Introducing six bps into dLight— To test the ability of our modified method for insertion of sequences, we opted to insert six bps (TGAATG) at position 50 or 120 at the first and second intracellular loop (ILI or ILII, respectively) of a fluorescent dopamine probe (dLight16). This insertion should translate into a stop codon followed by a methionine (Stop-Met). We designed SDM primers that containing the insertion (Table 1; bold) and tested our procedure to introduce these changes after having repeatedly failed to do so by standard SDM approaches in multiple trials. In fact, these failures could not be remedied by systematic modifications of annealing temperatures, steps’ durations, number of cycles, added reagents (e.g., DMSO), and commercial kits (see Suppl. Text).
Table 1
List of the primers employed for this work.
Primer name
|
Primer sequence (5’-3’)
|
ILI_SDM_F
|
GGTCTGTGCTGCCGTTATCTGAATGAGGTTCCGACACCTGCGG
|
ILI_SDM_R
|
CCGCAGGTGTCGGAACCTCATTCAGATAACGGCAGCACAGACC
|
ILII_SDM_F
|
CCTCTGTGTGATCAGCGTGTGAATGGACAGGTATTGGGCTATCTC
|
ILII_SDM_R
|
GAGATAGCCCAATACCTGTCCATTCACACGCTGATCACACAGAGG
|
ILIII_R
|
CGTTAATGAGTGAGCTCAGCATTCACTGTTTCTGAGCAATCCTG
|
hSyn promoter_F
|
CGCACCACGCGAGGCGCGAGATAGG
|
FP_R
|
CTTGTACAGCTCGTCCATGCC
|
hChR_203R_F
|
GCATATATCGAGGGTTATCATACTAGGGTGCCAAAGGGTCGGTGCCGCCAG
|
hChR_203R_R
|
CTGGCGGCACCGACCCTTTGGCACCCTAGTATGATAACCCTCGATATATGC
|
CAG promoter_F
|
GCAACGTGCTGGTTATTGTG
|
hChR_203All_F
|
GCATATATCGAGGGTTATCATACTTGANNNGTGCCAAAGGGTCGGTGCCG
|
hChR_203All_R
|
CGGCACCGACCCTTTGGCACNNNTCAAGTATGATAACCCTCGATATATGC
|
We first amplified the sequences flanking the site of insertion (sites A and B) by re-using the complementary SDM primers, even though these have failed by standard approaches (see Suppl. Text for details), with general primers from our inventory. Specifically, for amplification of fragment A, we employed the antisense SDM primer (ILI_SDM_R) along a sense primer annealing to the promoter of the plasmid (hSyn promoter_F) (Fig. 2a and Tables 1 and 2). Similarly, amplification of fragment B was obtained by the sense complementary SDM primer (ILI_SDM_F) and a standard antisense primer which we have previously used (ILIII_R) (Fig. 2a). For an identical insertion at ILII (flanked by sites B and C), we employed the same strategy; combining ILII_SDM_F/R with another general-use primer that anneals to the 3’ terminal of GFP; a sequence that is shared by many other fluorescent proteins (FPs)17 (Fig. 2a, FP_R, green, Tables 1 and 2). A standard amplification protocol yielded two sets of products at very high amounts (and rapidly, <1 hrs, see methods; Tables 2 and 3) and these could easily be distinguished and quantified on 1% agarose gel (~300 and ~550 bps, ~500 and ~650 bps for Parts A, B and C of ILI or ILII, respectively) (Fig. 2b). To assemble the fragments (A with B, and B with C), we used the Gibson Assembly mix, even though the extent of complementation between our fragments (obtained by the standard SDM primers) does not meet the requirement of Gibson primers (Suppl. 1)12. Therefore, we assumed that longer DNA-excision times by the T5-exonuclease may be required to remove matching sequences between the fragments, that would otherwise prevent the fragments from annealing to one another. In fact, the fragments need to undergo extensive digestion past the entire sequence of the primer to enable the ligation (Suppl. 1). We placed the isolated PCR products from step I within the Gibson reaction mixture for prolonged incubation (two hours; methods). Notably, the expected product of this assembly (~800 or ~1150 bps for ILI and ILII, respectively) could not be visualized on 1% agarose gel (see example below in Fig. 4c). We therefore could not assess whether the reaction succeeded and, if it didn’t, which step was faulty (for instance, whether the digestion of the overlapping sequence was insufficient). This did not meet our primary goal of providing added quality-control checkpoints throughout the process. We therefore opted to try to detect the potential Gibson-ligated product by amplifying it using primers employed in step I, namely sense hSyn promoter_F with the ILIII_R or FP_R antisense primer, for ILI or ILII, respectively. Importantly, we chose these pairs of primers as they can only amplify the ligated product, if extant in the tube. This rapid PCR reaction (1 hr) yielded easily detectable amplicons of the correct size (Fig. 2c, step III; ~1 Kbp). Next, amplicons and plasmid were digested, ligated overnight followed by transformation and plating (Fig. 2d). We isolated DNA from several colonies and visualized them on 1% agarose gel. Though a handful of colonies did not contain the right plasmid (‘negative’ colonies), all of the ‘positive’ colonies contained the desired substitutions (Fig. 2e). Together, we found that we could easily introduce the desired six bps at two distinct DNA regions after only four days by the Gibson method without Gibson primers.
