Team:EPF Lausanne/Safety



Safety Form

The safety form of our team can be found here.

Safety in Microfluidics

Potential biopad
A potential BioPad

The use of microfluidics throughout our project sparked an idea within our team: microfluidics could be used as a powerful tool to improve biosafety when handling Genetically Modified Organisms. It provides an enclosed space to culture and analyse engineered organisms, thus reducing the risk of releasing them in the environment.

To develop this idea, we brainstormed on how we could provide a safe and practical microfluidic context for synthetic biological systems. We imagined a few improvements to current existing devices and discussed them with 2 professors teaching in Lausanne, who are both experts in the interface between biology and microfluidics:

- Pr. John McKinney, who is the head of the Laboratory of Microbiology and Microsystems at EPFL (Ecole Polytechnique Fédérale de Lausanne) and uses microfluidics to study pathogenic strains such as uropathogenic E. coli or M.tuberculosis

- Pr. Jan van der Meer, who is the head of the Department of Fundamental Microbiology at UNIL (Université de Lausanne) and works on implementing biosensors in a microfluidic chip

Discover here what came out of these interviews and how we took advantage of microfluidics to design a completely safe device!

What are the safety issues related to synthetic biology?

Bacterial conjugation : Horizontal gene transfer

The commercialization of devices containing GMOs – just like our BioPad – must address biosafety related concerns that currently don’t always have solutions.

When brainstorming about this, we realized that one of the main dangers is the release of these organisms in the environment. Horizontal gene transfer could lead to the spread of new competences to other organisms. This is especially true for plasmids containing antibiotic resistance genes. Their release could favour the development of new antibiotic resistant strains – an omnipresent threat to modern medicine.

A second possible issue could arise when handling pathogenic strains. In this case, it is vital that the devices containing the organisms do not pose any threats to human health. Thankfully, synthetic biological systems based on pathogens are quite seldom made. We nevertheless think that it could be helpful to design devices limiting the risks of contamination.

What are the main advantages of microfluidics regarding biosafety?

The most significant advantage of using microfluidics is that it constitutes a closed environment: cells are hermetically trapped between the glass slide and the PDMS chip. This minimizes the manipulation of open liquid cultures, which reduces the risks of contamination and release of dangerous biological components.

The output flows into a container with bleach

One could still argue that cells can leave the chip if liquid is actively flowed through it (for example to continuously provide cells with fresh medium). In John McKinney’s lab, they solved this problem by connecting the outlet of the chip to a bottle containing bleach or ethanol; this allows immediate decontamination of any liquid leaving the chip.

The second advantage of microfluidics is size reduction. Only minute amounts of cells or media need to be manipulated. In comparison to analyses on 96-well plates, where typical reaction volumes are about 100 µL (approximately 107 cells for an overnight bacterial culture), microfluidics allows the manipulation of nanoliter-range volumes, corresponding to only 102-104 cells. This is of critical importance when handling pathogens as it limits the risk of contamination by drastically decreasing the exposure dose.

Size reduction is also crucial to commercialize biosensor devices containing GMOs. As Jan van der Meer explained to us, the current regulations concerning biosensor devices depend not only on the type of organism used but also on the amount of cells that the device contains. Designing biosensors based on microfluidics could therefore make acceptance of GMO-containing devices easier by the regulation authorities. For more information on regulation of biosensors in Switzerland, we encourage you to visit this website.

Are there drawbacks?

Since microfluidics is a very recent tool regarding biological studies, there are still some points that need to be optimized as far as biosafety is concerned.

Plasma treating the glass slide and the chip

First of all, the chip remains a hermetically closed space as long as the bonding between the glass slide and the PDMS is perfect. It sometimes happens that the chip delaminates, i.e. that the PDMS unsticks from the glass. This can be the case when liquids are flowed into the chip with a very high pressure or when cells form clots that block the channels, leading to local increase in pressure. Delamination is a real safety problem as it leads to release of cells in the environment and thus possible contamination. Yet it can be avoided by plasma-treating both glass slide and PDMS, which makes the bonding permanent. Another solution that is used in McKinney’s lab is to clamp the PDMS and the glass slide together but this is generally not suited for long-term applications or when the chip is intended to be commercialized as a biosensor.

A second issue is that the assembly and disassembly of the chip still needs to be done under a biosafety cabinet as it involves manipulation of open liquid. There are usually 2 options to load a chip with cells:

  • either one can bond the glass and the PDMS and then flow cells through inlets that are punched in the PDMS, which is the option we chose in our project.
  • or the cells can be spotted on the glass slide and the PDMS can then be bonded on top of them. This solution requires smaller amounts of cells but is not compatible with plasma-treatment as the PDMS chip has to be aligned very precisely to the cells whereas plasma-bonding is a one-shot process.

Depending on the kind of cells/media/pressures that are used, one therefore has to find a compromise between manipulating open liquids and reducing the risk of delamination.

