One of the most powerful aspects of microfluidics is the potential to sort microdroplets based on some measurable characteristic, such as fluorescence intensity. There are many different ways to do this (acoustic, electric, magnetic, optical and inertia), the choice of which depends on your given application.
Here, we are very happy to report the success of droplet sorting based on chlorophyll content. This was achieved through application of an electric field in response to chlorophyll concentration measured by a laser. This can be seen in the video below, droplets enter from an inlet channel to a Y-junction, if they do not contain cells they pass into the lower channel. If a droplet containing a cell passes under the detector, a short flash can be observed, and the electrodes at the top of the picture are activated pulling the droplet into the upper channel. To achieve this is was necessary to re-design the sorting chip (from an earlier chip made for working with cyanobacteria) to include a wider outlet channel.
As this phase of the project comes to an end I'm pleasantly surprised to say nearly all our goals have been achieved! The only part remaining would be to create opensource software for this project, but due to time constraints this has not been possible.
Moving forward, we have obtained funding for a follow-up project courtesy of the OpenPlant Fund. So in the coming months we will be concentrating on developing methods for on-chip transformation of protoplasts. The reason for this is currently very large volumes of DNA are required for transformation (>10ug), and this is the last remaining bottleneck before we will truly be able to perform high-throughput analyses.
We started this project with the aim of testing whether it is possible to use microfluidics to analyse plant protoplasts, and I think we now have the answer.
After numerous rounds of testing we have improved our working method and are now able to routinely isolate and encapsulate protoplasts. This has been done for two model plant species including A. thaliana, and everyone's favorite Bryophyte -Marchantia polymorphia, the workhorse of the OpenPlant Project for plant synthetic biology (Figure 1).
Figure 1: Encapsulation of protoplasts from model plant species
So the take home message from this project is - if you can protoplast you can encapsulate! But the story does not end here. To be of real use, this process needs to be coupled to transformation of protoplasts. As a result, we teamed up with Oleg Raitskin from Nicola Patron's group at the Earlham Institute. Oleg has been optimizing protoplast isolation and transformation using Nicotiana benthamiana and had a couple of tips for improving isolation, including the use of a cork borer instead of scalpel blade for cutting up tissues to minimize mechanical damage, cutting tissue when submerged in the enzyme mix and using a high ratio of DNA to protoplasts during PEG transformation.
This was a fruitful collaboration, Oleg managed to transform protoplasts with a nuclear targeted Venus reporter and these were encapsulated by Ziyi in the Chemistry department (Figure 2).
Figure 2: Encapsulation of N. benthamiana protoplats expressing a nuclear targeted Venus reporter
So putting all this work together, we have in hand a simple, but very powerful system that opens up a whole range of possibilities for rapid phenotyping in plants (Figure 3).
Figure 3: Schematic of microfluidic analysis of plant protoplasts and some of the potential applications.
One of the stipulations of the project was to pursue science in an open manner, so we have been putting up information on the website protocols.io. I highly recommend checking out the site if you haven't done so already, it has a great set up for disseminating protocols. Further we believe microfluidics is a great technique, so would encourage others to have a go as well!
Looking to the future there are still a few things we would like to work on, The project was briefly presented at Cambridge's Cafe Synthetique meet-up and we had some great feedback, such as trying Calcium alginate encapsulation as a means of improving protoplast viability. Sorting of protoplasts is the next major goal, and requires redesign of a new chip, and finally improving the efficiency of protoplast transformation by developing an on-chip procedure would be a big advantage. This round of our project has come to an end, but stay tuned for future developments.
Finally I want to finish this piece with a big thanks to Cambridge Synthetic Biology SRI for funding the work, it has been a great experience, and to encourage anyone who is interested in protoplasts or microfluidics to get in contact, we are always happy to chat!
In our last post we showed it was possible to encapsulate plant protoplasts. The next step is to couple encapsulation to analysis of fluorescence. This is important as it will allow us to measure the abundance of a given fluorescent protein if we want to test the strength of a promoter or the output of a genetic circuit. Additionally, the technique could be used to measure the amount of a particular chemical produced by cells in each droplet (such as ethanol) if we wanted to test novel metabolic pathways.
To do this we are using a laser setup in Chris Abell's lab shown in Figure 1.
Figure 1: Laser setup for analysis of protoplast fluorescence. Courtesy of Dr. Sara Abalde-Cela. Photomultiplier tube (PMT).
We use the laser beam to illuminate each droplet at a particular wavelength as it passes through the microfluidic device. The choice of wavelength is dependent on the molecule you wish to detect and the experimental conditions (such as pH), (a great online tool with detailed information about the parameters for fluorescent proteins can be found here).
