Floral symmetry genes and the origin and maintenance of zygomorphy in a plant- pollinator mutualism

Wenheng Zhang, Elena M. Kramer, and Charles C. Davis1

Department of Organismic and Evolutionary Biology, Harvard University Herbaria, Cambridge, MA 02138

Edited by Michael J. Donoghue, Yale University, New Haven, CT, and approved February 10, 2010 (received for review September 8, 2009) The evolution of floral zygomorphy is an important innovation in provides the bees access to oil glands, which are borne in pairs on flowering plants and is thought to arise principally from special- the abaxial surface of the sepals. The stereotypical floral mor- ization on various insect pollinators. Floral morphology of neo- phology of New World Malpighiaceae, despite tremendous var- tropical Malpighiaceae is distinctive and highly conserved, especially iation in vegetative and fruit morphology, led Anderson (9) to with regard to symmetry, and is thought to be caused by selection hypothesize that floral uniformity in the group results from their by its oil-bee pollinators. We sought to characterize the genetic specialization on these oil-bee pollinators. basis of floral zygomorphy in Malpighiaceae by investigating Interpreting the origin and maintenance of this unique floral CYCLOIDEA2-like (CYC2-like) genes, which are required for estab- morphology in a comparative evolutionary framework, however, lishing symmetry in diverse core eudicots. We identified two copies has remained elusive, in large part because of our lack of of CYC2-like genes in Malpighiaceae, which resulted from a gene understanding of the closest phylogenetic relatives of Malpigh- duplication in the common ancestor of the family. A likely role for iaceae. Fortunately, this problem has recently been resolved: the these loci in the development of floral zygomorphy in Malpighia- family is successively sister to two species-poor clades that pos- ceae is demonstrated by the conserved pattern of dorsal gene sess actinomorphic flowers, Elatinaceae and Centroplacaceae expression in two distantly related neotropical species, Byrsonima [Fig. 1; Malpighiaceae (1,300 spp.), Elatinaceae (35 spp.), Cen- crassifolia and Janusia guaranitica. Further evidence for this function troplacaceae (6 spp.)] (13–16). These findings, together with is observed in a Malpighiaceae species that has moved to the pale- morphology-based character state reconstructions of floral sym- otropics and experienced coincident shifts in pollinators, floral sym- metry in the group, demonstrate that Malpighiaceae evolved metry, and CYC2-like gene expression. The dorsal expression pat- from actinomorphic-flowered ancestors (Fig. S1). Moreover, tern observed in Malpighiaceae contrasts dramatically with their these results suggest that the origin of Malpighiaceae and their actinomorphic-flowered relatives, Centroplacaceae (Bhesa panicu- unique flowers appear to correspond with a dramatic shift in lata) and Elatinaceae (Bergia texana). In particular, B. texana exhib- speciation rates (14, 15). During the course of this radiation, its a previously undescribed pattern of uniform CYC2 expression, there appears to have been seven subsequent dispersal events suggesting that CYC2 expression among the actinomorphic ances- from the New World that gave rise to Old World Malpighiaceae tors of zygomorphic lineages may be much more complex than pre- (17–19). The Old World tropics lack the oil-bee pollinators that viously thought. We consider three evolutionary models that may visit New World Malpighiaceae (ref. 20, p. 913), and these have given rise to this patterning, including the hypothesis that geographic transitions have resulted in alterations in both floral fl oral zygomorphy in Malpighiaceae arose earlier than standard symmetry and petal morphology, as well as the loss of the oil morphology-based character reconstructions suggest. gland morphology among Old World members of the family (17, 18, 21). Thus, the majority of Old World species possess flowers CYCLOIDEA | development | gene duplication | Malpighiaceae | phylogeny that are either truly actinomorphic or zygomorphic in a manner that is very divergent from the pattern exhibited by New World ost flowers are either bilaterally symmetrical (i.e., zygo- species. These zygomorphic Old World species possess two Mmorphic) and have a single plane of symmetry or radially dorsal petals, two lateral petals, and one ventral petal. Mal- symmetrical (i.e., actinomorphic) and have several planes of pighiaceae therefore provide a rare opportunity to elucidate the symmetry (1). Floral zygomorphy has evolved independently at ways in which important morphologies originate and are alter- least 38 times (2–4) and is a hallmark feature of the most diverse natively maintained or remodeled following changes in a selec- angiosperm clades, including Asteraceae (23,600 spp.), Orchid- tive regime, such as a shift in pollination system. aceae (21,950 spp.), Fabaceae (19,400 spp.), and Lamiales One way to approach this problem is to investigate the floral (23,275 spp.) (5). Plant evolutionary biologists therefore propose developmental genetic basis for these shifts, both in terms of the that the origin of floral zygomorphy may have been a key inno- origin and maintenance of zygomorphy in New World Mal- vation for promoting speciation throughout the course of pighiaceae, and the secondary loss of this conserved morphology angiosperm evolution (6). The driving force behind the origin of among Old World species. Fortunately, the developmental fl oral zygomorphy has long been thought to be a consequence of genetics of floral zygomorphy is being elucidated at a rapid pace. selection by specialization on certain insect pollinators (1, 7), which has recently gained experimental support (8).

