Thinking about Biophysics

[Mr_Useless]

hi

 

Ive been fascinated by this subject but find its wide variety of aspects make it difficult to understand just what Biophysics is about, am i right in thinking its split up in to two parts like theoretical and experimental ?

 

is the theoretical part where all the mathematical modeling and computer simulation goes on ? like making predictions on how biological matter will behave

 

and the experimental side where people focus on developing methods to "look inside" biological matter..how it all interacts , how its actually structured.

 

i find the more experimental side appealing, but is that part of biophysics? or does that come under crystallography or microscopy ? or they part of biophysics to?

 

Thanks

[Greippi]

It's a pretty huge field. The biophysics I'm interested in is related to protein folding/unfolding processes - this can be studied using a wide variety of spectroscopic techniques such as NMR, CD, fluorescence etc. The main problem/goal in this field is fully understanding the weak interactions that "hold a protein together" such as H-bonds, van der waals... and how tehy all work together.

There's also other kinetics such as enzyme kinetics.

Key areas of study with this field would include understanding photosynthesis and the ribosome.

 

The work I would be doing if I go in to this field would be more to do with with experimental techniques to 'prove' predictions made by mathematics or computer models. You can't really separate the two.

[CharonY]

It is really becoming a multidisciplinary field. As a rule of thumb biophysics deals more with the small interactions, usually between single molecules. In the biophysical projects that I have conducted involves for instance force measurements between enzyme and ligands, or force measurements between subunits of a single protein. Other examples include the measurement of electron transfer between cytochromes and so on. Some biophysicists are also working with more complex biological interactions, often with more limited success.

In the biophysical area quite often no standard models exist for biological systems. The experimental measurements form the basis for the modification of existing models and thus help to create more accurate mathematical descriptions.

 

However, there is an area of overlap with protein chemistry, e.g. in the field of protein folding. Much of the work here is more often done by (bio) chemists and structural biologists. As a rule of thumb there is an increase in complexity of interactions from biophysicists, biochemists to biologists, but there are large overlaps. (I have worked on different layers as biologist myself, though I may be a bit of a special case).

[Mr_Useless]

Thanks for your very informative posts, they sure did bring home the vastness of the subject and its different areas like working with the very small to scaling upwards

 

@Some biophysicists are also working with more complex biological interactions

 

is that like the proteins interacting with each other?

 

 

@You can't really separate the two.

 

but doesn't the "spirit" of biophysics really lie in experimental side of things though? like the modeling side going in to systems biology / bioinformatics /mathematics area?

 

@ there is an increase in complexity of interactions from biophysicists, biochemists to biologists

 

in my ignorance can i say that a biophysicist is a person who makes the "special microscopes" and thinks up novel experimental techniques to help the biochemist look inside the cell and its molecular parts. like knowing what the jigsaw pieces look like (structure) and how they link up and with what?

[Greippi]

 

@Some biophysicists are also working with more complex biological interactions

 

is that like the proteins interacting with each other?

Could be. Could be enzyme-substrate interactions. Could be other things.

 

 

but doesn't the "spirit" of biophysics really lie in experimental side of things though? like the modeling side going in to systems biology / bioinformatics /mathematics area?

Well from my point of view, that's used for when something can't easily be shown experimentally (in vivo or in vitro as apposed to in silico). Empirical evidence if at all possible is always preferential. I see no reason why physics should be confined to computer models.

 

@ there is an increase in complexity of interactions from biophysicists, biochemists to biologists

 

in my ignorance can i say that a biophysicist is a person who makes the "special microscopes" and thinks up novel experimental techniques to help the biochemist look inside the cell and its molecular parts. like knowing what the jigsaw pieces look like (structure) and how they link up and with what?

Biology is the study of teh chemistry of life. Chemistry is basically physics. So it's all the same really. Why should there be boundaries? I wouldn't want to restrict myself to one area, when they can all work together.

[CharonY]

Complex interactions are more in the realm of biochemists and biologists, biophysicists are mostly concerned about highly defined, small interactions, rather than complex situations that are not easily described mathematically.

One major difference between biophysicists to biologists is that their approach is much more axiom-driven. This is where modeling becomes relevant. The data they produce are used to create mathematical descriptions of the given biological system (e.g. forces between or within molecules). A major challenge is often to translate that information into biological function (which was my role in those projects).

[Mr_Useless]

thanks again for your highly informative reply s,

 

@So it's all the same really. Why should there be boundaries? I wouldn't want to restrict myself to one area, when they can all work together.