Photons from the laser beam are absorbed by chemicals and re-emitted in a process known fluorescence, which can then be read by a detector. Fluorescence is often found in nature (Figure 2), and use of a protein from jelly fish (known as green fluorescent protein 'GFP') for visualization of biological processes has become so vital to biological research it was awarded the Nobel Prize for Chemistry in 2008.
Amazing Fish. By Sparks, J. S.; Schelly, R. C.; Smith, W. L.; Davis, M. P.; Tchernov, D.; Pieribone, V. A.; Gruber, D. F. - Sparks, J. S.; Schelly, R. C.; Smith, W. L.; Davis, M. P.; Tchernov, D.; Pieribone, V. A.; Gruber, D. F. (2014). "The Covert World of Fish Biofluorescence: A Phylogenetically Widespread and Phenotypically Variable Phenomenon". PLoS ONE 9: e83259. DOI:10.1371/journal.pone.0083259., CC BY 4.0, https://commons.wikimedia.org/w/index.php?curid=30540872
For our first test run we used the principle of chlorophyll autofluorescence (Figure 3). Chlorophyll is one of the molecules that photosynthetic organisms use to absorb sunlight, and when an excess amount of light is absorbed, chlorophyll molecules fluoresce.
Figure 3: Absorption spectra of chlorophyll a and b, the two main photosynthetic pigments in higher plants. http://www.austincc.edu/biocr/1406/labm/ex7/prelab_7_4.htm
This gives us a nice proxy for presence or protoplasts within a droplet, and using fluorescence intensity we were were determine which droplets were empty, contained intact protoplasts or ruptured cells (Figure 4).
Figure 4: Measurement of chlorophyll fluorescence as droplets pass through the laser set up (Time/s).
This was very exciting as it means we can now expand this approach to analyse fluorescence at a wide range of wavelengths, providing the basis for our test. Further, the fluorescence signal intensity can potentially be used to sort droplets.
We presented our project at the OpenPlant Forum in July. It was encouraging to see the excitement the work generated and we had a number of good discussions with people about improving the quality of our protoplasts.
Following on from this we will be teaming up with Dr. Oleg Raitskin at the Earlham Institute to couple fluorescence analysis to protoplast transformation.
Taking individual cells and encapsulating them inside liquid droplets. The idea is simple, but powerful. It opens up a host of possibilities for high-throughput, single cell analysis ranging from genomics to screening for metabolite production. It has been used widely in analysis of bacterial communities and cell tissue cultures, but has not yet been adopted by plant scientists.
I have previously written about the process of extracting individual cells from leaves by enzymatic digestion. The next step is to encapsulate cells in microdroplets - this is achieved by flowing a cell suspension into a 'carrier solution', the two are immiscible so droplets of buffer containing cells are formed (Figure 1). If you have ever eaten an oily soup before you will have observed a similar phenomenon where circles of fat can be seen floating on the surface.
Figure 1: A diagram of encapsulation of cells in water in oil droplets
However, protoplasts are extremely fragile as they lack a cell wall, and they are prone to rupturing. This can occur from excess agitation, shearing pressure from pipetting or even allowing protoplasts to drip from one solution to another (as during filtering). Although our yields have been okay, perhaps 30-50% of our cells rupture, so there is room for improvement. There is a recent method in Cold Spring Habour Protocols that suggests a means to remove ruptured protoplasts which we may be trying in the coming weeks.
Due to this fragility we were concerned that the pressures protoplasts encounter during passage through a microfluidic device might cause them to burst. So we needed to test whether it is possible to encapsulate cells without them rupturing. To help us do this we have been joined by Ziyi Yu on the project, who has previous experience working with algal cells.
The first attempts were unsuccessful, most droplets were empty, and the odd ones we did see encaspulated were irregular in shape, suggesting they were very unhappy or partially ruptured. Cell densities were too low, and we also experienced problems with protoplasts settling in the syringe used to inject the cell suspension into to the microfluidic device.
This project is split between the Plant Science and Chemistry Departments at the University of Cambridge. Transport of the protoplasts between the two caused a great deal of damage to the cells, so we instead performed the extraction in the chemistry department using wild type N. benthamiana leaves. W also added a product to alter the density of the buffer solution to stop cells pelleting in the syringe and made sure we loaded cells in the top of the syringe rather than sucking up through the needle to reduce shearing stress. After a couple of hours of fiddling we were able to effectively encapsulate cells as demonstrated in the following clip:
The droplets were collected and analysed under a fluorescence microscope (Figure 2).