The tropical plant family Malpighiaceae exhibits a strong Author contributions: W.Z., E.M.K., and C.C.D. designed research; W.Z. performed re- association between floral zygomorphy and insect pollinator search; W.Z., E.M.K., and C.C.D. analyzed data; and W.Z., E.M.K., and C.C.D. wrote the attraction. The floral morphology of the more than 1,000 New paper. World species of this clade is very distinctive and highly con- The authors declare no conflict of interest. served, especially with regard to symmetry and pollinator reward. This article is a PNAS Direct Submission. The single upright/dorsal banner petal is strongly differentiated Data deposition: The sequences reported in this paper have been deposited in the Gen- from other petals in the corolla whorl, and appears to help orient Bank database (accession nos. GU982187-GU982264). and attract an extremely limited suite of pollinators, principally 1To whom correspondence should be addressed. E-mail: [email protected]. female bees of the tribes Centridini and Tapinotaspidini (Fig. 1) This article contains supporting information online at www.pnas.org/cgi/content/full/ (9, 10–12). Furthermore, the very narrowed base of the petals 0910155107/DCSupplemental.

6388–6393 | PNAS | April 6, 2010 | vol. 107 | no. 14 www.pnas.org/cgi/doi/10.1073/pnas.0910155107 coding region were included in our analyses. The aligned CYC2 matrix included 78 sequences and was 384 bp in length; 78 of these bps were constant and 278 were parsimony-informative. The phylogenetic relationships inferred from the CYC2 homologues mirror our understanding of accepted species tree relationships (14, 15, 17, 19) (Fig. 2 and Fig. S2). Cen- Malpighiaceae Elatinaceae Centroplacaceae troplacaceae and Elatinaceae are successive sisters to Mal- pighiaceae with 100% maximum likelihood (ML) bootstrap Fig. 1. Comparative floral morphology of Malpighiaceae and their closest fl support (BP) and 100% Bayesian posterior probability (PP; actinomorphic owered relatives, Elatinaceae and Centroplacaceae. Banis- reported for simplicity here as percentages) and 70% BP/75% teriopsis argyrophylla, Bergia texana, and Bhesa paniculata represent Mal- pighiaceae, Elatinaceae, and Centroplacaceae, respectively. Dotted lines PP, respectively. A comparison of the CYC2-like gene tree with indicate planes of symmetry. Note: the flower of B. texana was forced accepted species tree relationships indicates that the origin of opened for illustration. The exact floral orientation of B. paniculata is Malpighiaceae coincided with a duplication in the CYC2 clade, unclear given the congested nature of their inflorescences. which yielded two major copies, CYC2A and CYC2B.Both copies receive moderate to high support (CYC2A, 57% BP/65% PP; CYC2B, 80% BP/99% PP). We inferred additional gene Genes belonging to the recently defined CYC2 clade (22) of the duplications in CYC2A in Galphimia, Hiptage, Tristellateia, and TCP [Teosinte Branched 1, CYCLOIDEA (CYC), and PCF] tran- Verrucularia, and in CYC2B in Banisteriopsis, Flabellariopsis, scription factor family have been shown to play a critical role in Hiptage, Janusia, Ptilochaeta, Ryssopterys, and Spachea. CYC2 the development of zygomorphy. In Antirrhinum, the two closely duplications were also inferred in the outgroup taxa Bergia tex- related CYC2-like genes, CYC and DICHOTOMA (DICH), ana and Elatine minima (Elatinaceae). These results are con- demarcate the dorsal region of the flower by differentially regu- sistent with previous analyses that have uncovered CYC2 lating the rate of cell growth in the developing floral organs (23, duplications across diverse phylogenetic groups (22). Fur- 24). CYC/DICH expression is continuous throughout floral thermore, they demonstrate that CYC2 gene duplications have development in the dorsal region of this species. In cyc/dich double occurred independently in Malpighiaceae and Elatinaceae, with mutants, however, zygomorphy is lost—the dorsal petals become the former duplication corresponding to the origin of Mal- ventralized, resulting in actinomorphic corollas. The emerging pighiaceae and their unique floral zygomorphy (Fig. S1). paradigm from this and numerous subsequent studies is that Importantly, this duplication is not likely to be the result of zygomorphy has evolved independently through the repeated polyploidization, as there is no evidence of genome doubling recruitment of CYC2-like genes in several phylogenetically diverse associated with the origin of