"When good complementation tests go bad, Part I"
When two mutations cause similar phenotypes and map to the same locus yet show complementation in heterozygotes. Note that although intragenic complementation is often cited as evidence for multimerisation (mechanism 2), it can also occur in monomeric proteins (mechanism 1). Of course, both mechanisms can apply in some cases.
[Aka 'intracistronic' or 'interallelic' complementation. Sometimes called 'complex complementation' when the pattern of complementation is nonlinear.]
(1) The protein has multiple independently mutable domains
The two alleles may affect different functional domains of a protein--as in complementation of deletion alleles. This is suggested when the pattern of complementation is simple, ie can be drawn as a linear map of complementation groups.
"alpha-complementation" of beta-galactosidase in E. coli.
Ullman, Jacob, and Monod (1967). J. Mol. Biol. 24: 339-343
[The basis for blue/white selection in cloning experiments. beta-Gal normally forms a homotetramer. The M15 deletion allele of LacZ makes a nonfunctional protein lacking residues 11-41 that can dimerise but not tetramerise. The deleted region ("alpha peptide") can be supplied in trans; ie it does not need to be covalently attached to the rest of the protein to promote tetramerisation. See Jacobson et al (1994) Nature 369: 761 for a structural explanation of alpha-complementation.]
HIS4 in yeast S. cerevisiae (Fink GR, 1966, Genetics 53: 445)
Genetic analysis initially suggested that HIS4 (called hi-4 in this paper) was polycistronic, analogous to the his operon of E. coli; later it was found to encode a single multifunctional polypeptide with 3 distinct, independently mutable catalytic domains. Note that the interpretation of the polar mutations does not distinguish between the two models!
Clifford R, Schupbach T (1994) Molecular analysis of the Drosophila EGF receptor homolog reveals that several genetically defined classes of alleles cluster in subdomains of the receptor protein. Genetics 137: 531-50
Some examples in this paper appear to reflect multiple functions (mechanism 1) and some likely reflect complementation by heterodimerisation (mechanism 2).
Ohya Y, Botstein D (1994). Diverse essential functions revealed by complementing yeast calmodulin mutants. Science 263: 963-966.
Intragenic complementation in a protein that almost certainly functions as a monomer.
Zabin I, and MR Villarejo. (1975) Protein complementation. Annu Rev Biochem. 44: 295-313
Crick, FHC and Orgel, L. (1964). The theory of interallelic complementation. J. Mol. Biol. 8: 161-165.
(2) Hybrid multimers
Complementation by restoration of the quaternary structure of a multimeric complex: i.e., mutant proteins may not be able to function as homodimers (due to folding or stability problems) but can compensate for each other's defects in heterodimers (Crick and Orgel 1964). In this case there may be no simple pattern of complementation corresponding to functional domains, and no linear map of complementation groups can be devised. (This is in fact a type of intragenic mutual suppression).
Alkaline Phosphatase (e.g., Garen & Garen 1963, JMB 7: 13-22)
Glutamic dehydrogenase (e.g. Fincham and Coddington, 1963, JMB 6: 361-373)
(3) Multiple protein isoforms
"Intragenic" complementation can also occur if single locus generates multiple products (by alternative splicing or alternative promoters). Mutations affecting alternatively spliced exons or alternative promoters may have not effect on the other exons/promoters. This should yield a pattern of complementation corresponding to the different domains present in each isoform. (This is really a variation on (1) above, except that the different domains are in different isoforms.)
(4) Pairing-dependent complementation ("transvection")
Two alleles display complementation when present on paired homologs, but fail to complement when pairing is disrupted, for example by a translocation.
First described in Drosophila by Ed Lewis. Mechanisms still not well understood. See review by Wu and Morris (1999; Curr. Opin. Genet. Dev. 9: 237-46) on this and other 'homology effects'.