|
Author
|
Topic: Intelligently Designing Immunity
|
charlie d.
Member
Member # 159
|
posted 12. April 2003 18:06
quote:
Yes frameshifts sound really cool and presents a problem for just that transition. The fact that it has a 2/3 chance of frameshift poses a boundary for getting to the human arrangement, (i.e. long strings of Vs Ds and Js) from the clusters that sharks have, fusing VDJ segments has a chance of a frameshift occuring.
Actually, I said that frameshifts must sound really cool to you (maybe they sound "engineered", uh?), but in fact they are just the entirely boring, mechanistic and necessary by-product of the genetic code: whenever nucleotides are added or removed from within a gene, 2/3 of the times they result in a frameshift (i.e. all the times that the number removed or added is not a multiple of a whole codon, or 3 nucleotides - get it?).
As I said, the chances of frameshifts per gene during VDJ recombination are exactly the same in sharks and mammals. The chances of each cell not making a functional receptor because of repeated frameshifts are much higher for a mammalian lymphocyte (which basically has 2 chances for heavy chain, and 4 for the light chain - or about 55% cell loss due to unsuccessful rearrangements) compared to a shark's, which has many more. Thus, from a frameshift point of view, the sharks' system is more efficient than ours. There is no "boundary" or "jump" between the systems (if there is, it's downwards). Of course, frameshifts are ultimately functionally irrelevant because all vertebrates make a vast excess of lymphoctes compared to what they need, and to have about half of them die during development because they make no functional receptors is just a drop in the bucket. That's the malthusian logic of the immune system; I am quite sure Darwin would have appreciated it. quote:
Secondly, to say that it was lost in any clade is more parsimounous than to say it evolved from the shark system, even if what you say is true. The question of one or a few antibodies with single specificities comes up, again it's more likely that they lost that instead of wasting resources or completely dying out because they had a few antibodies that were completely useless due to the fact that they had single specificities. Or the problems I describe above concerning useless antibodies altogether due to nonsense combination.
I am not sure this makes any sense, but anyway, let me repeat: according to what you say (not I), i.e. that affinity maturation must have existed in a common vertebrate ancestor (otherwise, why would have the Designer wasted fancy VDJ recombination to generate only low affinity antibodues, that are basically not much better than innate immune recettors?), affinity maturation must necessarily have been lost independently thousands upon thousands of times in each individual vertebrate clade that did not eventually give rise to mammals or birds. That's basic evolutionary biology, I am sure you understand. quote:
In response to Yersinia, no part of the IC system is being discussed here when I say the secondary response may have been lost, the VDJ recombinase/receptor system is still there. [i]As such I have no problem with either saying that another "big bang" had to occur when it came to mammals, and the evidence does show this, or that it was lost in sharks, and present day clades reflect that loss. Either one is fine with me and squares with the evidence. If saying that the mammalian immune system complete with affinity maturation and secondary responses occured abruptly is a more parsimounous scenario for Charlie, he can be my guest and go with that one. This is brainstorms after all.
Sure, let's go ahead. Now you can call it the "VDJrecombinase/receptor/affinity maturation" system, whose core and basic function is antibodydiversity/affinity/specificity/placentalcrossingetcetc that appears in this evolutionary "Big Bang" that however lasted a few hundred million years, and required the independent loss of the better half of this fancily designed system in all but a few vertebrate taxa. Now it's all clear. It's Brainstorms after all.
[ 12. April 2003, 18:09: Message edited by: charlie d. ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 13. April 2003 11:51
Note, since Yersinia's post was short, I'm going to reply to both Charlie and Yersinia in one post.
Yersinia,
For the IC system of VDJ recombinase/receptor system, which is only one facet of combinatorial immunity, single specificities are not required if you are making generic binders. As I've always stated, both in my OP and the "Organism GA" thread, the problem is the jump from generic binders to specific binders, which would most likely require a big bang in and of itself. The issue with single specificities is the issue of minimal function. One or a few antibodies isn't enough, the system has to start making a lot from the get-go.
