Viewing bacteria on a molecular level, how to do molecular modeling, and seeking beginner guidance

[coolcash2004]

Hi - I am new to biotech (though not to entrepreneurship in general).  I have a whole bunch of DIY questions that I am saving for later - for now I've just been doing a lot of research on my own trying to determine how I want to start and planning everything out.  Any help with the questions below would be greatly appreciated.  

1 - For a DIY lab, is it possible to view bacteria on the molecular level?  Separately, is it possible to view them while changing (pre treatment, during, and post) on a molecular level?  Is the equipment needed for this reasonably priced?  Is it ever possible to rent or borrow equipment?  (The microscopes I saw online seemed to cost $25,000+ but I can't imagine thats the only way - or I'm looking at the wrong machines).

2 - If I were to take the findings from part 1 (and have a detailed view of the bacteria in various states) is it possible to then design a new compound that will accomplish something specific with the bacteria WITHOUT taking existing compounds and without actually doing the lab testing for now?  And what would be the best way to approach this?  Is molecular modeling the best way to go?  Is it reasonable to learn as much as I can about various compounds, etc and then model a new compound from a problem solving approach rather than looking at existing compounds?  (Assume to date all existing compounds tested have failed).  Where should I look to begin learning the basics of how to engineer compounds?  Any links are appreciated.  School is not an option at this point.  I understand the process in general of looking at bacteria, finding targets, then design compound that will achieve goals within parameters - but I'm struggling to find info on how you actually design a compound (without just testing existing compounds for potential matches).    

3 - Is the NYC group I saw here still active?  It would be great to meet with other people from NYC at some point later on.

Thank you so much.  Any other feedback / resources are appreciated as well. 

  

[Mac Cowell]

Regarding 1: generally, no. Most biomolecules are smaller than the resolving power of conventional light microscopes, so you cannot see molecular structures.

Atomic Force microscopes could be used to probe molecular structures. And there's even a new DIY kit that is about the price of a fancy laptop - maybe you could raise funds for one with your local network / school / etc (simultaneously becoming the coolest kid on the block): Strømlingo Basic DIY AFM Kit + Strømnest ($3000).

Otherwise, if you are on a serious shoestring budget, then perhaps building your own super-pixel resolution lensfree optical tomopraphic microscope (what a mouthfull!) - you won't be able to see individual proteins, but in principle you will be able to resolve microbes in 3D in near-realtime. And the components are dirt-cheap off-the-shelf digital cameras (used would be fine), a computer, some carefully arranged LEDs, and software for deconvoluting the captured signals. See details below.

Lastly, you can get a very affordable high-power (binocular 2500x)  optical microscope from amscope for $300-500.

Add some filters and a wideband (halogen would be good) light source, and you might be able to DIY yourself an epifluorescence microscope, By using fluorescent dyes that only stain particular cellular biomolecules, you'll get pretty close to your goal of imaging molecules directly. Even though you won't be able to see them individually, you may be able to explore their activity and localization in realtime. Base your DIY design off the specs of one of AmScopes Epi models (http://www.amscope.com/special-microscopes/epi-fluorescence-microscopes.html), or just buy one straight up, they start at around $2500. But really you are just getting a $300 2500x microscope they've upgraded with a bright light source and filter set. I'm sure those components could be scrounged up and added to one of their $300 models for less than $2000. 

So there you have it!

I'll leave you with some more info re: lensfree tomography:

Multi-angle illumination with pixel super-resolution enables lensfree on-chip tomography

In conventional imaging, lenses resolve a sharp image of the target onto the sensor by focusing just the light emitted by the target within the lens' "focal plane" (a volume in front of the lens determined by the width of its aperture and the wavelength of the light passing through it) while blurring light from everywhere else into background noise. One can think of this system as amplifying certain information (the light from the focal plane) while filtering out other information (the light from everywhere else) from the total input light.

In other words, the images formed by lenses are lossy by design. This isn't practically a problem at macroscales, but at microscales the focal plane becomes extremely thin (micron-scale), oftentimes 1/100th as thick as the sample being imaged. To image large cells, for instance, a microscopist might be forced to capture tens or hundreds of images of the sample from different focal lengths (a "z-stack"), then deconvolve them to generate an image in which most/all of the sample is in focus.

These new techniques are interesting because they fundamentally attempt to capture more of the "information" present in the incident light. The lens does not capture an "image" that can be directly visualized, but rather a more full representation of the wave-field of all the incident light. Generating a conventional image from the raw data requires sophisticated additional processing.

