WASHINGTON — Technology in the lab has come a long way and brought many amazing advances, but it still has its issues — or roadblocks as Jennifer Lippincott-Schwartz calls them. Professionals dedicated to life sciences discovery and technology from around the world gathered Monday at SLAS2017, the annual international conference and exhibition from the Society for Laboratory Automation and Screening.

Lippincott-Schwartz, group leader at the Howard Hughes Medical Institute's Janelia Research Campus, delivered Monday's keynote address, in which she identified three main roadblocks in visualization technology in cell biology and how scientists and researchers may overcome the obstacles through new technology:

• multispectral imaging
• increasing spatio-temporal resolution
• increasing labeling density and resolution

"Even though there is a state right now where we can use all of these phenomenally exciting and innovative new imaging inventions, we still have many roadblocks that prevent us from doing even better imaging,” Lippincott-Schwartz said.

When it comes to multispectral imaging, she wondered why can't scientists look at organelles simultaneously within cells?

"You might think that we should be able to do that because we have a fluorescent toolbox of fluorescent proteins that spans the visual spectrum," she said. "But it turns out that spectral overlap among these different fluorophores limits the ability to look at two or three at the same time without spectral overlap."

How can this be roadblock be overcome? A technique known as fluorescent spectral imaging may solve the problem.

Fluorescent spectral imaging employs a matrix inverse operation for finding the best fit of known fluorophore spectra to that of the recorded spectrum at every pixel in a digital image and also allows many different spectrally variant fluorescent markers to be distinguished in a single sample.

In an experiment, Lippincott-Schwartz's team combined excitation-based spectral unmixing and lattice light sheet microscopy to visualize up to six organelles (i.e., ER, Golgi, mitochondria, lysosomes, peroxisomes and lipid droplets) simultaneously within cells. The linear unmixing helps to overcome the challenge of spectral overlap allowing system-level analysis of organelle interaction. It can also describe frequency and locality of two-, three-, four- and five-way interactions among six different organelles.

"We now have the ability to look at all these organelles at one time within a cell and it's very exciting. It's extremely exciting to be able to see this for the first time," says Lippincott-Schwartz.

Once linear and unmixing takes place, set hunting occurs. Set hunting allows researchers to measure the extent of different organelles overlapping with another and the extent of organelles moving (where, how fast, etc.). All of this information can be pulled from multispectral imaging.

Lippincott-Schwartz presented tips on how to overcome the roadblock for multispectral imaging through use of a commercial confocal microscope, but the microscope itself still presents a problem.

"The problem with a confocal, it's very hard to wrap a timescale to get this information across a full Z stack of your sample," she said. "So in order for us to get 3-D information on how these organelles are interacting in a time domain that's relevant for the behavior of these organelles, we need to turn to the last light sheet microscope."

By using a light sheet microscope, scientists are able to visualize individual organelles and get information about those organelles. The light sheet creates an ultrathin sheet as a result of a series of light that pass across the cell (260 planes per cell). It takes 9 seconds to create a 3-D stack. Since the sheet of light is elapsed, it has relatively low photo toxicity and is isotropic.

Combining the unmixing and lattice light sheet microscopy and the benefits of light sheet microscope over a confocal microscope are just a few examples of how to overcome the roadblocks for multispectral imaging.

"It's an exciting way to overcome this spectral overlap that has plague the community," Lippincott-Schwartz said.

To examine the last two roadblocks — increasing spatio-temporal resolution and increasing labeling density and resolution — Lippincott-Schwartz turned to the endoplasmic reticulum (ER) for answers.

The endoplasmic reticulum in many respects is the "master organelle of the cell." By a wide margin, the ER occupies the largest part of the surface area of the cell. It forms the nuclear envelope, extends out as a tubular array of membranes all the way out to the peripheral of the cell.

The ER appears to have two distinct organizational units.

Tubes are linked through protein in three-way junctions, and sheets are splayed/flattened out. By using improved microscopy, there could be a way to get better insight into the characteristics of these different forms of the structural units of the ER.

Current technology (spinning disc microscopy) gives scientists the base to look at the context structure of the ER in the peripheral of the cell. Newer technology (TIRF SIM) shortens the focal point depth (as low as 90 nm and up to 100-150 nm) using a leaky TIRF system. It increases the speed of how you are imaging a cell from 1-2 second per frame to about 25 milliseconds per frame.

"There's a lot more dynamics than what we had expected," Lippicott-Schwartz observed. "The ER tubes are jiggling around pretty quickly. The three-way junctions are moving, but are appearing to be clustered."

Once this information is learned, data setting begins to take place.

Looking at a higher resolution, the tubes are actually oscillating at a frequency faster than the shutter speed of a conventional microscope. This is due to an energy dependent activity rather than a thermal drive in the cell, perhaps from the exosite skeleton.

Three-way junctions are also changed in a higher resolution. Junctions are shift from clustered to loose patterns over millisecond time.

Having a fast ER model may be important in cell migration.

What does the future hold for visualization technology?

"I think what we're trying to do now is to do even more colors," Lippicott-Schwartz said. "But really it's pretty much a bottleneck right in [that] we have all this data and mining data, but we need people to help us develop processing approaches to be able to map out questions we want to pull out from this data. That applies to not only the last light sheet data, but also the TIRF SIM data as well."