Figure 2: Intact encapsulated protoplasts (20x magnification) showing bright field (left) and chlorophyll fluorescence (right).
After our initial disappointments we were very pleased to see this! However, there is more work to be done; most droplets were still empty and we still had quite a large amount of cells rupturing. Over the coming weeks we will be tweaking the protoplast preparation procedure, and optimizing encapsulation to increase the number of droplets containing intact protoplasts (which can be achieved by varying cell density) before going on to test the ability to analyse fluorescent proteins in transformed plant cells.
I will also be uploading some protocols with a bit more detail for those who are interested asap, however in the OpenPlant project we are currently in the process of discussing the best way to do this. More soon.
How do you isolate plant cells for high
throughput analyses?
Under normal circumstances plant cells are held together in tissues by crosslinked pectin
molecules. Enzymatic digestion of the
cell wall releases individual cells, or “protoplasts” which can be genetically
transformed and used for further analyses. Perhaps the most prominent proponent
of protoplasts is Jen Sheen’s Lab
at Harvard. Commendably on their website they provide detailed
protocols about how to isolate protoplasts from Arabidopsis and maize,
as well as answers to FAQs about preparing and transforming protoplasts and a
very helpful video about how to isolate healthy
Arabidopsis protoplasts.
In our initial trials (Figure 1) we have been relying heavily on these
protocols; I am not going to repeat them here as I didn’t deviate from the
Sheen Lab’s protocol for maize protoplast isolation.
Figure 1: Protoplast isolation (A) Maize seed germination - 48 h in dH2O (B) Maize seedlings grown for 9 d under long day conditions (C) the middle 8 cm of the second leaves were cut into strips and incubated in enzyme mixture (D) vacuum infiltrate the enzyme mix into leaf sections (E) gentle shaking at 40 rpm for 90 minutes (F) tissue was filtered through Miracloth and protoplasts harvested by centrifugation.
Marchantia polymorpha (Cam-strain) according to a protocol inspired by
M. Bopp et al., 1988, Plant Cell Physiol.: 10 day old thalli were harvested and
pre-incubated in 10 mL half strength Gamborg's B5 medium supplemented by 1 g/L
casamino acids, 0.3 g/L L-glutamine, 20 g/L sucrose, and 60 g/L d-mannitol
(Solution S1) for 30 min at room temperature. Subsequently, thalli were shaken
in 10 mL S1 supplemented by 2% Driselase at room temperature for 3 h.
Protoplast were then filtered through a nylon net of 80um mesh width, and
washed four times in S1 (centrifugation: 100xg, 4 min). Protoplasts were shaken
in S1 in light over night prior to processing.
As you can see in (Figure 2) the plant cell wall predominantly consists of
cellulose microfibrils in a cross-linked polysaccharide network of pectins (a
short review can be found here). To isolate Maize
protoplasts we have been using a combination of Cellulase and Maceroenzyme
R-10, a multi-enzyme mixture containing Cellulase, Pectinase and
Hemicellulase from Rhizopus sp. (common saprophytic fungi). Likewise the Driselaseused in
Marchantia protoplast isolation is also a multienzyme mixture,
this time from Basidiomycetes sp. (also
fungi) that contains Cellulase, Pectinase and Hemicullase activity.
Figure 2: Diagram of the plant cell wall
Source: LadyofHats - Own work, Public Domain, https://commons.wikimedia.org/w/index.php?curid=2881078
Once digested for several hours protoplasts are harvested by filtering to remove residual tissue, followed by centrifugation. Our initial attempts were successful (Figure 3),
although a fairly high amount of rupturing was observed in maize protoplasts,
so the next few weeks will involve optimizing the procedure (if anyone has any tips we would love to hear from you), whilst trying to
develop a transformation protocol.
Figure 3: Maize Protoplasts
PS: A tip for beginners –
don’t use a glass slide with coverslip to view your protoplasts – they will pop
leaving you with chloroplasts!
‘There once was a speedy hare who bragged about how fast he could run. Tired of hearing him boast, Slow and Steady, the tortoise, challenged him to a race. All the animals in the forest gathered to watch.
Hare ran down the road for a while and then and paused to rest. He looked back at Slow and Steady and cried out, "How do you expect to win this race when you are walking along at your slow, slow pace?" - Aesop’s fables
Working in plant science I sometimes feel like the tortoise in Aesop's fable. Animal biologists race on ahead making discoveries, before stopping to laugh, amazed plant science's ‘slow, slow
pace’. Speed
is a real issue when you work with plants, to illustrate the point a
colleague of mine recently obtained some seeds that came with the following advice - ‘germination may take between two
weeks and two years’. The challenge of time is one of the key
reasons why a lot of research is still done in model plant Arabidopsis
thaliana: it has a relatively short life cycle, so it can be quite quick to make a mutation and analyse the effect. However, even at 8 weeks
to seed, as Arabiodpsis is diploid, so you need to do through the process of growing plants, collecting seed and select mutants three times in order to generate plants suitable for analysis – as a result it is months before you can carry out an experiment.
