Kyle Klamka


This class led by Dr. Crowthers really embraces WPI's ideals of a hands-on learning environment. In STEM, we practice many skills which will be useful later in life: everything from interviewing, to public speaking, to research, and finally to writing. We participate in a multitude of projects which give you the ability to choose a topic and practice writing actual laboratory documents and papers on said research. Some of them are group work, while others are individual. The amount of variety which we encounter in STEM is amazing! STEM 1 mainly focuses on an individual project where we learn about common scientific practices and conventions.

Feeding Practices Change the Gait of C. elegans


It is well known that various senses can interact with each other. However, there are still unknowns in this area, especially in terms of other organisms. Caenorhabditis elegans is a model biological organism used in a myriad of scientific fields. Typically, research conducted with the feeding of C. elegans look mainly at food nutrition and only compare different foods or differing chemoattractants. However, the response to food typically plays off other senses in the natural world, as nematodes rely on a combination of chemosensation, thermosensation, and mechanosensation to guide them. First, a chemotaxis assay was conducted with Sphingobacterium multivorum and Escherichia coli, resulting in a greater choice index associated with the S. multivorum. Modifications were then made, comparing the S. multivorum and a weaker chemoattractant with the E. coli and a stronger chemoattractant. As a result, the worms tended to reside with the E. coli. After giving the less attractive food an enticing odour, the worms tended to aggregate around said bacteria, implying that other smells can mask C.elegans’ detection of food, and therefore influence their attraction to the food. This might help to explain why simply placing lesser liked foods–such as vegetables–on the plates of children tend to acclimate them to the food and increase their affinity with the food when they are older, as well as helping to explain unknowns about interactions between senses in general. This also develops future research with C. elegans, allowing for unbiased environments and further analysis on food selection.
Keywords: Behaviour, Ascarosides, Bacteria, Locomotion, Olfactory

Graphical Abstract

A Picture Summarising My Project!

Research Proposal

Research Question

How do the properties of the bacteria used to feed C. elegans impact their attraction to the bacteria?


I hypothesise that the olfactory senses will dictate the attraction more than the species of bacteria used, as C. elegans does not have visual senses and therefore mainly detects through chemoattraction.

Background Infographic

A Quick Summary of Known Information!


In recent decades, Caenorhabditis elegans has provided humans with a large variety of knowledge about the brain. What makes C. elegans so special is that the nematode is composed of 959 somatic cells, 302 of which make up its nervous system. This simplicity of sorts has allowed it to be the first animal to have a fully mapped out system (Gjorgjieva et al., 2014). Using this model organism, scientists have been able to test the neural processes of everything from responsive behaviours to communication and basic locomotive movement, with the goal of applying this knowledge to humans in the future.
These nematodes are commonly used for behaviour analysis due to their short lifespan and limited required resources. In order to measure these changes in behaviour, a large variety of mechanisms and measurement systems have been utilised. Many of these pertain to sensory and locomotive behaviours, as they are the easiest to identify differences (J. Srinivasan, professional communication, October 3rd, 2022). A significant area of research involving both of these systems is the taxis ability of C. elegans: A way of dictating the nematode’s gait based on external gradients. Chemotaxis, aerotaxis, and thermotaxis serve as ways with similar mechanisms in which the nematode’s seemingly stochastic gait may be influenced and recorded for abnormalities (Thiele et al., 2009). As a result, these methods of forced movement are typically used in laboratories to measure other factors.
For instance, locomotive behaviours were found to change under the presence of food depending on the nutritional status of the worm, denoted as Basal Slowing Response (BSR) and Enhanced Slowing Response (ESR) (Rivard et al., 2010). This study went beyond the normal laboratory usage of Escherichia coli strain OP50 in some cases. However, the study conducted focuses on the presence of food as well as the feeding history of the worms. This study suggests that OP50 may not be the best food source for the worms, especially as OP50 is mainly used due to its accessibility in laboratory environments and low lawn height (Samuel et al., 2016). Other studies’ results further this, presenting that the species of bacteria used as food can impact variables such as lifespan and development periods (Stuhr & Curran, 2020), as well as general attraction to the food (Shtonda & Avery, 2006).
As a result, the foods used are slowly shifting towards other bacterial groups such as CeMbio (Dirksen et al., 2020), meaning that the research into other possible food species is increasing in importance. Most past research has concerned itself with the nutrition of foods as well as basic attraction to them based on such. While nutrition alone is significant in bacterial attraction, as C. elegans tend to prefer healthier foods (Shtonda & Avery, 2006), there are also many other factors which can play a part in attraction. For instance, previous research has implicated that the wild strain, N2, harbours an intrinsic attraction to foods from its natural habitat (Abada et al., 2009; Samuel et al., 2016).

Procedure Infographic

The Procedures I used and may use in the Future!