Malpighiaceae (27, 28). clades (reviewed in refs. 25, 26). Similarly, we uncovered evidence for the loss of CYC2A or B The present study builds on this strong developmental frame- in various Malpighiaceae clades, which has also been observed in work and suggests that CYC2-like genes are similarly associated other groups (22). Nearly all species of New World Malpighia- with the origin and maintenance of zygomorphy in Malpighia- ceae possess CYC2A and CYC2B, but some species, especially ceae. We identified two forms of CYC2-like genes in Malpighia- Old World Malpighiaceae, appear to retain only one copy (Table EVOLUTION ceae, which resulted from a gene duplication in the common S1). We further verified the copy number of CYC2-like genes by ancestor of the family. Both gene copies, CYC2A and CYC2B, are Southern hybridization in a subset of taxa (Fig. S3 and S4). expressed in the dorsal region of the flowers in distantly related Bergia texana, Byrsonima crassifolia, J. guaranitica, and T. aus- New World species with the stereotypical floral morphology (i.e., tralasiae have six, two, four, and one copy of CYC2-like genes Byrsonima crassifolia Kunth and Janusia guaranitica A. Juss.). respectively, which is consistent with our initial assessment of Further evidence of the function of CYC2 in establishing floral gene copy number ascertained from PCR and clone screening. symmetry in Malpighiaceae derives from expression patterns in These results suggest that our PCR/clone screens provide a an Old World species that has lost its association with the New reliable estimate of copy number in taxa for which Southern World oil-bee pollinators and its stereotypical morphology. In the hybridizations were not conducted. zygomorphic-flowered species Tristellateia australasiae A. Rich., CYC2A expression mirrors a dramatic shift in the plane of floral Recruitment of CYC2 and the Maintenance of Floral Zygomorphy in symmetry, whereas CYC2B has been lost. The dorsal patterning of Malpighiaceae. To determine if the evolution of floral zygomor- gene expression we observed in Malpighiaceae contrasts dra- phy in Malpighiaceae is likely a result of changes in the regu- matically with their actinomorphic-flowered relatives, Cen- lation of CYC2 homologues, we investigated the pattern of CYC2 troplacaceae (Bhesa paniculata Arn.) and Elatinaceae (Bergia expression using locus-specific RT-PCR. RT-PCR was conducted texana Seub. ex Walp.). Results from the latter species suggest on J. guaranitica and B. crassifolia (New World Malpighiaceae), unique insights on CYC2 patterning in actinomorphic-flowered T. australasiae (Old World Malpighiaceae), B. texana (Elatina- species, and highlight the importance of investigating CYC2-like ceae), and B. paniculata (Centroplacaceae; Fig. 3). Each of these genes in other actinomorphic core eudicot clades beyond the taxa was carefully selected based on its phylogenetic affinities to model species Arabidopsis thaliana. elucidate changes in gene expression related to (i) the origin of zygomorphy in Malpighiaceae, (ii) the conservation of floral Results and Discussion morphology of New World Malpighiaceae, and (iii) the loss of Gene Duplication and Loss. Thirty-three species from 29 genera of the stereotypical New World floral morphology in Old World Malpighiaceae representing all major clades of the family, two Malpighiaceae. species from both genera of Elatinaceae, and one species from Several studies from phylogenetically diverse species indicate Centroplacaceae, were used for assembling the CYC2 genealogy that zygomorphy has evolved independently via the repeated (Table S1). Oxalis herrerae R. Knuth was used as an outgroup recruitment of CYC2-like genes (reviewed in ref. 25). CYC2 (16). One to six copies of the CYC2-like genes were isolated from expression in Malpighiaceae similarly suggests that these genes each taxon. As with previous studies (e.g., ref. 22), the TCP and are responsible for the unique floral symmetry of New World R domains were highly conserved, but the intervening coding Malpighiaceae (Fig. 3). During the later stages of floral devel- region was variable across all taxa. Partial TCP and R domains opment, CYC2A and CYC2B are differentially expressed along (i.e., 26 of 60 and two of 18 total amino acid sequences were the dorsoventral plane of symmetry in Byrsonima crassifolia and obtained for each region, respectively) and the entire intervening J. guaranitica, which mirrors their floral morphology. B. crassi-

