Super resolution microscopy has garnered much attention recently. Capable of generating resolutions 10 times greater than conventional methods, and with a patent for one of the techniques having expired in May, it is no wonder that the major players in the field are moving into the super resolution space.
Traditional fluorescence microscopy methods are diffraction-limited and so up until recently, the highest resolution that optical microscopes could reach was 200nm. Thanks to the development of super resolution techniques, however, this optical law can be overcome, and microscopes can now achieve resolutions of 20nm or less. ‘To pass beyond the legacy diffraction limit was the major achievement for super resolution microscopy,’ noted Matthias Schulze, director of marketing of OEM components and instrumentation at Coherent.
Increasing the resolution by a factor of 10 creates new possibilities for fluorescence microscopy. As most biological structures are much smaller than the previous 200nm limit, scientists can now image biology that could not be seen before. ‘In terms of cell biology research, it really allows the researchers to identify molecular interactions, with molecules being more on the size of 50nm,’ said Mike Szulczewski, vice president and general manager of the Bruker fluorescence microscopy business unit.
As the field of fluorescence microscopy advances, microscope manufacturers are starting to move into the super resolution market. At the end of July, scientific instrument manufacturer Bruker acquired Vutara, a specialist in high speed, three-dimensional, super resolution fluorescence microscopy for life science applications, with a plan to add Vutara’s super resolution technology to its existing portfolio of confocal fluorescence microscopes. ‘Bruker bought the technology because this is pushing the performance of instruments forward for cell biology research,’ noted Szulczewski. ‘We are selling into the fluorescent microscopy research space, and we think that Vutara products with the super resolution are the next innovative step for fluorescent microscopy.’
There are several techniques that can be used to pass the diffraction limit and achieve super resolution, all of which rely on the fluorescence emission properties of the fluorescent dyes used to stain the specimen. These fluorescent molecules can be ‘switched’ on and off, and the technique varies depending on the mechanism used for this on/off photoswitching.
One example of the technology acquired by Bruker is the Vutara 350, a video-rate super resolution fluorescence microscope. To pass the diffraction limit, the instrument first pulses a 1,000mW laser onto the sample, placing the fluorescent molecules into a deep depletion state. ‘You are almost bleaching all of your molecules − you’re putting so much energy into the molecules that all of the fluorescence is turning off instead of on,’ said Bruker’s Szulczewski.
After all of the fluorescence molecules have been ‘switched off’, the energy levels are adjusted carefully so that just a few molecules blink at any one point in time; therefore only a limited amount of light enters the detector. The position of the blinking fluorescence molecule can then be derived, explained Szulczewski: ‘We can look at these blinking pieces of light and we can find the centroid of that light and we can bring that back to an understanding of where the light was coming from.
‘In other words, the light is of a smaller resolution than we can detect, but it’s the same idea as when you look into the sky and see a star − you can relate to where that star is by seeing that one bright light and relating to the centre of it,’ Szulczewski added.
For researchers, a useful feature of the system is being able to obtain three-dimensional images of the sample. Vutara’s Biplane Imaging approach involves splitting the emission light that enters the 2D camera. Because half of the light is directed into one optical path, and the other half goes into another, the axial position of a specific molecule can be obtained. ‘We are getting two different focal planes of collection at any one time,’ Szulczewski described. ‘This allows us to understand where in the three dimensional space that is coming from, along with the XY resolution.’
The real-time capability also allows scientists to watch biological reactions as they happen. For this, collecting the information needs to happen incredibly quickly, which in part has been enabled by advances in processing power and camera technology, Szulczewski pointed out: ‘You need to collect hundreds of images very quickly to apply the mathematics and figure out where these points are in space. The reason it works is because technology is advanced so the cameras are very fast, the computers are very powerful, and we can use graphic processors to do the calculations very quickly,’ he explained.
‘That allows the microscope to be useful to researchers because they can get these super resolution images at a rate of 3,000fps, so they can watch things at the timescale of the technology, where years ago it might have taken an hour to get the same image,’ Szulczewski added.
Image sensors have come a long way in recent years, and have been ‘critical in enabling the development of such super resolution techniques that require thousands of images to be captured with very short acquisition times,’ said Jim Owens, sales manager at Hamamatsu Photonics UK. Previously, the only possible camera solution was to use an EMCCD, but the extra noise introduced by this type of camera meant that the effective sensitivity was limited.
Scientific CMOS sensors are the sensor of choice for super resolution applications, as they can deliver the required high speeds, ‘in all but the most light-starved situations’, Owens said.
Hamamatsu continues to develop and improve its range of Orca Flash cameras to meet the demands of super resolution, as speeds and amount of data is continuously on the rise.
To obtain even better results from the super resolution technique, data analysis and three-dimensional imaging are areas that are going to see much development in the not-so distant future, according to Owens. ‘There is now an increased effort to perform more sophisticated analysis of the data and to create algorithms that can more accurately identify the locations of the emitting molecules and deal with more dense labelling, potential artefacts etc,’ he noted. ‘3D super resolution is still only in its infancy, so there should be some further developments in this area in the next few years.’
