-Engineering Success-

Cycle 1: From protein level to genetic level

Design

Based on the research papers by Vlachostergios et al. (2021), we chose prostate-specific membrane antigen, PSMA, as the biomarker for our project. PSMA is a protein found on the surface of prostate cancer cells discovered in recent years. It is one of the newest and most accurate biomarkers used in the detection and treatment of prostate cancer. (For more, please refer to the history of prostate cancer on the description page). Given the fact that there is a high frequency of PSMA expression in all stages of prostate cancer (Minner et al., 2010; Tsourlakis et al., 2015), it has a higher correlation for prostate cancer than PSA, another biomarker commonly used.

Based on the experience shared by the iGEM team last year, (referring to our integrated human practice) protein purification typically results in a lower antibody yield, which makes it hard to verify the result. Apart from that, the concentration of PSMA in patients’ urine was unknown, making it tough to know the threshold value for detection. Thus, we decided against constructing a protein. Instead, we utilised another pathway with a similar concept to detect prostate cancer this year.

The pathway we used to detect prostate cancer uses plasmids, containing PSMA promoters, which will be transported to the cells in the prostate gland. To produce PSMA, prostate cancer (PCa) cells contain substances that activate the promoter in its copy of the organism gene, which allows the transcription of PSMA to be facilitated. The same substance activates the PSMA promoters in our plasmid and transcribes our required genes to detect and kill cancer cells.

At first, green fluorescent protein (GFP), a protein that exhibits green fluorescence when exposed to light in the blue to ultraviolet range, stood out to our eyes. It was a commonly used gene with easy-to-record results, which can be easily excited using blue light and observed under an inverted microscope. Utilising these characteristics, we can use GFP to evaluate the proficiency of the promoter of our chosen biomarker, PSMA.

For more information regarding the design of the plasmid, please refer to our Description page.

Aim: to build a plasmid (pENTR1A-PSMA-GFP) that allows us to detect PSMA-positive prostate cancer cells by measuring the fluorescence by GFP.

Build

We purchased the pENTR1A backbone and another plasmid containing the PSMA promoter and GFP gene. Both plasmids were digested and gel electrophoresis was performed.

We then purified the agarose gel and the two genes were ligated to form plasmid pENTR1A-PSMA-GFP. (for more information refer to cycle 3)

The ligated plasmids were then transformed into competent DH5a and extracted. They were digested to confirm that the gene length was correct.

After the ligated plasmids were extracted, we transfected different concentrations of them into multiple concentrations of PSMA-positive cancer cell line MLLB-2.

Test

Fluorescence given out by GFP was visualised by using an inverted microscope with a live cell-monitoring system.

First attempt

We collected the cell media of the cancer cells containing transfected plasmids, which were then photographed with an inverted microscope. At first, we recorded a bright image with a lot of light-emitting spots. However, it was suspected to be a false positive result, and the reasons are evaluated below. themselves.

First attempt evaluation

We have discussed the results with Dr Yu, who has extensive experience in GFP analysis. (For more, please refer to our integrated human practice page)

Possible reasons for the false positive result:

1. Most of the spots are a result of autofluorescence from the cell organelles but not GFP.
2. Only the cell media was photographed, which mostly consisted of dead cells that could not transcribe genes.

Possible reasons why fluorescence from GFP is not observed:

1. There was a noticeable delay between collecting the supernatant and photographing, which may lead to the degradation of GFP.
2. The concentrations of plasmids/cells may be too low, which resulted in not enough GFP being expressed and seen.

Cell media from cells with pENTR1A-PSMA-GFP photographed with a blue light filter

Second attempt

After evaluating the previous attempt, we made the following changes:

1. Photograph cells directly instead of cell media
2. Cells were kept in a 37°C incubator with 5% CO2 until minutes before photographing to prevent the degradation of GFP.
3. Cells were photographed in the same position with and without the blue light filter to confirm the fluorescence given out by actual cells.

We have successfully photographed the fluorescence given out by GFP genes transcribed by pENTR1A-PSMA-GFP.

