WALTHAM, MASSACHUSETTS
University of Massachusetts

The identity of Mga (maize glume architecture) on 4S confused with a multiple allelic series at the Tu (tunicate) locus
--Walton C. Galinat

The old defunct and discredited wild podcorn hypothesis of Mangelsdorf and Reeves, as the key part of their tripartite theory that controlled thinking on the origin of maize for more than 25 years, has left a legacy of prejudice against both Mangelsdorf and the tunicate locus. For example, Doebley and Kermicle (In: Dorweiler et al., Science 262:233-235, 1993) would allow only two alleles (Tu1 and tu1) at the tunicate locus, apparently because of this prejudicial barrier rather than actual facts. A multiple allelic series of tunicate alleles (Tu1 tu1-l, tu1-d, tu1-f, tu1-w, tu1) at the Tu locus was identified long ago (Mangelsdorf and Edwardson, MNL 27:24, 1953) and its components separable by mutation and reconstructible by recombination (Mangelsdorf and Galinat, PNAS 51:147, 1964). The various tunicate alleles tend to focus on the architecture of the female spikelets to different degrees. The strongest allele (Tu1), key to the wild podcorn hypothesis, has monstrous effects, especially when homozygous, with large amounts of developmental activity going into the foliaceous elongation of just the first and second glumes, with the lemmas and paleas usually left wanting. This hyperactivity of the strongest tunicate allele is not grounds for rejecting or ignoring the remainder of the allelic series, which are much more modest in their effects. The tu1-f allele from Chapalote and Reventador is close to normal (tu) for most modern maize, with slight increases in the foliaceous and length traits of the outer female glume. The tu1-f allele has domestic values by its interactions with the Vg (vestigial glume) on chromosome 1 with partial restoration of glumes that are important in the tassel for sunburn protection. The tu1-f allele was important in the past by its interaction with the teosinte allele mga linked to su. In the combination tu1-f tu1-f mga mga (Fig. 1E), the outer female glumes become more foliaceous and elongate, the rachilla more elongate and reflexed, the cupule more reduced. Together these modifications to the teosinte fruitcase made the teosinte grain easily threshable from its enclosure.

More important than the modifying interaction of the tu1-f allele with the teosinte mga gene was a dominant mutation at mga to Mga (maize glume architecture), which then made the teosinte spike even more threshable and opened the way for an even lower allele (tu) at the Tu locus (Fig. 1F). It is not clear at this point if the tu allele came by means of mutation from tu1-f or by backcross introgression from teosinte.

In attempting to deal with the tu1-f allele extracted from Reventador, Doebley would place it at the mga locus rather than the Tu locus. Such a switch could appear to occur by an accidental mixing of su gl3 stocks with su gl4 stocks. Thus, mga at point 48 just 14 units above gl4 could appear to be the location of the tu1-f gene instead of Tu, at just 11 units above gl3.

A comparison of some of the mga-Mga tunicate phenotypes is made here in Figure 1 with caption. It is important to note that the tunicates shown here do not include full (Tu1). All of the tunicate teosinte which Beadle grew and observed was entirely tu1-d tu1-d which he obtained from me. I had explained this to him but, in his enthusiasm, it somehow got overlooked.

Figure 1. Ear of teosinte (A) compared with its maize glume architecture (Mga or Tga) and tunicate (B to E) and non-tunicate derivative (F). Ears (A) and (B) are adapted from Dorweiler et al. (1993) in which the rachis segment (R) of teosinte (A) is fully developed with the cupule sealed by the outer glume (G) in forming a fruitcase type of protective device about the kernel. Ear (B) is claimed by Dorweiler et al. to be Tga teosinte. On a basis of comparison with the other teosinte derivatives (C to F), it is suggested here that it may be tga tu1-f. Ears (C) and (D) are tu1-d mga teosinte adapted from Beadle in 1971 and 1980, respectively. The 1980 line drawing may be made from the 1972 photograph. Ears (E) and (F) are from my large collection of connecting link stocks that I have developed from my studies still underway on the origin of maize. Ear (E) is believed to be mga tu1-f and similar to ear (B). This ear is typical of about 300 ears all from one plant with about 20 tillers, each 7 to 9 feet tall with the main stalk 6-1/2 feet tall. Ear (F) has a combination of the modern maize genes Mga and tu. All ears appear to be stable for single female spikelets. A common metric scale is on the right.


Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 68 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page