Comments/FAQ

Please request the reagents from Addgene and from the CGC (see links from reagent pages). The MTAs that University of Utah require are a nightmare for our administrative assistant to handle.

Yes, please see the list of additional insertion sites under MosSCI – Insertion strains.

Great idea! The more insertion sites the better.
1) Pick out an appropriate Mos1 allele on wormbase. We picked Mos1 insertions located between genes that were expressed based on EST data. You might have other/better criteria.
2) Order the allele from the NemaGENETAG consortium.
3) Cross Mos1 insert into appropriate background. We used unc-119(ed3). You may want to use another rescue marker.
4) Build cloning vector with approx. 3 kb homology to genomic insertion sites (centered on Mos1 insertion site).
5) Test for insertion and expression.

We don’t think there is anything magic about the ttTi5605 site, although it does seem to be particularly well behaved. Both of the initial sites we tried (ttTi5605 and cxTi10882) generated insertions at similar frequencies. We developed a new set of insertion sites and all tested sites worked for insertions. However, transgenes behave slightly different at different sites, particularly for germ line expression.
Please let us know if you notice obvous differences between any two sites.

We have used the following primer sets to outcross the Mos1 insertions. We mostly do a PCR that spans the Mos1 element (1.3 kb long) to determine homozygosity. In a few cases we’ve also verified the orientation of the Mos1 transposon by using an internal primer oJL103 (5′ – TCTGCGAGTTGTTTTTGCGTTTGAG -3′) which anneals inside the Mos1 element.

Site	Forward oligo	Reverse oligo	PCR fragment (no Mos1)	PCR fragment (w Mos1)	PCR w forward oligo and oJL103
ttTi5605	TTTCTCAGTTGTGATACGGTTTTT	CGCTACTTACCGGAAACCAA	0.6 kb	1.9 kb	0.6 kb
cxTi10882	ACCTTCCAATCCGCCATATCC	TCCCCATTTCACCAGAGAAC	0.9 kb	2.2 kb	1 kb
cxTi10816	GGCATTTGATGCGATGAGTA	CAGTAGGGCCCGTGTAAAAA	0.4 kb	1.6 kb	not tested
ttTi4348	CTCTCGTCTCTCCACGATTTACAC	GGCGCAAGAACTGCGATTAG	0.5 kb	1.8 kb	not tested
ttTi4391	TCCCTACAGTATCCCTACAGC	ACCCGTTCAGAATATACCCAG	0.4 kb	1.7 kb	not tested
ttTi14024	TCGAACACGTACATGACGACTG	CAACGTGCCAGTTGTTGACT	0.3 kb	1.6 kb	not tested

Hmm – it’s the most frequent question but also the most difficult to answer. Although many labs have told us they have good luck with the technique, then there are many things that could make insertions difficult: Your transgene could be toxic, the DNA you are injecting is not clean enough, you have limited experience with injections, you don’t like injecting into unc-119 animals…

First: Are you following the protocol we published? We are sorry to say this but most often when the technique is not working it is because one or more steps of the protocol have been changed. There are undoubtedly still aspects that can be improved (please drop us a line if you make improvements!) but for the most part the protocol works pretty well. One key part of the protocol for direct injections is to not pick F1 animals off the injection plates onto other plates. Some are used to doing this when they make “regular” transgenic lines and have kept that habit. Although we don’t have conclusive experimental data ourselves it really seems that the direct injection does not work (or at least works very poorly) if you pick rescued F1 animals. We have now heard from several labs that tried picking rescued F1 worms and it has not worked for them. If they don’t pick F1 animals, then the protocol works in their hands, too. It’s possible that MosSCI inserts often come from animals that are not phenotypically rescued and you would then throw these events out by picking F1 rescued animals.

Let all the injection plates (which each started with one individual injected animal) starve out completely. Don’t distribute F1s – forget about the plates until starved. It’s ok to put the plates at 25C. Actually, it appears that it may improve insertion frequency. Once the plates are starved, then screen the animals by fluorescence for animals that move like wild type animals but lack fluorescent markers. Then clone out 8-10 individual, non-fluorescent animals to individual plates. Several of these plates will typically breed true and have homozygous animals.

Second: It seems like the number of transformed F1 progeny per injected animal is the most important parameter for successful insertions. We get between 10-15 visibly rescued F1 animals per injected animal. Injecting young adults kept at 15C increases the survival rate and number of transformed F1 progeny. Also, growing animals on comamonas bacteria increases the brood size and health of unc-119 animals. Placing the animals at 25C after injection seems to be fine, though, and speeds up the process considerably.

If you are getting a good number of rescued F1s but no insertions, then we suggest you try the positive control plasmid pCFJ68 (Punc-122::GFP) which we’ve used to characterize the insertions sites with.

It may be more important than we previously thought. We are currently testing this but it seems that higher quality DNA (from a midi-prep kit, for example) gives higher insertion frequency than miniprep DNA. See the worm breeder gazette article from Morris Maduro. We have had several people tell us that following his protocol improved their insertion frequency.

The largest transgene we’ve inserted is 16 kb. We did not get successful insertions of fosmid sized transgenes (Frøkjær-Jensen and Sarov, unpublished).

Single copy insertions are dim. Some insertions are visible on a fluorescent dissection scope (Prab-3::mCherry and Pmyo-2::mCherry::H2B, for example) whereas others are only visible at 63x magnification on a compound microscope (Punc-122::GFP, for example). Your gene may be expressed at low levels… sometimes arrays are nice.

It’s not yet clear how well the different promoters express in the germline. If your promoter has a lot of complex sequences (simple or complex repeats) it may be disrupting the DNA repair machinery. See if there is anything on the germline page that can help you out.

