrown at the inflammatory/immunomoderated hypothesis, a better treatment may have been found a "cure" or at least a moderately effective treatment, but I would suggest that Mr. Summers rather myopic view does not guarantee as much. This is a very difficult disease, and the study of the plaques is hardly a wasted effort, in particular because as indicated in the Uniprot accession document under functions it says this:

The gamma-CTF peptides as well as the caspase-cleaved peptides, including C31, are potent enhancers of neuronal apoptosis.

Caspases are enzymatic hydrolysis agents - about 10 are known in mammalian species - that cleave at the amino acid asparagine (Asn), in contrast to the widely used trypsin and chymotrypsin utilized in analytical chemistry to cleave proteins at, arginine or lysine residues for trypsin, and aromatic residues for chymotrpsin.

Asparagine is a rather interesting residue. It's formally a derivative of aspartic acid, it is the ? amide of aspartic acid, to which it can, and does, revert via amide bond cleavage. The ? nitrogen has an interesting property; it can be glycosylated (functionalized with a very complex sugar). This is called "N glycosylation." Other oxygen functionalized amino acids can also be glycosylated, serine, somewhat more rarely threonine or tyrosine (O-glycosylation). Because of the relative ease with which the amide bond can be broken in comparison to the carbon oxygen bond, it is much simpler to characterize N-glycans than O-glycans, although significant progress with the latter has been made in the last few years. Glycans are like trees, often highly branched in various ways that affect the function of proteins. One important feature of a N-glycosylated asparagine is that it is probably unavailable for hydrolysis by a caspase. It can also undergo rearrangement to form was we call ? linkages.

I downloaded the sequence for the Amyloid-beta precursor protein. I count 31 asparagines in the sequence. The glycosylation of this protein has been the subject of 23,000 papers, many of which predate Dr. Lesné's famous or perhaps infamous 2006 paper in Nature. It strikes me as entirely possible - and I am not qualified to do any more than speculate wildly, perhaps not at the same level as Mr. Summers speculates - that the ultimate cause of Alzheimer's is connected with glycosylation. Again, wild speculation.

In any case, there is evidence in other papers that the Amyloid-beta precursor protein is involved in the cause of at least some cases of Alzheimer's, perhaps just one of the possible "more than one" cases of disease. This is because caspase inhibitors seem to effect the cognition of Alzheimer's model mice:

Flores, J., Noël, A., Foveau, B. et al. Caspase-1 inhibition alleviates cognitive impairment and neuropathology in an Alzheimer’s disease mouse model. Nat Commun 9, 3916 (2018). None of Dr. Lesné's papers are cited in this paper.

I'm not fond of Mr. Summers rhetoric and frankly, by way of even more full disclosure, I have been banned from Daily Kos for interpreting a scientific paper by the climate scientist Jim Hansen in a way that Mr. Kos did not like. I contend that Mr. Kos's view of how to address climate change is dangerous in the extreme - I think the world is aflame because his view has clearly experimentally failed - and I further contend that Mr. Summers' view of Alzheimer's research, while not quite as dangerous as Mr. Kos's dogmatic climate views, is not helpful, inasmuch as it incites contempt for a serious and hardly disproven scientific hypothesis, albeit one that seems likely to have been supported, partially, by fraud.

I have a low opinion of discussions of science at Daily Kos.

Mr Summers, by choosing his title, has engaged in hyperbole of the type that clearly sets the internet in flames, and may do serious damage to research that may work - it also may not work - to treat this serious disease by reducing public support.

I thought I should at least comment on this.

A little critical thinking would be in order.

Have a pleasant Sunday evening.

[elleng]

1. Thanks; gotta take time to read.

[niyad]

2. KNR and bookmarking. For later.

[MLAA]

3. Thank you for taking time to review, interpret and share your thoughts on the article.

[eppur_se_muova]

4. Thanks for going so deep into the weeds on this one.

A more appropriate summary of the situation should be "doubts have been plausibly raised" rather than "the sky if falling! the sky is falling!" but we all know which gets more eyeballs, more clicks, more advertising fees. I can't offer any more precise criticism than that; AFAIK you're the only one on this site who has such an appropriate background to argue the important details.