 

i agree with you totally,..a holistic kind of approach ,

in ignorance i was just curious at like say the time and expense spent trying to predict protein structure and details using supercomputers and other fancy "algorithms" could have been better spent on like developing the sure fire techniques like NMR etc

 

 

 

@biophysicists are mostly concerned about highly defined, small interactions

 

thanks for shedding light, i was confusing biophysics to be totally to do with microscopy,,, so its about finding the rules through experiment and then forming the more accurate mathematical model , like the goal ultimately being the mathematical description

 

 

 

 

 

@situations that are not easily described mathematically.

 

could this be where the imaging microscopy technology comes in the future , so like we can kind of see what links with what

 

 

@A major challenge is often to translate that information into biological function (which was my role in those projects).

 

i hope you had success! you mentioned that you were focusing on the small concentrated details (please forgive my ignorance) but can true function be ever found from such detailed specifics? like the elephant in the room kind of thing?!

[Greippi]

in ignorance i was just curious at like say the time and expense spent trying to predict protein structure and details using supercomputers and other fancy "algorithms" could have been better spent on like developing the sure fire techniques like NMR etc

A computer model might tell you what you're looking for though! And once you know how something works via experiments like that, you can then go back to computer models for example design of proteins with new function.

 

 

 

 

thanks for shedding light, i was confusing biophysics to be totally to do with microscopy,,, so its about finding the rules through experiment and then forming the more accurate mathematical model , like the goal ultimately being the mathematical description

The stuff I do, we have the mathematical models, then we go see what the protein actually does, and see how it fits with the model.

[CharonY]

i hope you had success! you mentioned that you were focusing on the small concentrated details (please forgive my ignorance) but can true function be ever found from such detailed specifics? like the elephant in the room kind of thing?!

 

It depends a bit. For instance, one could find out by modeling and measurements (also note that the measurements themselves need models to interpret the results) the binding forces between protein and ligands. One could find out that e.g. certain amino acids are the major elements of the binding pocket, however modeling could predict that the backbone structure also adds to that. Then you could change the amino acid sequence of the protein and repeat the experiments and compare it with theoretical values. Ideally you could then prove which amino acids have strong interactions with the ligand (and what type of interactions) plus you would know how the rest of the protein facilitates and enhances this interaction.

Ideally this model is then precise enough so that one could e.g. design a protein with enhanced (or otherwise modulated) properties.

Of course things are rarely that straightforward.

[Greippi]

Ideally this model is then precise enough so that one could e.g. design a protein with enhanced (or otherwise modulated) properties.

Of course things are rarely that straightforward.

Yeah haha, tell me about it. More often than not it just falls apart or is nowhere near as good as the original enzyme.

Hence another reason we need to improve our knowledge on/understanding of the forces holding a protein together/folding mechanism.

 

CharonY, have you done this/worked on this sort of thing?

[CharonY]

One of the things I worked on was protein-DNA interaction. However, my side of the story was mostly to relate it to biological function rather than trying to improve the system. Mostly the things worked so well that it was easier just to disrupt it instead.

[Greippi]

One of the things I worked on was protein-DNA interaction. However, my side of the story was mostly to relate it to biological function rather than trying to improve the system. Mostly the things worked so well that it was easier just to disrupt it instead.

Oh okay, coincidentally that's pretty much what I might be doing next year.

[Mr_Useless]

thanks again for your enlightening posts,

 

@ you can then go back to computer models for example design of proteins with new function.

 

sorry to sound foolish but was just curious as to why there was a focus on new protein design and function when so little is known of like the ones already in nature?

 

 

 

@One could find out that e.g. certain amino acids are the major elements of the binding pocket, however modeling could predict that the backbone structure also adds to that. Then you could change the amino acid sequence of the protein and repeat the experiments and compare it with theoretical values. Ideally you could then prove which amino acids have strong interactions with the ligand

 

 

that was very interesting, is that like finding the "certainties " in protein sequences, like what stays the same, ,(apologies again for my ignorance) but how we be certain that those amino acids will interact with the ligand in the same way under any amino acid sequence that could be found in existing proteins?

 

i mean like the "exact" amino acid sequence producing the "exact" same protein which interacts with the ligand in the "exact" same way.. but with any different combination (with those test amino acids included) might not give similar results all the time ?

(sorry to sound stupid!) but would you not have to repeat your experiment with every known protein sequence (with have those test amino acids) to prove that strong interactions with the ligand was always true?

[Greippi]

thanks again for your enlightening posts,

 

@ you can then go back to computer models for example design of proteins with new function.

