Limits...
A tunable zinc finger-based framework for Boolean logic computation in mammalian cells.

Lohmueller JJ, Armel TZ, Silver PA - Nucleic Acids Res. (2012)

Bottom Line: We describe 15 transcriptional activators that display 2- to 463-fold induction and 15 transcriptional repressors that show 1.3- to 16-fold repression.The split intein strategy is able to fully reconstitute the ZF-TFs, maintaining them as a uniform set of computing elements.Together, these components comprise a robust platform for building mammalian synthetic gene circuits capable of precisely modulating cellular behavior.

View Article: PubMed Central - PubMed

Affiliation: Department of Systems Biology, Harvard University, Boston, MA 02115, USA.

ABSTRACT
The ability to perform molecular-level computation in mammalian cells has the potential to enable a new wave of sophisticated cell-based therapies and diagnostics. To this end, we developed a Boolean logic framework utilizing artificial Cys(2)-His(2) zinc finger transcription factors (ZF-TFs) as computing elements. Artificial ZFs can be designed to specifically bind different DNA sequences and thus comprise a diverse set of components ideal for the construction of scalable networks. We generate ZF-TF activators and repressors and demonstrate a novel, general method to tune ZF-TF response by fusing ZF-TFs to leucine zipper homodimerization domains. We describe 15 transcriptional activators that display 2- to 463-fold induction and 15 transcriptional repressors that show 1.3- to 16-fold repression. Using these ZF-TFs, we compute OR, NOR, AND and NAND logic, employing hybrid promoters and split intein-mediated protein splicing to integrate signals. The split intein strategy is able to fully reconstitute the ZF-TFs, maintaining them as a uniform set of computing elements. Together, these components comprise a robust platform for building mammalian synthetic gene circuits capable of precisely modulating cellular behavior.

Show MeSH
Engineering and characterization of ZF-based OR and NOR Boolean logic gates. In the sensory module, input signals lead to expression of corresponding ZF-based transcription factors. In the computational module, transcription factors act on response promoters. (A) OR gate response promoters contain target sites for two different ZF activators, and the logical operation is computed as TRUE (CFP expression) when either one or both inputs is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun activators were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control. (B) NOR gate response promoters contain the binding sites for two different ZF repressors, and the logical operation is computed as TRUE when neither input is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun repressors were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3367183&req=5

gks142-F3: Engineering and characterization of ZF-based OR and NOR Boolean logic gates. In the sensory module, input signals lead to expression of corresponding ZF-based transcription factors. In the computational module, transcription factors act on response promoters. (A) OR gate response promoters contain target sites for two different ZF activators, and the logical operation is computed as TRUE (CFP expression) when either one or both inputs is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun activators were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control. (B) NOR gate response promoters contain the binding sites for two different ZF repressors, and the logical operation is computed as TRUE when neither input is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun repressors were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control.

Mentions: Within this framework, we began by generating response constructs that exhibited OR gate behavior. OR gates were developed by utilizing hybrid promoters consisting of various copies of binding sites for two distinct ZF DBDs. To determine the effect of binding site architecture on signal output, constructs containing either 2, 4 or 6 copies of the BCR_ABL and erbB2 binding sites were generated in multiple configurations (Figure 3A and Supplementary Figures S10 and S11). We characterize all of our Boolean logic systems using artificial inputs in which the CMV promoter drives expression of the computational proteins for a positive input or a negative control protein for a negative input. Each OR gate configuration was tested by co-transfection with either the BCR_ABL-1:GCN4 activator, the erbB2:Jun activator or both activators in tandem. All promoter architectures functioned as OR gates, with either a single activator alone or both factors present together resulting in signal output. The presence of both activators in tandem resulted in an additional 7-fold increase in output signal, likely due to the increased occupancy of reporter target sites by the transcriptional activators (Figure 3A). We also observed minor position effects, with activator constructs having corresponding binding sites more proximal to the TATA box displaying higher induction (Supplementary Figures S10 and S11). The varying outputs for different promoter architectures allow for an additional potential layer of tunability.Figure 3.


A tunable zinc finger-based framework for Boolean logic computation in mammalian cells.