Cancer cells photographed in a live cell monitoring system	The same cells photographed with a blue light filter

Learn

From the process of using a fluorescence microscope to visualise the results of transcribed GFP, we learnt that

1. Cells should be photographed directly instead of using cell media
2. Photographs must be done rapidly after the cells are taken out of the incubator

Moreover, we confirmed the success of converting from a protein level to a genetic level. Unlike protein synthesis, plasmid constructions are easily replicable, and we were able to not only discover issues earlier but also synthesise a high concentration of plasmids. By purchasing:

1. A plasmid that contains our gene but is not mammalian; and
2. Another plasmid that is mammalian but does not contain any promoter genes,

We constructed our plasmid by combining A and B mentioned above. Purchasing prefabricated plasmids (A+B) from firms like Genscript is ideal for long-term storage as it comes in powder form. However, this method is immensely expensive. Our approach of purchasing A and B separately is less costly, which is more practical for our project. This approach also allows for more flexibility in plasmid construction as we could mix and match these genes for use in other formulations, i.e. combining B with another gene that we would use.

Cycle 2: Changing reporter gene

Design

We initially chose GFP as the reporter gene for our project. However, it has quite a few disadvantages. For one, it is cytotoxic (Ansari et al., 2016) and requires external excitation light, which is harmful and not possible when it comes to in vivo studies. For another, autofluorescence from cells and tissues can interfere with the GFP signal, making it difficult to distinguish true GFP signal from background noise.

Knowing the disadvantages of GFP, we realised that another reporter gene must be used for future development. Hence, after countless literature readings, we selected Gaussia luciferase (Gluc) as our reporter gene. Gluc is a 20kDa protein from the marine copepod, Gaussia princeps, which is a bioluminescent enzyme that is highly secreted into urine and can be detected through it (Tannous, 2009). Gluc provides the following advantages:

1. It produces a strong bioluminescent signal, making it extremely sensitive for detecting low levels of activity.
2. Does not require external excitation light, making it suitable for in vivo studies.
3. Gluc is naturally secreted into the media when performing cell culture, which makes it less arduous.

As Gluc is secreted into the urine, in the future, we hope to be able to collect urine samples of patients, which will allow for longitudinal studies and repeated measurements over time on the level of cancer cells in a patient.

Build

Plasmid construction

We purchased the pENTR1A backbone along with another plasmid that contains a PSMA-Gluc gene.

After digesting both plasmids, we performed gel electrophoresis. We then purified the agarose gel and ligated the two genes to create the plasmid pENTR1A-PSMA-Gluc (for more details, see cycle 3). The ligated plasmids were transformed into competent DH5α cells and subsequently extracted. To confirm the correct gene length, we digested the extracted plasmids.

Plasmid construction

Following this, we transfected various concentrations of the plasmids into different amounts of the PSMA-positive cancer cell line MLLB-2, which were subcultured into 96-well plates one night beforehand. The cell lines are incubated for 2 more days in a 37°C environment with 5% CO2.

Plasmid map of pENTR1A-PSMA-GFP	Plasmid map of pENTR1A-PSMA-Gluc

Test

Following that, Gaussia Luciferase Flash Assay was performed to start the oxidation of coelenterazine with Gluc as a substrate. Cell media is collected to measure the secreted Gluc specifically. The luminescence in the cells was quantified with a plate reader.

Attempt 1

We set the plate reader to measure the luminescence given out at a specific wavelength 485nm. (Emission maximum of the assay)

Luminescence given out by different combinations of plasmid and cell concentrations. (P3>P2>P1, C3>C2>C1)

Attempt 1: Evaluation

Our team had the opportunity to interview Dr. Molly Wenqian Xue, a Technical Specialist and Assistant Manager at ThermoFisher Scientific. We learnt that as we are using a flash assay, luminescence at all wavelengths should be measured instead of 1.

Attempt 2

We fixed the problem from the previous attempt and measured luminescence at another setting.

Settings for the measurement from attempt 2. Blanks are set to wells with only cell media without any plasmids.