 

sorry to sound foolish but was just curious as to why there was a focus on new protein design and function when so little is known of like the ones already in nature?

 

 

There are 60415 protein structures in the Protein Data Bank. We might as well do something with this knowledge! Designing proteins with new function could help create new biocatalysis for industry, or even have a role in therapy for disease.

[pioneer]

The best path for biophysics research should be related to the hydrogen bonded water associated with bio-materials. This areas of research is under funded and staffed, but has the potential for the most futuristic gains, using one variable to tweak all the variables.

 

If you take away water from bio-materials, they don't behave the same way. One can not substitute other solvents in the cell, nor can we get the same bioactivity effects in the dehydrated state. Water creates a unique scaffolding needed for bio-activity.

 

For example, the DNA double helix has an double helix of water integrated into its structure. The DNA double helix is actually a quadruple helix; two DNA and two water, but due to lack of modern bio-physical insight, the DNA is still approximated as a double helix; 50 year old thinking.

 

Water can't just hook itself up into a double helix. Rather the water needs the DNA as scaffolding. Once the water is set up within the DNA, it can be used to identify the base identity, at a distance, through information transfer within the water. If we substituted another solvent, the information chain is garbled and nothing works the same way.

[CharonY]

Excuse me, but do you even try to make sense?

[Mr_Useless]

@or even have a role in therapy for disease.

 

it certainly sounds exciting to see how something newly designed could affect things, perhaps some kind of "silver bullet" may be found, but isn't that like trying to fix something without knowing what the components involved are or what they look like?

are diseases like cancer still waiting to be solved because we dont know the shape of the proteins and dont know how they interact?

 

 

 

@This areas of research is under funded and staffed, but has the potential for the most futuristic gains, using one variable to tweak all the variables.

 

that was a fascinating and inspiring, like water is the "architect" of it all, the water molecules interacting and guiding everything to correct structure..and linkage...so DNA, proteins and protein interactions are due to the interaction with water essentially..

 

i was just curious as to why this important field of investigation was so underfunded and under staffed? what is like the latest "fad" that is getting the staff and funding?!

[CharonY]

*sigh* that is why I hate those uninformed silly posts. Those without sufficient knowledge in the field could take them seriously.

 

Almost all biomolecule analyses take the (aqueous) medium into account. It is neither groundbreaking or underfunded (well, which branch isn't?) but it is common knowledge. Depending on the type of interaction additional parameters as ion strength, pH (surface charges) etc. are also part of it. The interaction is far more complex than simply hydrogen bonds. Also common knowledge.

 

The DNA part is utter nonsense. There is course well-known data on how the helix twist change in dependent on the ionic strength of the solution, though. I have to add that designing novel artificial proteins are (to my knowledge) only met with moderate success. Those in use tend to be modified proteins (e.g. adding linkers or anchors) rather than de novo structures. But this is not really my specialty.

[Mr_Useless]

just curious: can proteins fold up properly in any other liquid other than water?

 

 

@though. I have to add that designing novel artificial proteins are (to my knowledge) only met with moderate succes

 

the very thought of designing new proteins from scratch blows my mind, do you like pick and choose what amino acids you want and then wait to see what happens?

 

but the novel protein like keeps falling apart..was just thinking may be theres some kind of "rule of thumb" involved in selecting the amino acids in your sequence, like not to many hydrophobics ones or hydrophillic so theres some kind of balance in involved?

[Genecks]

Proteins are going to need an ionic solution. I'm thinking that's how life got onto evolving.

 

Seems like more than an aqueous solution can be used, though.

See this: http://cat.inist.fr/?aModele=afficheN&cpsidt=21264197

Titre du document / Document title

Protein folding in the protic ionic liquid milieu : from native conformation to fibril

Auteur(s) / Author(s)

BYRNE Nolene ; ANGELL C. Austen ;

Résumé / Abstract

Protic ionic liquids (pILs) stand as a versatile subclass of the ionic liquid family. They are low melting. (<100°C) liquids formed by neutralization of a Brønsted acid by a Brønsted base, and consist only of ions. Despite the absence of water, pILs have an acidity (an "effective pH") lying between those of acid and base components. We describe NMR methods of calibrating the acidity (or "proton activity" PA) of a given solvent, and we use composition control of the PA to create media that can either stabilize or destabilize the proteins we study. Here we show how PA tuning can stabilize the native state of simple globular proteins, like hen egg white lysozyme, to concentrations as high as 350 mg/ml, but can also destabilize it so that it forms amyloid fibrils (with stacked b-sheet structure) by self-assembly. Finally we report that, using pILs, the native state can be reformed with most of its original bioactivity.

