gain or loss of function?

if there is a mutation in a special worm strain… how could I figure out if it is a gain or loss of function for the affectet gene?

If you know what gene is affected by the mutation you’re studying, compare it to the RNAi phenotype. If it’s the same or very similar, it’s likely to be a loss of function mutation.

If you don’t know what gene is affected you’d obviously have to map the mutation first before being able to address the question of gain or loss of function.

The nature of the mutation can also give a clue, if it’s a big deletion taking out most of the gene/the start codon/etc. you can be pretty sure it’s a lof mutation.

As tobyb notes, if you’ve cloned the gene responsible then there are a lot of tools that may be able to tell you.

Assuming you haven’t cloned the mutation, and that all you know is the rough (to within a few map units) genetic map position of the mutation, your best approach is gene dosage experiments, and I’ll briefly outline some of the thinking involved.

First of all, does the mutation (which I’ll call “m”) act dominantly or recessively to cause the phenotype? Dominant phenotypes are far more commonly caused by gain-of-function than by loss-of-function mutations, although there are exceptions.
You should look for genomic deficiencies or duplications that span the position of your mutation, and take a look at strains carrying them. If your mutation acts dominantly, but a deficiency heterozygote does not show the same phenotype, then your mutation is not causing loss-of-function. If the +/+/+ or +/+/+/+ strains carrying the duplication show the phenotype you’ve identified, that would suggest the phenotype results from an increase in normal gene function.
Once you’ve got a deficiency and/or a duplication, you can combine them with your mutation and start looking at some comparisons:
If your mutation acts recessively, how does m/Df compare to m/m? If m/m is stronger than m/Df, this suggests gain-of-function. If m/Df is similar to m/m but stronger, or if it’s identical, that would be consistent with loss-of-function (partial or possibly complete, respectively) but could also be consistent with some forms of gain-of-function (see below).
Similarly, what does m/m/+ look like? If you can see your mutant phenotype in m/m/+ but not in m/+ then it’s gain-of-function (the same goes for m/+/+ compared to m/+). If m/+ is as strong as or stronger than m/m/+, or if both look wild-type for your phenotype, it’s consistent with loss-of-function – but, again, it doesn’t prove loss-of-function.

The reason I’ve been saying that some results are consistent with, but do not prove, loss-of-function is that there are multiple types of gain-of-function, and some can be difficult to distinguish from loss-of-function by simple dosage experiments. The types of gain-of-function are:
Increased wild-type function (“hypermorph”). The mutation can act recessively or dominantly (it could act recessively despite causing increased function because of the dose-response curve), but should not be antagonized by the wild-type (m/m/+ should be stronger than or equal to m/m, which should be stronger than or equal to m/+, which should be stronger than or equal to m/Df)
Altered gene function (“neomorph”). This can be tricky, because these mutations often act recessively and their effects are often antagonized by adding a wild-type copy of the locus. Further confusing matters, one subset of this class is the dominant-negative (“antimorph”) mutation, in which the mutant gene product actually interferes with wild-type activity to cause a partial loss of gene function, in a way that cannot be accounted for by losing one copy of the locus.

When you’ve done all the dosage experiments you can do, if you don’t find that your mutation causes increased wild-type function (which should be easy to spot) you may still find yourself unable to answer whether your mutation causes reduced or altered gene function . If you have cloned the gene, then getting a deletion or doing RNAi experiments can distinguish between these possibilities (in addition to asking whether the deletion allele or the RNAi treatment phenocopies your mutation, in some cases you can suppress the effects of an altered-function mutation by RNAi’ing the mutant gene; you can also do this without cloning the gene, by mutagenizing and identifying tightly linked suppressor mutations, but then definitively proving these are loss-of-function mutations in the same gene becomes tricky, and is greatly facilitated by cloning the gene).

These are genetic terms, and thus defined by genetic tests. A loss-of-function allele - let’s call it mut-1(al1) - produces the same phenotype over a deficiency as it does when homozygous. In the absence of dosage effects, the mut-1(al1) phenotype would be fully recessive. You would also predict that mut-1 RNAi of the homozygous mut-1(al1) strain would have no effect on the phenotype. The nature of the molecular lesion (e.g., alteration of an active site residue of an enzyme) can also support this designation, but is not sufficient on its own.

A gain-of-function allele mut-1(al2) usually behaves differently in these genetic tests. The phenotype of the homozygous mutant differs from that of the mutant over a deficiency. The phenotype is often dominant or semi-dominant. RNAi of the homozygous mut-1(al2) strain might produce a different phenotype.

Caveats exist for each these tests/interpretations, so multiple tests are recommended.

Hope this helps,
Harold