Team:UC Davis/PartFamilies

From 2011.igem.org

Revision as of 00:07, 24 October 2011 by Aheuckroth (Talk | contribs)

Our Sponsors

Start a Family

Got a favorite BioBrick? Check our our process for expanding basic parts into part families.

Criteria

View our judging criteria for iGEM 2011 here.

Make Your Own Part Families

Part families are a great asset to the Parts Registry. Our process for expanding basic parts into part families is quick, easy, and prioritizes the use of materials that most iGEM teams already have on hand. By following these steps, you can improve the utility of your favorite part and contribute to a strong foundation of useful parts for synthetic biologists around the world.

Our Method

Our part family expansion process is broken down into four major steps:

Part Selection: Choose your basic part of interest and obtain a DNA sample.
Mutagenesis: Create a library of mutants from your basic part.
Screening: Select a range of mutants that will offer the most utility for future projects.
Characterization: Collect detailed information on each mutant to document its behavior.

Selection

The Parts Registry has a wide selection of basic parts that are good candidates for part family expansion. This includes promoters, repressors, reporter proteins, and many others. For our process to work properly, selected parts must contain standard VF2 and VR sites around the region of interest.

Ideal parts are around 200 base pairs or more in length, as smaller parts show reduced levels of transformation success. Part selection is also key to the success of later screening and characterization steps: measuring the activity of parts often requires the use of a fluorescent reporter like GFP, so your parts must affect the expression of this reporter in some way. For example, promoter mutants can be used to transcribe different levels of reporter mRNA, and differences in the activity of reporter mutants can be measured directly.

We also recommend that parts be selected with their future utility in mind -- we chose to work with repressible promoters because they are frequently used in designing genetic circuits. This will help frame the paramater space over which mutants will be selected for final characterization, and helps judge the amount of mutagenesis required to achieve good mutant library resolution.

Mutagenesis

Mutagenesis creates the genetic variation that leads to changes in part behavior. This can be achieved in a number of ways, but we designed this process with the use of our suped-up error-prone PCR protocol in mind. This protocol has been optimized specifically for creating mutants of Registry parts: it uses standard VF2 and VR primers, produces a large number of mutations per reaction for rapid library generation and visual screening, and it uses materials that are already present in most iGEM team labs.

Since the degree of mutation required to produce desirable levels of phenotype variation is not generally known before this step has been completed, we suggest performing several rounds of mutagenic PCR in succession, saving the products from each reaction, then digesting, ligating and transforming the results separately. This will produce plates of mutant transformants at different degrees of mutation from wild-type.

We found that our protocol produces about 1 to 7 base pair mutations per round of error-prone PCR, which was enough to visually identify GFP mutants (~800bp) on transformation plates after just one round and LacI promoter mutants (~200bp) after three rounds. We also observed a qualitative increase in successfully transformed colonies when we used a PCR purification kit on our PCR products before cutting them with restriction enzymes.

Screening

This is perhaps the most difficult step in the process of generating part families. All parts present a unique challenge: each part has a different set of characteristics that will be modified by mutagenesis over a theoretical range of useful values.

First, a vector must be prepared into which the digested error-prone PCR products can be ligated for screening and characterization. This vector should associate reporter activity with the function of a mutant part. We have designed a generic screening plasmid, K611018, which can be used to assay promoters, repressors, and other regulatory sequences. It can be prepared for use with repressible promoters, for example, by ligating the wild-type repressor to the 3' end prior to digestion and ligation with promoter mutants.

Fluorescence-activated cell sorting is a terrific tool for selecting individual bacteria based on reporter expression, but the cost of owning or renting a FACS machine can be prohibitive to many teams. As an alternative, we found that visually screening for GFP expression is an efficient way to initially screen mutant libraries. The increased mutation rate of our mutagenic PCR protocol is high enough that a range of colonies, from nearly-white to bright green, can generally be seen on successful transformation plates. The use of a hand-held long-wave UV lamp my assist in this process. While colonies across a wide range of GFP expression can be selected, not all variations in part behavior will be visible based on GFP fluorescence under these conditions. For this reason, many more colonies should be selected than will be expected to become part family members.

Further rounds of screening must be performed to narrow the choice of mutants down further. We used a Tecan InfiniTe M200 plate reader to measure the fluorescence activity of 250 uL liquid cultures of cells containing our promoter mutants. Fluorescence plate readers are an excellent tool for quantifying the activity of wild-type and mutant parts, but they are also expensive to own and take at least 12 hours to produce data on a set of cultures. Initial visual screening helps to reduce the total amount of plate reader access time required to select final mutants, and our team is currently exploring alternative options for the screening and characterization of cell cultures.

While intermediate screening can be performed on single samples of cell culture, mutants should be run in triplicate ensure that deviation from wild-type activity is repeatable. All plate reader runs should contain a wild-type control containing the part from which the mutants were generated ligated into the same screening vector, an optical density negative control containing only growth media and no cells, and a cell fluoresence control containing cell culture without reporter expression.
..