Mild-mannered bacteria by day, glowing jellyfish by night

In the “Bioreactor Cell Culture and Protein Recovery” course at Laney College, my students recently transformed bacteria with two different sources of DNA to create green and blue glowing bacteria. Here are the before and after shots:

IPTG induced bacteria under ambient light (left) and UV light (right).

Why did we do this experiment? For one, it looks really cool when you put a gene from a jellyfish into a bacteria and then the protein makes it glow under UV light. But the real reason was to compare protein expression systems in E. coli, a commonly used “chassis” in biotechnology, biomanufacturing, and the new field of synthetic biology.

In past experiments, we have used a pGLO plasmid in which the gene for the glowing protein, green fluorescent protein (GFP), is induced (or “turned on”) by the presence of a sugar called arabinose. The plasmid also contains resistance to ampicillin, an antibiotic that inhibits the cell wall formation in gram-negative bacteria like E. coli. Here is a map of the arabinose induced pGLO plasmid:

Image from Bio-Rad.com

How the arabinose induced plasmid works:

Arabinose is a simple plant sugar and is used as food by bacteria like E. coli. The bacterial genes for arabinose digestion are not expressed unless arabinose is present. No need to waste all that energy on nothing, right? When arabinose is present, however, these genes get turned on and arabinose is broken down. When arabinose runs out, the genes turn off. Three structural genes (araB, araA, araD or BAD) encode for enzymes used in the digestion of arabinose in bacterial cells. Another gene called araC produces a protein that regulates the expression of BAD. The araC protein also acts to negatively control its own synthesis. In other words, the araC protein binds to its own operator and prevents more araC transcription. The presence of both the araC protein and arabinose cause the expression of BAD. Arabinose binds to the araC protein and causes a change in its shape, which then allows both to bind to the regulatory region known as the activator site. The binding of the araC protein-arabinose complex to the DNA activator site allows the binding of RNA polymerase to the promoter region upstream of BAD. These three genes can now be transcribed and the resulting RNA can be translated into protein. Thus, araC negatively controls its own expression and positively controls BAD. If you replace BAD with the GFP gene, you can then positively regulate the production of GFP using arabinose. Here is a summary figure of the arabinose operon:

Comparison of original arabinose operon with the GFP modified operon.

This expression system has given us excellent results. We follow up the GFP production with a purification step, which looks like this:

Purification of arabinose induced GFP.

The protein that was produced in the first picture above (green and blue glowing bacteria) used a different expression system involving the lac operon. The lac operon is induced by IPTG, an analog or similarly shaped molecule to lactose. Here is a diagram showing how the IPTG induction works:

Image from edvotek.com

How the IPTG induced plasmid works:

Lactose is another sugar that can be used as an energy source by bacteria like E. coli. Bacteria have three genes involved in metabolizing lactose when it is the only carbon source present (bacteria will use glucose first if both lactose and glucose are present): lacZ, lacY, and lacA. These three genes are considered inducible because they will “turn on” when lactose is present. Another gene, lacI (that is the letter “i” for “inducible”), produces a repressor protein. When lactose is absent, the repressor binds to the operator which overlaps the promoter region and prevents transcription of the XYA genes. When lactose is present it binds to the repressor protein and prevents it from binding to the DNA operator. Then, RNA polymerase can transcribe the XYA genes and lactose will be metabolized. This is another example of negative control. Here is a diagram of how the lac operon works:

downstream gene.

Comparing the two fluorescent protein expression systems:

As mentioned above, the lac operon can be induced by IPTG, which does not get broken down by the proteins encoded by the normal lac operon. Arabinose gets eaten as food by bacteria so the induction concentrations are typically much higher than with IPTG. Arabinose induction is done at a 2-6 g/L final concentration, whereas IPTG induction is done at 0.12-0.24 g/L final concentration.

