Defense of Bernard Lagat

Not to me ...

but that record was set in the Pre-EPO 1980s - and David certainly didn't have the dosh at hand to afford the more sophisticated forms of blood doping available at the time.

I should add the caveat of "these days" when I talk about the reality of suspicious minds, because EPO has, quite literally, changed everything.

Martin

Besides, David, as we here in the British Empire like to call him....

Is one of us upper crust white folks... not one of those heathens from some other country... that we used to enslave by the way..... so of course they must be the dirty ones.

Things were much better for us in the good 'ol days.

phonix ---f.o

belive no one,

this is dirty stuff

I have seen the A sample results of Bernard Lagat and really would appreciate that WADA provides these original whole gel pictures for the public. This Western blot results speak for themselves!!! Smears, dots, scratches, clouds ? you name the artifacts (which can sometimes show up, especially when using the ?double blotting? procedure of Lasne et al.)- the artifacts are all there and visible and more important these are also in the area of EPO protein bands.

To me it looks like spraying art of punks in the subway. (Sorry punks out there for this spontaneous association!). To call this result a ?positive A sample? was a scientific fault. Without immediate repeating the procedure in first place, the result should have named at least ?inconclusive?.

The further report to WADA on such a bleak basis and the leakage of this message (perhaps by purpose) have almost ruined carrier and reputation of the athlete.

Fortunately the result of the B sample speaks clear in favor of Bernhard Lagat.

This approved EPO test is complicated, using an antibody, not able to discriminate endogenous from recombinant EPO, and is obviously also dependent on the experience and ?daily performance? of the lab personal. The whole procedure is a nightmare, instead of a fair, repeatable analysis. It is almost a ?French roulette?.

Phoenix supporter, that's what worries me, the EPO test is so ridiculously long and complicated, that the potential for human error or even laziness is so alarmingly high that it has no credibility for me as a reliable drug test.

We have a way round the Lagat situation under Scots Law: the finding of 'not proven'. This means that we're pretty sure you're guilty, but can't prove it.

That said, the information on this posting has shed a lot of light on a very difficult area, with things not being as clear-cut as I'd first imagined. Keep up the knowledge and banter!

The EPO test has one major weakness. The applied test and the bands running at the same isoelectric point in parallel lanes, are by far not proven to be the same proteins. With the used antibody you might be able to detect specifically human EPO. So far, so good. But now the next and major question if one gets positive Western blotting signals: Which EPO is involved? Endogenous, normal or one of the artificial, recombinant EPOs? Again, to make it clear, with the monoclonal antibody used, which was raised against a specific peptide sequence common to all EPO forms, this discrimination is not possible! Same tallness (isoelectric point), does it mean it is the same person (same EPO form)! Definitely not. One needs better proof to come to this conclusion.

Phoenix supporter wrote:

. But now the next and major question if one gets positive Western blotting signals: Which EPO is involved? Endogenous, normal or one of the artificial, recombinant EPOs? Again, to make it clear, with the monoclonal antibody used, which was raised against a specific peptide sequence common to all EPO forms, this discrimination is not possible! Same tallness (isoelectric point), does it mean it is the same person (same EPO form)! Definitely not. One needs better proof to come to this conclusion.

This is not really an issue. The recombinant EPO runs at a different isoelectric point than natural due to its altered glycosylation. So the band appears at a different place.

The test, when run well, is a good, definitive test. HOWEVER, when the gel is as obviously bad as this one, it should indeed be labeled "inconclusive" and run again.

If one has some experience with protein sample handling, with isoelectric focusing and 2D gels, for sure this is a tremendous issue. Different variants of the same protein in different individuals and/or samples treated slightly different might occur more often at different pH positions as officially assumed. For EPO it is not shown that all, there are ten or so bands seen in a control lane and similar numbers of bands slightly shifted in the ?best of best show gels? for recombinant EPO. These natural modifications of proteins and enzymes in certain populations happen and definitely they are not excluded yet in the EPO test applied (Look for example on the alcohol dehydrogenase enzyme activity in many asian people, or the milk intolerance in some african populations etc.). One real peer reviewed epidemiological study was announced but I am still looking forward to read this missing paper. And if some exemptions exist, for genetic reasons or whatsoever, these athletes have to pay the price for an intellectual inappropriate assay.

