Integration of Transgene Arrays in C. elegans

Summary of information obtained from a C. elegans bionet poll.
Comments, additions, corrections:
Morris Maduro (
(October 26, 1998)

Methods for Integrating Extrachromosomal Arrays

General Guidelines

Oligonucleotide-Mediated Integration

Gamma Irradiation

X-ray Treatment

UV Crosslinker

MMS Mutagenesis

EMS Mutagenesis

General Guidelines


There is a detailed description of the rationale and strategy for making integrants written by Michael Koelle, and there is a thorough discussion of worm transformation in Mello and Fire (1995) Methods in Cell Biology 48, Chapter 19 (pp.451-482). Aside from the oligonucleotide protocol (see below), obtaining an integrant starts with a transmitting line which is treated with a DNA-damaging agent to induce integration. To make the screen efficient and recover multiple independent lines, many researchers use the following procedure.

Briefly, a transmitting line with a moderate precentage of transmission, around 20%, is convenient to use. Approximately 50 late L4 or young adult worms are treated. These animals are separated into groups to allow later assurance of independent lines. To screen for integrants, a two-step approach is used. First, approximately 250 F1 animals are singled out. The progeny are screened for broods which demonstrate >75% transgenic progeny. Then, about five F2 animals are picked per candidate to obtain 100% transmission (i.e. homozygous integrants). Lines are checked for expected attribute of the transgene (e.g. GFP expression) and lethality, backcrossed several times to the original genetic background, and given strain designations. Most researchers report that using their favorite procedure, around 1-5% of singled animals turns out to be an integrant.

If the original transmission frequency is too high to efficiently pre-screen the F1s, two F2s can be singled per F1, and the next broods examined for 100% inheritance.

Although the vast majority of researchers use rol-6D (plasmid pRF4) to mark transgenic arrays, there may be advantages to using other genes, as described below.

Mechanism of integration. The extrachrosomal sequences are presumably integrated as the result of chromosome breakage, followed by a healing event which includes part of the (broken?) transgenic DNA. The integrated sequences are 100% stable because replication and segregation are under chromosomal control.

Site of integration. Most integrants are apparently located randomly in the genome. It is possible to recover integrants at the chromosomal location of a gene in the array. This was first observed by Fire et al.1986 (EMBO J 5: 2673-2680), and has been observed many times since. For example, homologous integrants have been seen spontaneously (Broverman et al. wbg11.1p24); after gamma irradiation - WBG article by Stuart Kim (wbg12.2p22), and following Tc1 excision (paper by Plasterk et al. 1992. EMBO J 11: 287-290).

Markers for Extrachromosomal and Integrated Transgenes

rol-6D. The 'old faithful,' the rol-6D plasmid pRF4 confers a dominant roller phenotype, advantageous because for most applications no further genetic manipulations need be made to a starting strain. There are some disadvantages. First, for imaging adults, the Rol phenotype can make visualization of structures difficult; some dpy mutations can suppress the Rol, but these must be crossed in. In theory, RNAi with rol-6 DNA (because null mutations in rol-6 show no phenotype) or with a Rol-suppressing Dpy could also be used ad hoc to abrogate the Rol phenotype. Second, mating transgenic lines can be difficult. Some integrated rol-6 arrays, however, demonstrate a recessive nature to the Rol phenotype, allowing heterozygous males to mate efficiently. Third, in an extrachromosomal line, the Rol animals are at a slight growth disadvantage. Fourth, the Rol phenotype does not show until about 40 hours after injection.

unc-22 antisense. In the Andy Fire vector kit is a plasmid (pPD10.46) encoding an unc-22 antisense construct. The advantage is that this marker, like rol-6, is also dominant, but it probably cannot be suppressed by crosses or RNAi. (See Fire et al. Development 113: 503-514.)

GFP markers. These are (obviously) dominant, and an easy way to score for a transgene if the expression can be seen under a GFP-equipped dissecting scope.

Recessive marker rescue. Rescue of a homozygous recessive phenotype by addition of the wild-type locus as a transgene works well, especially if there is a selective disadvantage to untransformed animals (e.g. ts lethal). Because a mutant strain is needed, the mutation has to be introduced into the desired genetic background before injection.

unc-4 rescue has been described in Miller et al. 1992 (Nature 355: 841-845), as cited in Mello and Fire above.

cul-1 rescue (described in WCWM98e195) The advantage here is that unintegrated animals are unhealthy and suffer a strong growth disadvantage, allowing easy identification of transgenic and/or integrated progeny.

unc-119 rescue Since I distributed the strain (null alleles ed3 or ed4) and the rescuing DNA (pDP#MM016B) to over fifty labs several years ago, I have received encouraging comments from researchers who have used this marker. The original Unc animals are at a severe growth disadvantage: They are Dpy, Egl-d and Daf-d. The rescue phenotype can be scored immediately after hatching (i.e. within one day after injections) by following the L1 'trail' from near the immobile injected parent to the transgenic animal. For integration screens, putative F1 integrants can be picked by looking for simultaneous rescue of the unc-119 dumpy, Egl, movement and tap withdrawal defects, improving the odds of recovery.