Charlie,
Why don't you think frameshifts sound cool? It not only sounds cool to me but it may sound cool to other people too. But who cares? Whether you think I brought it up just because it sounds cool is irrelevant. My point, which has not been responded to, is that it presents a problem for the transition from the Shark pattern to the mammalian pattern. And even if it doesn't result in a frameshift you have the problem of producing useless antibodies. And then you have the specificity problem in which I still have gotten no response. Darwin would have not appreciated such a barrier when it means getting non-function or garbage antibodies. Let me explain (again).
That the chances of a frameshift occuring per gene is the same as it is for sharks and mammals supports my point. In fact, this is precisely one of the problems of getting from the shark version to the mammalian version.
We have large arrays of Vs Ds and Js that can rearrange. However, sharks do it in clusters. In other words, although we have long strings of Vs and Ds and Js, sharks have V1-D1-J1-C V2-D2-J2-C,etc. When you try to get something like what humans have, you'll probably end up with something like a V1-V2-J2-C which is garbage. In fact, you won't find this arrangement in the shark genome. In fact, I predict that you will not find a useful antibody anywhere with this kind of arrangement. That fusing these segments has a two-thirds chance of a frameshift excaberates the problem of getting from the shark version to the mammalian version step by selectively advantageous step .
With regard to your second point, as I also keep repeating, if you are making generic binders, then you don't need to make a boat load of specific antibodies. Thats why innate receptors work. The problem is the intermediate step between generic binding and highly specific binding. The intermediate step is absent from all organisms, and probably never existed, and this is for good reason, one or a few antibodies are not sufficient to make a difference. Thats why either you have boat load generic binders or a boat load of specific binders. Nothing in between. When it came to the mammalian system, either a big bang occured not unlike the big bang that occured to get sharks VDJ recombinase/receptor system, (because when it comes to the mammalian system the issue of minimal function renders the mere presence of these components insufficient), or the an entire kit and kabootle was present in the shark system at the beginning, and modern lineages lost them. Take you pick either one is better than the Darwinian story by far.
I'm not sure what the rest of Charlie's post was responding to. For certain, it wasn't anything that I wrote. Maybe it bears repeating what I think is IC about the combinatorial immunity.
The clonal selection mechanism of membrane-bound antibodies, the messenger, and the secreted form of the antibody is IC.
The VDJ recombinase/receptor system is required for antibody diversity, and therefore, it is IC.
With the VDJ recombinase/receptor system, when it comes to antibodies with single specificities, the issue of minimal function comes into play, where making one or a few antibodies is not sufficient to make a difference.
The complement cascade, when it comes to the switches, is IC.
Antibodies themselves may be IC, in that they require all three VDJ segments. In order to show that they are not, show me an antibody without one of these segments.
[ 13. April 2003, 15:07: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 13. April 2003 14:06
quote:
For the IC system of VDJ recombinase/receptor system, which is only one facet of combinatorial immunity, single specificities are not required if you are making generic binders.
So then, what is your objection to the natural origin of recombining receptors via insertion of RAG into a non-rearranging receptor? E.g. as published in this and many other scientific papers?
PS: Charlie can deal better with the rest of Nelson's most recent, but regarding the origin of the mammalian vs. shark pattern of VDJ,
(1) We're not talking about the origin of VVVVV DDDD JJJJJJ from VDJ VDJ VDJ VDJ, we're talking about the their common origin from VDJ.
(2) But it wouldn't make a difference anyway, because RAG figures out where to cut based on RSS (IS in above diagram), and these sequences would get copied along with the V, D or J segments as they get duplicated (mutant duplicate segments without those sequences would either become pseudogenes (and we have plenty of those in Ig regions of chromosomes) or (if harmful) get selected out of the organism population, like any "bad" mutations. This is why it is crucial to think in terms of population genetics).