I think you can agree with me, that a given antibody, specific for recombinant EPO alone, would not cause these kind of troubles and would be of enormous advantage. There would be clear-cut results, no natural EPO in any form could interfere and cause these doubts and misinterpretations. Even slightly degraded, shifted rHuEPO would not matter. It would be detected at almost any pH position, if there. This would give a clear YES or NO answer and not leave us puzzled in many cases with this semi-quantitive scanning and cutoff percentage dealings.

Yes, I do have quite a bit of experience, but I guess I misunderstood exactly what you meant there...sorry.

In general, I am not aware of a case where population-specifc gene polymorphisms change a protein enough to move it on such a gel. I'm not saying it doesn't occur, however I am unaware of it. I've never seen an antibody or human protein standard labeled as "asian" or "icelandic" origin, meaning that the genetic difference does not significantly change the protein. Such a change would be unusual since protein functions are almost 100% conserved within species...Asian alcohol dehydrogenase being an example of an exception (but does it differ enough to run differently on a gel??).

However, they should have done some testing to make sure that such polymorphisms did not shift the protein bands. This would not be a hard experiment, though it would be difficult to test every population in the world.

So I still say that it still isn't really an issue. The "B" was clean, so we're not talking about a protein identification issue but one of bad sample or bad technique.

But yes, CLEARLY, the best solution would be to develop a recombinant EPO-specific monoclonal antibody. I'm sure someone is already working on this, but this can be difficult especially when you're trying to raise it to different glycosylation states (direct protein sequence is easier).

AH/Phil/Phoenix can you email me. We are looking for expertise to overview testing at the UCLA lab. It would be a 2 day study, your would be compensated for your travel expenses and expertise. email gosub42003@yahoo.com

It may not have been an issue in the Lagat case, but to ensure that the test is generally reliable, isn't it necessary to determine whether variables like race affect the gel pattern? Lasne apparently thought so, since she herself said that a control population study is needed (see below). Do you know if/where the study she mentions was published?

Excerpt from Detection of isoelectric profiles of erythropoietin in urine: differentiation of natural and administered recombinant hormones. Analytical Biochemistry 311 (2002), 119-126:

"This method has been thus proposed for antidoping control after having been tested in a large control population study that included different athletes to assess the influence of ethnic origin, sex, age, physical exercise, and erythropoiesis-stimulating situations (altitude, hypobaric chambers) on the natural urinary EPO pattern. The results of this study and those of administration trials using the different recombinant hormones will be published at a later date."

Phil. wrote:

Unfortunately you provided none! We are not guessing about the testing procedures and handling they are documented!

In Phoenix's original argument, he uses the words if and likely several times to make his argument. How is that not a guess? If Phoenix knew what happened, I would think that the if and likely words would be replaced by was and did.

I provided no evidence because I made no factual argument.

And if I made a factual argument, I would provide some facts about what actually occurred. Not if they occurred or likely occurred.

So again, Phoenix makes a good case about how the test can go wrong, he makes no case about whether it actually did go wrong. So there is no defense of Bernard Lagat from the argument that I got. Give me a new more factual argument and I might change my mind.

Wouldn't the IAAF try to repair Lagat's reputation if they thought the A test was botched so bad that the result that was leaked caused him much undue harm. Instead, they treat him like a cheat that got off on a technicality. This is my "opinion" as well.

mud wrote:Wouldn't the IAAF try to repair Lagat's reputation if they thought the A test was botched so bad that the result that was leaked caused him much undue harm. Instead, they treat him like a cheat that got off on a technicality. This is my "opinion" as well.

And here's my opinion: At this point, I don't think the IAAF/WADA can admit that their test was flawed in this or any other case. Such an admission would open them up to appeals by those previously busted using the same test. There would be too many questions asked about the general validity of the test. All hell would break loose.

