September 6, 2018
by Carol Lynn Alpert
All hands in attendance today. Lots of good news.
In their quest to shrink both imaging targets and exposure time, Abhijit and Peter have acquired a new camera, an EMCCD, which is supposed to reduce their “read noise” (stray light) down to the required 0.1 e- level. They are now able to image smaller beads (100 nm) at 50 milliseconds. However, they are still experiencing some read noise. A recheck of the simulations showed there should be almost none. Yet when the laser is on, stray light is somehow getting into the camera. It could be from reflections, or possibly even fluorescence from something in the cardboard screen around part of their set-up. Abhijit used black duct tape to seal the inside plane of the cardboard. That helped a bit, but they will have to continue trying to get as close to a “black room” as possible. This is essential, since, as the diameter of the object being imaged shrinks, the read noise increases by a factor equivalent to reduction in volume; so, trying to go from 50 nm diameter beads down to 25 nm beads (our next target) will increase the noise by a factor of 8.
Peter and Abhijit have continued their collaboration with Daichi Kamiyama, using the new microscope to image microtubules. Daichi’s real goal is high-resolution imaging of neuromuscular junctions; however he is working with microtubules first in order to optimize his labeling process. These are easier to access. Abhijit and Daichi have produced some good images of microtubules using Alexa 405 and 647 dyes, but Daichi is not yet satisfied with the density of dye labeling, so he is adjusting the required histochemistry.
Both Abby and Kil Ho have gotten close to optimizing their respective approaches to QD-ssDNA conjugation: Kil Ho through the DNA embedding technique, and Abby through the Loops, Trains, Tails method and click chemistry.
Abby: Loops, Trains, Tails
Abby walked us through her blog post which details her further pursuit of click chemistry to attach single stand DNA oligonucleotides (ssDNA) to quantum dots. She is trying to optimize the number of ssDNA attaching to each QD in aqueous solution. The more conjugation sites, the greater the thermal stability and yield of the end-product: the QD – DNA – gold nanoparticle (AuNP) conjugates. This month she focused on adjusting incubation times as well as increasing the density of the intermediate activating molecule DBCO that is conjugated to the QD-PC3 in Step 1 of the click chemistry process. The more DBCO molecules that attach, the more DNA will conjugate in Step 2. Abby began by finding the optimal ratio of DBCO molecules to QD in solution, determined by measuring the absorbance peaks of dye molecules attached to the DBCOs. 1000:1 failed, with conjugates precipitating out of solution. Both 100:1 and 10:1 succeeded, however, with 100:1 giving a moderately higher yield. However, since increasing the proportion of DBCO tended to increase hydrophobicity of the conjugates, she explored using a sulfo-DBCO molecule instead. This is more hydrophilic than the plain DBCO, and proved successful.
The longer incubation times (12 hours rather than 7) also produced higher yields, so Abby will now continue using those and will substitute sulfo-DBCO for all future experiments. Her next step is to test the success of the click-chemistry-produced QD-ssDNA conjugates with “origami hinges” made of complementary ssDNA.
Kil Ho: DNA-Embedded QDs
Kil Ho reported on his latest round of systematic tests to optimize all the variables involved in producing QD-ssDNA conjugates through the DNA embedding technique. He found unexpected success when he reduced the speed of the centrifuge used in the purification process from 7000 rpm down to 4000 rpm. The slower speed prevented aggregation; the conjugates were well dispersed in solution. However, there was a marked red shift in the fluorescence of these conjugates, which might be explained by the increased diameter of the ZnS shell grown on the CdSe shell-core QDs. Kil Ho followed up by testing the effect of using a pH of 12 rather than 10. Although pH 12 had been suggested in the literature, it hadn’t worked as well on Kil Ho’s previous crops of less stable QD conjugates. This time, it did improve the results. The conjugates did not aggregate, they were highly soluble in water, and there was only a slight redshift. Kil Ho will doublecheck these for successful DNA conjugation using fresh DNA origami hinges, as soon as they arrive. Later, because of the slight redshift, the team may have to make additional adjustments to ensure overlapping of the narrow bands of absorbance wavelength suitable for triggering both QD fluorescence and FRET quenching through the azobenzene gold nanoparticle linker. Redshift thus far is 10-14 nm.
The team at MOS is launching an effort to develop a 3D animation of the holographic light sheet STORM process the Optics Team is developing. Megan Litwhiler, Research Communication Associate at MOS will be working with Peter and Abhijit to identify existing visual and educational resources. Megan is working toward achieving a good understanding of the techniques so she can translate them for broader audiences and so we can develop a storyboard for a new 3D animation. This work will contribute to the filmmaking project and the development of new educational resources for the website.
Two weeks ago, Carol Lynn gave an invited talk, “Science Communication Skills that Enhance Discovery and Innovation,” at the annual meeting of NSF Science-Technology Centers Directors at UC Berkeley, and participated on the panel “Bringing Science and Engineering to the Public.”
Jessica was feeling optimistic that the QDot Team might be able to bring quenchable QDs to Georgia by December. Peter noted that by mid-December, he will have to vacate his current lab space and move it to a building that currently houses plant biologists. Reportedly, the vibration level in that building is no more pronounced than in their current site. Since he’ll have a month or so of downtime during the move, we now hope to meet in January or February. Our goal is to be able to show QSTORM-AO proof-of-concept by spring 2019.
While the NSF IDBR program that funds our current work seems dormant, we discovered, during the course of this meeting, the launch of a new program for Infrastructure Innovation for Biological Research (IIBR), with a rolling proposal process. Our current program officer and a program officer who co-organized the Innovations in Biological Imaging and Visualization Ideas Lab (IBIV) in 2010 (where we all met and conceived QSTORM) are involved in the new program.
We scheduled our next meeting for 1 pm on Oct 11. Blog posts are due at 1 pm October 9.