Like many iGEM projects, we needed fluorescent microscopy to characterize our cells. In order to get our cells to lase, we need to shine a laser on the cells with different excitation intensities, without photobleaching the cells. A setup with a pulsed laser could be the outcome for this. However, the appropriate setup was not available in our department. Therefore, we decided to build our own microscope out of separate optical parts. By discussing our problem with photonics and optics companies, and show them our motivation to solve this difficult challenge, we were able to get all required components sponsored or borrowed, making the setup nearly cost-free. With our minimal set of available tools, we calculated and designed the optics in such a way we could image fluorescent cells while minimizing photobleaching. Our home-built optical setup did not only look impressive, but it was also crucial for characterizing the ability of our fluorescent cells to become a laser. We are using a 405nm laser and filters suitable for detection in the blue spectrum of our mCerulean (part BBa_K1890010) cells, but the setup can be easily customized to fit other projects. Each component in our setup is easily interchangeable to customize the excitation and detection capabilities for any project. We believe our specific design, as well as our approach towards solving a difficult hardware problem, can be of good use to future iGEM-teams.
Designing a custom setup
In order to make a cellular laser, we cover the cell in polysilicate to create a biological ‘optical cavity’. An optical cavity is a space consisting of different refractive indices, that has the special ability to 'trap' light. Additionally, we will express fluorescent proteins in the cytosol, which will form the 'gain medium'. The photons emitted by the fluorophores will be reflected within the optical cavity, allowing them to resonate within the cell or cell culture. Once an emitted photon hits an excited fluorophore, this fluorophore will emit a photon just like the incident photon, which we call stimulated emission. This concept was discovered in 1917 by Albert Einstein (Einstein, 1917), and is the theoretical basis of our biological laser. This theory also introduces one of the key concepts required for the cells to emit laser light: in order for the light to ‘accumulate’, the majority of the fluorescent proteins should be in an excited state. Only in an excited state, an incident photon can cause stimulated emission. Therefore, if we want the cells to emit laser light, it is essential to keep the majority of the fluorescent proteins exited. We call this population inversion (Gather & Yun, 2011; Svelto & Hanna, 1976).
The lifetime of the excited state of the fluorophores is usually a couple of nanoseconds (Berezin & Achilefu, 2010), but continuously shining a laser on the cells will cause photobleaching and the fluorescence will decay within 100 µs (Jonáš et al., 2014). Due to photobleaching, we will not be able to maintain population inversion in a cell for a long time, and therefore we are not able to observe processes in the cell for a significant timespan. Therefore, we need a system where population inversion is possible without bleaching the proteins. The solution to this is using a pulsed laser with a high intensity. This way, you immediately create population inversion and you can measure whether your cell is able to lase. By allowing a long time between the pulses, the fluorophores get enough time to fall back to their ground state before getting excited by a new pulse (Gather & Yun, 2011), which extends the time before the bleaching occurs (Jonáš et al., 2014).
This type of excitation is often not possible in a normal fluorescence microscope, since these microscopes are not equipped with pulsing lasers. The department of Bionanoscience of the TU Delft owns a confocal microscope with pulsed lasers, which we will use for our research. However, this microscope does not entirely fit our needs, as it’s not customizable, only focuses on a part of the cell and has limited pulsing capabilities. We need to excite the entire cell with a strong pulse to cause population inversion. Therefore, we have decided to build our own microscope.
Bridging the gap between synthetic biology & photonics
Unfortunately, the kind of setup our research requires does not allow a DIY setup with cheap materials. The required equipment is very specific, so there is only little room to improvise. Buying all this equipment was not an option, as it was way too expensive. However, we decided to see this as an opportunity instead of a problem, since building our own microscopes allowed us to collaborate and interact with one of the fastest growing fields within science: the field of photonics. Photonics companies are just now discovering their potential in the life sciences, as their classic applications were mostly material science and electrical engineering. Collaborating with these companies gave us the chance to narrow the gap between photonics or optics and the life sciences. None of these companies were familiar with iGEM and only slightly familiar with the field of synthetic biology, so it was quite a challenge for us to convince these companies that a collaboration would be of great value for both of us. Luckily, a lot of photonics companies acknowledged that the life sciences are an exciting new field for them to work in. Many companies were happy to collaborate and help us, by either sponsoring us with equipment, loaning equipment or helping us designing and building the setup.
Building the microscope
The microscopy setup
Before we started building, we thought about the requirements our homebuilt optical setup should have. The microscope, should be able to excite the laser cells using a pulsed excitation laser, and should be able to measure both an image, from which we can deduce intensity, and a spectrum of the cell, since the biolasers are expected to have improved spectral resolution. For reproducibility purposes, we wanted to keep the setup as minimal as possible. A schematic picture of the design of the setup is shown in figure 3.