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 13. April 2003 14:29
Yersinia, my objection is that there is no step by step pathway from "non-rearrangeing innate receptor" to the rag genes to the VDJ recombinase/receptor system we see in acquired immunity. Innate receptors look absolutely nothing like antibodies. Furthermore, antibodies themselves may be IC in that they need all three VDJ segments. This is why the literature shows a Big Bang of acquired immunity, consistent with IC.(with the possible VDJ recombinase being ushered in via lateral gene transfer, a completely non-Darwinian mechanism)
As far as I can see, the hypothesis that the VDJ recombinase/receptor system is IC (i.e. had to show up abruptly) and was therefore designed is concrete and locked in. Now we are just moving on beyond that. Anything posited just before that, like your picture, is just imagination.
Secondly, we are indeed talking about the origin of both VDJ segment patterns. This a required topic in our discussion, since the issue of minimal function comes into play.
Thirdly, the issue is not just one of "population genetics" where the "bad ones just get weeded out". You keep mentioning gene duplication as if it was just salt on a steak. Organisms have mechanisms to silence gene duplicates (and that is their most common fate). An organism making one or a few antibodies with specificities would be wasting resources, and most likely, due to the chance of a frameshift and the making of completely useless antibodies, and the quantum leap from generic to specific binding, it is quite reasonable to conclude that the issue of minimal function is as insurmountable a problem as the issue of getting the IC system of VDJ recombinase/receptor.
[ 13. April 2003, 15:11: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 13. April 2003 16:00
Nelson:
pay attention, this usually takes students 5 minutes to understand.
Whenever one antigen receptor segment rearranges to another segment (eg, a V to a D or a D to a J), a frameshift can occur in 2 out of 3 events, because during VDJ recombination nucleotides are chewed up and added to the edges of the rearranging segments, and if the addition/deletion is not by a multiple of 3, the reading register is messed up. This is an inescapable consequence of the triplet-based codons in our genetic code, which applies to sharks and humans alike. Thus, whether the locus arrangement is V1-V2-Vn-D1-D2-Dn-J1-J2-Jn-C or V1-D1-J1-C1-V2-D2-J2-C2-Vn-Dn-Jn-Cn makes no difference at all on a per gene basis. Clear so far? Good.
Now let's look at the total chances: imagine a human heavy chain locus rearrangement, for instance
V1-V2D1J2-Jn-C. If this rearrangement is out of frame (2/3 chances), there's pretty much nothing else this locus can do. The only chance for the lymphocyte is to go on and try to rearrange the second allele on the other chromosome; if that's out of frame too, the lymphocytes can't make any heavy chain and it apoptoses. Now let's look at the shark. The first try is, say, on cluster 2: V1-D1-J1-C1-V2D2J2C2-Vn-Dn-Jn-Cn. If this is out of frame, the lymphocyte has potentially (n-1) more tries on the same chromosome, plus n on the other to get it right. Thus, potentially, the shark system is much more efficient than ours as far as frameshift management goes. Hope it's clear now.
As far as getting from the shark arrangement to ours, there is nothing mysterious there. Indeed, if you look at human lambda light chains, you'll notice kind of an intermediate structure: V segments are clustered together, J and Cs are separated, organized in J-C repeated pairs kind of like in the shark system (structure VVVJCJCJC) (note, light chains don't have D segments). Even more interestingly, in T cell receptor genes, you get almost the whole assortment. Alpha chains are VVVJJJC, beta chains are VVVDJJJCDJJJC, gammas are VJJJCJJJC and deltas (which, peculiarly, are nested inside the alpha locus) are VVVDDDJJJC. Go figure. Maybe the designer was getting bored.
[ 13. April 2003, 16:03: Message edited by: charlie d. ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 13. April 2003 17:19
I'm not sure what the first and second paragraph of Charlie's post was responding to. I very much agree that the chances of a frameshift occuring is the same for both sharks and mammals. In fact, that supports my point.
