2017 YIQ Workshops

Last summer we conducted introduction to biotech workshops with students participating in the Young Innovators Quest, run by our friends at Health and Science Innovations.  In the workshop we covered basic biotech lab techniques including micropipetting, gel electrophoresis, and the polymerase chain reaction.  We also got a chance to brainstorm about the future of biotechnology and potential impacts on society.  Checkout some of the pictures below. 

Starting off the day with an overview of the different types of biotech industries.

Starting off the day with an overview of the different types of biotech industries.

The smell (of E. coli) that surrounds you

Anyone that has worked with the molecular biology workhorse, Escherichia coli, is well aware that the bacteria can create a foul odor.  The bad smell made by E. coli is caused by the reaction of an enzyme called tryptophanase that converts the amino acid tryptophan into the noxious chemical indole in the reaction shown below.

L-tryptophan + H2O ---> indole + pyruvate + NH3

Indole is actually responsible for the bad smell of human feces and it can be produced by E. coli living in the intestinal tract.  

Indole_structure.png

It's possible to genetically engineer E. coli to remove the gene (tnaA) that codes for tryptophanase and the bad smell.  This has been done as part of genome-wide functional studies like the Keio knockout collection and also IGEM teams have made use of a tryptophanase knockout to change the smell of E. coli.  This year I mentored a group of students in Hershey, Pennsylvania, as part of the Biobuilder Club, they were trying to make E. coli produce a vanilla smell.  The plan was to add genes that could produce vanillin to a tryptophanase negative strain.  

Vanillin is the compound responsible for the smell of vanilla.  It was first extracted from the seeds of the vanilla orchids and is still obtained from this source as a "natural vanilla".  More recently, chemical methods were developed to produce vanillin synthetically from petroleum byproducts.  The petroleum derived vanillin is sold as "artificial vanilla" and is usually less expensive than "natural vanilla".  A third means of production has been via microbes.  Microbes are capable of converting the naturally occurring plant phenolic compound ferulic acid to vanillin via a multi-step metabolic pathway (see below).  The microbes have no need for vanilla flavors -- the vanillin is just an intermediate in the conversion of ferulic acid to energy and carbon.

This schematic was copied from Di Gioia, Diana, et al. "Metabolic engineering of Pseudomonas fluorescens for the production of vanillin from ferulic acid." Journal of biotechnology 156.4 (2011): 309-316.

This schematic was copied from Di Gioia, Diana, et al. "Metabolic engineering of Pseudomonas fluorescens for the production of vanillin from ferulic acid." Journal of biotechnology 156.4 (2011): 309-316.

The genes for ferulic acid catabolism in Pseudomonas species are co-located in an operon in the genome.  To produce vanillin the activities of feruloyl-CoA synthetase and feruloyl-CoA hydratase/aldolase are needed and vanillin dehydrogenase should not be present.  If vanillin dehydrogenase is present then the vanillin will be converted to vanillic acid which doesn't have much smell.  

Ferulic acid operon from Pseudomonas fluorescens BF13.  Only the ech and fcs genes are needed for vanillin production.

Ferulic acid operon from Pseudomonas fluorescens BF13.  Only the ech and fcs genes are needed for vanillin production.

To produce vanillin in E. coli two genes from Pseudomonas were codon optimized for expression in E. coli and purchased as gene fragments.  These genes were originally identified in a 2008 publication.  The fragments were assembled in a stepwise fashion using Golden Gate assembly and the pGLB3 vector to form a lactose inducible promoter fused to the ech and fcs genes.  When this was transformed into the E. coli unfortunately no vanilla smell was generated.  The plasmid was sequence verified and confirmed to be correct.  

More careful inspection of the fcs gene open reading frame suggests that the authors who originally identified the gene may have mispredicted the start codon (see alignment below).  It appears that in the DNA sequenced by the original research group, the N-terminal region is truncated as compared to essentially all other similar Pseudomonas fcs genes.  To fix this potential issue, additional DNA will be synthesized that adds the coding sequence for the probably N-terminus of the genes.  Hopefully this will enable the recombinant E. coli to generate vanillin, and more importantly the pleasant smell of vanilla, in large enough quantities to be easily noticeable.  Expect another update on this progress soon.

