[3dem] Cryo-EM size limit

Thu Oct 28 10:35:37 PDT 2010

Hi, Michael, Sacha and Reza

It is fun to have different opinion in Science. I wish Michael could read the figure legend completely before making real comments. Figure 5, clear stated that the B, C, D, E, F, G images were from cryo-EM instead of negative-staining stated in Michael’s second comment.

I copied the figure 5 legend here for you, 

“FIGURE 5. Discoidal shapes of reconstituted apoA-I/HDL particles revealed by negative-stain and electron cryo-tomography. In each view the axis of tilt is vertical to the images (diagrammatically illu-strated in upper and bottom center of panel A). Selected titled images are linked by dotted arrows. Rela-tive tilt angles are indicated in each image. Scale bars represent 200 Å. A. Three selected tilted views of apoA-I/HDL particles from one negative-stain electron microscopic field. B-G. Selected tilted views of near native state apoA-I/HDL particles embedded in vitreous ice from six cryo-electron microscopic fields.”

Some background information for both of you for discussion of the molecular weight in HDL particle, HDL contains two copies of apolipoprotein A-I (apoA-I, 28KDa) and Phospholipids (POPCs). Each apoA-I contains 11-mer tandem amino acid sequence repeats in the periodicity of an amphipathic α helix, often punctuated by prolines. Two apoA-Is anti-parallel to each other and wrapped the around the hydrophobic edge of the discoidal shaped lipid-bilayer. To be noticed, HDL particles are highly dynamic and heterogeneity in size and amount of lipids.

If you are interested in more background knowledge about this HDL, following paper is a good one,

 <http://www.jbc.org/content/274/45/31755.full> http://www.jbc.org/content/274/45/31755.full

Comparing to lipids, protein portion in HDL particle is relatively high, the ring-shaped high density in the figure 5 represented for the protein portion (~56kDa), but may bound with small amount of attached POPCs.

Of course, you may argue that HDL particle is a special case for cryo-EM because the protein is not packaged tightly as most soluble protein, and lipids may enhance the protein contrast due to it less density than ice. 

Back to Reza’s original question, my option is that the protein with molecular weight less than 100kDa is possible to be determined by cryo-EM, I didn’t see any fundamental reason from physics and electron microscope that limited the small protein reconstruction by cryo-EM, however, I do know it is challenging project that highly depended on how to optimize the experimental conditions, such as sample preparation, cryo-EM operation and 3D reconstruction procedure as I stated before. Nevertheless, challenging the difficulty is part of fun for Science.

Best

Gary

----------------
Gang (Gary) Ren, PhD
Staff Scientist
Ph: (510) 495-2375; Fx: (510) 486-7268
Email: gren at lbl.gov; Web:http://foundry.lbl.gov/
Lawrence Berkeley National Laboratory
Molecular Foundry, Rm 2220
1 Cyclotron Road, MS 67R2206
Berkeley CA 94720-8197

  _____  

From: Michael Landsberg [mailto:m.landsberg at imb.uq.edu.au] 
Sent: Wednesday, October 27, 2010 11:47 PM
To: Sacha De Carlo
Cc: Reza Khayat; 3dem at ucsd.edu; Gang (Gary) Ren
Subject: Re: [3dem] Cryo-EM size limit

Dear Reza,

As Sacha has correctly pointed out, Gary's response is misleading and does not really answer your question principally because:

* The cited example images are claimed to be of a 56kda dimer attached to a presumably much larger HDL particle. This is a very different proposition to analyzing a 56kda particle in isolation.

* The cited example is (as stated in the figure legend) negatively stained, not cryo as you asked.

So to answer your question, my own experience is that cryoEM SPA becomes challenging at 200-300kda and below. Others may share differing opinions though and in my opinion this is because ultimately it comes down to whether you can confidently identify and acquire a homogeneous distribution of your target particle.

