Current Research
Investigating the developmental and genetic basis of nectar spur evolution
Researchers: Elena Kramer, Levi Yant, Joshua Puzey (past researcher)
    Petal nectar spurs, which have evolved many times independently, are thought to represent key innovations in numerous angiosperm taxa (Hodges 1997; Hodges and Arnold 1995). Despite this, we know nothing about their genetic control. In the 60-70 species of Aquilegia, spur length has been shown to change in association with pollinator shifts that occur in the context of speciation (Whittal and Hodges 2007). Aquilegia is the only member of its tribe to have true nectar spurs, although nectar cups are not uncommon, such as those in the sister genus Semiaquilegia. The very recent diversification of Aquilegia suggests that true spurs evolved on the range of ~5 mya (Wang and Chen 2007; HOdges and Arnold 1994; Hodges et al., 2004). Within the genus, spur morphology varies enormously in length (0-16 cm) and curvature (Hodges and Arnold 1994), including the derived species A. ecalcarata that has secondarily reverted to a nectar cup (Hodges 1997). Some authors have suggested that spur development derives from ‘meristematic’ bulges at the base of the petal (Tucker and Hodges 2005). This could be consistent with Antirrhinum mutants in which KNOX genes are over-expressed in petals, resulting in de novo spur-like outgrowths (Golz et al. 2002). However, our studies of Aquilegia KNOX genes have not implicated these loci in spur development (Collani et al. submitted). Thus, relatively little is known about the development of petal spurs, in either morphological or genetic terms.
    Our first step was to obtain a detailed understanding of the cellular processes – the balance of cell division and expansion – that underlie spur morphogenesis. In collaboration with our colleague in the School of Engineering and Applied Sciences, Prof. L. Mahadevan, we have used molecular markers, microscopy, live imaging, chemical treatments and modeling to elucidate the initiation, development and evolution of Aquilegia petal spurs (Puzey et al., 2011). Our studies demonstrate that spurs develop in two distinct phases corresponding to cell division (Phase I) and cell elongation (Phase II). Initially, cell divisions occur diffusely throughout the petal primordium but when it reaches ~1mm in length, the domain of cell division begins to contract, starting at the margins and progressing towards the presumptive nectary (Fig. 1A, D). The effect of this contraction in the cell division domain is that divisions persist in the spur region for a longer period than in the limb. Modeling shows that this alone may be sufficient to produce the initial out-pocketing of the spur but we also cannot rule out a role for differential cell divisions along the abaxial/adaxial axis. This region does not resemble a discrete meristem, however, and no localized cell divisions are associated with the previously suggested ‘meristematic knobs’ at the base of the petal (Fig. 1A). Another critical feature of Phase I is the orientations of cell divisions in the spur zone. SEMs of very early petals show dramatic patterns of cell walls being positioned radially around the presumptive nectary – as if in response to some kind of organizing center (Fig. 1B-D). Although the phytohormone auxin would seem to be an obvious candidate for the organization of this focal point, our experiments to date with auxin and auxin transport inhibitors have not affected spur development. It is important to note, however, that Phase I occurs at a stage that is difficult to access experimentally, so it remains possible that auxin plays some role at the earliest stages. This period of localized, oriented cell divisions completely ceases by the time the petal is ~2mm in length, representing only 3% of its final length in A. coerulea.
