The Calling

Looking for an attractant

When we were considering setting up a transport for bacteria, we thought it was necessary to have a 'destination', a place to go. That destination would be a 'hot spot' on a heterogeneous substrate where Caenorhabditis elegans should carry the bacteria to.
Leveraging the powerful smell of the nematode, it was decided to try a number of attractants from various lists from the web www.wormbook.org that could work as 'hot spots' in our experiment. Thus, using C. elegans chemotaxis, we would be able to direct transport.

The attractants experiment

The test was carried out by creating our own plates in which half plate would be unmodified NGM and the other half would include soluble compounds before solidifying, or after solidification (in the case of volatiles). The list of modifications can be found in Fig. 1.

To place C. elegans on the plate, different cuts were made on a fresh plate of NGM and the resulting small pieces were placed in the exact center of the 50% - 50% plates to determine which side of the plate the nematode preferred, one per Petri plate.

The results after counting 2 replicates per attractant can be seen in Fig. 2.

From Table 2 many of the attractants that were thought viable were discarded.

Volatile attractants were not a good choice because they evaporate quickly (which would also be a limitation for the field experiments).

Another impediment arose. Once we had already performed the experiments, we decided that the promoter that would control the production of interference RNA in Escherichia coli would be regulated by nitrogen, so amino acids had to be discarded as attractants; because it would modify the controlled expression of E.coli.

That left MgSO₄ and hypoosmotic media as potential attractants.

Attractant final choice

Because the hyposmotic medium could interfere with the proper growth of bacteria (our nematode's food), the final decision was to choose MgSO₄ as our attractant.

The question at that time was: How can we multiply the amount of MgSO₄ to increase the efficiency?

MgSO₄ efficiency

Once we had selected the most feasible attractant for the ‘transport’ during our experiments, we needed to know which could be the largest concentration of MgSO₄, in order to optimize the nematode's attraction.
To choose the concentration, attraction was tested in a battery of increasing concentrations (regarding the initial medium of 1 ml / L). Bearing in mind the results of the x2 factor, we decided to test other factor concentrations: x3, x4, x5, x8 and x10; covering a wide range that did not exceed the concentration that would affect the life of C. elegans or bacteria.
Moreover, experiments with E. coli and Pseudomonas putida were carried out. These trials were performed in order to obtain the final media: half plate with non-altered NGM and half with the medium for PHA production by Pseudomonas and interference RNA production by E. coli.
Not knowing which fatty acid could activate the transcription process better, we tested two possibilities: oleic acid (named in the tables as PHAol) and octanoic acid (PHAoc). The concentration for each fatty acid was tested in E. coli and we obtained best results with:

• 1.28 µl/ml of Octanoic acid.
• 2.58 µl/ml of Oleic acid.

You can see our results in figures 3, 4, 5 and 6. The first two counts were made after 3 hours and the next ones after 6 hours. We supposed that after that time, the worms were already capable of selecting their "favourite" side of the plate, get to the desired side and keep moving in the same part of the petri dish. We prepared control plates with the same composition but without MgSO₄ in PHA media.

Final concentration choice

Once the tests were performed and after observing the results obtained with the different selected fatty acids, we then observed a scattered result(there was probably a repellent effect at high concentrations in the medium with oleic but reversed in the octanoic) and in the end, we decided to use the following MgSO₄ concentrations for each medium:

• If oleic acid is used: Best results with 4ml/L MgSO₄.
• If octanoic acid is used : Best results with 10 ml/L of MgSO₄.

Octanoic was discarded as a transcriptional activator of the iRNA of clumping after other results for E. coli were obtained, so finally we selected the PHA medium with oleic acid only.
The biggest problem while trying to have an effective attractant was knowing how effective could an attractant, in the presence of E. coli, be. It was therefore necessary to test the tradeoff between the value of 4ml / L MgSO₄ and pair it with different concentrations of bacteria, high enough to feed the nematode but low enough to allow the attractant effect of MgSO₄.
To find this point of commitment we prepared experiments in which the concentration of 4ml/L of MgSO₄ was faced against different ODs from serial dilutions of a preculture of E. coli DH5a.
Table 7 shows the results obtained. The count took place 3h after the passing of fresh nematodes.