There are advantages to using arabinose as well. Arabinose induction is titratable, meaning that the higher the concentration of arabinose, the stronger the expression. Therefore, protein expression can be adjusted which may be important if the protein has toxicity or solubility problems. Another advantage is that arabinose is also less expensive than IPTG.

One thing I noticed with the IPTG induced green and blue fluorescent protein fermentations was that there appeared to be a little glowing before IPTG induction. It could be an indication of background fluorescent protein production that can happen with this system. Another E. coli strain called BL21(DE3)pLysS has a T7 lysozyme engineered into it. The T7 lysozyme decreases the background expression level of genes behind the T7 promoter but does not block target protein expression following induction by IPTG. Perhaps this would be something we would look at using if background expression was becoming a problem.

Both systems of fluorescent protein expression have their advantages and disadvantages. The real test will come later when we scale up our production volumes from 100 milliliters (mL)(shown above) to 1 or 3 liters (L). Also, we will want to test both systems in downstream processing performance. Which will give the higher protein titers? Will insoluble protein inclusion bodies be a problem with the higher expression levels? We will find out later this semester!

 

 

March 9th, 2014 | Category: Announcements

Have you heard about the hot biotechnology job market in the Bay Area? Interested in getting the hands-on training biotech employers want? The Biomanufacturing Program at Laney College in Oakland offers hands-on training in the skills needed to land entry-level biomanufacturing andbiotech jobs. We currently have graduates working at Bay Area companies like Genentech, Life Technologies, BioMarin, Medivation, Bayer, ForteBio, and biofuel startups.

Enroll in a one- or two-semester biomanufacturing certificate program. We also offer a two-year A.S. degree.

Certificates:

Certificate of Proficiency in Biomanufacturing Skills (9 semester units)
Certificate of Achievement in Biomanufacturing (17 units)

Courses offered for the spring 2014 semester:

BIOL 75 — Fundamentals of Biotechnology (Lab)
BIOL 76 — Principles of Biomanufacturing (Lecture)
BIOL 77 — Business and Regulatory Practices in Biomanufacturing (Online)
BIOL 79 — Bioreactor Cell Culture and Protein Recovery (Lecture/Lab, Biol 75 or Biol 3 prerequisite)

For more information on the Biomanufacturing Program at Laney College: http://www.laney.edu/biomanufacturing/

http://web.peralta.edu/apply-enroll/

January 10th, 2014 | Category: Announcements

Event agenda

There is still time left to register for the Biomedical Manufacturing and Technology Networking Summit this Friday, October 4th in the Laney College theater (maps and directions here). Please invite your colleagues, students, and especially industry personnel who might be interested. Cost is free and it’s open to all those interested in the medical device and biopharmaceutical fields. Lunch will be provided.

Go to http://mfgdaynorca.org/ to register. Scroll down to the bottom of the page and you will see the Laney event listed. Click on the “Register for this event online” link.

October 2nd, 2013 | Category: Announcements

Checking trypsinization of C2C12 cells in second passage.

In early July, in the midst of teaching my summer biomanufacturing course, the North Valley Biotechnology Center (NVBC) announced a two-week intensive cell culture and stem cell techniques course with Dr. Jason Bush at Fresno State University. I leapt at the opportunity and was selected to join five other students and faculty from around California to participate in the course.

My cell culture skills needed some brushing up, and I have never had the opportunity to work with stem cells. I previously worked in a neuro-oncology research lab doing cell culture and fluorescent in-situ hybridization (FISH) of cancer cell chromosomes. That was more than a few years ago now, so I would have a lot of re-learning to do. Aside from intellectual curiosity, I wanted to take this class because we are beginning to offer our Associates of Science (A.S.) degree classes in the biomanufacturing program at Laney College. This Spring we will offer a bioreactor cell culture and protein recovery class, and I wanted to be up-to-date with the latest cell culture technology.

Down to Basics

We spent most of our time in the TC room.