The Ozcan lab has published some really neat (and eminently DIYable) work in this area:

"Lensfree On-chip Tomographic Microscopy Employing Multi-angle Illumination and Pixel Super-resolution", Serhan O. Isikman, Waheb Bishara, and Aydogan Ozcan. J Vis Exp. 2012; (66): 4161. doi: 10.3791/4161. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3487288/

"Existing 3D optical imagers generally have relatively bulky and complex architectures, limiting the availability of these equipments to advanced laboratories, and impeding their integration with lab-on-a-chip platforms and microfluidic chips. To provide an alternative tomographic microscope, we recently developed lensfree optical tomography (LOT) as a high-throughput, compact and cost-effective optical tomography modality. 7 LOT discards the use of lenses and bulky optical components, and instead relies on multi-angle illumination and digital computation to achieve depth-resolved imaging of micro-objects over a large imaging volume. LOT can image biological specimen at a spatial resolution of <1 μm x <1 μm x <3 μm in the x, y and z dimensions, respectively, over a large imaging volume of 15-100 mm3, and can be particularly useful for lab-on-a-chip platforms."

Also see http://spie.org/newsroom/technical-articles-archive/3979-mul...

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/50b2d439-4ab9-40d3-b298-1b84bd22f898%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[coolcash2004]

Thank you so much for the info and for the quick reply!  That first AFM looks amazing!  Do you know of anyone that has used it / it worked okay?  Do you think the images from that microscope would be clear enough to put together a full structure of a bacteria? 

Thanks again,

David

[Mega [Andreas Stuermer]]

For producing compounds, it is not neccessary tosee the bacteria. Is your goal to produce compounds or watch bacteria?

Making synthetic compounds is usually harder, or rewuires more computional power I guess. If you want to produce a pore forming enzyme that kills bacteria by leakimg them out, where would you start? Probably taking a domain that is known to bind to an E. coli surface protein and a known pore (like a Colicin). Designing a pore yourself from scratch is hard I guess, where to start. There are countless possibilities, and some of them are effective. On the other hand, nature has shown that X,Y are effective

[Nathan McCorkle]

To see molecules sounds a lot like wanting to see atoms. Currently the people are doing things with scanning electron microscopes (and also scanning transmission electron microscopes) where you scan a surface and analyse the xrays that are emitted as a result of the electron beam impact. There are also people that look at the energy loss of electrons that bounced off (EELS).

Then there are environmental SEMs that can operate at high enough vacuum that cells can survive (so imaging can take place, then the cells returned to their normal growth chamber).

Then there is combination FIB +SEM serial block face sectioning... You end up with a 3D image dataset. In this technique an image is acquired, then the FIB is used like a machine-shop belt-sander to grind off (ablate) a layer, then another image is acquired, and the process repeats until the sample is gone through. This has to happen at cryo freezing temperatures, otherwise the molecules will move and the images will be blurred, not to mention the sequence of images in the serial block face operation.

And there's also cryo TEM tomography, where a cell or section/slice of a cell is in a frozen state and put into the TEM and an image is acquired. If the slice is too thick (which is the use-case for this technique) it will be blurry, but with math this can be unblurred into a 3D volume dataset. This is kind of like diffraction crystallography.

[Brian Degger]

I reccomend going to your local biohackers space (like genspace nyc) growing your first colorful bacteria and talk about your ideas.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2B82U9LB1KZYHU-OcdhUbB1yteLO61PTiR2pLzvUsbr8%3Ddq4_w%40mail.gmail.com.

[Bryan Jones]

Regarding #2, there are some in silico modeling options to look for chemical interactions that you can work on without getting your hands dirty or spending any money.
If you are trying to find/design a compound to cause a specific effect in bacteria, do you know the target? Are you trying to interact with a specific protein? If you have a specific protein target, you can start to use some molecular modeling type software to at least make guesses as to what type of compound will give you the interaction you want. In general, you can't get the atomic resolution of a protein you'd need to design interactions with any type of microscopy. You'd need to use X-ray crystallography or NMR to get that type of resolution, both of which are still out of reach for the DIY biologist. However, if the structure of your protein of interest has already been found (search pdb.org for your protein), then you are in business. Alternatively, if your protein of interest has enough sequence similarity to proteins with known structures, you can build a homology model (e.g. http://swissmodel.expasy.org/).
Once you have a structure model, you can start o try to dock various compounds computationally to try to find a hit. A lot of the software to do this type of modeling is ridiculously expensive. Thankfully, there are some free/open-source options available. Check out http://autodock.scripps.edu/ for one option.