Zhang et al. PNAS | April 6, 2010 | vol. 107 | no. 14 | 6389 Oxalis OhCYC2-1 O folia has two CYC2 genes: CYC2A (BcCYC2A) and CYC2B Oxalis OhCYC2-2 (BcCYC2B; Fig. 2). J. guaranitica has four CYC2 genes: CYC2A Bhesa BpCYC2 C (JgCYC2A) and three copies of CYC2B (JgCYC2B-1, -2, and -3). 99/100 Elatine EmCYC2-1 Elatinaceae Elatine EmCYC2-2 BcCYC2A of B. crassifolia and JgCYC2A and JgCYC2B-3 of J. 81/98 88/89 Bergia BtCYC2-2 guaranitica are expressed broadly in the dorsal region of the Bergia BtCYC2-4 flower, i.e., in the dorsal and lateral petals and sepals, and weakly Bergia BtCYC2-1 99/100 Bergia BtCYC2-3 in the gynoecium (Fig. 3). In contrast, BcCYC2B of B. crassifolia 75/100 Bergia BtCYC2-5 and JgCYC2B-1 and JgCYC2B-2 of J. guaranitica are expressed Bergia BtCYC2-6 only in the single dorsal banner petal, and weakly in the 100/100 60/98 Byrsonima BcCYC2A Diacidia DiagCYC2A gynoecium. Whether differential expression of these CYC2-like 66/98 Spachea SeCYC2A genes affects the symmetry of the calyx is unclear. In J. guar- 99/100 Lophanthera LpCYC2A anitica, CYC2-like genes are differentially expressed in the dorsal Lophanthera LlCYC2A and lateral sepals with oil glands, but not in the eglandular -/69 93/100 Verrucularria VgCYC2A-1 Verrucularria VgCYC2A-2 ventral sepal. This raises the possibility that CYC2 might be 57/65 96/100 Galphimia GmCYC2A involved in the development of the dorsoventral asymmetry of Galphimia GgCYC2A-1 B. crassifolia 92/100 Malpighiaceae CYC2A the calyx. However, all of the sepals of are identical 70/75 Galphimia GgCYC2A-2 and bear oil glands, despite the fact that CYC2 expression was Acridocarpus AsCYC2A * -/86 Lasiocarpus LasCYC2A detected only in the dorsal and lateral sepals. Ptilochaeta PnCYC2A Overlapping expression patterns of duplicated CYC2 homo- 90/98 -/91 Dinemagonum DingCYC2A logues, as observed here in Malpighiaceae, are common in Dinemandra DeCYC2A fl – Echinopterys EeCYC2A numerous zygomorphic- owered groups (23 25, 29, 30). With- 98/100 Heladena HbCYC2A out functional assays, however, it is difficult to discern if the 64/91 68/81 Henleophytum HeCYC2A CYC2A and 2B paralogues have significant functional redun- Tristellateia TaCYC2A * dancy, as in Antirrhinum (23, 24), or have more differentiated Tristellateia TafCYC2A-1 * 83/99 Tristellateia TafCYC2A-2 * functions, as in Fabaceae, where one copy is more broadly 69/98 fi fl -/88 Tristellateia TafCYC2A-3 * expressed and the other is speci c to the dorsal ag petal (29, Mascagnia MbCYC2A 30). Regardless, duplication and functional divergence of CYC2- Hiptage HdCYC2A-1 * 97/100 73 like genes may have contributed to the evolution of the unique /99 Ryssopterys RtCYC2A * 100/100 Janusia JgCYC2A zygomorphy in New World Malpighiaceae. The two species 64/92 Banisteriopsis BlCYC2A examined here, B. crassifolia and J. guaranitica, span the basal 87/100 Malpighia McCYC2A node of crown group Malpighiaceae, and diverged from one Madagasikaria MaCYC2A * another at least 64 Mya (17), yet their pattern of gene expression 69/84 Hiptage HdCYC2A-2 * -/81 Flabellariopsis FaCYC2A * is identical. This highly conserved pattern provides a likely Byrsonima BcCYC2B molecular explanation for the stereotypical floral morphology in 87/100 Spachea SeCYC2B-2 New World Malpighiaceae, which is in turn the basis of a well Spachea SeCYC2B-1 documented plant-pollinator mutualism. 91/100 Lophanthera LpCYC2B CYC2 fl 62/86 Lophanthera LlCYC2B An important test of our hypothesis that controls oral Verrucularria VgCYC2B zygomorphy in Malpighiaceae is to examine gene expression in 82/83 Galphimia GmCYC2B those clades that have lost the stereotypical New World mor- 99/100 Galphimia GgCYC2B 80/99 Acridocarpus AsCYC2B * phology. T. australasiae is an Old World species with zygomor- fl 86/85 Echinopterys EeCYC2B phic owers that has lost its association with the oil-bee Heladena HbCYC2B pollinators, and the characteristic New World morphology (Fig. 52/61 Henleophytum HeCYC2B Malpighiaceae CYC2B 3). Instead, T. australasiae possesses two dorsal, two lateral, and 98/99 Dinemagonum DingCYC2B 86/100 fi 77 Dinemandra DeCYC2B one ventral petal. Although this species super cially resembles a fl /99 91/99 Lasiocarpus LasCYC2B resupinate New World ower, its altered symmetry does not Ptilochaeta PnCYC2B-1 result from resupination. If this were the case, the innermost 99/100 Ptilochaeta PbCYC2B-1 77/100 petal of T. australasiae, which is homologous with the dorsal flag/ 86/100 Ptilochaeta PnCYC2B-2 97/100 Ptilochaeta PbCYC2B-3 banner petal of all New World species (31), would initiate in the Ptilochaeta PbCYC2B-2 dorsal position and terminate in the ventral position during 83/98 Hiptage HdCYC2B-1 * development. This is not observed: T. australasiae produces an Flabellariopsis FaCYC2B-1 * Sphedamnocarpus SphCYC2B * innermost petal that begins and ends in the dorsal position. Banisteriopsis BlCYC2B-1 Hence, the shift of floral zygomorphy in T. australasiae is best -/66 Banisteriopsis BlCYC2B-2 explained as a 36° rotation in the plane of floral symmetry, rather Ryssopterys RtCYC2B-2 * than a 180° resupination (cf. ref. 32; Figs. S5 and S6). This is Malpighia McCYC2B Flabellariopsis FaCYC2B-2 * likely a developmental feature that is established very early in 50/96 Mascagnia MbCYC2B floral development. How this happens remains an open question Flabellaria FpCYC2B * that we will seek to address in future studies. Janusia JgCYC2B-1 72/99 Janusia JgCYC2B-3 We have shown that T. australasiae has lost the CYC2B gene 63/80 Janusia JgCYC2B-2 and that CYC2A (TaCYC2A) expression closely mirrors its dra- Stigmaphyllon SpCYC2B matic shift in the dorsoventral plane of floral symmetry (Fig. 3). -/80 Ryssopterys RtCYC2B-1 * Hiptage HdCYC2B-2 *