Designing a low cost molecular biology platform

Functional DNA is at the heart of modern biotechnology.  At the Great Lakes Biotech Academy our key educational goal is to teach students how to find and create functional DNA.  Once a functional DNA sequence is discovered it may be challenging to produce in the lab -- especially if you don't have access to sophisticated (and expensive!) scientific equipment.  To address this challenge most of my time in the academy lab the past few months has been focused on finding ways to simplify and reduce the cost of DNA assembly.  This post will provide an overview of my progress thus far.

Example of functional DNA at the biotech academy.  The flask contains E. coli expressing functional DNA for a green fluorescent protein.

Example of functional DNA at the biotech academy.  The flask contains E. coli expressing functional DNA for a green fluorescent protein.

A brief history of DNA assembly

The modern era of molecular biology was enabled by two key inventions: the polymerase chain reaction (PCR) and restriction enzymes.  PCR was developed in the early 1980's and allows the facile copying of small to medium sized DNA molecules(typically 100 to 5000 base pairs).  With PCR it became relatively easy to amplify and purify many types of functional DNA.  Restriction enzymes are part of bacterial immune systems that recognize and cut short sequences of DNA.  Some restriction enzymes create "sticky ends" after they cut and these sticky ends can be used to join different pieces of DNA together.  For 20+ years most molecular biologists used PCR and a plethora of different restriction enzymes to build functional DNA molecules.  The techniques worked, but most DNA molecules required different restriction enzymes and it was especially difficult to build large DNA sequences.   In the early 2000's new DNA assembly technologies were developed that allowed much easier synthesis of large DNA sequences.  One of these technologies, The Gibson Assembly method, was used to synthesize an entire genome and create a synthetic life form.  Gibson Assembly is an extremely efficient way to produce large DNA sequences, but it is inherently expensive because relatively large homology regions must be added to every DNA fragment in the assembly.  

Hierarchical cloning and the Golden Gate Assembly method

With Gibson Assembly any large DNA sequence can be synthesized, however the expensive DNA fragments used to assemble the large DNA sequence generally can't be reused for related designs.  When assembling multi gene pathways researchers often want to test the effect of different promoters, coding sequences, and terminators for every gene in the pathway and Gibson Assembly is not a cost effective way to explore that parameter space.  In 2008 a new method for hierarchical DNA assembly, termed Golden Gate cloning was published by Engler et al.  For Golden Gate cloning, the Type IIS subclass of restriction enzymes was used in a clever way to join multiple DNA fragments very efficiently and in the process remove the restriction enzyme recognition site from the product.  This approach worked because type IIS restriction enzymes cut outside their recognition sequence and can thus be used to create custom "sticky ends".  By nesting two different Type IIS restriction enzymes a molecular biologist can put together larger and larger pieces of DNA.  The only caveat of the Golden Gate Assembly method is that every DNA fragment in the assembly can not contain the Type IIS restriction site used in the Golden Gate reaction.   The commonly used Type IIS restriction enzymes BsaI, BmsBI, and SapI have either 6 or 7 base pair recognition sites so for most genes this is not an issue.

Recent improvements on the Golden Gate Assembly method have involved exploiting the sensitivity of the BsaI restriction enzyme to DNA methylation and I am incorporating this approach into the biotech academy's molecular biology platform.  This will allow us to follow a very simple and low cost procedure to assemble functional DNA several thousand basepairs in size.  

progress towards a molecular biology platform

The first experiments related to the molecular biology platform were performed in May and were done to determine if a subset of BsaI sites could be blocked effectively by a DNA methyltransferase.  Two commercially available methyltransferases, MspI and HpaII both recognize the sequence CCGG and methylate the external or internal cytosine, respectively.  By adding two cytosine bases immediately 5' to the BsaI restriction site a MspI and HpaII methylation site can be created that overlaps the BsaI recognition sequence.  The schematic below summarizes the general strategy where the methylation sensitive mBsaI site is on the outside of the insert (on the vector backbone) and the methylation insensitive site is on the interior and linked to a LacZalpha marker.

