Gibson vs SLiCE (single part recommendations)

[Koeng]

Recently I've begun using SLiCE for cloning and have got some of my labmates to jump aboard. From about 1-2 weeks of trials, here's what I've found with single part SLiCE assembly. In the future, I hope to try 1. PEG8000 to increase SLiCE efficiency and 2. Multipart assembly 3. using homemade comp cells.

Important stuff: SLiCE is approx. 1/3 the efficiency of homemade gibson assembly. Gibson increases efficiency with additional homology while SLiCE decreases. The optimal homology length for gibson is ~40 while the optimal homology length for SLiCE is ~20. SLiCE recommended time is 15 minutes @ 37c while gibson is 1 hour @ 50c. SLiCE mix can be produced cheaply with Triton-X100/Tris-HCl buffer, the only solution required.

Experiments: I produced the SLiCE enzyme mix using the protocol outlined in (1). I took them out at about OD 2.7 because I was getting tired of waiting for the cells to grow. If you put the cells in fairly late at night, you might be able to get them in the morning. An important point that I have not verified is that efficient SLiCE without lambda red requires refrigerated centrifuge during purification.

Immediately after making the mix, I did a single part cloning with ~30bp homology with gave me 1/3 efficiency of 13 colonies (SLiCE) vs 42 colonies (homemade gibson) with both at their optimal temperature of 37c and 50c for 30 minutes. When I sequenced the DNA, I found no mutations within the overlap of either one of the constructs, both being completely correct (only sequenced 2 from each and got proper sequence data)

Some other people in lab tried the SLiCE mix and found that it gave roughly 1/2 to 1/4 and 1/3 the efficiency in 2 separate cloning experiments using the same amount of DNA in both coloning reactions (10µl total for SLiCE and 20µl total for gibson with ~100ng of insert for both) and transformed 5µl of that. That means that they transformed twice as much DNA for SLiCE, (per ng) but still got ~1/3 the efficiency. They also used 30 minutes for SliCE, which is optimal (2), but 1 hour for gibson.

I did another experiment that was more controlled. I used 100ng of vector and 200ng of insert for this cloning. Reaction times were 37c for 15 minutes for SLiCE and 1 hour at 50c for gibson. The reaction volume was 10µl for SLiCE and 20µl for gibson and I transformed 5µl. I was cloning a CDS, and decided to test overlap length. The design would be [vector - [flank 1] - [insert] - [flank 2] - vector]. Flank 1(SLiCE) had a tm of 58c overlap of 19bp, while Flank 2(SLiCE) had an overlap length of 24 and a tm of 55c. Optimizing the other side, I had 2 flanks which simply used up the entire 60bp possible because empirically that increased the efficiency of gibson. Flank 1(gib) had a tm of 71c overlap of 41bp, while Flank 2(gib) had an overlap length of 37 and a tm of 61c. When the SliCE optimized flanks were used for SLiCE, I got 28 colonies while when I used the SLiCE optimized flanks for gibson I got 35 colonies. Remarkably, when I used the gibson optimized overlaps for SLiCE I got only 3 colonies while when I used the gibson optimized overlaps for gibson I got 64 colonies! It seems like going over a certain length can actually hurt your SLiCE reaction. More is not always better. This data matches (2) fairly well. 

I’m going to do more experiments, so I’ll keep this post updated in case there is interest. I plan on doing the above experiments soon. Note that I did do all this cloning with lab made comp cells using the zymo kit, https://www.zymoresearch.com/e-coli/transformation-kits-accessories/transformation-kit-buffer-set , so I don’t know how good this would be with homemade cells.

-Koeng

(1) http://www.sciencedirect.com/science/article/pii/S2405580815000850

(2) http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4453199/

[burnrome]

Interesting - keep us updated! Haven't had much luck with homemade SLICE from JM109s using Celllytic.

[Nathan McCorkle]

On Mon, Nov 16, 2015 at 7:06 AM, Koeng <koen...@gmail.com> wrote:
> Recently I've begun using SLiCE for cloning and have got some of my labmates
> to jump aboard. From about 1-2 weeks of trials, here's what I've found with
> single part SLiCE assembly. In the future, I hope to try 1. PEG8000 to
> increase SLiCE efficiency and 2. Multipart assembly 3. using homemade comp
> cells.
>
> Important stuff: SLiCE is approx. 1/3 the efficiency of homemade gibson
> assembly. Gibson increases efficiency with additional homology while SLiCE
> decreases. The optimal homology length for gibson is ~40 while the optimal
> homology length for SLiCE is ~20. SLiCE

Does you last sentence mean SLiCE won't work with more homology? Have
you verified this?

[Koeng]

Try Triton X100! Worked very well for me. 

It does work, just with severely compromised activity (at 40bp of homology). I haven't tested in between 20 and 40, however.

