Note: This post, which I wrote the summer after my freshman year at Duke, was originally published on the Howard Hughes Research Fellows Program blog. I ported it over here as an homage to my past life in biology research. Also, from a conceptual standpoint molecular cloning remains pretty interesting.

My goal this summer has been to generate an animal cell line that can give off a measurable signal when a certain cancer-related gene is expressed, or made into protein. Specifically, I am trying to place the regulatory sequence, or “promoter”, of the ARF gene before the DNA sequence which codes for the fluorescent protein mCherry. When a certain wavelength of light is shined on cells expressing mCherry, the protein emits another wavelength of light which appears red in the visible spectrum. Cells which contain this new DNA and have ARF activated will respond to excitatory radiation by shining red. Those without ARF will return a weak signal or no color at all.

The process of making this so-called “reporter gene” involves a standard molecular biology method known as molecular cloning. Cloning in razzmatazz headline science is a little different than what I’m doing in the lab, which essentially involves cutting up pieces of DNA, pasting them together into a circular piece of DNA with a key retrovirus gene, introducing this into bacteria to isolate the right construct, throwing this new DNA into one type of animal cell to make an infectious virus, and then letting the virus loose on another set of mammalian cells.

Why retrovirus?

The virus’s DNA is injected into those cells and incorporates itself into the cells’ genome. This way, the reporter gene can be inherited by the daughter cells once the initially infected cells replicate, eliminating the need to keep reintroducing it as the cells divide in culture. Inclusion in the genome also subjects the reporter to conditions that mimic experiences of the actual ARF gene, such as access to regulatory proteins. As the reporter indirectly gauges how much ARF is being made in the cell by using an additional promoter, the general idea is to keep everything as similar to what’s happening to the actual gene as possible.

Borrowing tools from biology

Many of the technologies for manipulating the DNA are based on properties that were discovered in bacteria or viruses. For example, some bacteria are thought to have evolved machine-like proteins called restriction enzymes as a defense mechanism against invading viruses. Restriction enzymes recognize specific sequences of nucleotides on foreign DNA and cleave the sugar backbone of each strand of the double helix. This can generate ends that are staggered, or “sticky” because they can glue back together again more efficiently than “blunt” ends, which are cut so that neither strand sticks out more than the other. If the concept of sticky ends resembles conventional puzzle pieces, blunt ends are like Triazzle triangles.

Some scientists recognized the enormous potential of having this kind of selective cutting power and began combining it with other techniques to make new DNA constructs. Once the sequences specific to many restriction enzymes were known, it was possible to design stretches of DNA with a high number of unique restriction sites. Such a sequence is commonly called a multiple cloning site (MCS). Since the restriction sites in the MCS are generally very rare, an MCS restriction enzyme is likely to cut the plasmid only once, in a well-defined spot. When put into a plasmid (a circular piece of DNA separate from the bacterial chromosome which can be transferred not only from parent to daughter cell but also between cells in the same generation), an MCS can greatly simplify the task of inserting a gene of interest into a desired spot on the plasmid.

Thanks to such discoveries, I can now purchase purified restriction enzymes from a series of vending machine-like freezers in the hallway and use them to cut out the gene I want to insert out of its storage plasmid, as well as the “backbone” viral plasmid which will house my final construct. Because the restriction sites I’m using create staggered ends which don’t match the staggered ends of the backbone plasmid, it is necessary to make the ends flat or “blunt” so they can be stuck together.

To do this, I add an enzyme called a “Klenow fragment” to the cut DNA. The Klenow fragment is a subunit of a bacterial enzyme, DNA polymerase I, which manufactures strands of DNA. The special thing about Klenow, though, is that it can only add nucleotides in a certain direction (5’ to 3’), limiting its manufacturing capacity to one of DNA’s two strands. It can also trim away nucleotides from the other strand. The end result is that sticky ends are either trimmed or filled in until both strands are matched in length, as the Klenow fragment can’t make new DNA from scratch once the overhanging template strand is all paired up. Neither can it keep eating away at double-stranded DNA once a protruding length of single strand has been removed.

Aside from restriction enzymes and the Klenow fragment, a few other enzymes discovered by biologists are used in the process of making the construct, including a phosphatase (a phosphate-removing protein) and a DNA ligase which uses the energy-rich molecule ATP to create bonds between DNA strands. DNA ligase is the glue which puts the cut and blunted fragments together, making a circle which ideally includes the so-called “insert”, either your desired promoter or gene. The problem is that with blunted fragments, or with sticky-ended fragments with the same restriction site on either end, the insert could be added in backwards. So it becomes necessary to later test the plasmid constructs to find out which way the insert has been added in each one (if at all).

Depending on the experiment, it can be sufficient to simply transfect this into bacteria which are competent, or ready to accept new plasmids. Then, it’s just a matter of isolating a colony of identical bacteria which contain your desired construct. For reporter genes, which have two parts (the promoter and the detectable signal), this entire process needs to be repeated to get the next part in.

But getting to this stage isn’t enough to reach my particular goal. Bacteria, like all single-celled organisms, don’t develop cancer and lack the genes/genetic networks (such as ARF) that we want to investigate. So another step (or series of steps) is required: the introduction of the reporter gene into mammalian cells.

Gotta love them pathogens

Now this is an especially fascinating example of how scientists turned a discovery of a seemingly esoteric biological process into a powerful genetic engineering tool. They’ve managed to exploit the hallmark activity of retroviruses like HIV, which are able to insert their genetic code into the genome of host cells. All retrovirus genomes must contain variants of three key retroviral genes (gag, pol, and env), in addition to a gene whose expression causes the virus’s protein capsule to assemble (necessary for infection).

Working with functional retroviruses can be pretty hazardous to the researcher, so biotech companies like Clontech have removed the gag, pol, and env genes from the plasmid which receives the reporter gene inserts, instead putting them into the genome of an intermediate cell line where they are expressed continually. The retroviral plasmid still retains the assembly gene which encodes the packaging signal for making a functional virus. When the plasmid is introduced into the intermediate cell line and starts producing packaging signal, infectious retroviruses are assembled. However, they only enclose the plasmid (in RNA form) and not the critical genes contained in the intermediate cells, so viruses produced in this fashion can’t make new infection-capable viruses even after incorporating into the final cell line’s genome. Although researchers are still at risk for having some of their cells infected, the virus will (most likely) not breed and infect others.

BioBricks TM

In short, many of the techniques that are now common in molecular biology and genetic engineering were built on discoveries in lower-order microorganisms (and non-organisms, if we’re discussing viruses). Although many aspects of cloning are fairly standard already (in terms of ordering plasmids, enzymes, and reagents from biotech companies), my efforts to built the construct tell me that the process could still be made simpler and more efficient. More predictability and simplicity in cloning could accelerate both the study of existing genetic networks and invention of new ones.

Some scientists are working to address one aspect of this problem by maintaining a Registry of Standard Biological Parts, called BioBricks (trademarked). By providing a framework for consistent documentation of DNA segments and their functions, they are trying to make DNA easier to mix and match to make useful or interesting circuits, much like how engineers mix and match resistors, capacitors, and other parts to make functional electronic devices. The BioBricks project is a super exciting endeavor (and probably fodder for a future post).