It was known in the literature that methylation could severely inhibit BsaI, but it hadn't been reported if the methylation pattern catalyzed by either MspI or HpaII could block BsaI.  To test this I PCR amplified a portion of the Neurospora crassa genome that contained the mBsaI site.  The PCR amplified product was treated with MspI and HpaII and then subjected to digestion with a high fidelity version of BsaI.  The methylation by MspI and HpaII appeared to completely protect the DNA from cleavage (see below).

Both MspI and HpaII treatment prevented cleavage by BsaI at the CCGGTCTC site. 

Both MspI and HpaII treatment prevented cleavage by BsaI at the CCGGTCTC site. 

Building the pglb vectors

So the next step was to incorporate this methylation sensitive BsaI site into a useful cloning vector like pUC19.  The series of vectors created at the Great Lakes Biotech Academy are called pGLB vectors, where X designates the number of the vector.  The design of the pGLB vectors includes blue/white screening using a modified LacZalpha fragment, a modified version of beta-lactamase which reduces the production of satellite colonies, and the incorporation of the methylation sensitive and insensitive BsaI cloning sites.  The other modifications in the pGLB vectors will be highlighted in future posts.  

This plasmid map shows the features of pGLB2.  pGLB2 is a pUC19 derivative with a redesigned LacZalpha, no BsaI sites in the backbone, and no multiple cloning site.  It incorporates the methylation sensitive (light green) and insensitive (yellow) sites to allow hierarchical golden gate assembly with a single restriction enzyme.

This plasmid map shows the features of pGLB2.  pGLB2 is a pUC19 derivative with a redesigned LacZalpha, no BsaI sites in the backbone, and no multiple cloning site.  It incorporates the methylation sensitive (light green) and insensitive (yellow) sites to allow hierarchical golden gate assembly with a single restriction enzyme.

After building this vector tests were performed to judge the effectiveness of the cloning strategy.  The pGLB2 vector is produced in E. coli like any other plasmid.  After miniprepping the plasmid it is treated with HpaII or MspI methyltransferase and then is ready for use in the Golden Gate cloning reaction.  At the Great Lakes Biotech Academy we use a modified protocol for Golden Gate Assembly that uses high fidelity BsaI restriction enzyme, T7 DNA ligase (instead of T4 DNA ligase), pGLB2, and the Golden Gate compatible DNA inserts.  For the initial test of pGLB2 we performed three component assemblies with the Lac promoter (part 1), a green fluorescent protein gene (part 2), and methylated pGLB2 (part 3).  

Possible outcomes for this reaction when the transformed cells are plated onto LB agar supplemented with ampicillin and X-gal are as follows:

  1.  Green fluorescent colonies -- a successful reaction!
  2.  Blue colonies -- a failed reaction, indicating re-ligation of the vector
  3.  White colonies -- a failed reaction, indicating loss of LacZalpha and mis-ligation of vector
  4.  No colonies -- a failed reaction, inserts with improper overlaps
Transformation of DH10B E. coli with Golden Gate assembly reaction containing pGLB2, a promoter, and a green fluorescent protein.  The higher the ratio of green to blue colonies indicates a more efficient assembly reaction.

Transformation of DH10B E. coli with Golden Gate assembly reaction containing pGLB2, a promoter, and a green fluorescent protein.  The higher the ratio of green to blue colonies indicates a more efficient assembly reaction.

 The initial test showed thousands of green fluorescent colonies and just a handful of blue colonies (can you find them in the image?).  Control experiments where no GFP insert was included showed no colonies.  Together these results suggest that the pGLB2 cloning strategy works very efficiently.  Soon we will use this system to rapidly assemble several pieces of functional DNA into a biosynthetic pathway. 