The class was designed to be mostly laboratory-based so we could get as much hands-on instruction as possible. We started with learning basic techniques of cell culture, such as thawing cells, freezing and cryopreservation of cells, changing media, passaging cells, cell counting, media formulations, and sterile technique. Our tissue culture (TC) room had two hoods and we also had access to advanced equipment at Fresno State’s RIMI facility.

We worked mostly with C2C12 and HEK293 cell lines. C2C12 is a mouse myoblast cell line and HEK293 is a human embryonic kidney cell line. We also cultured MDA-MB-231 cells, mouse embryonic fibroblast (MEF) cells, and J1 mouse embryonic stem cells (mESCs). As our instructor Dr. Bush noted, “If you can culture HEK293’s, you can culture just about anything.” The amazing thing was that even though almost all of us had little experience in culturing eukaryotic cells we did not have a single instance of contamination over the two weeks!

Passaging, cell count, and cryopreservation of on day 3.

One thing that was unique about this class as compared to others I have taken is that we had the opportunity to practice each technique we learned multiple times. We had cultures running continuously throughout the two weeks and, as we quickly learned, you cannot leave them alone for more than a day before they demand your attention. Cultures need continuous care and monitoring to prevent overpopulation, starvation, and contamination. Certain cell lines will even start to form structures if you let them get too crowded. C2C12 cells will produce muscle proteins and form myotubes if left too long at high density. HEK293’s will form kidney tubules when packed too closely together.

Cell Culture Applications

Setup of MTS assay for effect of two pesticides, toxaphene and methoxychlor, on breast cancer cells, MDA-MB-231.

Once we had gotten the basics down, we were able to move on to more advanced topics. Since contamination is a constant threat in cell culture, we learned how to do a mycoplasma test. There are various methods for testing for mycoplasma, such as complement fixation or DAPI staining. We went with a PCR mycoplasma detection kit. Testing for mycoplasma is routine in academic research and pharmaceutical production, so it was good to be exposed to this test.

Now that we could culture cells successfully and keep them free of contamination, we started doing things with our cells. The first application we learned was the cell-based assay. Cell-based assays are the true power of working in vitro. For our cell-based assay, we tested the effects of two pesticides, toxaphene and methoxychlor, on proliferation of a breast cancer cell line, MDA-MB-231. We used an MTS assay in which tetrazolium salt gets reduced into a purple product, formazan. The reduction is accomplished by the cell’s mitochondria, which makes the assay a good indicator of cellular proliferation.

MTS assay results for the effects of toxaphene and methoxychlor on breast cancer cells (MDA-MB-231).

We ran the MTS assay on a 96-well dish. The first row is our control showing normal cellular proliferation. The second row is the highest concentration of toxaphene, and the third and fourth rows are successively lower concentrations. Clearly, these pesticides are toxic to cells but these are all at very high concentrations — high enough to cause organ failure and death if someone drank it. The point was to show the principle of a MTS assay. One could easily swap pesticides with a potential chemotherapy agent and test its effectiveness against this aggressive breast cancer cell line.

Light micrograph of HEK293 cells from 3D culture.

The next exciting application we learned was 3D cell culture. This new approach is gaining popularity in cell culture research because growing cells in three-dimensional space rather than on a flat, two-dimensional plate surface more accurately represents the in vivo condition. As our instructors put it, “Where do we have perfectly flat surfaces on our body?” We started with a gel-based culture using 20% methocel and our HEK293’s grown in a standard flask. The procedure was much like a regular cell culture except we used a 96-well plate to form our spheroids. It was easy to tell if a spheroid culture had grown in a well because they were large enough to see with the naked eye. To visualize the structures better, we fixed the cells with paraformaldehyde and stained them with DAPI and phalloidin 488, which interestingly comes from the death cap mushroom that grows in the Bay Area.  We also stained 2D cultured HEK293’s for comparison. The images we generated were astounding — easily the best fluorescent microscope images I have ever made!

HEK293 cells from 3D culture stained with DAPI and phalloidin 488.

