BioScaffold notation, {w,x;y,z}

Since other BioScaffolds beyond the circuit tuning and N-terminal protein engineering application (or for the same function (tuning circuits) where PsrI restriction enzyme sites or RFP reporters are present) might be desired, we created a notation to make it simpler to compare two BioScaffolds. We define the following notation {w,x;y,z} to describe in condensed form the positions of the cut sites relative to the BioScaffold: w is the position of the left cut on the forward strand, x is the position of the left cut on the reverse stand, y is the position of the right cut on the forward strand, and z is the position of right cut on the reverse strand. The position numbering begins on the right side of the BioScaffold part with 0 as the first position before TACTAGAK, where numbers increase to the right, and on the left side with -1 as the first position after the K of TACTAGAK, where numbers increase to the left--see Additional file 1 Table S1 for examples. The cut sites of the prototype BioScaffold are given by {0,5;15,10} (see Figure ​Figure22 where the excision positions are marked above the cut sites). When referring to a BioScaffold in BioBrick format, the notation {w,x;y,z} assumes that the BioScaffold is in its most common form when used with BioBricks (i.e., surrounded by a scar sequence on either side).

Testing the BioScaffold

As a demonstration, we positioned the prototype BioScaffold in a synthetic circuit between a promoter and GFP (Figure ​(Figure3).3). Replacement of the BioScaffold with an RBS caused this test circuit to become a GFP reporter (containing a promoter, RBS, and GFP) that expresses GFP within the cell (Figure ​(Figure4).4). Alternatively, replacement of the BioScaffold with a RBS-(maltose-binding protein)-(glycine-serine) (RBS-MBP-GS) sequence created a circuit that produced a MBP-GS-GFP fusion protein (Figure ​(Figure5)5) that is fluorescent green and binds to amylose resin. The prototype BioScaffold has been designed to contain specific cut locations on either side of its sequence (Figure ​(Figure2).2). After excision (Figure ​(Figure3),3), the BioScaffold will be replaced with one of several RBSs that are designed to drive well-defined levels of expression in the downstream gene, demonstrating how BioScaffolds can be used to facilitate the optimization of circuits. Alternatively, replacement of the BioScaffold with the MBP-GS fusion protein part will demonstrate how BioScaffolds can be used to create protein fusions. The visual markers used in the prototype system help track the presence of the BioScaffold (which contains an RFP reporter) and the performance of the optimization process (effect of different strength RBSs or a MBP-GS fusion on the GFP reporter). Figure ​Figure33 shows graphically the proposed replacement of the BioScaffold with repair oligonucleotides that contain an RBS sequence, which will affect the performance of the final GFP reporter circuit (Figure ​(Figure4).4). Figure ​Figure55 shows that the BioScaffold can successfully be replaced with an expressed MBP-GS fusion, which causes the final protein fusion to gain the property of affinity to an amylose column as well as maintaining the fluorescence of GFP.

Open in a separate window

Figure 4

GFP expression levels created by the different RBSs. Sequence verified clones of the form "promoter-scar-RBS-GFP" (a) were analyzed visually (b) and using flow cytometry (c). The inserted RBSs were expected to drive relative translation initiation rates of 10, 100, 1,000, 10,000, and 100,000 for the downstream sequence GFP with the cell fluorescence proportional to the translation initiation rate (a). Qualitative visual assessment revealed green fluorescent intensity commensurate with the expected values (b), although colony thickness can influence perceived intensity. Simultaneously transformed colonies of RBS10, RBS100, and RBS1,000 appeared light green, while RBS10,000 appeared green and RBS100,000 appeared bright green (b). Quantitative assessment using flow cytometry data revealed GFP intensity levels commensurate with expected values, except for the higher than expected translation initiation of GFP driven by RBS10 (c).

Open in a separate window

Figure 5

Properties of GFP with an N-terminal MBP-GS fusion. Lysate supernatants from cultures expressing GFP alone (a) or the MBP-GS-GFP fusion (b) were applied to amylose columns. The green fluorescent product was not retained on the "GFP" column when column buffer was applied, but rather appeared in the first wash eluate (c). In contrast the green fluorescent product was retained on the "MBP-GS-GFP" column when column buffer was applied (c). The protein purification process for MBP-GS-GFP was monitored using SDS-PAGE: Lane 1, cell lysate ("load"); Lane 2, first wash eluate ("wash"); Lane 3, maltose eluate ("elute") (d). Mass spectroscopy confirmed that the band at ~67 kDa was MBP-GS-GFP and identified the minor contaminant at the top of Lane 3 ("elute") as a dimer form. The dimer is known to be a normal occurrence with this construct [69,70].