3D Printing in the Lab

Many types of molecular biology lab equipment are very expensive and thus hard to get access to.  This is because the market for such equipment is relatively small, and when that is coupled with limited competition among suppliers, the result is a reduced incentive to lower prices or innovate.  In competitive mass markets product quality goes up and prices drop dramatically.   For instance, flat screen televisions and smart phones have increased dramatically in quality and decreased dramatically in price over the last 10 years.  In contrast , the cost of most lab equipment increases on a yearly basis. Companies also do not want to accrue the cost of storing unordered products, adhering to a “just in time” production schedule. This can increase the time a lab needs to wait to use the product while it is being manufacture and shipped. 

We have been using the Taz 5 "open source" 3D printer for most of our custom lab equipment.

We have been using the Taz 5 "open source" 3D printer for most of our custom lab equipment.

With the arrival of increasingly more affordable 3D printers, and better public access to high-end 3D printers commonly found in maker spaces, laboratories on a limited budget can now design, print and use, quality equipment quickly and at a reasonable cost. 

The Great Lakes Biotech Academy is focused on making biotechnology research more accessible and needs to cleanly extract DNA from various types of isolated fungi.  Fungi have complex cell walls made of chitin and beta-glucans, that need to be disrupted in order to release DNA.  Mechanical shearing utilizing  rotor-stator homogenizers can do this, but the tip which generates the high shear forces, needs to be cleaned between each extraction.

Currently any laboratory processing more than just a few samples at a time needs to use a tube based “bead beater” homogenizer, that is coupled with silica based DNA absorption and elution.  To break down fungal cell walls efficiently you need to place fungal tissue in a tube along with hard beads and use the bead beater to shake the tubes at very high speed to completely disrupt the tissue and release the DNA.  Typically, such bead beaters can cost anywhere from $800 to $15,000 dollars, so this was a good target for using 3D printing.

Homogenizer chambers that were freely available on the internet.

Homogenizer chambers that were freely available on the internet.

As 3D printers have become more affordable, a number of websites, such as Thingiverse and Tinkercad, have begun to act as repositories of third party 3D designs that are shared free or at a low cost.  These are good sites to check to determine if someone else might have already designed the type of equipment you need.  After an internet search, a homogenizer design was found to be available on a NIH website.  However, it was only available as a WRL file, which is hard to work with relative to the more common STL file.

The design (red) could hold a microcentrifuge tube snuggly within the T shaped section but utilized a plastic piece which extended from the homogenizer and to which two adapters were used to attach to a reciprocating saw that could shake the tube at greater than 3000 rotations per minute.  The tubes fitted into the device and depending on tolerances of the printing they had the potential to either fall out during homogenizing or alternatively be very difficult to remove.

We therefore wanted to modify the original NIH design to not only increase the number samples that could be processed, but also more securely attach the homogenizer chamber to the reciprocating saw. Soon after we started to craft this new design, an updated version to the original file became available.  The new design could process up to 6 samples, but it still relied on the plastic attachment point, that could weaken and snap off (tan).  To address these remaining issues, we used an iterative development process that included: design, print, test, then redesign/modify, and repeat as needed.

Early GLB designs were similar to the designs available on the internet, but included a custom attachment point for a metal hex shank.

Early GLB designs were similar to the designs available on the internet, but included a custom attachment point for a metal hex shank.

Our initial designs started out very similar to the NIH design.  After some analysis and consideration, we decided that our homogenizer design should be able to process multiple samples and different tube types.  We also incorporated a lid that can be bolted shut to safely contain the sample tubes during the homogenization process and inserted an integrated metal hex shank.  This should make the  attachment to the reciprocating saw more secure and unlike the original design,  should lower the possibility of the homogenizer breaking away from the reciprocating saw.  