-Koeng

[Koeng]

(meaning I got ~10x more colonies with 20bp vs 40bp)

On Tuesday, November 17, 2015 at 11:12:27 AM UTC-8, Nathan McCorkle wrote:

[Bryan Jones]

I've also been doing some SLiCE experiments with SLiCE made from DH5alpha E. coli with and without pKD46 plasmid that I got from the-odin.com. SLiCE was used to successfully ligate two pieces of DNA together. The efficiency was fairly poor when using SLiCE from E. coli with pKD46 (12 colonies from 108ng of DNA, where the commercial kit NEBuilder produced ~500 colonies). The SLiCE from E. coli without pKD46 produced many more tranformants, but some/many of these seem to be original vector that was able to religate after being cut (vector was neither gel purified nor phosphatase treated, which I have found to be unnecessary when doing Gibson/NEBuilder assemblies). Consistent with this, no SLiCE controls also contained many colonies that sequenced as the original re-ligated plasmid. Spontaneous re-ligation seems to occur frequently, yet the lambda exonuclease in pKD46 (and in NEBuilder) seems to prevent religation of digested vector, yet it is not clear from these results that pKD46 aides in efficiency of ligation.

Full experimental details and results/pictures are in this Google Doc.

[Alexandra Johnson]

I think that pKD46 requires arabinose for expression of the lambda proteins, and I didn't see it in your protocol. Maybe that's part of the reason for the low efficiency? 

[Cihan AYDIN]

Just learned about the plethora of information starting from Gibson assembly and came across this post.

If this is not you, someone have already tried the PEG8000 approach and pervailed :-) http://openwetware.org/wiki/SPLiCE

What I wonder is however this protocol also do not mention about inducing the pKD46 plasmid, is it expected (kind of D'uh! moment?) or it just leaks enough to make a difference (since it is araB, one should not expect leakage).

[Koeng]

Yep, I saw that, that's how I got the idea to try it :) Unfortunately haven't gotten the time to try it with PEG8000 yet

In my protocol I used JM109 cells without pKD46. It shouldn't be needed, and I don't know if expressing it helps. 

-Koeng

[Cathal Garvey]

Just an FYI: you can often substitute PEG3350 for PEG8000 in protocols. Just watch for molarity, substituting it gram-for-gram is probably unwise. Someone with a better grasp of how PEG affects osmolarity might chip in, here. :)

In cases where you *can* substitute, it's really handy, because PEG3350 is sold as an over-the-counter laxative under various brand names. In the US, it's known as Miralax, and I've successfully used it with other protocols that called for PEG3350.

-- @onetruecathal / @cat...@quitter.is

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/e9de1efd-ef20-4985-9520-a5aabbb8560a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[John Griessen]

On 10/19/2016 12:55 PM, Cathal Garvey wrote:
> Just an FYI: you can often substitute PEG3350 for PEG8000 in protocols. Just watch for molarity, substituting it gram-for-gram is
> probably unwise.

Isn't the number after the PEG acronym for length or molecular weight? Its property of interest is absorbing/adsorbing water, so
the much bigger molecule at the same number of moles would turn a broth into a pudding. Right?

[Cathal Garvey]

Very fair point, yes! This is outside my comfort zone, in terms of calculating the impacts. It's an observation I've made that many protocols allow substitution, but honestly I can't remember whether they specify different molarities.

It would seem to me that yes; longer polymers would mean more soupy buffers. But the difference between PEG50 and PEG3350 is probably orders of magnitude larger than the difference between PEG3350 and PEG8000?

Anyone know good back-of-envelope ways to calculate the relative effects of polymers by chain length? :)

-- @onetruecathal / @cat...@quitter.is

--

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en Learn more at www.diybio.org --- You received this message because you are subscribed to the Google Groups "DIYbio" group. To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com. To post to this group, send email to diy...@googlegroups.com. Visit this group at https://groups.google.com/group/diybio. To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/7f7b1f9e-1cab-b5f5-5b43-25a9d460e1d7%40industromatic.com. For more options, visit https://groups.google.com/d/optout.

[Bryan Jones]

I haven't tried with the PEG. I recently did another side by side comparison with SLICE made from DH5a cells without pkD46, with pkD46 uninduced and with pkD46 induced. This was just a single shot, so not significant results, but the pkD46 made a big difference, but it didn't matter a much whether it was induced or not. Both SLiCE mixes made with pkD46 gave me about half as many colonies as Gibson Assembly. The reaction was a single gene insert into a plasmid (~1kb and ~5kb) with 25bp overlap on either end.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/1476948978.1963.4%40mail.1984.is.

[Koeng]

What are the numbers like with and without the pKD46 plasmid?

[Nathan McCorkle]

On Oct 19, 2016 11:21 AM, "John Griessen" <jo...@industromatic.com> wrote:
>

> Isn't the number after the PEG acronym for length or molecular weight?  Its property of interest is absorbing/adsorbing water,

I've heard it can also be used for 'molecular crowding'. The longer the molecule, the thicker the molecular crowd in that area of volume, because the more would-be monomer can't diffuse from the rest of the polymer.