Regards

Michael

--

Dr Michael Landsberg

Senior Research Officer

Division of Chemistry & Structural Biology

Institute for Molecular Bioscience

The University of Queensland, St Lucia, QLD 4072, Australia

Email:  <mailto:m.landsberg at uq.edu.au> m.landsberg at uq.edu.au

Phone: +61-7-3346 2010

On 28/10/2010, at 3:05 PM, Sacha De Carlo <sachadecarlo at yahoo.com> wrote:

Your "particles" from Fig. 5 are almost as big as your scale bar. From the legend one can read it's corresponding to 200A, or 20nm. That's huge for a 56kDa protein!!!

Euk. RNA polymerase II is about 16nm in its "largest" extension, and it's a 514kDa protein.

Please refrain from using preposterous statements such as "it should be no problem to get an intermediate resolution reconstruction on protein that is less than 100kDa by single-particle cryo-EM".

Any object that is below 200 kDa and more or less globular in shape will be a "tricky" one in terms of obtaining very good alignment parameters (from unstained, frozen-hydrated samples) leading to a 3D map showing 15A details...

Thank you, and a good night,

sacha

================================================
Sacha De Carlo, Assistant Professor
Chemistry Department, Marshak Science Building &
New York Structural Biology Center
The City College of New York
Convent Ave & 138th Street
New York, NY 10031
(212) 650-6070
e-mail:  <mailto:sdecarlo at ccny.cuny.edu> sdecarlo at ccny.cuny.edu
URL:  <http://www.planetesacha.com> http://www.planetesacha.com
================================================

--- On Wed, 10/27/10, Gang (Gary) Ren <gren at lbl.gov> wrote:

From: Gang (Gary) Ren <gren at lbl.gov>
Subject: RE: [3dem] Cryo-EM size limit
To: "'Reza Khayat'" <rkhayat at scripps.edu>,  <mailto:3dem at ucsd.edu> 3dem at ucsd.edu
Date: Wednesday, October 27, 2010, 10:10 PM

Hi, Reza,
It should be no problem to get an intermediate resolution reconstruction on
protein that is less than 100kDa by single-particle cryo-EM, even by
cryo-electron tomography. An example is that the 56kDa protein (2 copies of
apoA-I, each is 28kDa) in high-density lipoprotein (HDL) can be clearly
visualized by electron cryo-tomography, even using non-FEG microscope. For
details, see Fig 5 in following paper
 <http://www.jbc.org/content/early/2010/10/25/jbc.M110.187799.full.pdf+html> http://www.jbc.org/content/early/2010/10/25/jbc.M110.187799.full.pdf+html
It is really depended on how you handle your sample preparation, cryo-EM
operation and 3D reconstruction procedure. 

Best wishes,
Gary

----------------
Gang (Gary) Ren, PhD
Staff Scientist
Ph: (510) 495-2375; Fx: (510) 486-7268
Email: gren at lbl.gov; Web: <http://foundry.lbl.gov/> http://foundry.lbl.gov/
Lawrence Berkeley National Laboratory
Molecular Foundry, Rm 2220
1 Cyclotron Road, MS 67R2206
Berkeley CA 94720-8197

-----Original Message-----
From: 3dem-bounces at ncmir.ucsd.edu [mailto:3dem-bounces at ncmir.ucsd.edu] On
Behalf Of Reza Khayat
Sent: Wednesday, October 27, 2010 12:34 PM
To: 3dem at ucsd.edu
Subject: [3dem] Cryo-EM size limit

Hi,
    I was wondering if anyone had an opinion on what the smallest  
macromolecular sample (in size or mass) could be for successful  
single particle cryo-EM and image reconstruction?  I'm hoping for a  
resolution better than 15Å.  This would be without the use of phase  
plate technology.

Thanks,
Reza

_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu
 <https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem> https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu
 <https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem> https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu
https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mail.ncmir.ucsd.edu/mailman/private/3dem/attachments/20101028/672821e1/attachment-0001.html