The bacterial strains used in this experiment are the typical food source Escherichia coli strain OP50 as well as Sphingobacterium multivorum strain BIGb0170. OP50 was used for the worms’ growth, however the OP50 used in assays was started at the same time as the BIGb0170. They were both made of a single colony placed in LB broth. The bacteria colonies were incubated at 37°C for 24 hours before being kept in 4°C until usage.

Binary Bacterial Choice Assay
In the choice assay, the bacteria and worms were grown as previously mentioned. Assays were done on 100mm petri dishes, filled with a slight diversion from NGM medium due to its NaCl content. Two circles were placed at opposite edges of the plate about 2 cm in diameter, each filled with 20μL of a bacterium. The plate was then dried for 30 minutes under a hood. 1µL of 1M Sodium Azide was then pipetted into each circle to paralyze the animals. The worms were washed 3 times in Chemotaxis Buffer prior to being placed on the plate. The assay was conducted for 1 hour at 20°C in a dehumidified room. Afterwards, the number of L4 and older worms were counted to calculate a Choice Index (CI) as in (1):
CI = (Worms on BIGb0170 - Worms on OP50)/(Total worms which moved)

Binary Bacteria and Chemotaxis Assay
In the mixture choice assay, the procedure is the same as the previous with only minor exceptions: 1M Glycerol was pipetted onto the BIGb0170, and 1% Acetone was pipetted onto the OP50 prior to placing the worms onto the plate.

Statistical Testing and Convolutions
The prior assays were done 9 times each unless mentioned otherwise. After each assay, an unpaired t-test was calculated using the resulting CIs. These were done as there were two “unrelated” groups made up of averages and distinct values instead of probabilities. They would also serve to fortify the differences or unearth the true similarities observed in the data. Normal model equations were used to give shape to the data, as in (2):
f(x) = 1/(σ√(2π)) e^(-1/2 ((x-μ)/σ)^2)
Wolfram Alpha was then utilised to assist in preforming continuous convolutions to the control data as well as the individual chemical data obtained from Liao et al. in 2010. The resulting equation was then compressed to an area of 1 and shifted by twice the domain to account for the natural convolution manipulation.

Figure 1

Figure 1

A comparison of the average percentage of worms which grouped towards each source of bacteria. Error bars indicate 2 standard deviations from the mean; p < 0.001.

Figure 2

Figure 2

A comparison of the choice indices between the bacterial assay results and the bacteria with attractant results; p < 0.001.

Figure 3

Figure 3

A graphical representation of the equations formed from the control data which I gathered from Liao et al. in 2010 and myself. This was compared to the equation from the experimental group data I gathered.

Figure 4

Figure 4

An example of the plates used for the Binary Choice Assays.

Results & Analysis

Natural Over Healthy Attraction
In order to establish control values for both the general experiment as well as the mathematical modelling, a preliminary set of tests was conducted with only the bacteria, as the comparison of common attractants and repellents on the worms is more widespread than that of two specific bacterial species. The two species selected, OP50 and BIGb0170, both had potential to be more attractive, as OP50 caused faster development and longer lifespan, while BIGb0170 was from the worm’s natural microbiome (Dirksen et al., 2020). However, the BIGb0170 ended up in the lead with an average CI of +0.14, and a median of +0.13. This number is positive, indicating greater attraction than OP50 (Figure 1). A t-test was run—comparing this to a distribution centred at 0—as the CI was quite low, yet the p-value was about 0.001, indicating a strong probability that this distribution is not simply a fluke. While C. elegans may not have a strong preference for the natural bacteria, it is still a preference nevertheless.

Chemical Beats Bacterial Attraction
The next step was to test how the addition of a chemoattractant and chemorepellent would change the results. Since the chemical attraction gradient was placed opposite to the bacterial attraction gradient, this experiment served as a direct comparison of the two. By measuring the CIs, a quantifiable change in attraction could be found. Unlike the control tests, the worms showed a preference to the OP50 with an average CI of -0.29, and a median of -0.36. These results are not only inverted but are also more than double the magnitude of the previous set of tests (Figure 2). Again, a t-test was run comparing this to a base distribution, and this resulted in a p-value less than 0.001, indicating a very strong probability that this result was not a fluke. A t-test with the data from the last experiment gave a p-value of less than 0.001, suggesting that the difference between these two distributions is quite high.

Complex Attraction Assays can be Predicted
One of the main goals of the tests was to determine if there was a connection between the resulting CIs. All means and standard errors of the individual parts—the control, the acetone, and the glycerol—were used to form bell curve equations, and then were merged as mentioned in Methods. After adjusting the resulting equation and comparing it to the equations derived from the experimental data, an R-value of 0.85 was gathered (Figure 3). This suggests that the created equation via convolution is a strong fit for the data gathered from the experimental data. If this distribution is then reverse engineered, a mean and standard deviation can be gathered and utilised for future comparison.

Discussion & Conclusion

Not Working? Click Here to View the File


Not Working? Click Here to View the File

February Poster

Not Working? Click Here to View the File