In fact this is exactly what I pointed to when it comes to getting from the shark system to the mammalian system. Whether there is a D segment in the light chain is irrelevant. The interaction between the heavy chain and the light chain may contribute to the ICness of the antibody. What has still not been responded to is how the shark system gets to the mammalian system. Which brings me to the erroneous illustration in Charlie's post of the light chain, beta chain, etc of both Igs and TCRs, attempting to make it look like the shark system.
The human lamda light chain actually has 40 V segements and 4 J segements. The T cell alpha chain has 100 V segments and 50 J segments. The beta chain has 50 V segments, 2 D segments, and 13 J segments. A similar assortment applies for the gamma and delta chains. I don't think that looks like the shark system at all. Lets go back to that shark system. Lets say that it tries to get a V1 fused to a V2. First, as already mentioned, you have the problem of a two-thirds chance of a frameshift, this problem is important because we are talking about 50-100 segments of Vs, and multiple Js, and mutlipe Ds, to get to the mammalian system. Having to do this while enduring all those frameshifts is would characterize your handling of the frameshifts themselves as inefficient no matter what you are (which is why Charlie's discussion of shark's handling of frameshifts is also irrelevant). Then you have the problem of useless antibodies, where a V1-V2-J2-C might be made, it's garbage. All the while we are trying to evolve a VDJ recombinase/receptor system that does antibodies with single specificities. Again, the issue of parsimony not only comes into play, but the issue of life and death. The now specific VDJ recombinase/receptor system is a whole can of worms that also needs to be explained, which if it makes just a few antibodies it wouldn't be sufficient to even be selectively advantageous.
[ 13. April 2003, 17:36: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 13. April 2003 17:46
quote:
In fact this is exactly what I pointed to when it comes to getting from the shark system to the mammalian system. Whether there is a D segment in the light chain is irrelevant.
Not if you're claiming that VDJ is IC.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 13. April 2003 18:12
Hopeless. Nelson, you have a lot of confusion in your mind (you've neen reading too much immunology in too short a time, perhaps) quote:
Lets say that it tries to get a V1 fused to a V2.
It can't, anymore that a human system can get a V1 fused to a V2 in its Ig clusters. V segments have RSSs only on their 3' end, and they all have the same spacers (for steric reasons related to RAG structure, recombination only occurs between RSSs that have different sized spacers), so that recombination cannot occur. What can occur is a V from one shark cluster recombining with a D in another. Just like in himan loci where D and/or J are clustered with their Cs, but Vs are all together. quote:
First, as already mentioned, you have the problem of a two-thirds chance of a frameshift, this problem is important because we are talking about 50-100 segments of Vs, and multiple Js, and mutlipe Ds, to get to the mammalian system. Having to do this while enduring all those frameshifts is would characterize your handling of the frameshifts themselves as inefficient no matter what you are (which is why Charlie's discussion of shark's handling of frameshifts is also irrelevant).
AAARGH! Frameshifts have the same frequency for every rearrangement! It does not matter how many segments you have. You could have one million segments, all in one clauster or in a hundred thousand clusters, and your chances of frameshift would still be 2/3 per rearrangement. I can't believe you don't get it still. quote:
Then you have the problem of useless antibodies, where a V1-V2-J2-C might be made, it's garbage.
As I said, it can't be made, either in humans or sharks, for well known molecular reasons. quote:
All the while we are trying to evolve a VDJ recombinase/receptor system that does antibodies with single specificities.
LOL. Single specificities again! Single specificity, high affinity antibodies have nothing to do with VDJ recombination. Once VDJ recombination is in place, due to transposon intergration in an innate immune receptor, that's it for VDJ recombination, nothing more of much significance happens (a few duplications here, a few deletions there, but the structure of the loci is very dynamic and "plastic", as shown by the assortment of locus organizations present in humans). Wait patiently for a few hundred million years, and an efficient method to increase specificity and affinity during the immune response emerges in birds and mammals - but again, it's unrelated to VDJ recombination.