Attempt 2: Evaluation

100,000 cells/well

At 100k cells per well, change in plasmid concentration did not show significant differences for luminescence measured. Moreover, the luminescence measured is lower than the blank samples, which shows signs of contamination.

Evaluation (100,000 cells/well)

The cell concentration may be too high, leading to nutrient competition in the medium of cells. As a result, the final number of cells may not be accurate and thus the correlation between plasmid concentration and luminescence cannot be correctly shown.

10,000 - 50,000 cells/well

At lower concentrations of cells per well, no matter the plasmid concentration, the luminescence measured is always higher than the blanks.

Evaluation (100,000 cells/well)

This shows a positive result of detection for PCa using our method, for all concentrations of PCa will be able to be detected with the minimum concentration of plasmids used in the test.

Learn

96-wells have a limiting capacity for growing cells, implying that more cells do not necessarily relate to higher levels of luminescence. The cells may compete for nutrients and die as a consequence, leading to inaccurate numbers of cells, thus the correlation between the number of cells and plasmid concentrations could not be visualised. Therefore, we realised that we should use lower concentrations of cancer cells in future experiments.

Moreover, our minimum concentration of plasmids was able to detect the lowest PCa concentration used (10,000 cells/well), demonstrating a high sensitivity for our product, which could allow early stages of prostate cancer to be detected and cured promptly.

Cycle 3: Killing cancer cells

Design

After the success of cycles 1 and 2, we moved on to the core part of our project: killing prostate cancer cells promptly after detection. At first, referencing results from other teams (Tsinghua - IGEM 2023, n.d.), we planned to use PD-L1 nanobody, an antibody to target immune checkpoints, enhancing T-cell response and mediating antitumor activity by cancer cells. However, this method relies on the immune system, which is an additional unknown factor for patients.

Therefore, we finally settled on the apoptosis regulator BAX, also known as bcl-2-like protein 4. The BCL-2 family of proteins, including pro-apoptotic proteins like BAX, play a crucial role in regulating apoptosis (programmed cell death) (Cory et al., 2016). They control mitochondrial dysfunction and cell killing, and we utilised this characteristic of BAX to kill cancer cells in our project.

As concluded from the results of cycles 1 and 2, low to medium concentrations of cells are the easiest to detect. Therefore, we transfected different concentrations of plasmid into 20,000 cancer cells for each well.

Build

The transfected cancer cells were continuously incubated in a 37°C incubator with 5% CO2 throughout the experiment in RPMI 1640 medium. We monitored the cell viability over time, including a control without any cells transfected, using a CyQUANT™ MTT Cell Viability Assay. It employs the widely-recognized MTT reagent to assess mammalian cell viability. In active mammalian cells, the redox potential reduces MTT to a vividly coloured formazan product, which can be measured using a microplate absorbance reader.

Test

Cell viability across three days after pENTR1A-PB-Gluc-Bax transfection into MLLB-2 cells

During the three days of measurement, we successfully recorded a decrease in cancer cell proliferation rate, compared to the control which has no plasmid transfected.

Learn

Expressing Bax is a successful method for inhibiting growth of prostate cancer cells and even reducing their number. Refer to our Results page for more information regarding the result.

Cycle 4: Plasmid construction (Web-Lab) on Practical Basis

Design

We followed a set of standard protocols, found in manufacturers’ product sheets, during our process of plasmid construction.

Build-1

During the first step of plasmid extraction, we incubated the colonies for 8 hours and centrifuged the bacterial cells for only 1 time. However, the pellet was not obvious with only a tiny size. The extracted plasmids had a deficient concentration, and further operations such as digestion failed with poor banding observed.

Poor result in gel electrophoresis

Test-1

After consulting other more experienced teachers in the field, we tried to centrifuge the colonies multiple times, by discarding the supernatant after each centrifuge step and then adding more bacterial culture. This resulted in a noticeable increase in the size of the pellets, and we used a Multiskan SkyHigh Microplate Spectrophotometer to measure the concentration of plasmids afterwards. We successfully achieved an incredibly concentrated volume of plasmids, which made downstream operations much easier.