For our setup we were able to borrow a blue laser with an excitation wavelength of 405 nm, which matched perfectly with the excitation requirements of our mCerulean cells. The cells have their emission maximimum at around 475 nm (Shaner, Patterson, & Davidson, 2007). Of course, we must prevent that we measure any reflection of the excitation laser, since we are only interested in the emission of our fluorescent cells. Therefore, the setup is designed in such a way that the light of the excitation laser is filtered out after it excites the sample, so we are sure that the only output signal we measure comes from fluorescent cells. We use a dichroic mirror and a filter to filter out the excitation laser. The dichroic mirror is a 425 nm longpass mirror, which means it reflects all light of under 425 nm, so including the wavelength of the laser, while letting light with a wavelength above 425 nm pass through, which is the light emitted by our mCerulean (part BBa_K1890010) cells. Because the dichroic mirror has an efficiency of ~90%, we use an additional filter to filter out the remaining light coming from the laser. This way, we will only measure the fluorescence output of our cells, as shown in figure 4. Furthermore, we use short exposure times of the laser to minimize photobleaching. The beamsplitter in the end enables us to measure one signal in both in the CCD camera and spectrometer at the same time.
Building such a setup is not as easy as it seems. Each component should be perfectly aligned in the path of the laser beam in order for the light to be guided correctly. If a mirror is only 1° off, the path of the light will be distorted. Each misalignment amplifies this effect. Therefore, all components need to be in line with each other, which is a precise and tedious work. All parts are mounted on an optical table, which is a shock-free table that allows parts to be screwed into it. The first building session lasted for 9 hours, but we summarized the building with a time lapse video of only 1.5 minute!
A more detailed view of the setup:
Because the setup consists of separate parts that are essential for fluorescence microscopy, it is very versatile. In our case, the setup is designed in such a way that it can excite and measure within the blue range of the spectrum. However, since every part of the setup is very easily interchangeable it can easily be modified to fit other requirements. If another iGEM team would require a custom fluorescence microscope, for example one that can measure in the red part of the spectrum, changing three components (the dichroic mirror, the filter and the laser) is all it takes to do so. Because the setup is so minimal and flexible, it can be easily adjusted to any project.
Our setup consists of the following parts, hover over the parts to find out more:
The laser cells will be placed on a stage using an ordinary microscope slide. We will use a very dilute sample of our cell culture to obtain a single-cell approach.
We will use a 50x objective of Nikon. This objective will both focus the light beam of the excitation laser on the cells, and it will magnify the laser light coming from our biolaser cells.
Since the cells have to be in a horizontal plane, we have to direct the light from the excitation laser in a vertical direction. We use a broadband dielectric mirror that reflects over 99% of the incoming light with an angle of incidence of 45°, so almost all of the photons of the excitation laser are directed towards the cells and all light emitted by the cells are directed into the setup.
For the optical setup we use a OBIS LX 405nm laser, kindly provided by Coherent. The laser is suitable for exciting the fluorescent protein mCerulean, which is one of the three fluorophores we use. This protein emits the shortest wavelength of the four fluorophores we use and is therefore more suitable for lasing. If the wavelength is close to the size of the cell, it will resonate less.
The BNC 745-4C pulse generator, which was provided by Laser2000, will allow our laser to pulse. When the laser is attached to a continuous power source, the laser will emit a continuous beam. When one attaches a pulse generator, one is able to modulate the output beam of the laser. By varying the input power between 0 and 5V we can change the output power of the laser, and changing the power inlet from continuous to short pulses will cause the laser to pulse. The pulses will only be a few nanoseconds.
The dichroic mirror is used to filter out any residual light coming from the excitation laser, so we don’t measure this as output. The dichroic mirror is a 425 nm longpass mirror, which means that is fully reflects light of all wavelengths below 425 nm, including our excitation laser, which is 405 nm. The light emitted by the fluorophores will have an emission at 475 nm, so this light will pass through the mirror. This way, we only measure the light coming from our cells, but not from the excitation laser.
To make sure that the signal we are measuring is only originating from our laser cells, and is not light of the excitation laser that was not filtered out by the dichroic mirror, we use a filter. This filter only transmits light of certain wavelengths. We use a filter that reflects all light at 405 nm, which is the wavelength of the laser.
To determine if a cell is producing laser light, we would like to measure both the intensity and the spectrum of the emitted light, while also making a picture of our cells. These measurements should be done at the same time, so we are sure that the cell has all the desired properties at once. However, we do not own a device that can do these measurements at the same time. Therefore we use a beamsplitter, that will split the emission light of the cell in a 50/50 ratio, and will direct one half to the CCD camera and the other half to the spectrometer, so we are able to measure the output from one cell with two devices at the same time.