Spheroids are very interesting structures. Certain cell lines will form them naturally if given the right supporting matrix. Many studies have shown how gene expression changes in 3D cultures versus 2D. Other factors such as cell polarity, growth, morphogenesis, motility, and differentiation differ in 3D versus 2D cultures. One thing that really fascinated me was that 3D culturing of cells causes cellular differentiation without the use of inducing factors. Spheroid bodies naturally form and the cells within them differentiate to form more complex extra cellular matrix (ECM). The center of the spheroid bodies will become necrotic and hollow, as can be faintly seen in the fluorescent microscope image.

HEK293 cells from 2D culture stained with phalloidin 488.

There are some disadvantages to 3D cultures. They vary in their ability to mimic in vivo conditions. Some experimental techniques use scaffolding and co-cultures to get closer to in vivo conditions. Some day soon it may be possible to introduce vasculature and transport of small molecules. Microfluidic devices can achieve some of this now. Another disadvantage is that 3D cultures tend to model static or short-term conditions, which of course is not how in vivo systems work.

Hanging drop cultures of HEK293 cells.

One of the most popular ways to get a 3D culture going is the hanging drop method. The name says it all: cells are literally suspended in a drop on the underside of a petri dish lid. We utilized this technique with our HEK293 cell cultures. It was a lot of fun to set up, although the final step of inverting the lid over the petri dish was a little nerve-racking because it is very easy to jiggle the lid too much and ruin the perfect dome shape of the drops. We checked these a few days later and found spheroid structures growing in our hanging drops. One of the cultures I transferred to a dish randomly formed a letter “H” (picture at end).

Stem Cell Techniques

Mouse embryonic stem cell (J1) embryoid bodies on fibroblast feeder layer.

Next, we were on to stem cells. Like I said earlier, I have never had the chance to work with stem cells so I was looking forward to this part of the course. We cultured mouse embryonic stem cells, specifically the J1 line. J1 cells were first grown in the lab in 1981 and have proved to be a useful model to study embryonic development and cell fate. Stem cells will naturally differentiate when grown in culture so Leukemia Inhibitory Factor (LIF) is added to maintain cultures in an undifferentiated state. (By the way, LIF is quite expensive. Only one company, Millipore EMD, supplies LIF and it costs over $1000 for a million units!) Also, stem cells do best when grown on a feeder layer. We treated our mouse fibroblast feeder layer cells with mitomycin C, which is highly toxic to cells and prevents them from dividing. Technically, stem cells grown on a feeder layer represent a co-culture since more than one cell line is in the same culture. We ended up getting some nice J1 embryoid bodies (EBs) forming on our feeder layer, as can be seen in the photo of the microscope computer screen. We had planned to coax these cells into the cardiac lineage using calcium treatments, but we had some difficulty passaging the EBs so we did not complete the task before the end of the second week. The problem was that our EBs were sticking to bacteriological plates and we did not know why. It was too bad, but the reality of working in the laboratory is that not everything works out!

ELF assay of mouse embryoid bodies time series.

One thing we did use our embryoid bodies for was an enzyme-labeled fluorescence (ELF) assay. This assay measures endogenous phosphatase activity, which is correlated to pleuripotency in stem cells. Thus, the brighter the EBs appear the more undifferentiated they are. We had some trouble with detection using our Zeiss Axio Imager A1 and it may have been due to using the wrong filter. The workaround was to do a time series of the EBs to see the change in fluorescence. The image to the right shows an EB at the end of the time series.

Flow Cytometry

Attune flow cytometer output for HEK293 culture with cytox green and dyecycle violet.