STED technology becomes available
Apart from the clear improvements that super resolution brings to the field, another reason that companies are moving into this area is that in May, the patent for stimulated emission depletion (STED) microscopy expired. ‘The fundamental patent of STED fell in May this year − before that, it was licensed only to Leica,’ said Uwe Ortmann at PicoQuant. ‘Now that the patent has run out, PicoQuant and other companies are able to produce these systems.’
Instead of developing entire new systems based on the STED technique, PicoQuant has gone another way and in September launched an add-on which can be used with its existing confocal microscopes. The company’s MicroTime 200 confocal microscope employs a similar laser scanning technique to other confocal microscopes, whereby a laser sweeps over the sample repeatedly to build up an image point-by-point. With the new system, however, scientists can switch on the STED feature, allowing them to further investigate areas of the sample with resolutions of less than 50nm. ‘You can switch on the STED add on, and instead of having a normal fluorescent image that is basically imaged by the confocal resolution, you get the super resolution,’ said Ortmann.
The components that were added into the STED unit consist of a pair of lasers − a pulsed fluorescence excitation laser and a 765nm depletion laser − and phase plates to create a doughnut-shaped laser beam. The laser pulses are synchronised so that the excitation pulse is immediately followed by the depletion pulse, which is red-shifted to the emission spectrum of the dye. Because the STED beam is set to a doughnut shape, only the fluorescence molecules at the edge of the excitation focus are depleted via stimulated emission, and the fluorescence molecules in the centre of the doughnut remain unaffected. The fluorescence is then detected by single-photon sensitive detectors.
According to Ortmann, the availability of STED systems will provide scientists with improved instruments to better demonstrate their research results: ‘It will be the cutting edge in microscopy and biology,’ Ortmann noted. ‘Before the STED, scientists published papers where you would have normal images. But now, because the systems are getting more and more affordable and available in the field, if you want to publish something that is really cutting-edge, you could use a super resolution image.’
Lasers and fibres
To achieve super resolution, lasers of a higher power are needed for the photodepletion or photoswitching of the sample. Laser manufacturers are also starting to notice an increased demand for more power among their customers. ‘What we’ve noticed is the trend for higher powers in that application,’ added Fiona Evans, laser technology product manager at Qioptiq.
But as the powers increase from typically milliwatts for conventional confocal microscopy to multi-watts of power, how does this affect the beam delivery? There are two ways to integrate the laser, according to Coherent’s Schulze: ‘You can [input the laser] directly and image the sample with conventional mirrors and beam steering. The other opportunity is to use a laser with a direct fibre that feeds through the microscope.’ However, as powers increase, it becomes more of a challenge to use fibres to deliver the beam. ‘There is a limitation when I talk about multiple Watts − you would not necessarily get the multiple watts easily through the fibre without facing damages,’ Schulze added.
Recognising the trend for higher powers and the strain that it puts on the fibres, Qioptiq has developed a high power version (HPV) of its Kineflex single-mode polarisation maintaining fibre. The optical design of its input and output has been altered to withstand powers of up to 500mW. The company can also accommodate powers above one Watt. ‘We have worked with researchers that are really pushing the envelope of super resolution microscopy and are using lasers with power coming up to a Watt or more,’ noted Evans. ‘We can offer custom KineFlex single-mode polarisation maintaining fibres up to several Watts input laser power.’
Another way that laser manufacturers are catering for super resolution is by providing a single instrument that uses different wavelengths. This is required because new dyes are being developed by the scientific community, and different wavelengths are required for the on/off photoswitching of the fluorophores. ‘As the biochemistry continues to expand, new dyes come about which require different wavelengths,’ said Schulze.
Coherent’s Optically Pumped Semiconductor Laser (OPSL) technology can be used to produce lasers over a wide wavelength range, with higher powers. The gain medium in an OPSL consists of a diode-pumped thin semiconductor chip, and by altering the design of this chip different wavelengths in the visible spectrum between 458nm and 594nm, as well as in the UV at 355nm, can be achieved. Therefore, the laser is capable of producing laser wavelengths targeted at the new dyes.
According to Schulze, the system also accommodates an interest he has noticed among customers wanting to eliminate acousto-optical modulators (AOM), as they can be complex and costly in super resolution applications that require rapid switching. ‘OPSL allows direct switching, so you just change the diode current and you get optical emission modulated without having an AOM,’ he explained.
Qioptiq has also identified the need for different wavelengths, and also provides a technology for this. Its iFlex-Viper laser engines contain up to five wavelengths all delivered through a single-mode polarisation maintaining fibre so that the beams are co-linear.
Previously, researchers would have to take time to build similar systems, which would be very susceptible to movements and temperature fluctuations.
The laser engine, however, contains all of the wavelengths inside a single box, so it can be moved around the laboratory and attached to different microscope instruments. ‘Condensing all of that into a box that you can carry around is a huge headache out of the way and people can just concentrate on the science that they want to do,’ said Evans.
These features, along with the ability to fire the lasers in any sequence, and adjust the laser powers and modulation rates independently make it well suited to super resolution applications.