Table 2
Reaction mixture for the KAPA HiFi HotStart ReadyMix PCR Kit.
Component
|
Step I
|
Step I
|
2X KAPA HiFi HotStart ReadyMix
|
12.5 µl
|
12.5 µl
|
Template
|
1 µl (10 ng)
|
1 µl (Gibson mix)
|
SDM_Primer_F (10 µM stock)
|
0.75 µl
|
0.75 µl
|
SDM_Primer_R (10 µM stock)
|
0.75 µl
|
0.75 µl
|
ddH2O
|
Up to 25 µl
|
Up to 25 µl
|
Table 3
PCR settings for KAPA HiFi HotStart ReadyMix PCR Kit
|
Step I
|
Step III
|
Step
|
Cycles
|
Temperature/ Duration
|
Cycles
|
Temperature/ Duration
|
Initial denaturation
|
1
|
95°C /3 minutes
|
1
|
95°C /3 minutes
|
Denaturation
|
30
|
98°C /20 seconds
|
30
|
98°C /20 seconds
|
Annealing
|
60/56°C /20 seconds
|
60/56°C /20 seconds
|
Extension
|
72°C /1 minutes
|
72°C /2 minutes
|
Final extension
|
1
|
72°C /5 minutes
|
1
|
72°C /5 minutes
|
Deletion and replacement of residues in ChR2 using SDM primers— We were interested in testing whether the modified procedure could support a slightly more challenging procedure, namely to remove six bps and replace them by three other bps (‘TGAATG’ to ‘AGG’) at the ILIII (residue 203) of a humanized Channelrhodopsin2-mCherry variant (hChR2-mCherry). We intentionally designed two standard complementary sets of SDM primers (51 bps each, at 53% GC content) instead of Gibson primers (Table 1). Again, under standard SDM conditions (and by use of different commercial kits), we could not obtain the final product (see Suppl. Text). We, therefore, similarly amplified the two segments of hChR2 using each of the SDM primers in two separate PCR reactions, with a standard sense primer annealing to promoter (CAG promoter_F), and a general antisense primer annealing to FP (FP_R) (Figs. 3a); yielding correct amplicons sizing at ~700 and ~1040 bps (Part A and B, respectively) (Fig. 3b; step I).
Fragments were assembled by the Gibson reaction mixture but, this time around, at various incubation times (15 min to 2 hours), immediately followed by PCR amplification of the potential ligated fragment by use of CAG_promoter and FP_R antisense primers (1.5 hrs, see Tables 2 and 3) (Fig. 3c). Surprisingly, even the shortest Gibson assembly reaction (15 min) yielded the expected ligated product, easily visualized on 1% agarose gel (Fig. 3c, ~1.8 Kbp). We then added another quality control step by sequencing the amplified ligation-product, a feat that requires a sufficiently large amount of product as obtained here. Indeed, we find the desired changes in the DNA the following day in the sequencing results (Suppl. 2a). Then, insert and plasmid were digested, ligated, transformed, and plated (methods). We sequenced DNA from only three colonies and find the desired modifications in all three (Fig. 3d).
Gibson assembly using degenerate primers to evolve a single residue in ChR2—
We next examined whether we could evolve a single residue within the third intracellular loop of hChR2-mCherry (residue M203). We designed degenerate primers targeting three bps for evolution (i.e., a mixture of 64 different primers, each 50 bps long; Table 1). Here too, standard SDM conditions repeatedly failed in over 20 different trials (Suppl. Text). We then applied our procedure to amplify the DNA flanking the site to be mutated using the degenerate primers separately, combined with sense CAG promoter_F and antisense FP_R primers (Figs. 4a). Standard amplification (1 hr, see Table 3) yielded the expected amplicons (~700 and ~1040 bps for Part A and B; Fig. 4b, Step I). Again, the product of the Gibson-assembly of the fragments could not be visualized on gel (Fig. 4c; Step II, dashed region) without amplification (Fig. 4c; Step III, arrowhead, ~1750 bps).
The next day (following digestion, ligation and bacterial transformation), we isolated DNA from 38 colonies, all of which contained different mutations at the desired site (two colonies contained a mixture of two DNAs); thereby yielding >95% efficiency (Fig. 4e, dashed regions). Thus, we have rapidly evolved residue M203; resulting in a small library of 16 different amino acid substitutions. Interestingly, though beyond the scope of this work, we noted that proline was the most abundant substitution and that none of the colonies contained the original M203 (either from residual template DNA or by mutagenesis) (Suppl. 3).