Nearing the end of the second week, we had time to learn one more advanced technology: flow cytometry. Flow cytometry uses lasers to count, sort, and detect proteins or biomarkers in cells. It relies on fluorescence and a variety of exclusion dyes (e.g., DAPI, Trypan blue, membrane dyes, etc.) and antibody stains. Fresno State’s RIMI facility has a Life Technologies Attune Flow Cytometer, which uses sound instead of fluid pressure to focus cells. We treated our HEK293 cells with two basic stains: Sytox Green and DyeCycle Violet. We ran our samples on the flow cytometer and set it to count 50,000 events. Once the data was collected, we got a crash course in gating and interpretation of results. What did we find? Well, this was not an experiment so we were just looking at untreated HEK293 cells in culture. DyeCycle Violet allowed us to see a distribution of cells in the cell cycle. As you can see from my results to the right, most of my cells were not dividing (bottom right graph), probably because they were over confluent. We did see interesting differences in cell cycle distribution depending on whose culture was being analyzed.

This cell culture techniques class was brought to you by the letter H and HEK293.

Give Me an H, Vasily. One H Only, Please.

What an incredible two weeks! I am amazed at all Dr. Bush and his graduate students were able to cover in such a short time. Plus, we spent the majority of it working in the laboratory, rather than sitting in lecture. It was one of the most hands-on classes I have ever taken. I gained a lot of confidence in my cell culture technique. If someone handed me a cryovial of adherent cells, I know I could culture them and keep them free of contamination. I am looking forward to taking this knowledge back to the biomanufacturing program at Laney College.

A special thanks is in order to Dr. Heather Brown at the North Valley Biotechnology Center for organizing this class. I hope NVBC offers more classes like this in the future!

August 7th, 2013 | Category: Announcements

What a week! All that work making green fluorescent protein (GFP) comes down to this: manual fill and labeling in BTEC’s aseptic processing suite. Sterility is of paramount importance because this is when the product is put in bottles that will go to patients. Contaminated product could lead to patient harm or death. Just look at a fairly recent example at the New England Compounding Center in which 48 people were killed by a fungal meningitis contamination of a steroid.

Fully gowned for aseptic processing.

The basic idea of aseptic processing is that the sterile filtered product needs to be protected from dirty people and dirty environments. Operators are carefully trained in gowning and cleanroom behaviors. Cleanrooms are extremely clean environments as compared to an average laboratory. Ultimately, the product will come out of the biomanufacturing facility completely sterile and ready to use.

After preparatory lectures we entered the gowning area to be trained in proper gowning procedure. Gowning is a choreographed, multi-step process that, when done properly, completely separates the operator from the cleanroom environment. No microbes, shed skin or hair, dust, debris, etc. will ever come in contact with the product under proper gowning and cleanroom conditions.

Filling out BPRs. Documentation never stops!

We worked in a cleanroom that met Class 10,000 (ISO 7) air quality standards and laminar flow hoods that met class 100 (ISO 5) air quality standards. We manually filled bottles with purified GFP, placed stoppers in the bottles, crimped our caps over the top, and did our weight checks and final fill calculations to determine product loss. Our group was within the acceptable range of product loss during fill. We had one double fill that led to marginal product loss. Our final reconciliation of percent product recovery included weight checks, GFP in Cozzoli manual filler assembly, and other places the GFP came in contact with. The calculation showed that we attained nearly 99% recovery. Not bad for our first time in a fill operation!

EM sets up shop in your space.

Working in a cleanroom was a lot more difficult than I thought it would be. Movements must be slow and deliberate. My goggles kept fogging up on me because my mask was not on right. Conditions are cramped in the laminar flow hood, especially when Environmental Monitoring (EM) comes in right in the middle of your work and sets up a particle counter and active air monitor. I kept leaning my head into the hood to get a better look, but that is a big no-no since I would be introducing particulates and other contaminants into the Class 100 environment. Also, cleanrooms are loud places. There are somewhere around 40 changes of air in a Class 10,000 cleanroom per hour, and the laminar hoods have 60 changes of air per hour. That means lots of fan noise from the HVAC system. We were told that some companies need their entire aseptic processing in Class 100 conditions. It must sound like a wind tunnel.