Best E. coli OD choice

An OD of 1 (minimum concentration of cells / volume) gives an attractive effect even better than expected (subsequent experiments try to check if there is synergy between E. coli and attractive factors MgSO4).
To improve the approximation we decided to repeat the experiment with MgSO₄ concentration and the chosen bacterial OD showing that the system worked correctly. The results can be seen in figure 8.

The Clumping

The main microorganism of this part is our engineered E.coli with a biobrick that is capable of changing the worm’s behaviour.

Our C. elegans strain (N2 strain) has two feeding behaviours: Social and solitary feeding. Social feeding is known as clumping. You can watch it in this video: http://www.youtube.com/watch?v=jCNsmcVVUVY

Clumping is known to be induced under some conditions like temperature or starvation, something that we had to consider during the entire project. However, it is also known how clumping is controlled from a genetic perspective and the main genes involved have already been described. Therefore, we thought about using this genetic approximation to control clumping under specific and controlled conditions. (view “Natural variation in a neuropeptide Y receptor homolog modifies social behavior and food response in C. elegans.” Bono M, Bargmann CI).

How do we induce clumping?

Our initial thought was to interfere the principal gene involved in this route: NPR-1. It is known that mutations in NPR1 convert a solitary strain into a social strain. Another option (the one we finally chose) was FLP-21, a gene that positively regulates NPR-1. We selected this one instead because it is not involved in so many vital processes. FLP-21 encodes a single FMRF amide-related neuropeptide that serves as a ligand forNpr-1, a G protein-coupled receptor that regulates social versus solitary feeding behavior in several Caenorhabditis species (see fig. 1).

Therefore, this would be the final process: engineered E.coli that expresses FLP-21, whose RNA would interfere with the one codified by the worm because of the ingestion of bacteria. The formation of a complex of dsRNA would induce the elimination of Flp-21 transcripts by the RNA silencing pathway and the final result would be the induction of clumping (see figure 2). We also had to take into account that the induction of clumping by this mechanism is almost immediate, in comparison with natural clumping which takes longer depending on the external factors.

Why do we induce clumping?

The main objective of the synthetic symbiosis that we have designed is to detect hotspots of interest in irregular substrates and the induction of clumping was a good tool to do so. With this natural mechanism worms are kept in the desired places, then bacteria transported by C. elegans (Pseudomonas putida) are concentrated where their action is needed, in order to generate value-added products such as bioplastic PHA and the increase of worms in a specific location improves the image-based detection mechanism (go to the Devices section)

Controlling the mechanism

The social feeding behaviour needs to be controlled; it can be induced by the mentioned conditions of interest as a tool for detecting hotspots in irregular substrates. We decided that fatty acids were going to be the inductors of the promoter (fadBp), which regulated the expression in the FLP-21 antisense sequence, in order to induce RNA interference (Fig.3). We decided to use this regulation, because fatty acids are also used for the PHA production, our biotechnological process of interest.

Choosing the right fatty acid

E. coli’s biobrick promoter is induced by fatty acids, but these also induce the promoter of Pseudomonas putida in order to produce PHA. Some investigations show that the induction of fadBp promoter by oleic acid is the most efficient (view “Regulation of fatty acid degradation in Escherichia coli: analysis by operon fusion” Clark D. et al) but this is a long chained compound and that could affect the production of bioplastic (PHA). We decided to test the growth of E. coli and the production of bioplastic by P. putida using PHA media with oleic acid and octanoic acid in order to get the best results in both activities.

Both species of microorganisms where grown in PHA production media with different concentrations of oleic and octanoic acids, maintaining a global concentration of fatty acids of 8mM, because it was the one where we found the highest growth of both E. coli and P. putida. Table 2 (below) summarizes the assays done.