Lohmueller JJ, Armel TZ, Silver PA - Nucleic Acids Res. (2012)

Engineering and characterization of ZF-based OR and NOR Boolean logic gates. In the sensory module, input signals lead to expression of corresponding ZF-based transcription factors. In the computational module, transcription factors act on response promoters. (A) OR gate response promoters contain target sites for two different ZF activators, and the logical operation is computed as TRUE (CFP expression) when either one or both inputs is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun activators were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control. (B) NOR gate response promoters contain the binding sites for two different ZF repressors, and the logical operation is computed as TRUE when neither input is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun repressors were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3367183&req=5

gks142-F3: Engineering and characterization of ZF-based OR and NOR Boolean logic gates. In the sensory module, input signals lead to expression of corresponding ZF-based transcription factors. In the computational module, transcription factors act on response promoters. (A) OR gate response promoters contain target sites for two different ZF activators, and the logical operation is computed as TRUE (CFP expression) when either one or both inputs is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun activators were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control. (B) NOR gate response promoters contain the binding sites for two different ZF repressors, and the logical operation is computed as TRUE when neither input is present. For the response data shown BCR_ABL-1:GCN4 and erbB2:Jun repressors were used as ZF-1 and ZF-2, respectively, and the response promoter contains six copies of the BCR_ABL target site upstream of 6 copies of the erbB2 target site. CFP expression was measured by flow cytometry and expressed as fold change over an off-target expression control.
Mentions: Within this framework, we began by generating response constructs that exhibited OR gate behavior. OR gates were developed by utilizing hybrid promoters consisting of various copies of binding sites for two distinct ZF DBDs. To determine the effect of binding site architecture on signal output, constructs containing either 2, 4 or 6 copies of the BCR_ABL and erbB2 binding sites were generated in multiple configurations (Figure 3A and Supplementary Figures S10 and S11). We characterize all of our Boolean logic systems using artificial inputs in which the CMV promoter drives expression of the computational proteins for a positive input or a negative control protein for a negative input. Each OR gate configuration was tested by co-transfection with either the BCR_ABL-1:GCN4 activator, the erbB2:Jun activator or both activators in tandem. All promoter architectures functioned as OR gates, with either a single activator alone or both factors present together resulting in signal output. The presence of both activators in tandem resulted in an additional 7-fold increase in output signal, likely due to the increased occupancy of reporter target sites by the transcriptional activators (Figure 3A). We also observed minor position effects, with activator constructs having corresponding binding sites more proximal to the TATA box displaying higher induction (Supplementary Figures S10 and S11). The varying outputs for different promoter architectures allow for an additional potential layer of tunability.Figure 3.

Bottom Line: We describe 15 transcriptional activators that display 2- to 463-fold induction and 15 transcriptional repressors that show 1.3- to 16-fold repression.The split intein strategy is able to fully reconstitute the ZF-TFs, maintaining them as a uniform set of computing elements.Together, these components comprise a robust platform for building mammalian synthetic gene circuits capable of precisely modulating cellular behavior.

View Article: PubMed Central - PubMed

Affiliation: Department of Systems Biology, Harvard University, Boston, MA 02115, USA.

ABSTRACT
The ability to perform molecular-level computation in mammalian cells has the potential to enable a new wave of sophisticated cell-based therapies and diagnostics. To this end, we developed a Boolean logic framework utilizing artificial Cys(2)-His(2) zinc finger transcription factors (ZF-TFs) as computing elements. Artificial ZFs can be designed to specifically bind different DNA sequences and thus comprise a diverse set of components ideal for the construction of scalable networks. We generate ZF-TF activators and repressors and demonstrate a novel, general method to tune ZF-TF response by fusing ZF-TFs to leucine zipper homodimerization domains. We describe 15 transcriptional activators that display 2- to 463-fold induction and 15 transcriptional repressors that show 1.3- to 16-fold repression. Using these ZF-TFs, we compute OR, NOR, AND and NAND logic, employing hybrid promoters and split intein-mediated protein splicing to integrate signals. The split intein strategy is able to fully reconstitute the ZF-TFs, maintaining them as a uniform set of computing elements. Together, these components comprise a robust platform for building mammalian synthetic gene circuits capable of precisely modulating cellular behavior.

Show MeSH