[ 13. April 2003, 18:15: Message edited by: charlie d. ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 13. April 2003 18:13
Nelson, what is this babble about an immune system producing "not enough specific antibodies to make a difference"????
Way back on the other thread, charlie pointed out to you:
quote:
Antibodies do not necessarily have single specificities. Indeed, the vast majority of our circulating immunoglobulins (the so-called "natural" antibodies) are low affinity, broad specificity antibodies directed towards common antigens (bacterial wall moieties, for instance).
Antibodies become highly specific and gain high affinity only late during an antigen-specific immune response, through a process of mutation/selection called affinity maturation. This however has nothing to do with the VDJ recombination process we are discussing here, which takes place, irrespective of antigen, during B cell differentiation in the bone marrow.
"Naive", newly generated B cells, as they emerge from the bone marrow, carry antibodies that are mostly of low affinity. For insatnce, the antibodies produced early during an immune response (which reflect the naive repertoire) bind antigen with a Kd in the 10^-5-10^-6 M range - compared that with the high affinity, "matured" antibodies of late immune responses, which have a Kd of 10^-8-10^-9 M. Most antibodies in the primary repertoire are also not very specific – in fact, polyspecific antibodies abound (which goes along with their low affinity for antigen). As an aside, the vast majority of antibodies in the primary repertoire do not in fact recognize anything at all, and the B cells that make them die after a while without ever seeing any "action" (one of the drawbacks of the darwinian approach of the adaptive immune system – high, widespread wastefulness for rare but exceptional returns).
As for innate immunity receptors, again you are mistaken. While some of them do indeed have broad spectrum, many have quite subtle specificities, for instance TLR4 binds very specifically to the lipid A moiety of the very large bacterial lipolysaccharide (LPS) molecules. Their binding constants also actually compare quite well with those of most primary response antibodies (in the 10^-6-10^-7 M range).
The fundamental difference between adaptive and innate immunity receptors is in fact neither in their affinity nor in their specificity, but in their logic. The adaptive immune system, using VDJ recombination, can generate an almost infinite variety of specificities, and thanks to clonal selection can pick any extremely rare, low affinity antibody molecule and turn it into close to a “magic bullet” (this however has again nothing to do with VDJ recombination). The innate immune system, on the other hand, can count on only a limited array of receptors, which must focus on a few abundant antigens (sometimes classes of antigens) commonly found on pathogens (often, like LPS, molecules that we ourselves do not produce); moreover, the binding of the ligand has to be good to start with, because these antigens cannot undergo mutation and selection processes.
So we have the following facts:
1) Many innate receptors have similar specificity to "naive" (first-generation B-cell) recombinant antibodies. Most of the antibodies in your and my blood, right now, are therefore "not specific enough to make a difference". According to you.
2) Specificity is not produced just by having cells that mysteriously produce lots of specific antibodies, it is produced by the selective replication of those very few cells that happen to match whatever the antigen is. Further somatic mutation and selection is what produces many copies of the very few antibody phenotypes that are "specific enough to make a difference".
3) Therefore at no point are huge numbers of diverse, "single specificity" antibodies produced.
So speaking crudely, phylogenetically, we have this sequence of organisms:
(a) invertebrates, with many non-rearranging receptors of moderate specificity (similar to the specificity of "naive" antibodies)
(b) cartilagenous fish, which add diverse rearranging receptors of moderate specificity, genes in VDJ VDJ VDJ arrangement
(c) "lower" vertebrates, which have diverse rearranging receptors of moderate specificity in a VVVV DDDD JJJJ-type arrangement
(d) mammals, like lower vertebrates except that a few of the rearranging receptors, which happen to match the antigen, get replicated and gradually improve from moderate specificity to high specificity via somatic mutation & selection.