The results of this assay showed that E. coli growth was better on PHA production media + 8mM oleic acid and the P. putida production of bioplastic was also acceptable in these conditions. However, the highest production of bioplastic was on PHA production media + 8mM octanoic acid. So then, we made a Colonization assay and finally we determined it was better to keep using the PHA production media + 8mM oleic acid. Furthermore, the ODs for plating both organisms were established as a consequence of this experiment. The “Choosing the right fatty acid” and the “Colonization” assays are explained at Building the bioplastic section.

Observing clumping induced by fatty acid rich media

We could observe how clumping was induced by the presence of 8mM of oleic acid (2.58 µl/ml) which regulated the FadBp promoter of E. coli and as a consequence the expression of the complementary sequence to the worms’ FLP-21 transcript, producing interference by RNA. In the plates, there was a high concentration of E. coli from the centrifugation of a preculture of 4ml XL1-Blue, 2 min at maximum rpm (2 times). The following images show the change in the behaviour of C. elegans, first eating alone and then in groups, just as we planned (Fig. 5-7).

We confirmed that the clumping was induced because of the expression of the iRNA of engineered E. coli. Two populations of worms were fed with engineered E. coli. The control population ate E. coli that was grown in a culture medium without fatty acids. The experimental population was fed with E. coli that was grown in PHA media (culture media with fatty acids). The ingestion of E. coli expressing FLP-21 iRNA did result in a clumping behaviour in comparison with the control, where one could not observe this immediate social feeding response, due to the expression of Flp-21 (Fig. 8)

The Building

We decided to include a regulated production of a value-added product in the hotspots of interest (where C. elegans would be attracted to) in the mechanism design. Therefore, we took advantage of the natural capacity of Pseudomonas putida of producing bioplastic PHA (polyhydroxyalkanoates) by bacterial fermentation of sugar or lipids (fatty acids in this case). We found this material interesting due to its easy detection (complexes of PHA with Nile red are fluorescent) and its bright future in the field of biomaterials.

The mechanism

A cluster of genes is responsible for the production of PHA, the phaC operon of P. putida which consists of 4 ORFs that are transcribed in the same direction: phaC1 and phaC2 genes, codify for two PHA synthases; phaZ gene, codifies for a depolymerase; and phaD gene, that codifies for a protein of the TetR family. In the opposite direction, two genes that codify for fasinas (phaF and phaI) and structural proteins are found (Fig.1).

The role of the operon in our project

Our 2013 project is a proof of concept of the benefits that an artificial synthetic symbiosis between bacteria and nematodes can offer. Pseudomonas play basically two roles: firstly, they have been engineered to “ride” on C. elegans by the formation of a biofilm, but not satisfied with that, we stimulated the production of PHA (a value-added product) under the promoter glnA. As mentioned before, PHA has several uses and applications such as:

• For short disposable packaging items (personal hygiene products, surgical clothes).
• In upholstery.
• In the photographic and printing industry
• For textile industry since PHA can be processed into fibers.
• For biofuel production, from PHA obtained from sewage sludge. This would combine two major advantages: the wastewater treatment and the generation of energy.

Choosing the right media

On the plates, where we were going to have P. putida , E. coli and their promoters’, response towards fatty acids was different, so we had to create a media where E. coli could induce clumping and P. putida could produce bioplastic. In E. coli, the briobrick promoter for the expression of FLP-21 iRNA engineered in order to induce the social feeding behavior of C. elegans (Clumping) is induced by fatty acids, but not all the fatty acids are able to induce the same level of expression (Fig. 2); oleic acid produces the highest activation of the promoter.

Whereas the standard medium for the production of PHA by Pseudomonas putida (composition reflected in Fig.3) employs octanoic acid (a short-chain fatty acid) as inducer. This fact made necessary to check if the modification of the fatty acid used, octanoic by oleic, would modify the bioplastic (PHA) production.