This is one of the basic workflows that large pharmaceutical companies use to find new drug candidates. It works, but they end up investing millions of dollars to make it work. It's important to understand that computational modeling of macromolecules like protein is still very crude, and there is no guarantee that the results are accurate. You might go through a million compounds in the computer to find 100 compounds that look good in the models to find one that performs well in live cells.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CACc%3DpRK0g300KVRXzo-%2B4VWursa5v2cSUfkNT-7etnP7vHYPpw%40mail.gmail.com.

[Simon Quellen Field]

Actually, you can see individual molecules using light microscopes.

with STORM (see the attached paper), you play tricks to make only one molecule light up per pixel of resolution.

You can also do Raman spectroscopy and image biological specimens according to their molecular makeup.

The LOT microscope Mac referred to is only claiming to resolve to "less than a micron", which is rather vague. The microscope here in my lab routinely gets 250 nm resolution, and can get higher than that in the ultraviolet. It was not all that expensive ($2500 or less if you shop around -- the newer Chinese microscopes are quite economical). Nevertheless, I love the idea of a DIY LOT microscope.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2BiLdW31_5wwEyAf6yPr16yuuGpcrNF25tjLevx%2B6r5S1HgfAQ%40mail.gmail.com.

[Simon Quellen Field]

Rather than the $3,000 AFM, you could try building your own STM.

See attached papers from The Physics Teacher and American Journal of Physics.

Also see this, and this.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

On Wed, Dec 9, 2015 at 9:35 PM, Mac Cowell <m...@diybio.org> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2BiLdW31_5wwEyAf6yPr16yuuGpcrNF25tjLevx%2B6r5S1HgfAQ%40mail.gmail.com.

[coolcash2004]

Thank you all for the incredible information!  

Bryan, do you think that the methods that Nathan mentioned or the machine simon mentioned would provide a clear enough resolution to be able to design interactions? Or is there really not a suitable DIY machine at this point to get the resolution needed? 

The bacteria I want to look at is borrelia burgdorferi http://www.rcsb.org/pdb/results/results.do?qrid=930DD4CE&tabtoshow=Current .  Many proteins have been identified, but from what I can tell more are being discovered, so I don't think we have a full picture yet.  Also, I want to specifically look at persister bacteria with biofilm that has undergone treatment, and that is a new area of study.  I do not yet know the target, so I'm hoping to get a full idea of the structure first (and I think its incomplete at this point).  All the existing targets of pharmaceutics have just recently been proved to be ineffective against the persister bacteria, so unless we aren't hitting the right known targets in the right way, I think I'l need to find a new target.  To find a target I'm hoping to look at changes in the bacteria from non-persister to persister state to see what mechanisms convert the bacteria to persister state and enable its survival.  The only research I have found at this point, other than testing specific treatments (none worked fully), was a report about the genes that were expressed differently from non-persister vs persister state and the corresponding proteins - but it includes the known proteins only and its unclear which of the genes or proteins was responsible for the survival of the bacteria.  

Sorry for the long message / extra info.  Any thoughts are appreciated and thank you again for the guidance. 

David

[Bryan Jones]

The techniques mentioned by others (electron microscopy, atomic force microscopy, and STORM) are all great techniques. The budget AFM looks impressive. However none of these techniques is going to give you the resolution to see interactions between a protein and a small molecule. These techniques would be great for looking at the level of a whole bacterial community, like a biofilm, or even at the scale of single bacteria cells or virus particles. You might even be able to see a single protein with some of these techniques, but it will be just a blob, you won't have the detail to look at and characterize the chemical interactions that might be present.

If your goal is to rationally design small molecule drugs, I think you will really need atomic resolution models of specific target proteins. Looking at the differences in protein expression between the two states of B. burgdorferi might be a good place to start. That might shed some light on what proteins are or aren't viable targets, or on why the persister state is so hard to kill. Rationally designing a antibiotic is hard because you need to know 1) what proteins are essential, 2) which proteins are accessible, 3) what compounds will interact tightly with the protein, 4) what interactions will actually inhibit the protein, and 5) whether the interacting/inhibiting compounds will be toxic to the host (i.e. humans). Because of all these obstacles, the approach of randomly trying chemicals on the bugs and seeing what kills them can sometimes seem like the easier option.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/f224302f-ef20-4f19-84b2-094e6229f5c9%40googlegroups.com.

