"
Team:MIT/miRNA
| Home |  Our Project
|
Lab Work
|
Outreach
|
About Us
|
Medals |
miRNA DETECTOR MODULESUBGROUP MEMBERS: Gary Burnett, Jing Wei "Raymond" Liu, Raashed Raziuddin, Jiaqi Xie Attributions: Jing Wei "Raymond" Liu (Description), Gary Burnett (Experiments), Kathryn Brink (Animations) sensing Alzheimer's through multi-input miRNA-based logic
DescriptionmiRNAs (microRNAs) are short, noncoding strands of RNA that facilitate gene silencing - a single miRNA degrades an mRNA through a process involving complementary base-pairing between the miRNA and part of the mRNA sequence. A cell’s miRNA profile comprises the relative levels of all the miRNAs produced by that cell. Because miRNAs play a key role in regulating gene expression, it ought to be expected that a liver cell’s miRNA profile would differ significantly from that of a neuron. But more surprisingly, miRNA profiles can discriminate between identical cells in different conditions. Neurons afflicted with Alzheimer’s disease display an miRNA profile significantly different from that of healthy neurons (A blood based 12-miRNA signature of Alzheimer disease patients, Leidinger et al, 2013). The miRNA subgroup aimed to use this difference as an approach to detecting Alzheimer’s disease. Our goal was to build a set of genetic sensors to specifically detect the miRNA profile of a neuron with Alzheimer’s and initiate a specific biological response upon doing so. Our strategy took its inspiration from a similar detection circuit demonstrated to respond to cancer onset (Multi-input RNAi-based logic circuit for identification of specific cancer cells, Xie et al, 2011) Through existing research, we identified six miRNAs that are critically up- or down-regulated in Alzheimer’s neurons. Using the inverting logic inherent to miRNAs, we designed detection circuits to release a response factor upon sensing either heightened or lowered levels of their target miRNA, and customized each circuit to use one of the six miRNAs as its input. Using the principles of combinational logic, we can integrate the inputs from all six of our miRNA sensors, and actuate our response only when all six miRNAs meet their critical threshold concentrations. This ensures excellent specificity for our circuit.return to top OutcomeThe miRNA detection team built individual sensing constructs for each miRNA. We determined input-output relations for our sensors using flow cytometry and found that our sensors respond to miRNA levels by modulating the production of a fluorescent reporter, exactly as we had predicted. Future work on our sensors will focus largely on implementation concerns - tuning as well as integration. Although we have shown that our sensors respond on a digital level, this does not accurately model the dynamic chemical conditions of the intracellular environment. We would thus like to refine our sensor control. In the ideal case, a small shift in a critical range of miRNA concentration will result in a large output signal, so that the treatment response is both specific and substantial. We only tested binary combinatorial inputs for our sensors (one high and one low, or two of each). The ultimate goal is to use all six sensors in tandem with one another. When we use more sensors, we achieve greater precision, but as a tradeoff we gain more variables that require keeping track. There is also the complication that the various miRNAs are not biologically present at the same concentrations, meaning that each of our sensors must be individually tuned for optimal response to its own miRNA. Because all six of our sensors actuate the same response, we must also ensure that one sensor does not become overstimulated, activating our treatment even in the absence of input from the other sensors. These are all issues that can only be answered through extensive iterative testing. The miRNA sensing team has established a conceptual grounding for a detection mechanism that responds to cellular conditions in the fashion of a true biological system. It is worthwhile to note that our strategy is not Alzheimer’s-specific, and can be implemented with any disease with a characteristic miRNA profile. This can be a novel approach for diseases with poorly understood etiologies, such as Parkinson’s (MicroRNA profiling of Parkinson's disease brains identifies early downregulation of miR-34b/c which modulate mitochondrial function, Minones-Moyano, 2011)return to top Experiments return to top Low Sensor ConstructionBy cloning an miRNA target site 3’ to a gene coding a reporter protein, we can easily create a sensor that produces reporter protein only when miRNA levels are low enough to permit translation. In our experiments, we used a fluorescent reporter as a placeholder for rtTA, which would activate our treatment circuit.
return to top High Sensor ConstructionBecause miRNAs naturally silence genes, for our high sensor design we cloned miRNA target sites to a repressor protein that would block transcription of response protein at the low sensor. We chose to use the L7ae/K-turn to eliminate the possibility of crosstalk with other cellular activities.
return to top Repression of L7aeBefore coming to any conclusions about the success of our constructs, we needed to make sure that the L7ae/k-turn system worked correctly. To do this we expressed k-turn:eGFP with and without the presence of constitutive L7ae. We used eBFP as our normalizing transfection marker.
return to top Single-Input miRNA Sensor ExperimentsWe tested the sensitivity of our sensor constructs individually for each of the six miRNA that are upregulated or downregulated in Alzheimer's neurons. Our experiments were conducted by transfecting HEK293 cells with our sensor constructs. Since there were no cell lines in existence that could naturally reproduce the exact miRNA expression profile of an Alzheimer's neuron, we needed to artifically alter the miRNA profile of an existing cell line. We discovered through research in literature that HEK293 cells do not endogenously express any of the miRNA's that we planned to sense for, thus they provided the perfect low miRNA environment. In order to create a high sensor environment in this cell line, we needed co-transfect siRNA along with our sensors.
return to top Multi-Input miRNA Sensor ExperimentsIdeally, we would put all of our miRNA sensors together into one, large classifier circuit that is capable of testing all the miRNAs in the Alzheimer's profile at once. However, before we test 6 sensors in tandem, we decided to try some pairwise combinations of high and low sensors. This would give us important information that we could use to tune our system before making to jump to testing all of the sensors at once.
| ||||||||||||||||||||||||||||||||||||||||||||||||||
return to top
Parts
full parts list available herepENTR_hEF1a
MAV1212-RFP
pENTR_L7ae
MAV1212_hEF1a_L7ae
MAV1212_hEF1a_eBFP2
Low sensor: eBFP2_miR-144
Low sensor: eBFP2_miR-181c
High sensor: L7ae_miR-30d
High sensor: L7ae_miR-146a
High sensor: L7ae_miR-125b
High sensor: L7ae_miR-let-7f
Low sensor: eBFP2_miR-144_miR-181c