In order to get an image of the cells we need a CCD (Charge Coupled Device) camera. A CCD camera has a sensitive charged chip. When a part of this chip is exposed to light, an amount of electrons gathers at this place. The ‘pattern’ of electrons on the chip is then translated to a picture. The image can also be used to determine the intensity of the emitted light. The CCD camera we use was sponsored to us by DeltaPix.
The main difference between a fluorescent cell and a laser cell is the spectra the cells produce. A laser has a smaller spectral range than fluorescent light. To determine if our cells also expose this behaviour, we include a spectrometer in our setup. The spectrometer was donated to us by Thorlabs.
After hours of hard work we managed to build the full setup! Unfortunately, we couldn't connect the pulse generator, so the exposure time was manually regulated. A picture of the finished setup is shown in figure 5 and figure 6 shows the setup with an animated light path.
Testing our setup
In order to confirm whether the setup was working, we used E. coli BL21 cells that were transformed with our constitutive mCerulean BioBrick. We have previously confirmed that these cells are able to fluoresce and that they can be excited at 405nm, the wavelength of our laser. To make sure the only output we were measuring was fluorescence, and not any ‘leakage’ of light from our laser beam, we also tested cells that were not transformed with the mCerulean (part BBa_K1890010) BioBrick (OmpA-silicatein, part BBa_K1890002). These cells are not able to fluoresce at excitation at 405nm, so if the setup is working properly we should not get a signal from these cells.
The cells were adhered to the microscope slides using 3% agarose pads and imaged at an excitation intensity of 0.5 mW. This energy is low enough to not instantly photobleach the proteins, but observe fluorescence clearly. Focusing on the cells was done manually. The results are shown in figure 7.
As we can see in figure 7B, the cells transformed with a gene for the fluorescent protein mCerulean are clearly visible. When using a strain transformed with a plasmid that is known to cause no fluorescence in the cells, in this case OmpA-silicatein, no light was observed, as shown in figure 7A. From this we can conclude that we successfully built a setup that is able to observe and measure fluorescence in a cell. There is no leakage of light of our excitation laser in the camera, since we do not observe anything when we use non-fluorescent cells. Also at a higher excitation energy (50 mW) we did not observe anything on the camera. Therefore we can conclude that our setup works as expected as it is indeed able to measure fluorescence without measuring other light sources.
More information on the results of experiments with the laser setup can be found at the Project page.
Working with lasers has some serious risks. Exposure of your eyes or skin to a laser can cause burns or loss of vision. As an example: a laser pointer can already cause harm after staring at it too long, and the power of such a laser pointer is only 1 mW. Since we were working with a laser up to 100 mW, we needed to take some serious safety precautions
As a start, we all received basic laser/microscopy safety training. Furthermore, the team members that were involved in building the optical setup for the hardware component of our project also received an advanced laser safety course prior to their work in our optical lab. The laser safety of our project was supervised by Jeremie Capoulade (Head of the microscopy facility of the Department of Bionanoscience, see acknowledgements). We wore the appropriate eye protection at all times, to protect ourselves. Since our faculty moved to a new building, our laser lab at first was not classified as ML-I (hence we are not wearing lab coats in the building video). In this time, we did not test any living material in the setup. As soon as our lab got an ML-1 status, we also wore lab coats and took the right ML-1 measures.
Our lab was an optical lab, which means it was fully isolated from any light going in (to protect our measurements) or going out (for the safety of other labs). We made sure that our laser was not pointed at the door or that there were reflective components that could reflect a harmful laser beam out of the lab. Furthermore, the lab has all the requirements for laser safety, such as a warning sign and an interlock that will shut everything down when the door is opened or in case of emergency.
- Berezin, M. Y., & Achilefu, S. (2010). Fluorescence lifetime measurements and biological imaging. Chemical reviews, 110(5), 2641-2684.
- Einstein, A. (1917). Zur quantentheorie der strahlung. Physikalische Zeitschrift, 18.
- Gather, Malte C., and Seok Hyun Yun. "Single-cell biological lasers." Nature Photonics 5.7 (2011): 406-410.
- Jonáš, Alexandr, et al. "In vitro and in vivo biolasing of fluorescent proteins suspended in liquid microdroplet cavities." Lab on a Chip 14.16 (2014): 3093-3100.
- Shaner, N. C., Patterson, G. H., & Davidson, M. W. (2007). Advances in fluorescent protein technology. Journal of cell science, 120(24), 4247-4260.
- Svelto, Orazio. Principles of lasers. Ed. David C. Hanna. London, New York, Rheine: Heyden, 1976.