[Mega [Andreas Stuermer]]

That's what I meant. It should be easier to take a pore forming domain from colicin and attach it to a domain that binds a - very conserved - receptor of your target bacterium. So it should bind your bacteria, make a hole in the membrane (many molecules -> many holes) and the bacterium dies.

Sounds easier than de-novo design. And you don't have to watch bacteria for that. You can test protein interactions with various methods, and if you take a very conserved protein (usualy phages choose them as their target), you might end up with somethingthat works.

I hope that's helpful!

[Mega [Andreas Stuermer]]

You actually don't have to test protein-protein interaction in the early stage if a phage and its binding site has been characterized. Take the opportunity to stand on the shoulders of giants.

[Brian Degger]

Such a great thread you have started David!

If you are looking for more techniques a Scanning Tunneling Microscope was built at this years berlin science hack day. 

----------

Documentation concerning the attempt to build an STM, a scanning tunneling microscope, at Fab Lab Berlin. With the microscope it is possible to see individual atoms and enter the world of quantum mechanics.

https://wiki.fablab.berlin/STM

On Fri, Dec 11, 2015 at 8:47 AM, Mega [Andreas Stuermer] <masters...@gmail.com> wrote:

You actually don't have to test protein-protein interaction in the early stage if a phage and its binding site has been characterized. Take the opportunity to stand on the shoulders of giants.

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/a7227db2-2565-4eff-9e8a-0c07d42040fe%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

----------------------------------------
Brian Degger

twitter: @drbrian

http://makerspace.org.uk

http://transitlab.org

----------------------------------------

[coolcash2004]

Okay so at this stage we don't really have enough understanding of proteins to be able to look at a protein and create something, based on how the protein is structured, that would be guaranteed to effect it in a certain way.  I'm guessing the computer models are as advanced as they can be given the existing combinations that have been tested - so just curious what is it that is holding them back? Is it just that proteins are so complex that we haven't viewed enough to put together a fully accurate set of rules (ways to resolve those 1-5)?

For the phage - Sorry if this is redundant or really straightforward but I'm not sure I understand fully.  So what you are saying is that if I figure out which protein to target and a specific receptor, then figure out what will bind that receptor, then I could attach the colicin domain to the binder and it would potentially enter and kill the bacteria?  Also this would assume that the protein targeted is the one that was responsible for persister survival, right?  Or it could really be any conserved protein?  So this would go on the assumption that the mechanism for persister survival would not be triggered by this attack?  I guess I'm trying to understand if this approach could be used to disable or circumvent the survival mechanism of the persisters.  Or would I really need to look at the atomic level structure to figure out the exact persister mechanisms and how to disable them?  So far we know how to kill the bacteria - to a point- and then the persisters survive all treatments tested.     

That STM looks great! I'm not sure yet if I'l need to look at atoms to be able to accomplish this or not.

Thanks again sorry if I'm repeating things I'm just trying to understand everything.

David

[Bryan Jones]

"Is it just that proteins are so complex that we haven't viewed enough to put together a fully accurate set of rules(ways to resolve those 1-5)?"

Partially, but the bigger problem is that we just don't have the computational power. We definitely don't have a perfect set of rules to govern protein interactions, and as we try to model with more complex & accurate rule sets it gets too computationally expensive. The more rules and parameters you include in your computer simulation, the more computing power you need. If you use rigid models with no chemical properties (e.g. charges, hydrogen bond donors, etc) you can do the modeling rather quickly, but it's not very accurate. When you start to add in parameters of increasing complexity like chemical properties, movement of amino acid side chains, movement of the protein back bone, solvent interactions, and quantum mechanical effects your modeling will become more and more accurate, but your computational load will grow and grow. 

For example, I use a super computing cluster to model fairly complex models of a single protein with solvent and a single small molecule that's already docked to the protein, and it takes hours to days to run a simulation that simulates a couple nanoseconds. There are ways to simulate even finer levels of detail that do take into account quantum mechanical effects, but these simulations are usually limited to only a small part of a protein, and an even smaller amount of time.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/b7670a13-a0af-4c27-85ba-0300cacf3e33%40googlegroups.com.