    From this point, the spur enters Phase II, which is entirely dependent on highly anisotropic cell elongation. This oriented cell expansion can be measured by the ratio of the long axis of the cell divided by its short axis, termed epsilon. We have measured this value for thousands of cells along continuous limb-to-nectary transects across multiple developmental stages from four different species of Aquilegia. For example, close to the nectary in A. coerulea, epsilon starts at 1.5 when the petal is 5 mm long and increases to 4.5 by the time it is full grown at 60 mm. Perhaps the most surprising finding is that when we compare across species ranging in spur length from 20-140 mm, we find that the initial and final cell areas do not differ significantly but the B/A values – the degree of oriented cell elongation – are highly correlated with spur length (Fig. 2). These findings, combined with other results not described here, demonstrate that variation is Aquilegia spur length is primarily driven by variation in degree of cell elongation (epsilon) rather than cell number. Even in A. longissima with 12-14 cm spurs, the increase in spur length appears to be almost entirely explained by an extreme increase in epsilon rather than increased cell proliferation (i.e., spur cells of A. longissima are much longer and narrower than those of A. vulgaris). This difference is generated primarily by a longer period of cell elongation rather than a faster rate. In summary, Aquilegia petal spurs initially form due to a localized region of prolonged cell division in which cell wall formation is radially organized around the presumptive nectary. This lays the ground pattern of the spur, which is then realized through rapid, anisotropic cell elongation that is the major determinant of spur length and shape. Having this detailed model now in hand, we can pursue a series of experiments that will significantly elucidate the genetic basis of spur initiation, elaboration and evolution.
    We are taking multiple approaches, including analyses of candidate genes such as TCP4, RNA-seq of very early spur cups, biophysical modeling, hormone applications and physical manipulations of spur development.
Golz, J.F., E.J. Keck, and A. Hudson, 2002. Spontaneous mutations in KNOX genes give rise to a novel floral structure in Antirrhinum. Curr Biol, 12:515-522.
Hodges, S.A. and M.L. Arnold, 1994. Columbines - A geographically widespread species flock. Proc Nat'l Acad Sci, USA, 91:5129-5132.
Hodges, S.A. and M.L. Arnold, 1995. Spurring plant diversification: Are floral nectar spurs a key innovation?
          Proceedings of the Royal Society of London Series B-Biological Sciences, 262:343-348.
Hodges, S.A., et al., 2004. Verne Grant and evolutionary studies of Aquilegia. New Phytol, 161:113-120.
Hodges, S.A., 1997. Floral nectar spurs and diversification. Int’l J Plant Sci, 158:S81-88.
Puzey, J. P.*, S. Gerbode*, S. A. Hodges, E. M. Kramer1, and L. Mahadevan1. (2011) Evolution of spur length
           diversity in Aquilegia petals is achieved solely through cell shape anisotropy. Proceedings of the Royal
           Academy of Science, London, Series B, 279:1640-1645.  *co-first authors, 1joint corresponding authors.
Tucker, S.C. and S.A. Hodges, 2005. Floral ontogeny of Aquilegia, Semiaquilegia, and Isopyrum
            (Ranunculaceae). Int’l J Plant Sci, 166:557-574.
Wang, W. and Z.D. Chen, 2007. Generic level phylogeny of Thalictroideae (Ranunculaceae) - implications for
            the taxonomic status of Paropyrum and petal evolution. Taxon, 56:811-821.
Whittall, J.B. and S.A. Hodges, 2007. Pollinator shifts drive increasingly long nectar spurs in columbine
            flowers. Nature, 447:706-710.
Fig.1. A. HIS4 expression in 1.25mm petal. K=knob, L=limb. Arrows mark the extent of the cell division domain, arrowheads indicate extent of region to be sampled for RNA-seq. B-C. SEM of 2 mm petal. Arrow indicates position of attachment point. C shows the same petal as B but with cell wall orientations highlighted in green. D. Schematic describing Phase 1 of spur development. All asterisks = nectary.
Fig. 2. Cell anisotropy plays an essential role in spur length diversity. (a) Petals from four diverent Aquilegia species. From left to right: A. vulgaris, A. canadensis, A. coerulea, and A. longissima. Scale bar equals 1 cm. Insets for each species show a cellular region of identical width. (b) The ratio of final to initial spur length Lf/Li versus the fractional increase in cell area Af/Ai is plotted to show that changes in spur length are not correlated with changes in cell area (R2 = 0.233, Pearson's r = 0.482). (c) Lf/Li is plotted versus the fractional increase in cell anisotropy, epsilon(final)/epsilon(initial), indicating that spur length diversity is characterized by cell anisotropy (R2 = 0.990, Pearson's r = 0.995). (d) Total petal length Lp is plotted versus time, demonstrating that all species follow the same growth curve but diverge in developmental duration. Puzey et al. 2011.