Fig. 2. Phylogeny of CYC2-like genes inferred using maximum likelihood Inferred gene tree is reflective of accepted species tree relationships (13-16, (ML) and Bayesian analyses. Majority rule consensus shown depicting clades 19) (Fig. S2). Accessions shown in boldface were additionally analyzed for with >50% bootstrap support and >60% Bayesian posterior probability gene expression using RT-PCR. Asterisks indicate Old World clades. Species depicted above lines, respectively. ML bootstrap support <50% indicated identities and voucher information can be found in Table S1. (O, Oxalida- with a hyphen. Black circles indicate inferred gene duplication events. ceae; C, Centroplacaceae.)

6390 | www.pnas.org/cgi/doi/10.1073/pnas.0910155107 Zhang et al. A dp ds ls vs dp lp vp AG ds ds BcCYC2A lp lp dp lp vp BcCYC2B ls ls ds ds actin vs ls ls vp vp vs os vs dp lp vp A G S1 S2 S3 S4 LV B dp JgCYC2A JgCYC2B-1 (New World)

Malpighiaceae os os JgCYC2B-2 lp lp dp lp vp JgCYC2B-3 os os actin vs vp vp ds ls vs C 1 2 a b c d e 1 2 3 4 5 AG b a c TaCYC2A 5 ed 3 13245 actin (Old World) Malpighiaceae 4 abcde

D dsdp ds ds ls vs dp lp vp A G S1 S2 LV BtCYC2-1 lp lp dp lp vp BtCYC2 (2/4) ls ls BtCYC2 (3/5/6) vp vp actin Elatinaceae vs ds ls vs

E EInfl LB ADF ABF G BpCYC2 actin Centroplacaceae

Fig. 3. CYC2-like gene expression in Malpighiaceae, Elatinaceae, and Centroplacaceae. RT-PCR was examined in whole flower buds at early developmental stages, dissected floral organs at later stages, and leaves. The stereotypical flowers of New World Malpighiaceae are represented here by Byrsonima crassifolia (A) and J. guaranitica (B). T. australasiae (C) is an Old World species of Malpighiaceae that has lost the stereotypical New World morphology and instead has two dorsal petals and eglandular sepals. Flowers of Bergia texana (Elatinaceae, D) and Bhesa paniculata (Centroplacaceae, E) are actinomorphic and eglandular. Gray highlighting on flower diagrams indicates the spatial pattern of CYC2 expression in the corolla and calyx. Actin-specific primers were used as EVOLUTION a positive control. ds, dorsal sepal; ls, lateral sepal; vs, ventral sepal; dp, dorsal petal; lp, lateral petal; vp, ventral petal; AG, androecium and gynoecium; os, sepal with oil glands; A, androecium; G, gynoecium; S, developmental stages of floral buds (S1, <1 mm in diameter; S2, 1–2 mm; S3, 2–3 mm; S4, 4–5 mm); LV, leaves; b, dorsal sepal; a, c, lateral sepal; d, e, ventral sepal; 1, 2, dorsal petal; 3, 5, lateral petal; 4, ventral petal; Elnfl, early inflorescence; LB, large buds (2–3 mm); ADF, adaxial flower (including sepals, petals, and stamens); ABF, abaxial flower. Note: our sequence analysis confirmed that the larger fragment amplified from the ventral sepal of J. guaranitica is not a TCP family gene. (Scale bars, 5 mm.)