(charlie can refine the above if I garbled things)
And yet, Nelson, you've been proclaiming for endless pages that there is some sort of requirement somewhere to produce large numbers high-specificity antibodies with different specificities.
Please, Nelson, can you help us out here?
[ 13. April 2003, 19:01: Message edited by: yersinia ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 13. April 2003 18:49
charlie said,
quote:
It can't, anymore that a human system can get a V1 fused to a V2 in its Ig clusters. V segments have RSSs only on their 3', and they all have the same spacers (for steric reasons related to RAG structure, recombination only occurs between RSSs that vave different sized spacers), so that recombination cannot occur. What can occur is a V from one shark cluster recombining with a D in another.
Thanks, that's what I was trying to explain when I wrote,
quote:
But it wouldn't make a difference anyway, because RAG figures out where to cut based on RSS (IS in above diagram), and these sequences would get copied along with the V, D or J segments as they get duplicated [...]
Yet another supposed reason for the unevolvability of VDJ recombination hits the dust!! It's like shooting skeet.
All Nelson has left is that:
(1) the exact non-rearranging ancestor receptor has not yet been identified, and (2) that he thinks a transposon insertion, a well-known natural event that is happening all the time (it causes some cancers for instance) is "non-Darwinian" and for some mysterious reason therefore an intelligent intervention.
Regarding (1), (a) Ig domains are common, (b) we know that non-rearranging receptors would work because we have a bunch of them, and (c) due to selection for immune system diversity sequence homology will decay very fast.
To emphasize the Ig domain point:
...all those circles are Ig(-like) domains.
Regarding (2), there are a multitude of transposon types and events; there is no need to postulate intelligent intervention in order to explain a transposon inserting into a non-rearranging receptor. Plus, there is the published literature and experiments which the immunologists view as having tested and strengthened the hypothesis.
[ 13. April 2003, 18:53: Message edited by: yersinia ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 14. April 2003 13:42
Charlie writes:
quote:
It can't, anymore that a human system can get a V1 fused to a V2 in its Ig clusters.
Well considering the way Charlie opened up his last response, this is ironic, since charlie d. completely misunderstood my argument here. I'm talking about the arrangement in the germ-line and how hard it would be to fuse V1 to a V2 through the use of mechanisms that cause tandem repeats. Keith Robison makes a similar argument:
quote:
Also note that these same processes could take the shark arrangement and generate: V5-V6-D6-J6-C V7-D7-J7-C by a single step which looks a little like the human case.
http://www.talkorigins.org/faqs/behe/review.html
Notice how he fuses V5 with V6 here. Now, we know that sharks have clusters of V,D,J and C that rearrange only within those clusters. So even if a frameshift doesn't occur, which as already mentioned, there is a two-thirds chance of it, we might get something like this V1-V2-J2-C, useless. Of course, here is the area that Charlie is talking about where a V might be added to a fused D-J but is largely irrelevant to my argument above.
As an aside, there are other problems that we can get into with respect to this. The large arrays of Vs have a non-random "Wu-Kabat" pattern. And this pattern has been maintained despite the fact that we have all those frameshifts and other random genetic events that would degrade the transition from clustered to large arrays of Vs in the mammalian system. Another astonishing characteristic is that they have extremely different DNA sequences.
quote:
A most striking feature of the large tandem array of V-genes in the germline configuration is the fact that they can exist in large numbers (several hundred) yet unlike the large families of identical housekeeping genes previously described above , the germline V-genes of a particular large family are very similar (say seventy percent or more similar in base sequence) yet distinctly very different. How did this pattern of sequence diversity evolve and how has it been maintained over evolutionary time? This is a very different problem to maintaining absolute sequence identity in a large set of conventional housekeeping genes. Indeed the more the problem is pondered upon, the more profound the problem becomes. Thus the differences between sequences are not randomly distributed - they are located in such places typical of somatic mutation and selection events (that a researcher can readily document in experiments where mice have been vaccinated with a foreign microbe or antigen).