Later GLB designs that incorporate important features like variable tube size and easier tube insertion and removal.

Later GLB designs that incorporate important features like variable tube size and easier tube insertion and removal.

After testing several designs, we chose two as “final” versions that should work well in the laboratory.  One design is very small,  but can only accept standard 2 mL screw top tubes.  The second homogenizer design is bigger, but we added the ability to insert adapters if the researcher would want to use either a standard 2 mL screw top tube, or snap top microcentrifuge tubes.

The final homogenizer chamber design relies on five elements. 

  1. A 3D printed lid that covers the tubes
  2. A 3D printed homogenizer chamber that the tubes sit in
  3. A 3D “cap” which is used to secure the hex shank/nut driver  in the homogenizer, and a
  4. 1/2 hex shank magnetic nut driver that facilitates the attached to the  
  5. A hex shank to reciprocating saw adapter

The hex shank magnetic nut driver can be obtained from a hardware store or bought online.  The nut driver is magnetic and a short bolt is inserted into the hex shank driver and is permanently attached to the bolt by superglue.  After the glue sets, the hex shank/bolt is then inserted into the homogenizer, such that the hex shank extends out of the homogenizer.  You should then test to make sure that the reciprocating adapter can be securely engaged to the hex shank which is now extending out of the homogenizer. 

When the adapter is attached to the hex shank at the base of the homogenizer it should be as flush to the 3D printed piece as possible.  If your tolerances are off, you may notice that you have space between the adapter and the homogenizer.  If this is the case it would be advisable to fill in that space with a couple washers. 

Once the hex shank is attached to the homogenizer chamber, at the other end you will need to insert or twist the small 3D printed cap over the bolt and push it until its flush with the homogenizer. Superglue should then be used to seal it.  After the sample tubes are placed in the homogenizer the lid is lowered over the bolt and is secured sequentially with a washer, a lock washer and then with at thumbscrew.  A nut could also work but the thumbscrew allows for easier access. 

The adapter is attached to the reciprocating saw model.  We used a 12 AMP unit which can reach ~3000 RPMS.  This is powerful enough to lyse even the very strong fungal cell wall.  Depending on the power of your reciprocating saw, you may need to change the homogenizing parameters to extract DNA from you process efficiently. 

We have used these designs to extract DNA from filamentous fungi, spores, and mushrooms with great success.  Checkout some representative images of the homogenizer in action on mushroom tissue.

Please note : 3D printed objects do not go through the same rigorous quality and safety testing that commercial products do.  Our design is based on printing with ABS filament and we have not tested this design with other filaments.  You should be extremely careful when using any 3D printed design in conjunction with power tools.  Always use safety glasses during the homogenization process and make sure the homogenizer is aimed towards the ground over a Styrofoam or cardboard box.    

YIQ Visit

On June 27th a group of students participating in the Young Innovator's Quest visited the biotech academy.  We spent some time talking about functional DNA and the future of biotechnology.  We then performed experiments to determine the genus and species of several non-hazardous mushrooms.  

To determine the species of mushroom, we performed PCR reactions on genomic DNA isolated from each mushroom at the biotech academy.

Lane 1, specimen #1; Lane 2, specimen #2; Lane 3, specimen #3; Lane 4, specimen #4; Lane 5, Ladder; Lane 6, specimen #5; Lane 7, specimen #6.

Lane 1, specimen #1; Lane 2, specimen #2; Lane 3, specimen #3; Lane 4, specimen #4; Lane 5, Ladder; Lane 6, specimen #5; Lane 7, specimen #6.

Participants got to choose some of the parameters in the PCR reaction, including the primer concentration, the genomic DNA template concentration, and the PCR program.  The best looking amplicon for each specimen will be sent for sanger sequencing and used to identify the genus and species later this week.

The three PCR programs used for the mushroom amplification.

The three PCR programs used for the mushroom amplification.

Most of the PCR reactions turned out nicely.  The most intense band for each specimen will be sent for sequencing.