TaCYC2A is still expressed in the dorsal region of the flower, but development (Fig. 3), which is similar to the well studied actino- now the domain encompasses the two dorsal petals and one morphic-flowered rosid A. thaliana (L.) Heynh. For the Arabi- dorsal and two lateral sepals. We hypothesize that changes in the dopsis TCP1 gene, dorsal gene expression is detected in the regulation of TaCYC2A expression contributed to the shift in young floral meristem but greatly declines before the floral organ flower morphology of Tristellateia, and may reflect adaptations to primordia are initiated (33). It remains to be determined a different pollination strategy. Tristellateia is visited by xyloco- whether early BpCYC2 expression is similarly localized in the pine bees whose reward for visiting the flowers appears to be dorsal region of the floral meristem in Bhesa. pollen, rather than oil (14, 15, 21). In contrast, a unique temporal and spatial pattern of CYC2- Additional expression studies of the CYC2 homologues across like gene expression was discovered in the actinomorphic flowers the six remaining Old World Malpighiaceae clades, plus two of Bergia texana (Fig. 3). The six CYC2-like genes in B. texana small New World clades that have lost the stereotypical New exhibit a high degree of sequence similarity (i.e., 78–95%), sug- World floral morphology, will provide further insights into the gesting that they are most likely recently evolved paralogues (Fig. genetic and developmental basis of how floral zygomorphy has 2). As sequence variation among some of these paralogues was been altered or lost in this group. This will enable us to begin to especially minor (i.e., ≤5%), we distinguished three groups for determine whether convergent floral morphologies have evolved RT-PCR analysis: (i) BtCYC2-1,(ii) BtCYC2-2 and BtCYC2-4, via convergent genetic changes in what appears to be a broadly and (iii) BtCYC2-3, BtCYC2-5, and BtCYC2-6. All six BtCYC2 conserved developmental program. genes exhibit no sign of early expression, but are expressed uniformly across the floral organs during later stages of devel- CYC2 and the Actinomorphic-Flowered Relatives of Malpighiaceae, opment. Given the apparent uniformity of gene expression in Centroplacaceae, and Elatinaceae. Our results demonstrate that these paralogues, it is most likely that the expression pattern we the two actinomorphic sister clades of Malpighiaceae, repre- observed in B. texana evolved before gene duplication. In terms sented by B. paniculata (Centroplacaceae) and B. texana (Elati- of the temporal pattern of CYC2 expression, our result most naceae), have different patterns of CYC2-like gene expression, closely resembles that of the zygomorphic-flowered species Iberis suggesting that different processes may be involved in controlling amara L. (Brassicaceae), in which early expression of IaTCP1 has similar floral symmetries. The single CYC2-like gene of B. pan- been lost but late-stage expression is present. Spatial expression iculata (BpCYC2) is expressed only during early stages of flower in I. amara is differential, however, which is consistent with the

Zhang et al. PNAS | April 6, 2010 | vol. 107 | no. 14 | 6391 fact that its flowers are zygomorphic (34). In this respect, the CE M uniform spatial expression of BtCYC2s throughout the corolla of I. B. texana is similar to the pattern in Cadia purpurea Forssk., an actinomorphic-flowered member of the otherwise zygomorphic- flowered Fabaceae (35). The uniform expression of LgCYC1B in 2 2 2A 2B Cadia is thought to be the result of a homeotic transformation, which accounts for the apparent morphological reversal to acti- nomorphy, i.e., gene expression in the lateral and ventral petals has become dorsalized. The main difference between CYC2-like gene expression in Cadia and Bergia is that LgCYC1B expression appears to be corolla-specific whereas expression of BtCYC2sin II. Bergia is not restricted to the corolla, and is instead expressed throughout all of the floral organs examined (Fig. 3). 2 2 2A 2B