http://www.erim.org/qas2001/quadrant.html
and a very provocative statement:
quote:
When we began to compare our sets of mouse derived V-gene sequences with those sets in other species, this pattern kept repeating itself. But then another feature became apparent . The pseudogenes amongst the V-gene germline families of mouse and man were in an apparently pristine condition! Indeed this exposed another feature of the data. Under Kimura's random genetic drift model of mutational change a certain proportion of "stop" codon mutations would be expected on statistical grounds. But on inspection germline V-gene families are largely devoid of stop codon mutations. They are dead V-pseudogenes and should be accumulating them! Kimura's clock has stopped for these genes, they look as if they have been recently created albeit with minor mistakes.
In other words, there are rare instances of stop codons within V genes and V pseudogenes. This distrubution contradicts any step by step process of getting from the shark pattern to the mammalian pattern since it does not look as if it got there through point-mutations/gene duplication/selection. It looks like it was recently put there. They offer a "Lamarckian" type mechanism as the solution, that seems interesting, but might be beyond the scope of this thread.
Again, that frameshifts occur at the same frequency between sharks and mammals supports my point.
Innate immune receptors look absolutely nothing like antibodies, and as I have stated, antibodies themselves may be IC in that they need all three segments (with interaction between the light chain and heavy chain). Secondly, the single specificity situation does indeed matter. Having the VDJ recombinase isn't enough. You need to start making hundreds of thousands of antibodies with single specificities from the get-go. A "few duplications here" or a "few deletions there" doesn't cut it. Organisms have mechanisms to silence gene duplicates, and more often than not, they are silenced. There is a large gap between generic binding and specific binding that has not been responded to. Even if there are some innate receptors or something like them, with single specificities, a handful of them wouldn't be enough. Which is why Behe says that a system with only a few antibodies are "not sufficient to make a difference". Bacteria would likely be able to evade, or develop resistance to those few antibodies. It has to make a lot and it has to make them fast. An animal making those few specific antibodies would be wasting resources.
In otherwords in order for a highly specific antibody to be useful it has to be made in large amounts. A few antibodies are not sufficient to make a difference.
[ 14. April 2003, 15:45: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 14. April 2003 14:00
Yersinia,
Heavy chains do have D segments and they do interact with the light chain. Show me an antibody devoid of a D segment in both the light and heavy chain and you will have shown that the antibody is not IC because of the requirement of the VDJ segments.
As for Charlie's comments, the secondary response is always highly specific, and necessarily so.
I disagree with Charlie that the vast majority of antibodies are low affinity. In fact, once the secondary response kicks in, low affinity binders are usually reduced and/or disappear completely.
This has everything to do with the VDJ recombinase/receptor system we are discussing. Through this mechanism, a large repetoire of antibodies with single specificities are made, about 1,200,000. Having the VDJ recombinase/receptor system components isn't enough in this case, if you're going to have highly specific response. You need to make a lot, and you need to make them fast. A receptor with a single specificity can be bypassed or adapted to by a bacteria. There is also the problem of the loss of those receptors with single specificities through stochastic evolution. Which means if a pathogen whom that specific receptor was acting as an antibody for disappears from the population for a while, a mutation occurs where loss of that specied innate receptor or group of specified innate receptors are now gone, this is inherited in a population, the pathogen comes back, they all die.
With those organisms that just have innate receptors, or even one or two antibodies, are in an arms race that may be too much for them. Just as the Red Queen said, "Sometimes it takes all the running I can do to stay in the same place."I have no problem repeating this argument until it is responded to. You keep asking for it as if I never outlined my argument, I'll keep showing you that I did indeed outline my argument about it until you respond.
Something that has my head scratching is this statement:
quote:
Specificity is not produced just by having cells that mysteriously produce lots of specific antibodies
You argue against this as if this was somehow my argument. Here's a good rule of thumb for you Nic. If you can't quote me, then most likely, what you think I said, isn't what I actually said.