John McKinney also pointed out 2 other shortcomings that could be problematic in some specific cases. First of all, microfluidics are not very well adapted for long-term experiments (for example studies of slowly growing strains), since PDMS is known to lose its mechanical properties with time, thus increasing the risk of delamination. Second, bleach or ethanol can be inefficient to decontaminate some spores, which could make it difficult to run microfluidic experiments with strains such as anthrax, as they could leave the chip without being killed.

How were these issues tackled?

To tackle the problems mentioned above, one can think of several solutions. The professors presented some dispositions they took in their lab and we also brainstormed on our side on how biosafety could be improved.

John McKinney first explained to us what they are doing in his lab to prevent any cell from leaving the chip, even through the outlet tubing. They actually add a dialysis membrane between the cells and the PDMS chip, so that cells are trapped between the glass slide and the membrane. The latter consists of nanometer-range pores, which prevents cells and even large proteins from leaving the chip. This is interesting when handling sporulating species or toxin-producing strains as it avoids any contamination by spores or harmful toxins.

To tackle this issue of keeping every cells on the chip, we thought of adding 0.2 µm filters at the end of the array of chambers, so that only medium can flow out of the chip. This also seemed to be a proper solution to deal with sporulating species. But it then appears that cells could form clots just before the filters, thus blocking the liquid and finally leading to delamination.

An interesting solution to this problem was engineered in van der Meer’s lab: instead of directly flowing cells through the chip, these are embedded in agarose beads which size is much larger than the size of the filters.

Bacteria in agarose gel beads that are stuck in a chamber

Therefore, cells are trapped in agarose and can neither clot the filters nor leave the chip. Another advantage of this technique is that the chip can be pre-filled with beads and then stored at -20°C before running the experiment, which is of great interest if one thinks of engineering a completely portable device. The only issue when handling cells in biomaterials is to be able to feed the cells. Indeed, bioreporter cells need a continuous energy input to produce the desired reporter proteins. When cells are trapped in a biomaterial, the reaction time can become extremely long as the diffusion of medium through the gel is hampered. Once again, one has to find a compromise between safety and efficiency.

What possible improvements did we imagine?

After doing these interviews, we thought of what treatment can be used on GMOs and pathogens and came up with the following table :

Bacteria Toxins Spores
High Temperature
Lysis Buffer (30mM NaOH, 12%w/v SDS)
Dialysis Membrane
Guanidine hydrochloride (denatures protein)

We then focused our attention on how these processes can be implemented in a microfluidic chip to increase safety:

  • do a continuous lysis buffer, ethanol or bleach flow that will mix with the cells which should leave the outlet.
  • add a suction device on the chip to assure the bonding between the chip and the glass slide, for experiments that need to do spotting.
  • add a microheater to kill any cells that would leave the chip and also some spores.

The CleanColi

the CleanColi : an on-chip waste treatment

With this in mind, we came up with a new chip design. This chip, CleanColi, is actually an extension for any kind of chip and is intended to be put before the outlet. It includes several decontamination steps to provide total on-chip waste treatment, thus avoiding any leakage of potential harmful organisms in the environment.

The process we created is the following :

The first step is mixing the cells with lysis buffer. You would think that the only thing that is needed for it to mix is to add the lysis buffer in the solution. This would be correct if we were not in microfluidics. Indeed at this scale the flow of liquid is laminar and so if two channels become one, the two flows will not mix but flow right next to each other. To assure that the two liquids mix, we have to create turbulence in the flow. For this, we created a serpentine pathway, which will slowly mix the liquids together. One turn is not enough as it will only create a small turbulence so the system requires that there are several turns, hence our serpentine design for this first step.

The first step weakens or kills most cells. To assure their destruction, another serpentine circuit was introduced to add a flow of a bleach that will take care of the more resistant cells.

Step 1 and 2. The cells leave the array of chambers and are flowed in a serpentine circuit with lysis buffer or other chemicals to mix them with the cells.

After the chemical treatment, the cells enter a chamber. This big chamber has the property to have filters throughout its length. The cells will be confronted to smaller and smaller filters to keep the debris and potential surviving cells in the chip. The goal of this is to ensure that only liquid will be flowed out of the chip. The three first filters (5, 2 and 1μm respectively) have a goal to hinder the progression of the cells and to continue exposing them to the chemicals as they advance slowly in the chamber. The last filter is an array of 5μmx5μm blocks separated from each other by a distance of 0.5μm to prevent the cells from getting out of this chamber as E.coli has an average radius of 0.5μm.

Step 3. The waste is retained in a last big chamber where they are confronted to smaller and smaller filters to keep the debris in the chamber so that only liquid will be flowed out of the chip.

If the cells release toxins or a cell survived and passed through the filter, they will enter a final smaller chamber that is located above an ITO (Indium-Tin Oxide) microheater inducing a local temperature of 95°C. This will denature the proteins and kill the remaining cells.

Step 4. The cells enter a big chamber located above a microheater inducing a local temperature of 95°C.