Specifying sex in plants: the Q factor

Flowers and fruit are a great way for plants to take advantage of the mobility of insects and animals for their own purposes [and even to encourage moderately keen gardeners like me to spend significant amounts of money on propagating them].  It’s not quite clear when flowers first evolved, but in the mid-Cretaceous period (some 100 million years ago) they suddenly became so prevalent in the fossil record that poor Darwin, struggling desperately to explain the whole history of life on the planet, called their rapid rise “an abominable mystery“.  The mystery is gradually getting less abominable: several transcription factors involved in flower development have been identified, and the relationship between the changes in these transcription factors at the molecular level and the changing shapes of flowers is becoming clearer.  The transcription-level innovations required to fuel the “great angiosperm radiation”, as the abominable mystery is called, may have been relatively modest.  Indeed, a new paper (Airoldi et al. 2010 Single amino acid change alters the ability to specify male or female organ identity PNAS doi: 10.1073/pnas.1009050107) traces a fairly important functional change in a transcription factor — the ability to specify carpels, the female organs, or stamens, the male organs — to a single amino acid insertion.

Some results of the "abominable mystery", growing in my garden.

The key event in the development of flowering plants seems to have been a proliferation of the MADS box transcription factors.  Four types of MADS genes, labeled A, B, C and E, are sufficient to transform a leaf into a flower; A + B + E specifies a petal, B + C + E a stamen, and C + E a carpel.  So to get both stamens and carpels in the same flower, you need tightly localized expression of B.  Some plants use two different C’s for carpels and stamens, presumably to make precise localization easier: Arabidopsis uses only one C, called AG, but Antirhinnum has two: FAR, which can only promote the development of stamens, and PLE, which can be used for either carpels or stamens.  Expression of AG in the wrong place in a developing Arabidopsis flower causes conversion of other structures (petals and sepals) into both male and female organs, while similar expression of FAR only leads to the conversion of petals into male organs.  This must be due to changes in the coding region, not changes in regulation: the functional distinction between the two transcription factors is clear even when both proteins are expressed under the same promoter.

Using domain swaps between AG, FAR and PLE, Airoldi et al. identified a single amino acid, Q173, as

More flowers from my garden. Spring seems so far away.

essential for restricting FAR function to male organ specification.  FAR that lacks Q173 becomes able to promote carpel development; conversely, AG with Q added at the appropriate position becomes restricted to stamen development.  Actually, all you need is a positive charge; instead of glutamine (Q) you can use arginine (R), with the same result.  The position of Q173 is at the boundary of a domain suspected to be involved in protein–protein interactions; Airoldi et al. therefore looked for differences in such interactions using yeast three-hybrid experiments.  They found that indeed AG and FAR interact with different subsets of the E-type proteins present in this system; while AG interacts with three different E-type proteins, SEP1, 2 and 3, FAR interacts only with SEP3.  Is this the crucial difference between the two proteins?  It seems so: SEP3 function is clearly important for FAR function, because when the authors expressed FAR in a SEP3-deficient Arabidopsis plant, they no longer saw conversion of petals to stamens.  But when they deleted the Q from FAR — which should now allow it to bind to SEP1 and 2, and support the development of both male and female organs — they indeed saw the development of both stamens and carpels, even in the SEP3-deficient plant.  So the presence or absence of a single amino acid alters the network interactions among the transcription factors responsible for flower development, switching the developmental fate of the organ it is expressed in from male to female.

One thing that’s interesting about Q173 is that it’s an insertion, not just a change in the identity of the side chain.  The sequence it sits in is highly conserved:

AG:   …MQKR(-)E…

FAR  …MQKRQE…

PLE  …MQKR(-)E…

The sequence encoding the Q, CAG, seems to have come from a duplication of the three base-pairs at the end of the intron sequence at the splice site for exon 7.  Not such a difficult thing to imagine happening on the molecular level, but a change with rich possibilities on the developmental level.  Suddenly you have more ability to control the location or timing of the development of male and female organs, by controlling the localization or timing of expression of the E-type proteins as well as the C-type proteins. It’s a surprisingly large result for such a small difference in DNA sequence, and another example, perhaps, of the potential power of single mutations.

Airoldi CA, Bergonzi S, & Davies B (2010). Single amino acid change alters the ability to specify male or female organ identity. Proceedings of the National Academy of Sciences of the United States of America, 107 (44), 18898-902 PMID: 20956314