[Bryan Jones]

In my most recent test, without pKD46, I got 10 colonies from the SLiCE reaction, which was the same as the no SLiCE control. SLiCE with pKD46 resulted in 50-60 colonies. For comparison, NEBuilder resulted in about 100.

--

-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2B82U9Kq2VqO_yxZb8dgfGu6ox53FOZg%3DMKsAaDaeUDmVCLueg%40mail.gmail.com.

[RBC]

Is there any worry that the amp resistant pKD46 plasmid will be extracted into the SLiCE lysate, in which case this plasmid will carry through to the final transformation? To me it would seem you can't use it to make amp resistant vectors..or maybe I'm missing something. Thanks for the clarification.

RC

[Bryan Jones]

I would imagine there is some risk of that, but I haven't seen that happen in my experiments. I carried out a negative control using just SLiCE extract with competent cells (No insert DNA), and didn't get any colonies on amp plates. Also the colonies I sequenced from my tests were all correct, no pKD46 carryover. It might be that DNAse breaks all the DNA from the SLiCE cells. Normally when you do a plasmid prep, you add EDTA to prevent this, but no EDTA was added here. I'm not positive why, but empirically, I can say that it does not seem to be an issue. 

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/ed3ec371-5de0-49e0-a8e5-db6dbdae3fda%40googlegroups.com.

[RBC]

Bryan, thank you very much for sharing your experiences. Sounds like you did all the right controls to get a proper assessment of the system. Logically it is a little surprising that there's no carry over of plasmid, but of course very good news and just goes to show the best way to answer a question is to just do the experiment. I tend to see too many people over thinking things to a point where they will convince themselves something is not feasible before they've even tried it, so the project stops before it even begins. :)

Cheers.

[Rob C]

Not too long ago we obtained the PPY strain as a kind gift from the original makers (Zhang at Albert Einstein). 

In our hands making extracts from this strain works WAY better than traditional DH10b, DH5alpha, etc (using the CelLytic lysis B reagent...it's cheap for what you get). 

Although I don't have exact numbers to report, we have done several tests side by side with Gibson, and PPY based SLICE gives at least equal transformation efficiency. Accuracy is equally as good, and in some isolated cases it has been better. 

After several month of heavy cloning, our lab has now dropped purchase of all Gibson reagents. 

I'm convinced that if more people implement the PPY-based SLICE cloning in their lab, Gibson sales will go out of business. 

[Cihan AYDIN]

I also got an MTA for the PPY strain but the shipping costs (for cold chain) was too much for me so I opted for getting a pKD46-deriven plasmid (pREDIA from Addgene). The only difference is in PPY gam and redb is integrated into the DH10B genome and reda in plasmid whereas in pKD 46 it is on the plasmid. I really would like to make a side-by-side comparison for the PPY and pKD46/DH10B in my lab.

Also there is this japanese guy Motohashi who does SLiCe from regular bacteria extracts (recA-) and has data on its efficiency with 15-19 bp overlaps between parts (https://link.springer.com/protocol/10.1007%2F978-1-4939-6472-7_23). He is using JM109 as his primary strain and uses a few more alternatives as far as i remember.

On the note, Gibson is too pricey for any lab who's on a budget, especially DIY labs and labs in developing countries. SLiCe and even CPEC has become better alternatives for synthetic biology and even simple cloning reactions - basically your only cost is bacterial reagents and primers (and PCR reagents in some cases).

On a last note, also read this article - www.ncbi.nlm.nih.gov/pubmed/26463009 - where they combine CRISPR with Lambda phage recombination system to efficiently modify e. coli genome.

Cheers!

C

[Rob C]

Thanks for your input Cihan. I would also be particularly curious to see how the plasmid-based recombinase extract works compared to PPY/genome-based.

The Motohashi papers on using traditional extracts are quite interesting. Thanks for sharing. I've read through his original paper (https://bmcbiotechnol.biomedcentral.com/articles/10.1186/s12896-015-0162-8) and he certainly gets very different numbers for non-recombinase extract than Zhang did, in both cloning efficiency (up to 2-orders of magnitude better) and ideal homology overlap length (20bp compared to 50bp from Zhang). Reasons for the discrepancy are not well discussed. The reported cloning accuracy when combining two inserts into a vector was kind of low as well. In our lab now with the PPY strain we are putting in 1-4 fragments at a time with nearly 100% accuracy, and with good enough efficiency to assemble large libraries. 

This is not to say the methods of using traditional cell extracts don't work well or aren't perfectly suitable for many labs and many applications...I suppose my point is that I am a believer at this point that making extracts from cells expressing recombinase (whether PPY or plasmid based) really does make a difference. And when students clone more efficiently, projects move forward faster. 

One of these days I'll get around to doing a more systematic side-by-side comparison of using extracts with and without recombinase.

Just my two cents. ;)