MOLECULE, Single localization Webinar Science Nov 11, 2015

this is a Webinar at 9AM Pacific  sponsored by Science Mag.

It originates in Europe.

moderator is Sean Sanders  in Wash DC...

Markus Sauer begins  (Wurzburg)

shows Chinin and Fluorescein...

shows multicolor images

shows Ernst Abbe 1873  and Rayleigh  (RAY -lee)...

1/2  wavelength of light... limit of resolution.

it is ON the GRAVESTONE of Abbe!!!  d=  lambda over 2 etc.

Nobel in Chem 2014  Betzig,  Stefan Hell, et al.

separate fluor

2500  squares  (50 by 50)  in one  square micron.

where each is  50 nano on a side.

now switch  teh tiny squares off and on  over time async...

and now can see them  even tho  way below  lambda over 2.

dSTORM eg

shows phalloidin  a toxin from death cap  that binds to F actin.

shows  video....

after 30,000 frames...

get incred. resolution...

shows microtubules filaments...

reconstruct  MTs.

dSTORM of mitochondria... eg  inner MT membrane

and can see folding of inner membrane and christae!!

studying plasm membs using click chem and dSTORM.

Glycan imaging on cell membrane...  5M  glycans on plasma memb.

SYNAPSES:  Andreska in Frontiers in Cell Neurosci 2014...

see presynap and post sites...

BDNF is located in single vesicles within glutamatergic presynapes of cultured HC neurons.

but how reliable is this technique... false + and false - ...  reliability?

use nuclear pore complex eg  to correlate  dSTORM and scanning EM = SEM.

and overlay them.

=>  only  0.1% of all  NPCs = nuclear proe complexs are not seen in

dSTORM images  and

99% of NPCs are imaged.

yes... do correlateive microscopy and SEM...

eg  SNC  =  synaptonemal complex... (homologous chromosomes)

see 3d model  of  SNC...  chromosomes together during meiosis.

in his group...  lots  of fly  neuron  imaging.

crash pilot  (in German)  if elim this protein then fly cannot fly  (drosoph)

see diffs in vesicle recruitment... ST  synaptic  plasticity...

active zones in brain slices

how is function encoded at the molecular level?

in tauma, stress, stroke, etc.

see Bassoon and Homer  molecules  in R and L hemisphere...

he has a massive group!!!  German  govt and German res fnd.

2nd Speaker... is  Christian Soeller  from

Univ of Exeter  in UK.

biomed physics... pron  solar.  (German origin)

will talk abt  SMLM  single molecule localization microscopy...

shows the Gartner hypecycle

tech trigger... then inflated expectations and then disillusionment  then slope of enlightment

the platau of productivity...

The HYPE cycle...

hard to overhype  SMLM... it's great!

one of the most robust!  eg of the dSTORM variety.

shows cardiac myocyte in a tissue slice...

focus here on antibody stained smaple  (usual stuff)

nice to start with other techniques and get high contrast image.... gives hope of doing dSTORM.

eg  glycosylated proteins...

how to get  robust  dSTORM working...

make molcules switch  (on and off)

shows  fluor when transitions occur   ... to get to photo blinking.

need some primary thiols  and helps to reduce free O2.

(Markus  was orig a chemist... most of us are not)

shows fiber probe  to check  local O2 in specimen  in liquid.

thiols are labile... Ellman's reagent.

make sure buffers are working...

we use  open slide  with a well in it  (cavity in the middle of it)

trying to shift colors...  (this is directd to beginners with this)

shows test slides  (to get going with this)

Gattaquant super-resolution slides  (buy these to trouble shoot your techniques)

RyR clusters... seen with dSTORM

heart muscle cells...contraciton in heart.

how well do the contract.  can see receptors.

can see make up of clusters....  SMLM => example is dSTORM.

can go down to 25 nm !!!  res.

can label  RyR clusters  label them

eg from  Soeller and Baddeley in JMCC 2013.

how many receptors are present in these clusters...  shows model  with 30 nm  pixel.

see relationship btwn RyR and alpha-actinin.

as in Hou in JMCC 2014...

sCMOS camera for super-res...

pixel properties  unfortun change from pixel to pixel... some have poor readout.

he shows the variance when imaging test beads.

so  can get bias in localization... must correct for this...

last point:  emCCD vs  sCMOS   compare quantum efficiency...

CMOS effic.  is perhaps 10% better.  but some apps for emCCDs.

conclusions:

progress in camera tech will incr  ease of use, speed and reduce cost of SMLM

Orla  Hanrahan  in Belfast, ireland

she works for ANDOR  an Oxford Instr  company

her title is fast and sensitive solutions for super res  Microscopy

PALM/ STORM  is a single molecule deteciton appraoch... so must be v sensitive...

cuz v low intensity signal...

and  localalizaiton width is  proport. to  1 over  SNR.

speed  want to acquite multipel iamges as fast as possible for summation.

detectors...

2 key params:  10  quantum efficiency  (how effic.  at converting

photons  into  electrons in the detectors)

used to be  CCD was camera of choice but moving to  CMOS.

convert photon sto photoelectrons...

2nd para is noise floor:  want a low noise floor  to high SNR.

can get  4T sCMOS  4 transistors per pixel  vs

5T  (which ha sless efficienty  cuz less light gets thru)

shows peak effic around 550 to 600 nm.

noise...   made up of  read noise,  floor noise and shot noise. and dark noise.

v fast readout...

if high QE  quant effic..  that's good.

sCMOS = scientific  CMOS is uqinue in simult offering

extreme low noise... down to  0.9 electrons.

rapid  frame rates  100 fps

wide dynamic range...

shows diff btwn  CCD and CMOS

CCD   must transfer  translated electrons row by row  during readout...

but in CMOS  each pixel is like a separ sensor  (that's good but => variance)

shows blow up view...

microlens...  amplifier  transistor and reset xsistor

photodiode/ potential well and silicon substrate...

sCMOS  fast acquis speeds...   v low read noise

large field is 2048 by 2048  to get  53 fps...

but  at  512 by 512  can get 406 fps.

Diskovery... a new platform from ANDOR  (her company)

can do confocal  ,  TIRF, SMSM

Borealis  their instrument...

can look from 400 to 800 nm...

shows zoom illumination

look at partic ROI.

* on to  audience questions in Webinar audience

Q1:  localize center of mass of photon point spread function...

size of antibody  eg  is 10 nm each ...

so size of label is

Nyquist theorem... to get  res  at 20 nm... must attach at every10 nm  by Nyquist.

Q2:  original  STORM vs  dSTORM...

original had to have  2 difft dyes on same  molecules... so that one would stim theother

but needed in molec  proximity...

but with  dSTORM ... just buy it... one fluor... simplified it enormously.

Q3:  how do photofluors  actually flash?

in nl  imaging... all are one..

with STORM   keep most OFF...  so that they are not excitable

but then stochas  visible...  switching on and off.

Q4:  speed of  camera  in regard to phototoxicity

(ie  laser or  LED  can bleach the sample... destroy it...

really want short exposure times)

so, yes... speed is important...

short exposure is good.

MOLECULE, Single localization Webinar Science Nov 11, 2015

Attachments