Increase in size of pellet

Figure 5: Result of measuring the concentration of plasmid

Build-2

For another, when we first performed restriction digestion, no measurement of the concentration of extracted plasmids was taken. We only followed the suggested volumes given by the manufacturer, resulting in a very dim band after gel electrophoresis. After gel purification, only a very small amount of DNA remained (measured with a spectrophotometer). Ligation of the said DNA resulted in no visible colonies after transformation.

Dim banding in gel electrophoresis after restriction digestion

Test-2

With the measurement from the spectrophotometer, we were able to calculate the volume of DNA required for a specific mass, leading to us being able to add a suitable amount of DNA for enzymes to digest. Also, we scaled up the reaction to prevent possible pipetting errors. After making the adjustments, we were able to obtain bright bandings in gel electrophoresis which were suitable for gel purification. However, as the difference of base pairs of the plasmid before and after cutting was a mere 100bp, we failed to separate digested and undigested plasmids and had to cut the entire banding.

Bright banding in gel electrophoresis after restriction digestion

Build-3

After purifying the DNA and then ligating it, we transformed the resulting plasmids into bacteria. Colonies were able to grow, which indicated that either a) our plasmid has been successfully ligated or b) undigested pENTR1A were transformed.

pENTR1A-PSMA-Gluc	pENTR1A-PSMA-Gluc	pENTR1A-PSMA-GFP	pENTR1A-PSMA-GFP	+ve control (plasmid)	-ve control

Bacterial transformation of different plasmid

We performed colony PCR screening to ensure the GOI had been successfully implanted into the plasmid.

Test-3

However, when we performed gel electrophoresis, we found out that the method could not successfully show if we had successfully combined our GOI with the plasmid. We could only see a very dim band that remained stuck in the well after neglecting the bands formed by the primers.

Result of colony PCR, visualised by performing gel electrophoresis

We believe that one of the following reasons occurred: (i) incorrect extension time; (ii) incorrect primer; and (iii) faults in the PCR machine. However, we cannot figure out which of them is the real reason.

Build-4

We have chosen to use another method to verify our results in plasmid construction, which is to extract the plasmids and then digest them to visualise their size.

Test-4

Gel electrophoresis were run to visualise the size of plasmids. Then we use a restriction enzyme to cut the plasmid to check whether the desired plasmid is being cloned successfully. All of them had the correct size, which means that the colonies contained successfully ligated plasmids.

Undigested plasmids (supercoiled) Proved that at least some plasmid being ligated successfully	Digested (Using 1 restriction enzyme to cut the plasmid to check the base pair) pENTR1A-PSMA-GFP and pENTR1A-PSMA-Gluc	Digested pENTR1A-PB-Gluc-Bax

Digestion results of ligated plasmids

Learn

From this cycle, we have learnt that the protocols from the manufacturer may not fit our project. We may need to modify them to obtain success in constructing our plasmid, including changes in:

1. Size of bacterial pellet during plasmid extraction
2. Add DNA for digestion in terms of mass not volume

We also have to change our methods for verifying the ligated genes because of faults. For example, colony PCR was changed to a more traditional way of restriction digestion since we cannot confirm the reason of it failing.

Cycle 5: Cell culture

This year, we have chosen two cell lines, MLLB-2 and PNEC30, to verify the result of our project. Hence, this cycle focuses on how we handle the cells before ascertaining the result of plasmid construction.

Design

Due to the decision to apply plasmid to cells, we initially decided to purchase several human cell lines to investigate the fate of the plasmid and the response of cells. However, due to safety concerns and regulation issues, we have finally chosen two mammalian cell lines, originating from rats, to be our crucial components of result verification.

We have found a PSMA-positive mammalian cell line, MLLB-2, from the website of the American Type Culture Collection (ATCC). However, we could not find a PSMA-negative cell line at the beginning. After searching for more information, we finally figured out that neuroendocrine prostate cancer cells do not have PSMA expression (Bakht et al., 2019). Therefore, we have chosen PNEC30, which is a neuroendocrine prostate cancer cell line, as the other cell line. We have decided to follow ATCC’s protocol to handle our cells.