 

This page eventually lead me to that information: http://www.public.asu.edu/~caangell/currentabstracts.html

 

Also, read about Top7: http://en.wikipedia.org/wiki/Top7

 

Top7 has been in my Internet bookmark "Coffee Table" folder's Read Later subfolder for some time.

 

The problems with protein folding and its interactions in a solution can seem like a... Rubik's cube in some ways. The hands are the solution. The protein is the rubik's cube. And it's all about unlocking their pattern. I'm not sure if Top7 should be considered iconic, though, of future biohacking possibilities.

Edited May 19, 2010 by Genecks

[CharonY]

De novo design tends to utilize folds found in nature and thus are often not completely artificial. However, using protein models one they can have totally different sequences.

Proteins require the correct medium in order to fold properly and often also require additional factors besides the solution (e.g. chaperones or co-factors).

[Mr_Useless]

thanks for the very informative internet links,

 

@protein folding and its interactions in a solution can seem like a... Rubik's cube in some ways. The hands are the solution. The protein is the rubik's cube.

 

that was a very interesting analogy, like the protein folding problem is one of those

combinatorial explosion problems which need more more time to solve than the age of the universe!

 

but like a person doing the rubiks cube has to keep trying and trying, one combination after another ... but with the protein like theres no "trial or error" kind of thing involved, it does it super fast every time, ..

 

..like the goal of the rubiks cube is the same colored sides

but is it 100% certain that the goal of the protein is to form a structure with the minimum energy? or could the sequence that encodes the protein fold up to in to some kind of equilibrium which may or may not be the lowest energy?

 

 

 

@often also require additional factors besides the solution (e.g. chaperones or co-factors).

 

are the chaperones like really vital for folding or do they become vital only due to the crowded environment of the cell?(like bouncers at a night club, not really needed for the club to operate but necessary when crowds get out of control?!) wasn't there things like test tube folding experiments which showed you only need water and the denatured protein?

[CharonY]

The Rubik cube is not a fitting analogy as it implies one final structure that has to be reached actively. Protein folding is dynamic and usually self-organizing, however (requiring interactions within proteins as well as between protein and solvent).

Essentially for each medium condition the protein will acquire the most stable thermodynamic state spontaneously. In some cases this is not necessarily the active form though. Algorithms involved in modeling (which is more accurate than "solving") protein structures essentially calculate interactions between the residues in a given medium and thus determine the most likely state. Again, code solving or pattern detection is not a good analogy to describe that. It is more like making a huge amount of thermodynamic calculations. For the most part the interactions are well known, but the challenge is the sheer amount of individual interactions and cross-interactions. However, the final state is defined.

 

The role of chaperones is generally to overcome energy barriers and guide the fold into the active form (if they cannot reach it spontaneously) , or, if several stable states exist, it serves to acquire a specific one. It has little to do with being crowded or not (although in specific cases there may be modulations due to whatever is around).

Edited May 20, 2010 by CharonY

[Greippi]

like a person doing the rubiks cube has to keep trying and trying, one combination after another ... but with the protein like theres no "trial or error" kind of thing involved, it does it super fast every time, ..

Kinetic studies have shown that finding the native conformation isn't the rate limiting step. The rate limiting step is actually the exclusion of water from the hydrophobic core of the protein.

 

Algorithms involved in modeling (which is more accurate than "solving") protein structures essentially calculate interactions between the residues in a given medium and thus determine the most likely state.

Are you referring to programs that will calculate 3D structure from amino acid sequence? As far as I was aware such algorithms are pretty much rubbish, although they're getting better.

 

It is also worth noting that each intermediate in the folding pathway is not necessarily a step towards being more ordered - the structure does not necessarily get more native. For example, the protein BPTI containing 3 disulphide bonds. Its folding pathway is not a simple sequential process, the kinetically preferred mechanism involves intramolecular rearrangements using non-native disulphide bonds. This throws a bit of a spanner in the works for computer models that try and predict the folding pathway of a protein, and shows that the process is pretty complicated to understand.

[CharonY]

Nope, I was referring to thermodynamic models which are up stepping from known folds. Simulations on naked sequences have been unsuccessful, I agree.

 

There are also those that actually calculate atomic interaction bottom up, but there are so computationally expensive that at best short peptides have been modeled (to my knowledge). They are mostly used for lipid models and similar. However, they appear to be fairly accurate (after running for a year or so....)