Otherwise, the medium for PHA production is tuned for the development of Pseudomonas putida, but not for the E.coli, so we had to test if these bacteria were able to grow in it. Once we had checked this, we realized that the development of these bacteria did not occur in the PHA production media. However, with the addition of acetate (carbon source), the removal of the iron salt (added to the media to avoid the synthesis of a siderophore by Pseudomonas sp.) and (due to serendipity) with double the concentration of trace elements, we obtained the media for the enteric bacteria growth.

Choosing the right fatty acid

Firstly, we tried to grow E. coli in the new PHA production media with different concentrations of octanoic acid. The same assays were carried out in parallel with P. putida in order to check which were the right conditions where both microorganisms could grow and where P. putida could produce enough bioplastic.

After overnight incubation, the cultures were washed using sterile PBS and ODs₆₀₀ were measured (Fig.4.). We also checked the PHA production by centrifuging the precultures, resuspending the pellet on PBS and adding Nile Red dissolved with DMSO. We could then observe the emission of fluorescence due to the interaction between the Nile Red and the PHA through an UV transilluminator (Fig.5). According to results of the assays we chose to work with a 8mM concentration of fatty acids (Fig.4).

At this point we found that E. coli and P. putida could grow at 8mM of octanoic acid and that bioplastic (PHA) was also correctly produced. However, because E. coli grows better in oleic acid and we wanted to check if P. putida could produce PHA if its phaC operon was induced by it, we grew both bacteria in different concentrations of octanoic and oleic acid but maintaining a constant concentration of fatty acids of 8mM. The assays we carried out are explained in the table below (Fig.6).

We saw that E. coli had grown better with the increase of oleic acid, getting at its maximum on experiment 9, even so, the quantity of pellet corresponding to 1 mL of preculture was scarce (and not visible at experiments 1 to 5). E. coli needed more time to grow in the PHA production media.

P. Putida had grown too, however we wanted to measure the production of bioplastic, we measured the ODs₆₀₀ from the precultures of P. putida by centrifugating 2 mL of them and then washing them on 1 mL of PBS and resuspending the pellet in 500 μl of PBS. Then, dilutions on PBS were adjusted to the lowest OD_600,corresponding to the experiment number 9 (8mM oleic acid). The growth of P. putida was lower in PHA production media with a concentration of 8mM of oleic acid (Fig. 7-8).

The results showed that the growth of E. coli on 8mM of oleic acid was better than in any other condition, then experiment 9 conditions would be the best ones for the induction of clumping. However, even P. putida had produced bioplastic at all oleic acid concentrations, the highest production levels corresponded to experiment 1 conditions (that is, in the absence of oleic acid), just on 8mM of octanoic acid. Because of that, following assays were done under the conditions of experiments 1 and 9 (Fig.9).

After the optimization of the media, kinetics of the production of PHA and biomass generation were obtained in relation to time (Fig.10).

Pre-field test

Final Test: PHA Production in an Heterogeneous Substrate (soil)

Finally, we designed an assay in which all the activities presented till this moment were carried out by the synthetic symbiosis that we have been working with. In this experiment we achieve the goal of our project: the detection of hotspots of interest in irregular substrates and the production of bioplastic (PHA) in those places. Our project works in a pre-field test!

A soil ground was prepared with an artificially distributed substrate-rich medium in discrete points (Fig.1). Once the medium was ready, we inoculated the soil with a C. elegans-Pseudomonas suspension and with transformed E. coli expressing the iRNA of interest. As a control, we took a picture of the soil ground the first day the microorganisms were added by exposing it to UV light (Fig.2). After a day we added Nile red to reveal the PHA production and to see if the experiment had finally worked (Fig. 3). It is possible to see bioplastic nodules which are red when combined with Nile red.

In addition, we took an electron microscopy image of one of the discrete points of medium (Fig. 4) to study which organisms were actually near the hotspots. More electron microscopy images were taken to prove that C. elegans dragged Pseudomonas to the points of interest (Fig. 4).

We were finally able to conclude that our synthetic consortium worked: Worms managed to reach the hotspots of interest, the worms dragged Pseudomonas with them and Pseudomonas finally made bioplastic, detected by the added Nile Red.