Build-1

After the arrival of cell lines, we thawed and cultured them in several T-flasks. MLLB-2 was cultured in RPMI 1640 medium while PNEC30 was cultured in neural progenitor maintenance medium (by Lonza)

Test-1

After a day, we observed the two cell lines under an inverted microscope. While MLLB-2 grew well in the T-flask, PNEC30 grew with a much lower cell confluency in the flask. We even saw a blurry object floating on the surface of the PNEC30 medium. Therefore, after discussing with our instructor, Ms Ng, we have decided to aspirate the object away to another flask and grow it in another incubator with the same settings to investigate what the object is.

A day after removing the blurry object from the flask, PNEC30 grew normally while the blurry object grew much larger, and some of the objects stuck to the bottom of the flask, forming a thick layer. We suspected that the presence of blurry objects was due to contamination, and we believed that it was due to fungal infection after seeking professional advice.

MLLB-2 a day after thawing

A blurry object floating on the surface of the PNEC30 medium.

Suspected fungal contamination in PNEC30

The blurry object in an isolated flask after a day

PNEC30 after separation of the blurry object

Learn-1

As we found that the presence of blurry objects is due to fungal infection, we have paid more attention to our aseptic techniques since then. We then used 70% alcohol more frequently to sterilise every component that we put into the biosafety cabinet in detail and have also sterilised our arms more thoroughly before operating in the biosafety cabinet after this accident.

Build-2

We have then handled the cell lines with care. We changed the medium every two days and subcultured the cells to another flask when the cell confluency was above 85%.

PNEC30 with a cell confluence of 85% after recovery from contamination

Test-2

MLLB-2 was ready for subculturing. However, we decided to wait for a day more. On the next day, the confluency of MLLB-2 decreased and they started to detach from the bottom and clump up, we instantly carried out the first subculture of MLLB-2 and also cryopreserved a portion of it. After a few days, we observed the subcultured cells with an inverted microscope and discovered that the cell line grew rapidly in number but did not stick well to the bottom of the flask and had many cell clumps.

MLLB-2 before subculture (detached and clumped up)

MLLB-2 few days after subculturing

Learn-2

With the decreased cell confluency before subculturing, we learnt that we should subculture cells instantly when the confluency reaches 85%. As for the cell clumping and poor attachment issue, We rethought our procedures and concluded that it was because we did not pipette up and down thoroughly enough to respond to the cells after centrifugation. Through this incident, we have learnt that we need to pipette up and down more thoroughly in the next subculture.

Build-3

After seeding MLLB-2 cells into 96-well plates for the experiment, we transfected plasmids pENTR1A-PSMA-Gluc into the cells, experimenting with the effect of different plasmid concentrations on different concentrations of cells.

Test-3

After measuring the luminescence of the reaction between coelenterazine(as substrate) and Gaussia luciferase(as an enzyme), the result shows that the plasmid was transfected in excess, causing the cell concentration to become the limiting factor of the experiment. However, we are investigating the optimal concentration of dna for the most efficient and accurate detection of prostate cancer cells, so concentration of dna should be set as the limiting factor in the experiment.

Learn-3

With the result, we learnt that we should reduce the concentration of Gluc dna in transfection, this causes the cell concentration to become the limiting factor. Therefore, in the next transfection of Gluc, we decided to reduce the amount of dna added while diluting it to the same total volume as before. A correct graph was obtained.

Build-4

At the first reading of the luminescence of Gluc, we used plate reader to measure our data.

Test-4

In the reading process, we used a wrong setting for the plate reader. Therefore, the result obtained was not in the optimal format and was inaccurate

Learn-4

After searching for the manual of the plate reader, we found the correct mode for the luminescence detection of Gluc. We then repeated the transfection of Gluc into newly seeded MLLB-2 cells, and read the luminescence of Gluc again, correct data were obtained.