To speed things up, plant scientists can use transient assays, whereby it is possible to switch on or off a gene of interest for long enough to analyse the effect. The most common methods in plants include agroinfiltration
(video), biolistic
bombardment (video)
or viral
induced gene silencing (VIGS). Agroinfiltration uses agrobacterium to infect plants, this gram negative bacterium is able to transfer DNA from itself into a plant genome in a process that has been adapted to introduce a genes of interest. This process works great in Nicotiana benthamiana, but there are
many species in which it is far from optimal, and we have not much success using
either grasses such as maize, or our model species Gynandropsis
gynandra. Biolistic bombardment involves coating metal particles with DNA, which are then introduced into plant cells using a genegun (sounds cool, but believe me it gets tedious after a while!). This works okay, and for my purpose it is
probably the best option currently available, but biolistics is not really suited to high
throughput analyses. VIGS can only be used for switching off genes. The
alternative is to take an approach that is analogous to using cell lines, the use of which gives animal research
its speed.
Maize mesophyll protoplast
Techniques have been established to create plant cell tissue cultures as well as isolated plant cells known as protoplasts (which do not propagate) from mature leaves, or plant embryonic tissue (known as callus) (if you are interested in the history of protoplasts a personal account is provided by Prof. Edward Cocking who pioneered their use here). However, these
approaches have their limitations, the most significant question is how representative of ‘real’ systems are these cells? Plant cell tissue
cultures are suitable for biotechnological purposes, but as the gene activation patterns are so far removed from ‘normal’ systems they are not always suitable for scientific study. Protoplasting
can introduce stress responses that must be considered when interpreting experiments, but protoplasts are a close enough to real cells to be a useful tool in preliminary analyses.
Therefore, as previously mentioned, we hope to couple the use of protoplasts to developments in microfluidics
to generate a system suitable for high throughput analyses and help give the
Slow and Steady plant scientists their running shoes. In
the next post we will be give a bit more information about protoplasts and provide an update on our initial attempts to isolate them, as well as preliminary tests about how they behave in a microfluidic system.
This blog is an experiment in open science. I am a postdoctoral researcher in Prof. Julian Hibberd's lab at the University of Cambridge, and moonlight as a editor for the PLOS Synthetic Biology Community. In my day job I work as part of a team that is seeking to understand an adaptation possessed by high yielding crops like maize, with the aim that we may one day be able to use this knowledge to boost yields of species such as rice. Our lab is part of a couple of international consortia pursuing this goal including the C4 rice project and the 3to4project.
Our research is focused on identifying DNA regulatory elements which limit gene activity to specific regions of a leaf in order to aid the design of synthetic circuits. Conventional approaches involve the fusion of a reporter (often the E. coliuidA gene) to a promoter that is being tested. This reporter is then inserted into plants, and the regions where the promoter is active are visualized by staining (see Kajala et al. for an example).
Plan for chip to sort protoplasts courtesy of Dr. Sara Abalde-Cela
The problem with these methods is that they are slow and low throughput. With genomic technologies starting to provide vast numbers of candidate promoters, novel methods are required to screen them. I am a molecular biologist, and so is my colleague Ivan Reyna-Llorens who works with me on this topic. We realized that to tackle the problem we needed to bring together a team of experts from different disciplines, so we managed to convince Christian Boehm, Dr. Sara Abalde-Cela, Dr. Paul Bennett to join us. The result was a proposal to use a combination of plant protoplasting, differential fluorescence analysis and microfludics to sort cells based on fluorescence intensity.
We recently received funding from the Cambridge Strategic Research Initiative for Synthetic Biology to develop this device. This came with the stipulation that all outcomes are open source. Additionally, as my fellow colleague Richard Smith-Unna has been encouraging Cambridge scientists to start implementing open practices in their research, I though this project was an excellent opportunity to learn how to do science openly (hence the blog).
Funding for the project lasts for six months, and we will start in January 2016. We are very excited, and if this pilot project is successful we hope to utilize the device to tackle a range of additional problems. In the spirit of open science we will be reporting regularly and making all our data available online in public repositories. Additionally we recognize science is a collective effort, so we will be keen to hear from anyone who has any bright ideas or novel applications for our device that might be interested in getting involved, comments are welcome.