Excision of the BioScaffold and replacement with the RBS and the RBS-MBP-GS sequences

Excision of the BioScaffold from the "promoter-scar-BioScaffold-scar-GFP" composite part and ligation of the RBS insert was expected to create a "promoter- scar-RBS-GFP" composite (where one RBS was inserted in each of five parallel reactions). The plated transformations of each reaction contained a mixture of red colonies and green colonies. It was assumed that the red colonies resulted from clones where the BioScaffold was not excised. At the optimal concentration of vector in the cutting reaction, less than 3% per 1,000 colonies were red (i.e., still contained the BioScaffold.) We picked 25 non-red clones for sequencing (5 for each RBS insert). For 80% of the non-red clones, sequencing showed that the RBS was inserted in place of the BioScaffold part at the correct location with the correct sequence. For 12% of the clones, sequencing showed that the entire test circuit was completely mutated out. For 8% of the clones, the sequencing result was noisy, preventing interpretation. We speculate that none of the errors are due to the cutting properties of the enzyme, but rather to selection against the circuit or inadvertent picking of multiple colonies. Using the same starting circuit, selection, and circuit verification protocol, ligation of the RBS-MBP-GS insert was expected to create a "promoter-scar-MBP-GS-GFP" circuit where the ATG start codon of the GFP was removed by the BioScaffold.

Expression levels of the RBSs

Colony and sequence verified clones that contain RBSs in place of the BioScaffold part were assessed qualitatively and quantitatively (Figure ​(Figure4).4). Visual analysis demonstrated that the clones exhibited a range of green fluorescent intensities: the lowest predicted expression levels appeared light green and the highest expression levels appeared bright green (Figure ​(Figure4).4). Visual analysis is only a rough guide to intensity since the number of cells present in a colony could influence its appearance. Overall, the flow cytometry data confirmed the results from [36] using our method to introduce the RBS. Flow cytometry data indicated that the predicted expression levels were commensurate with the actual intensity values except for sequence RBS10, which was expected to produce the lowest intensity of protein fluorescence (Figure ​(Figure4).4). The actual expression level for RBS10 was higher than expected; however, similar deviations are observed by Salis, et al. [36] for low RBS translation initiation strengths, and therefore appear to be unrelated to the use of the BioScaffolds. If it was important to create a circuit with low expression of GFP, we could have performed another optimization round using RBSs with similar but slightly different relative translation rates, such as 8, 9, 11, and 12.

Assembly of the BioScaffold test circuit: part BBa_J70423

The test circuit consists of a promoter, the BioScaffold, and GFP assembled into the circuit "promoter-scar-BioScaffold-scar-GFP." The assembly was performed in two rounds using the BioBrick Assembly kit (New England Biolabs, Inc.) for three antibiotic (3A) assembly. First, the promoter (BBa_R0010 in BioBrick vector pSB1A2), the BioScaffold part (BBa_J70399 in BioBrick vector pSB1AT3), and the destination vector pSB1AK3 were digested. The upstream part, the promoter (BBa_R0010), was digested with EcoRI and SpeI. The downstream part, the BioScaffold part (BBa_J70399), was digested with XbaI and PstI. The destination vector pSB1AK3 was disgested with EcoRI and PstI. The digests were mixed together and ligated with T4 ligase, to form a composite part (BBa_J70405 in the vector pSB1AK3). The ligation mixture was transformed into chemically competent TOP10 cells as above and plated on LB agar plates containing 100 μg/ml ampicillin and 50 μg/ml kanamycin. 5 colonies that appeared red were verified by colony PCR and sequencing as described below.

Next, the composite part (BBa_J70405 in BioBrick vector pSB1AK3), the GFP (BBa_E0040 in BioBrick vector pSB1A2), and the destination plasmid pSB1AC3 were digested and ligated in the same manner as above, where BBa_J70405 was the upstream part and BBa_E0040 was the downstream part, to form the test circuit composite part (BBa_J70423). The ligation mixture was transformed into chemically competent TOP10 cells as above and plated on LB agar plates containing 100 μg/ml ampicillin and 35 μg/ml chloramphenicol. 5 colonies that appeared red were verified by colony PCR and sequencing as described below.

Design of RBSs

The RBS calculator [71] was used to design the RBS sequences. The downstream GFP (BBa_E0040) was entered as the protein coding sequence. TACTAGAG was used as the presequence. Five different target translation initiation rates (10, 100, 1,000, 10,000, and 100,000) were entered into the calculator to generate the five different RBS sequences. The sequences generated by the calculator (see Table ​Table1)1) were then utilized to create forward and reverse oligonucleotides, taking into account the cutting profile of the BioScaffold. The forward oligonucleotide has the sequence "G-RBS sequence-ATGCGTAAA" and the reverse oligonucleotide has the sequence reverse complement of "CTAGAG-RBS sequence-ATGC." After the replacement of the BioScaffold with the annealed oligonucleotides, the downstream scar disappears, while the upstream scar remains.

Design of the RBS-MBP-GS insert

The RBS-MBP-GS insert was created using the plasmid pMAL-p5e (New England Biolabs) as a template and the primers RBS-MBP-GS-GFPstart-f (5'-gaattcgcggccgcttctagagacgaacgctctctactagatgcctagagtcgccccctaagggcggaggtaggagaaactcaaatg aaaatcgaagaaggtaaactggtaatctg -3') and RBS-MBP-GS-GFPstart-r (5'-ctgcagcggccgctactagtagtaatatatgttcgatagattttacgagaaccagtctgcgcgtctttcagg -3') using the same PCR and gel extraction protocol as for BBa_J70399.