It's a lot more convenient for IAA/WADA to let this die and let Lagat struggle with the implication, however unfair, that he got off on a technicality. That way the IAAF/WADA covers its bureaucratic ass, Lagat continues to race, and Templeton continues to work as an agent. Life goes on. Everyone gets something. The only real loser is the truth -- and any poor slob whose career is injustly cut short by this test in the future.

AH wrote:

In general, I am not aware of a case where population-specifc gene polymorphisms change a protein enough to move it on such a gel. I'm not saying it doesn't occur, however I am unaware of it. I've never seen an antibody or human protein standard labeled as "asian" or "icelandic" origin, meaning that the genetic difference does not significantly change the protein. Such a change would be unusual since protein functions are almost 100% conserved within species...Asian alcohol dehydrogenase being an example of an exception (but does it differ enough to run differently on a gel??).

However, they should have done some testing to make sure that such polymorphisms did not shift the protein bands. This would not be a hard experiment, though it would be difficult to test every population in the world.

In literatur and sequence data bases one finds many minor protein changes, single or few amino acid exchanges, different phosphorylation, glycosylation and other VARIANTS of the SAME PROTEIN looking within same species e.g. different persons / patients / human populations. In most cases the altered protein is fully functioning, nothing wrong, nothing special reported for these variants. Please make a quick test, go to the NCBI website, follow Entrez and Protein and search some proteins by name which come to your mind, human erythropoeitin EPO, for example.

FIRST SURPRISE, under acession number NP_000790 one finds for erythropoietin an amino acid variation, position 114 replace P/L.

SECOND SURPRISE, if one searches a little further we find under human mutant erythropoietin (Accession number gi415460 and the publication Funakoshi et al. 1993, BBRC 195(2),717-22) an erythropoietin which differs to normal EPO within three amino acids. Here S,L and P are replaced by N, F and Q respectively.

THIRD SURPRISE, with chain A and chain C of the crystal structure of human erythropoietin (Accession number gi6137383 and gi5822016) differs within six amino acids. N, N, N, R, P and P are replaced by K, K, K, G, N and S respectively.

FOURTH SURPRISE, human fetal erythropoeitin (Acession number gi31229) varies with one exchange, G replaced by R! This would definitely change the IEF pattern.

FIFTH SURPRISE with accession numbers AAF23133 and AAF17572 one finds a human erythropoeitin which is derived from a gene sequence in the Quechua, a high altitude native population in South America. Here, like in the fetal erythropoeitin, G is replaced by R.

These are only modifications of the human EPO amino acid sequence found by a short data base search. Unbelievable for me at the moment, why this was not mentioned during all the heated discussions on variants and charge shifts before. Here are some real data, literatur and reasons for band shifts, charge differences, seen with the EPO test.

EPO discrepencies solely detected by different isoelectric pH or charges is not enough anymore to be sure with this Lasne-EPO test!

As above outlined, with all the already known possible variations of human erythropoietin, different people might have different natural EPO protein modifications. Therefore the criticism of the current urine EPO-test is not based on pure speculation, there are scientific facts. These possible modifications of natural EPO might easily obscure the bands in the pH area of rHuEPO and leading to misinterpretation.

Therefore this test should be used solely, as the current blood EPO test too, as a pre-screen to sort out for the clear negative cases.

A COMPLETELY NEW and DIFFERENT SPECIFIC TEST should be established and used to verify for the unclear cases.

test...

sorry for that....

Phoenix supporter, you get the research award. Nobody has mentioned it simply because nobody (including me) took the initiative and time to actually do the simple search.

If you'll look, you'll see that MOST of the polymorphisms you noted are fairly innocuous. By this I mean that they do not change the charge or polarity of the amino acid. The reason is simple; as I said, FUNCTION is nearly 100% conserved. Some small variations can exist as long as they don't affect the active site/binding site of the protein.

Yes, a couple of the changes you mention are significant in the pH profile of the protein (the embryonic doesn't count because we aren't testing fetuses). However, this does not necessarily mean the test is completely junk. It is a 2D test, first on isoelectric focusing and second on mass. So a small shift will not likely move the band to the "recombinant" spot. What is more, if this were to occur in such an individual the "normal" band would be absent, and it would be a red flag that the test is inconclusive (both should be visible on a positive).