Also, did you wait for 3-5 generations after getting the insert? Hereditary silencing from the injection carries over for several generations but sometimes goes away. Re-check you inserts every couple of generations.

For a more thorough characterization of silenced MosSCI inserts see this paper from the Mello lab.

Jonathan Hodgkin has determined that MosSCI inserts should have a new designation “Si”. So, transgenes integrated by MosSCI are designated by italicized names consisting of the laboratory allele prefix, the two letters Si, and a number.

For example, inserts from the Jorgensen lab will be named oxSi31[Punc-122::GFP cb-unc-119(+)] II.

Mike Nonet sent us the following advice from his lab: “We have been growing the EG5003 and EG4322 strains on DA1877 Comamonas bacteria at 15C and under these conditions the strains grow larger, more healthy, have larger brood sizes and larger gonads for injections.” In our own hands, we find that HB101 bacteria is preferable. The animals are healthier than on OP50 bacteria but the HB101 strain is much easier to grow and maintain. Also, the animals are healthier when grown at 15 – 20C.

If you would like to verify that the insertion is at the correct locus you can use the primers: 5’-tctggctctgcttcttcgtt-3’ (anneals outside recombination region) and 5’-caattcatcccggtttctgt-3’ (inside cb-unc-119(+)). A transgene inserted in place of the ttTi5605 mos1 transposon is predicted to give a 1.7 kb PCR fragment.
So far, we have not detected any “off-site” insertions in > 40 inserts we tested.

It’s an unfortunate limitation that approximately 30% of the inserts are not full length. So, depending on your scientific question you should verify that the transgene is there. Sequencing is probably the best verification. Also, we can’t stress enough that having more than one insertion of each transgene is important to validate your interpretations.

You will have to do a Southern blot. Chemical southerns work well in our hands. And you don’t need to use the huge gels used for Tc1 mapping. In our experience a medium size PCR sized gel tray works well with 1 ug of cut DNA.

Mike has developed a detailed protocol for detecting transgene copy number and verifying full length inserts. He’s made a pdf with all the details of the approach, including universal primers, PCR conditions and examples of what the results look like from a couple of strains. All the info can be found here:
http://thalamus.wustl.edu/nonetlab/ResourcesF/PCR%20of%20MosSCI%20transgenes.pdf

pCFJ150 will only grow in ccdB tolerant bacteria.

Unfortunately, Invitrogen sells several different versions of their multisite Gateway kit. The three-fragment kit is most often used in the worm community because the Promoterome (Dupuy et al., 2004), ORFeome (Reboul et al., 2003) and 3’UTRome (Mangone et al., 2007) collections were made with that system. We don’t have experience with the newer four-fragment kit. You can buy the kit (cat no 12537-023, Invitrogen) with the reagents or you can find a colleague at your institution that uses it and get aliquots of the vectors. The Multisite Gateway Pro 3.0 Kits are incompatible with our vectors (cat nos 12537-102,, 12537-103, 12537-104, 12537-100). [Thanks to Aurora Kerscher for pointing this out].

If pCFJ150 is not grown under both Amp and Chloramphenicol selection then the plasmid recombines and loosed the ccdB selection. This results in a lot of background. Please see the Invitrogen website for additional information on Gateway cloning.

Yes, but it appears to affect efficiency. We had (low) success with reducing the homology arms from 1.5 kb to 0.5 kb. You should probably consider alternative cloning methods (recombineering or low copy size plasmid that allow larger plasmids) if you are limited by the targeting vectors.

It turns out that it really is necessary to use all three markers – or at least other markers at high concentration. We’ve now heard from several other labs that using only the Pmyo-2::mCherry marker is not effective. You will end up with a large number of false positive MosSCI insertions – seemingly rescued worms without fluorescent markers but that cannot be homozygosed. We think these are arrays animals that have not incorporated any of the fluorescent co-injection markers. When we use three markers at the recommended concentrations we don’t see this.

We have not tried but it should be possible. Pick the markers that work best with the filter set on your fluorescent dissection scope.

That is a reasonable concern. And yes, several people in the lab have used the transgene they are inserting as the only insertion marker. It works fine if the phenotype you are trying to rescue is reasonably severe. We have not built strains or targeting vectors with any of the other usual insertion markers (lin-15 or pha-1, for example). Consider sharing the reagents if you make them.

Yes. We have generated insertions with a low as 1 ng/ul of the targeting vector. We don’t have good data for how much it effects insertion efficiency.

Yes, even if the negative selection is not perfect you can generate insertions at a reasonable frequency. Inserts will out-compete a relatively paralyzed strain.

We see that sometimes. It’s not clear if the heat-shock does something to the array or if the animals just become better at moving even with the negative selection. The same thing happens when the animals are starved, which in some respects resemble the state after heat-shock. Usually, you can still get insertions at a decent frequency although you may have to look at little closer/screen through more plates.

No, we don’t.

We have not tried to heat-shock the worms at 37C but expect that it will harmful to the animals. The standard temperature for heat-shock of C. elegans is 33C or 34C and we recommend using the higher temperature if you are using an air incubator.

We often put in ~50 plates in stacks to heat-shock at a time. The “middle” plates in the stack generally heat up a lot slower than the top and bottom plates. Empirically, the worms are not harmed by two hours of heat-shock and we get a lot less false positives than if we do it for shorter time.

How long did you wait to screen the animals after the heat-shock? It takes at least a few hours to kill the animals and we find that we get the tightest selection if we screen the plates 4-6 hours after heat-shock.

Yes. In fact, we do that for efficient heat-shock to mobilize Mos1 elements from an array. It works more efficiently for killing animals than the air incubator. However, if you have a lot of plates it’s very inconvenient to parafilm and unwrap all the plates so generally we use a heated air incubator for the heat-shock.