Excision of BioScaffold part from the test circuit and replacement with RBSs

The oligonucleotides containing the RBSs were prepared by combining 8 μl of 100 μM forward oligonucleotide, 8 μl of 100 μM reverse oligonucleotide, 10 μl annealing buffer (100 mM Tris HCl pH 7.5, 1 M NaCl, 10 mM EDTA), and 74 μl milliQ water. They were then annealed by heating to 95°C for 2 minutes, ramping from 95°C to 25°C over 45 minutes, and then cooling to 4°C. We diluted 10 μl of oligonucleotides into a final volume of 1000 μl.

We digested the test circuit by combining 100 ng DNA, 1× Buffer Y (SibEnzyme, Inc.), 100 μg/ml Bovine Serum Albumin (SibEnzyme, Inc.), and 0.5 μL PsrI (SibEnzyme, Inc.) into a final volume of 50 μl. The restriction digest reaction was incubated for 1 hour at 30°C followed by 20 minutes at 65°C. In our experience digestion of more DNA dramatically increases the number of undigested products yielding red colonies upon transformation.

We performed the ligation by combining 5 μL of the PsrI digestion reaction (10 ng DNA), 0.2 μL of the diluted annealed oligonucleotides, 1× T4 DNA ligase reaction buffer, 200 units of T4 DNA ligase into a 20 μL total volume and cooling to 18°C for 30 minutes. The ligation mixture was transformed into chemically competent TOP10 cells as above and plated on LB agar plates containing100 μg/ml ampicillin and 35 μg/ml chloramphenicol. 5 non-red colonies per reaction were verified by colony PCR and sequencing as described below.

Replacement of the BioScaffold with the RBS-MBP-GS insert to create BBa_J70631

Both the test circuit and the gel purified RBS-MBP-GS insert were digested with PsrI as described above in separate reactions. We performed the ligation by combining 5 μl of the test circuit digestion reaction, 5 μl of the RBS-MBP-GS insert digestion reaction, 1× T4 DNA ligase reaction buffer, 200 units of T4 DNA ligase into a 20 μl total volume and cooling to 18°C for 30 minutes. The transformation and sequencing of BBa_J70631 occurred as described previously for a test circuit containing an insert.

Verification with colony PCR and sequencing

Colony PCR and subsequent gel electrophoresis were performed according to [12] except that the BioBrick primers BioBrick-f (BBa_G1004) and BioBrick-r (BBa_G1005) were used. The PCR protocol was the same as described for the BioScaffold part above, except that the extension step at 68°C occurred for 3.5 minutes instead of 2.5 minutes. The Massachusetts Institute of Technology Biopolymers Laboratory performed DNA sequencing, using the verification primers VF2 (BBa_G00100) and VR (BBa_G00101).

Flow cytometry measurement of final circuit fluorescent intensity

Expression data were collected using a Becton-Dickinson FACSCAN flow cytometer with a 488-nm argon excitation laser and a 515-545 excitation filter [72]. Cells were grown in M9 media [73] (1× M9 salts, 2 mM MgSO4, 0.5% glycerol, 0.2% casamino acids, 2 mM thiamine) with ampicillin and chloramphenicol. The cultures were grown according to [74], except that the samples were measured in M9 medium [73].

Purification of the MBP-GS-GFP protein using an amylose column

The BBa_J70631 construct, which contains a MBP-GS-GFP expression circuit, was expressed in E. coli Top10. The culture was grown overnight at 37°C in rich media (10 g typtone, 5 g yeast extract, 5 g NaCl, 2 g glucose) containing 100 μg/ml ampicillin and 35 μg/ml chloramphenicol, expanded 1:100 the next day and grown for 3 hours after the OD₆₀₀ reached 0.6. The BBa_J04430 construct, which contains a GFP expression circuit, was grown in the same manner except the media did not contain chloramphenicol. Purification of a 40 ml culture of each construct was performed according to [75] using amylose resin (New England Biolabs), expect that the GFP and MBP-GS-GFP lysates were diluted until they exhibited equivalent fluorescent intensity and an equivalent volume of each cleared lysate (containing less MBP than the binding capacity of the amylose resin) was applied to each column.

SDS-PAGE of MBP-GS-GFP

Samples were run on a 4-12% Bis-Tris gel (Invitrogen) and stained with Coomassie blue or SimplyBlue (Invitrogen) stain [76]. The contrast of the image was adjusted uniformly in Adobe Photoshop to simplify the visualization of the bands.

Mass spectrometry of MBP-GS-GFP

The lanes from a SimplyBlue stained SDS PAGE gel were excised and submitted for analysis to the Proteomics Core Facility of the Koch Institute for Integrative Cancer Research at MIT. The gel bands were subjected to in-gel protein digestion with trypsin, following standard protocols. LC-MS/MS analyses were carried out using a nanoflow reversed phase HPLC (Agilent) and an LTQ ion trap mass spectrometer (Thermo Electron). Protein identifications were carried out by database search using Sequest software (Thermo Electron) against an E. coli protein database, generated from the Uniref100 protein database. The protein sequence of the MBP-GS-GFP fusion protein was added to the E. coli protein database.