Interpreting the Origin of Floral Zygomorphy in Malpighiaceae. Dif- ferential expression of the CYC2-like genes during the late stage of floral development is required for establishing zygomorphic flowers in several phylogenetically diverse rosid [Iberis, Lotus, Lupinus, and Pisum (29, 30, 34, 35)] and asterid [Antirrhinum, III. Gerbera, Linaria, and Senecio (23, 24, 36–38)] clades, suggesting that the recruitment of CYC2-like function at later devel- opmental stages is necessary for evolving zygomorphic flowers. 2 2 2A 2B To better understand the evolution of gene expression that likely contributed to the origin of zygomorphy in Malpighiaceae, we reconstructed the evolutionary pattern of late stage CYC2 or expression in New World Malpighiaceae [i.e., which represents the ancestral floral morphology for the family (9, 19, 27)] and their closest actinomorphic-flowered relatives (Fig. S7). Our results indicate that the most recent common ancestor of Inferred transition of CYC2 gene expression and the origin of Centroplacaceae–Elatinaceae–Malpighiaceae likely exhibited the floral zygomorphy. pattern of gene expression observed in Centroplacaceae (B. Inferred transition of CYC2 gene expression associated with actinomorphy. (NOTE: A concomitant shift to actinomorphy from paniculata), which is also similar to the actinomorphic-flowered zygomophy is inferred only in scenario III.) rosid, A. thaliana (Fig. 4 and Fig. S7). CYC2 expression in the most – Fig. 4. Hypothesized scenarios for the origin of floral zygomorphy in Mal- recent common ancestor of Malpighiaceae Elatinaceae, how- fl ever, is equivocal. pighiaceae. Gray highlighting in the ower diagram indicates the spatial pattern of CYC2 expression. Orange coloring on the tree indicates the infer- In view of the uncertain pattern of CYC2-likegeneexpres- fl – red transition to differential CYC2 expression and oral zygomorphy; green sion in the most recent common ancestor of Malpighiaceae indicates the transition to actinomorphic CYC2 expression in actinomorphic- Elatinaceae, we propose three alternative, and equally parsi- flowered species. In scenario (I), floral zygomorphy evolved in stem lineage monious, scenarios to explain the origin of floral zygomorphy in Malpighiaceae from an actinomorphic-flowered ancestor with no late stage Malpighiaceae (Fig. 4). In the first scenario, the most recent CYC2 expression. In scenario (II), floral zygomorphy evolved in stem lineage common ancestor of Malpighiaceae–Elatinaceae had actino- Malpighiaceae from an actinomorphic-flowered ancestor with actinomorphic morphic flowers with no late-stage CYC2 expression. Under late stage CYC2 expression. In scenario (III), floral zygomorphy and differ- fl this scenario, floral zygomorphy and differential CYC2 ential CYC2 expression evolved from an actinomorphic- owered ancestor expression evolved in the common ancestor of Malpighiaceae with no CYC2 expression along the stem lineage leading to the most recent common ancestor of Malpighiaceae and Elatinaceae. (C, Centroplacaceae; E, and the broad expression in Elatinaceae must have evolved Elatinaceae; M, Malpighiaceae.) independently. The second scenario proposes that the most recent common ancestor of Malpighiaceae–Elatinaceae was actinomorphic and exhibited uniform CYC2 expression across actinomorphy in Elatinaceae may instead be better explained as the floral meristem. Under this scenario, floral zygomorphy in a result of a homeotic transformation from a zygomorphic- Malpighiaceae originated through the loss of ventral gene flowered ancestor, which ultimately gave rise to Malpighiaceae. expression. The third scenario proposes that the most recent Overall, this complex set of alternatives demonstrates that common ancestor of Malpighiaceae–Elatinaceae possessed analyzing molecular data, and gene expression data in particular, zygomorphic flowers with differential CYC2 expression. Under in an explicit phylogenetic framework enhances comparative this scenario, floral zygomorphy and differential CYC2 studies and can add a level of richness to our understanding of expression evolved in the common ancestor of Malpighiaceae– morphological evolution. Determining which of these models is Elatinaceae. Actinomorphic flowers in Elatinaceae would most likely, however, will require additional comparative expres- therefore have been secondarily derived from a zygomorphic sion data, particularly from other actinomorphic rosids, as well as ancestor. Support for this model derives from the Cadia functional studies of CYC2-like genes in Elatinaceae and Mal- example discussed earlier, which is what we hypothesize here to pighiaceae. It is important to note here that most studies of CYC2 explain the gene expression pattern observed in Elatinaceae. homologues have focused exclusively on zygomorphic-flowered This scenario is especially intriguing because it is one that we groups (e.g., Asteraceae, Dipsacales, Fabaceae, Lamiales), mean- would not have previously predicted given our character state ing that we actually know very little about comparative patterns of reconstructions of floral evolution based on standard morpho- CYC2-like gene expression in actinomorphic-flowered groups. This logical characters (Fig. S1). Those analyses clearly indicate that is especially relevant given the diverse actinomorphic-flowered zygomorphy arose in the most recent common ancestor of ancestors that gave rise to these zygomorphic clades. Our results, Malpighiaceae and that actinomorphy in Elatinaceae was ple- as well as those from Busch and Zachgo (34), indicate that further siomorphic. The precedent established by Cadia,combined expression studies are needed to determine if early dorsal with these analyses, however, raises the distinct possibility that expression followed by gene repression is the ancestral condition

6392 | www.pnas.org/cgi/doi/10.1073/pnas.0910155107 Zhang et al. for many actinomorphic angiosperm clades as is commonly 1.4.1 (http://tree.bio.ed.ac.uk/software/tracer/), trees were sampled from the assumed (cf. Arabidopsis), or if the ancestral condition is, in fact, posterior distribution to calculate clade posterior probabilities. highly variable across the angiosperms. RT-PCR. Floral organs were dissected directly in the field from flower Materials and Methods buds ranging in size from 1 mm in diameter to open flowers, and preserved immediately in cryodewars or RNAlater (Ambion). For the samples pre- Sequence Alignments and Phylogenetic Analyses. CYC2-like gene sequences we pared in RNAlater, the floral organs were finely chopped on sterile Petri obtained from Malpighiaceae–Elatinaceae–Centroplaceae–Oxalidaceae (SI dishes to facilitate penetration of the preservative into the tissues. Materials and Methods) were assembled with Sequencher 4.7 (Gene Codes) These materials were processed in our laboratory using the RNAqueous and aligned by eye with reference to the translated amino acid sequences kit (Ambion). RT-PCR was performed as described (43) using locus-specific using MacClade 4.06 (39). We applied the WAG + G model of amino acid primers to examine the expression of each CYC2-likegenecopy evolution to the aligned CYC2 data set as determined by the Akaike Infor- (Table S2). The sequence identity of RT-PCR fragments was confirmed mation Criterion in ProtTEST (40). One hundred ML bootstrap replicates were by sequencing. conducted using RAxML-VI-HPC (41). Bayesian analyses were implemented in MrBayes version 3.1.2 (42) under the same model using default priors for the ACKNOWLEDGMENTS. We thank L. Ahart, W. Anderson, B. Bartholomew, rate matrix, branch lengths, and γ-shape parameter. A Dirichlet distribution M. Bartlett, D. Boufford, M. Davenport, J. Davit, L. Gómez, P. Griffith, L. was used for the base frequency parameters and an uninformative prior was Holappa, D. Howarth, D. Lee, S. Manickam, R. McFarland, C. Morse, M. Opel, used for the starting tree topology. Four chains were initiated with a random C. Specht, E. Voss, K. Wong, S. Yee, and members of the Davis and Kramer starting tree and run for two million generations sampled every 1,000 gen- laboratories. This work was funded by National Science Foundation Grants erations. Following a burn-in of 1,000 trees as determined by Tracer version DEB-0544039 and AToL EF 04-31242 (to C.C.D.).