You also say:
quote:
Therefore at no point are huge numbers of diverse, "single specificity" antibodies produced.
I'm afraid that you are very wrong. Everyone makes an enormous number of different Ig each of which can act as a receptor on the B-cell surface, each of which has a specificity for a particular epitope.
For some odd reason you quote Charlie's erroneous statement:
quote:
As for innate immunity receptors, again you are mistaken. While some of them do indeed have broad spectrum, many have quite subtle specificities, for instance TLR4 binds very specifically to the lipid A moiety of the very large bacterial lipolysaccharide (LPS) molecules. Their binding constants also actually compare quite well with those of most primary response antibodies (in the 10^-6-10^-7 M range).
And yet, as I already showed (about three times), Charlie is wrong. TLR4 is extremely generic. Yes it binds to LPS, but also an RSV-fusion protein, it also binds Hsps, a lipid from TB, and many more. This was Inlay's problem as well.
In fact, innate immunity is often called "non-specific" immunity. So yes, a major difference between acquired immunity and innate immunity is in it's specificity.
Hope this helped you out. Maybe you can help me out. You write:
quote:
Regarding (2), there are a multitude of transposon types and events; there is no need to postulate intelligent intervention in order to explain a transposon inserting into a non-rearranging receptor. Plus, there is the published literature and experiments which the immunologists view as having tested and strengthened the hypothesis.
Can you please reference, discuss, and quote these papers so that we can argue it? Thanks.
[ 16. April 2003, 20:03: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 14. April 2003 14:51
But Nelson, if that's what you meant you are using the wrong terminology. Duplicated V genes are not fused. What would you make you think so? They are well spaced, each with a promoter and leader exon at the 5' and an RSS at its 3'. Inbetween are non coding segments sometimes kb long, which are simply deleted upon recombination. Thus, whether V region segments are duplicated or not makes no difference regarding what frame they are in before or after rearrangement.
All the rest of your inaccuracies hopefully I'll take care when I have more time.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 14. April 2003 15:32
Charlie,
Yes maybe that is what caused the confusion, using the word "fused". My bad. There are also other factors that degrade such arrays that are not seen with the mammalian system, it's almost as if it was put there recently.
With respecting to "fusing", the two topics are interrelated here, which is why I mistakenly kept using the same word. When trying to get the long arrays of Vs and Ds and Js from the shark system to the human system, you have the problems I discussed with respect to getting the germline DNA at that stage. While it's doing this, you have the problem that I discussed with respect to get a useless antibody, and the chance of a frameshift while doing so. For example, say we have Keith Robison's arrangement in the germ-line, a kind of intermediate step,
V1-V2-D2-J2-C V3-D3-J3-C
You can have an antibody that looks like this:
V1-V2-J2-C
While the array in the germline is trying to get to that long arrayed mammalian version, you have it rearranging within that cluster and thus the chance of making useless antibodies, while still having that 2/3rds chance of a frameshift, while trying to evolve antibodies with single specificities by try to bypass the useless intermediate stage of making a few to the useful stage of making a lot.
[ 14. April 2003, 15:52: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 14. April 2003 15:45
No. In a V1-V2-J2-C2 system (where "-" are intersegmental, noncoding intervals, or introns in the J-C case) you'd simply get V1 or V2 independently rearranging onto J2, giving rise to V1J2 or V2J2 rearrangements, both equally functional in 1/3 or the cases (and frameshifted in 2/3).
I think you simply are not getting something very basic about how the system works, although I can't put my finger on exactly where your misunderstanding is from here.
That's why online teaching has been such a bust so far: it's never going to work without eye contact! ![[Big Grin]](biggrin.gif)
[ 14. April 2003, 15:46: Message edited by: charlie d. ]
IP: Logged
|
|
|
Printer-friendly view of this topic