Grade A, Class 100, ISO 5 fill.

With filling accomplished we left the cleanroom with our bottles of GFP, de-gowned, and headed for the final step – labeling. Labeling was done in a Class 100,000 room. “Labeling is every bit as important as manufacturing,” our instructor told us. If a drug is made perfectly but is labeled improperly it is deemed “misbranded” by the FDA. Misbranding is a prohibited act under FDA regulations and could leave a company subject to fines or steeper penalties like being shut down. The FDA approves most drug labels. Even changes to existing labels must be approved. Therefore, labeling and label reconciliation are carefully documented and checked by quality control, just like every other step in the biomanufacturing process. One thing I learned was that it is very important that the QA representative be independent of manufacturing to limit potential conflicts of interest.

Capping off an amazing week.

I must say that I got a great deal of satisfaction seeing my finished product on the table once my group completed reconciliation and got the appropriate signatures from the manufacturing manager and QA. But we were not done yet! We had one last quality control step to do. With the day nearly over, we raced back to the lab with several of our bottles to test our product’s pH, conductivity, and concentration.

I had the most incredible time at BTEC this week. I am looking forward to taking everything I have learned here back to the biomanufacturing program at Laney College. We are starting our Associates of Science (A.S.) degree in biomanufacturing classes this Fall. So, if you have enjoyed reading my blog and would like to work towards a career in biomanufacturing, I encourage you to join our program at Laney College or find a program near you on the Bio-Link website (www.bio-link.org).

July 16th, 2013 | Category: Announcements

Downstream processing typically begins after harvest of your cells from the bioreactor. In our case we needed to disrupt or break open our cells because green fluorescent protein (GFP), our product, is made inside the bacteria, not secreted externally. Internally produced proteins have a few more processing steps than externally excreted ones because you have to deal with all the debris and immunogenic material of the disrupted cells.

GE Healthcare’s AKTA Prime Plus.

Purification operations are a critical part of a downstream processing because this is where the product is separated from all the other stuff that makes up bacteria. This step is what gets the product to fill and finish and into the hands of doctors and patients. Purification can occur with a variety of methods and technologies, but chromatography is often preferred because it has very high resolution. Our purification process workflow was the following: dilute the supernatant containing GFP from the previous centrifugation step to reduce conductivity, adjust pH to 8.2, syringe filter the product with 0.45 micron sterile filters, and then run strong anion-exchange chromatography using the AKTA Prime Plus system. (For a great animation of this process on the AKTA Prime Plus, go to this link: http://www.ncbionetwork.org/interactive-e-learning-tools-iet. You need to download the web player to view it.)

Chromatogram showing GFP peak in dark blue.

Anion-exchange chromatography uses resin inside a column that traps proteins based on their overall charge. Positively charged resin beads (we used a HiTrap Capto Q 5 column from GE) attract negatively charged proteins in low salt conditions. We first conditioned the column by running buffer through the system. Next, we loaded the GFP onto the column with a low salt buffer. This step is referred to as binding. Our chromatography system was programmed to run a step gradient increasing the salt conditions over time. Our elution buffer contained high salt concentration and was mixed with the starting buffer in increasing amounts. Under low salt conditions proteins other than GFP come off the column and are collected in the fraction collector. Once the salt concentration increased to a certain level, GFP came off the column and was collected. We could watch this happen with our black lights. In addition, the UV detector in the AKTA measured all protein coming off the column. We saw a very large peak near the end of our chromatogram, which represented the GFP coming off the column. We collected almost all the GFP in three fractions, although there was just a bit of tailing.

GFP glow gradient after elution.

We now had highly purified GFP. We were told that this step typically yields over 80% purity. Ideally, we would have followed this step with hydrophobic interaction chromatography (HIC) and a desalting diafiltration to push the purity even higher, but we only had one day left for manual fill and labeling in the aseptic processing suite at BTEC.

July 14th, 2013 | Category: Announcements