Oligonucleotide-Mediated Integration

I obtained few replies about this procedure. Briefly, single-stranded oligonucleotides are included in the original injected DNA mixture to induce integration. People who use this approach have said that although the overall percentage of F1 transgenics is significantly lower, a good proportion of transgenic progeny (e.g. 10%) will turn out to be integrants. There is a WBG article (wbg11.2p28) on the original work by Mello et al.

Gamma Irradiation

This procedure has worked well for many. The only inconvenience may be obtaining access to a gamma cell (Cobalt-60 or Cesium-137). The standard dose is between 2000 and 4000 rad. (I've used 3500 and 4000 rad with good results, about 1-2% F1 integrants.)

X-ray Treatment

There were mixed replies about X-rays. For some, this had worked equivalently well as gamma rays, but for others it was unsuccessful. As X-ray sources are artificial (i.e. generated by X-ray tubes) this may be the result of calibration problems.

" If you want to use an equivalent to gamma rays, why don't you use X-rays. We used to use a medical X-ray machine in Cambridge back in 1972, it worked fine for making mutations and rearrangements. I used Doses in the same dose range as with gamma (500-5000 R) with success. We use [an] X-ray machine here at SFU. It seems fine, although we have never used it for integration. The local hospital was also helpful in allowing us to use the X-ray machine for raying things, maybe you can do the same." (Dave Baillie)

UV Crosslinker

This procedure was first suggested to the worm community by Mitani (wbg14.1p22). The WBG article suggested a dose of 300 J/m2. In the Rothman lab, we use a Stratagene UV Stratalinker 1800 at a setting of 300-700 (the measurement is in 'microjoules x 100', and the machine self-monitors until the dose is complete -- usually less than 20 seconds). The rates of integration range from 0% to 5% depending on the line. A high rate of lethality is observed in the treated animals (e.g. >50%). Other replies are summarized below.

First, many reported that they had tried the procedure and did not obtain any transformants after screening as many as 800 F2s. Three replies addressed potential problems with the procedure:

"I've used the UV Stratalinker 1800, with the energy setting at 300. Initially I had great successes and got a frequency of about 1% (for 3 or 4 integrants; all rol-6); I have lately, however, had a hard time to get an integrant. Might be the array, though (or the UV lamp getting old)." (O. Hobert, Ruvkun Lab)

"I imagine that one critical factor will be the wavelength of the radiation. I've seen people talk about using UV radiation and mean 254nm (classically germicidal) while others will mean 365nm. These produce very different spectra of DNA damage." (Phil Hartman)
Comment: The Stratagene machine comes equipped with 254 nm bulbs standard (although others are available for protein-protein or protein-nucleic acid crosslinking), and the original Mitani WBG suggests using a 254 nm source.

"Make sure there is absolutely no E. coli around. I tried several times to do a dose-response curve with food on the plate and got wildly different results each time. I finally got repeatable results when I washed the worms out of all the bacteria and put them down on an unseeded plate. The bacteria absorb the UV -- they must act like sunscreen!" (Eric Moss)
Comment: This is a great suggestion! I had assumed that as long as the plate lids were removed, the UV treatment itself would be fine. This is an important observation about UV, and may explain why initial attempts were not successful.

MMS Mutagenesis

I received more than a few replies reporting the use of the mutagen MMS for integrations. It would seem that an alkylating agent doesn't cause double-stranded breaks in DNA, and shouldn't result in integrants. My guess is that alkylation products caused by MMS produce replication blocks, resulting in single-stranded gaps. In the next round of replication, double-stranded breaks result. As the mitotically proliferating gonad is the target of mutagenesis for integration, it may be that MMS treatment overwhelms the capacity for germ cell nuclei to repair the adducts before they become replication blocks. I have obtained MMS (Sigma # M4016) but have not tried this protocol yet.

"We have gone to inducing integrants chemically with MMS. The procedure is fairly simple." (Laura Mathies)
Note: The protocol is reproduced below.

MMS is very dangerous: wear gloves and follow the safety procedures below carefully.

  1. Wash plates with 2-3 ml M9 and transfer to sterile 15ml screw cap tube. Spin down worms for 2 min at 2-3 on the clinical centrifuge. Resuspend in 4ml of M9.
  2. Add 4 ul (microlitres) MMS (0.1%) to side of tube, cap, and immediately invert to mix.
  3. Incubate for 10 minutes then spin down in clinical centrifuge.
  4. Rinse 2-3 times with several mls of M9. MMS has a short half-life in aqueous solution.
  5. Transfer to seeded plates and single L4's and young adults.

This concentration and incubation time produces about 1% integrants.

All MMS solutions should be handled in a fume hood.

EMS Mutagenesis

From the reply below, EMS is capable of causing integration. As EMS is also used to recover small deletions from sib-selection PCR-based screens, and deletions have been recovered after EMS treatment in conventional forward mutagenesis experiments, EMS alkylation products must also result in double-stranded breaks.

"We have (unintentionally) used EMS to integrate transgenes. This comes up as an undesired background event in certain mutant screens, and from our numbers, it's probably about the same frequency [with] the Brenner standard mutagenesis (50 mM, 4h) as after gamma." (Leon Avery)

[Top] [Resources Page] [Main Page]