Finally, while I don't have the time to search, it is likely none of these changes will affect glycosylation in such a way as to mimic recombinant EPO. glycosylation sites are most likely to be highly conserved..once again, if they are different, this would cause the "normal" spot to be missing on the gel and it would be a red flag.

So while I still believe this doesn't invalidate the test it does suggest (as I said before) that there is a lot of room for improvement. A specific monoclonal for one (though we could run into a THG-like situation where someone changes the glycosylation to escape detection).

Better would be to go to GC/MS or NMR. I did some research in oliosaccharide characterization by NMR and GC/MS, but the sample size needed is likely beyond what would be available.

AH wrote:

If you'll look, you'll see that MOST of the polymorphisms you noted are fairly innocuous. By this I mean that they do not change the charge or polarity of the amino acid. The reason is simple; as I said, FUNCTION is nearly 100% conserved. Some small variations can exist as long as they don't affect the active site/binding site of the protein.

Yes, a couple of the changes you mention are significant in the pH profile of the protein (the embryonic doesn't count because we aren't testing fetuses). However, this does not necessarily mean the test is completely junk. It is a 2D test, first on isoelectric focusing and second on mass. So a small shift will not likely move the band to the "recombinant" spot.

1. Remark: There are some major changes of overall charge of EPO to be considered. Let´s go to the crystal structure chain A and chain C, three uncharged N are replaced by three charged K and charged R is replaced by uncharged G.

These changes will for sure change isoelectric points!

2. Remark: Embryonic forms of proteins, does not mean, this modifications do not exsist in adult life. It might be just a major form in fetal and kept in adults (like stem cells), so to say as reserve. Some adults might have kept more, some less of this early modification. I think the same modification as the described fetal form, replacing uncharged G by charged R, is obviously also found as major form in adult Quechua indios. So if the first researcher, who found this variant, names this "fetal form", this does not mean, this is the only and final description. Like here he might have had overlooked certain other circumstances with adult samples.

3.Remark: Each phosphorylation of a protein causes charge differences, and individuals have quite different amounts of phosphorylation, depending on certain situations and exposures. Erythropoietin (Accession number NP_000790) has one possible Protein kinase C phosphorylation site (PKC Pos. 164-166) and three possible Casein kinase II phosphorylation sites (CK2 Pos. 61-64; 67-70, 147-150). Interestingly several of these sites are no more available (i.e. amino acids are exchanged), when looking through the other EPO variants formerly mentioned from me. In one case (gi_415460) CK2-site SPP is switched to SPQ and you mention this polymorphism as fairly innocuous and not important. Again, as an example, when there is no charged amino acid replaced by uncharged, it might well be that charges differ through phosphorylation clearly possible or absolutely not.

Last remark: This EPO-Test is NO 2D-gel test as you noticed ("It is a 2D test, first on isoelectric focusing and second on mass"). It is ONLY a one-dimensional separation, a focusing within a very narrow pH-range. There is nothing with mass separation. Only charges are involved and thatis why this changes of charges are so extremely important for results obtained.

The urine EPO-Test is in the real sense one-dimensional and even when properly done, might overlook many natural occurring variants and declare them as rHuEPO positive.

Martin wrote:

Yes, but improvements, by nature, narrow as one moves furhter into world class territory.

If you're talking about the youngster from Stanford, you must remember that he's much younger than B.L. - at least in athletics terms.

For instance, a college athlete runing 9:00 for 3000 metres might improve to 8:30 in a given year and very few would accuse him of cheating.

However, a runner approaching physcial maturity (let's say in his mid 20s) that drops from 7:50 to 7:30 in a single season is going to raise some eyebrows.

I'm not saying this is definitively the case with B.L., but as others have noted, huge PR gaps these days are going to be digested with a large amount of proverbial salt.

This might not be fair, but it is reality.

Martin

Martin I might agree with you if Lagat was a steady 3:35 preformer for 4 or 5 years consecutively during his peak years but he wasn't. He was an athlete that had a reasonably steady drop in performances which was not all that unusual. If he hadn't attended Washington State most Americans would have viewed him when he first hit the international scene just like any other Kenyon who bursts on the scene.