1. Neal PR, Dafni A, Giurfa M (1998) Floral symmetry and its role in plant-pollinator 22. Howarth DG, Donoghue MJ (2006) Phylogenetic analysis of the “ECE” (CYC/TB1) clade systems: terminology, distribution, and hypotheses. Annu Rev Ecol Syst 29:345–373. reveals duplications predating the core eudicots. Proc Natl Acad Sci USA 103: 2. Donoghue MJ, Ree RH, Baum DA (1998) Phylogeny and the evolution of flower 9101–9106. symmetry in the Asteridae. Trends Plant Sci 3:311–317. 23. Luo D, et al. (1999) Control of organ asymmetry in flowers of Antirrhinum. Cell 99: 3. Stebbins GL (1974) Flowering plant: evolution above the species level (Harvard Univ. 367–376. Press, Cambridge, MA). 24. Luo D, et al. (1996) Origin of floral asymmetry in Antirrhinum. Nature 383:794–799. 4. Westerkamp C, Classen-Bockhoff R (2007) Bilabiate flowers: the ultimate response to 25. Preston JC, Hileman LC (2009) Developmental genetics of floral symmetry evolution. bees? Ann Bot (Lond) 100:361–374. Trends Plant Sci 14:147–154. 5. Stevens PF (2001) Angiosperm Phylogeny Website; version 9, June 2008. Available at 26. Kramer EM (2007) Understanding the genetic basis of floral diversity. Bioscience 57: http://www.mobot.org/MOBOT/research/APweb/. Accessed January 5, 2010. 479–487. 6. Sargent RD (2004) Floral symmetry affects speciation rates in angiosperms. Proc R Soc 27. Cameron KM, Chase MW, Anderson WR, Hills HG (2001) Molecular systematics of Lond B Biol Sci 271:603–608. Malpighiaceae: evidence from plastid rbcL and matK sequences. Am J Bot 88: 7. Endress PK (2001) Evolution of floral symmetry. Curr Opin Plant Biol 4:86–91. 1847–1862. 8. Gomez JM, Perfectti F, Camacho JPM (2006) Natural selection on Erysimum 28. Anderson WR (1993) Chromosome numbers of neotropical Malpighiaceae. Contrib mediohispanicum flower shape: insights into the evolution of zygomorphy. Am Nat Univ Mich Herb 19:341–354. 168:531–545. 29. Wang Z, et al. (2008) Genetic control of floral zygomorphy in pea (Pisum sativum L.). 9. Anderson WR (1979) Floral conservatism in neotropical Malpighiaceae. Biotropica 11: Proc Natl Acad Sci USA 105:10414–10419. 219–223. 30. Feng X, et al. (2006) Control of petal shape and floral zygomorphy in Lotus japonicus. 10. Vogel S (1974) Ölblumen und ölsammelnde Bienen. Trop Subtrop Pflanzenwelt 7: Proc Natl Acad Sci USA 103:4970–4975.

283–547. 31. Anderson WR (2004) Malpighiaceae (Malpighia family). Flowering plants of the EVOLUTION 11. Sigrist MR, Sazima M (2004) Pollination and reproductive biology of twelve species of Neotropics, eds Smith N, Mori SA, Henderson A, Stevenson DW, Heald SV. (Princeton neotropical Malpighiaceae: stigma morphology and its implications for the breeding Univ Press, Princeton, NJ), pp 229–232. system. Ann Bot (Lond) 94:33–41. 32. Vogel S (1990) History of the Malpighiaceae in the light of pollination ecology. Mem 12. Araújo VA, Antonini Y, Araújo APA (2006) Diversity of bees and their floral resources N Y Bot Gard 55:130–142. at altitudinal areas in the Southern Espinhaço Range, Minas Gerais, Brazil. Neotrop 33. Cubas P, Coen E, Zapater JMM (2001) Ancient asymmetries in the evolution of Entomol 35:30–40. flowers. Curr Biol 11:1050–1052. 13. Davis CC, Chase MW (2004) Elatinaceae are sister to Malpighiaceae; Peridiscaceae 34. Busch A, Zachgo S (2007) Control of corolla monosymmetry in the Brassicaceae Iberis belong to Saxifragales. Am J Bot 91:262–273. amara. Proc Natl Acad Sci USA 104:16714–16719. 14. Zhang W, Kramer EM, Davis CC (2009) Exploring the developmental genetic basis of 35. Citerne HL, Pennington RT, Cronk QCB (2006) An apparent reversal in floral symmetry independent reversals in floral symmetry in Malpighiaceae. Botany & Mycology 2009: in the legume Cadia is a homeotic transformation. Proc Natl Acad Sci USA 103: Annual Meeting of the Botanical Society of America in Snowbird, Utah. Available at 12017–12020. http://2009.botanyconference.org/engine/search/index.php?func=detail&aid=851. 36. Broholm SK, et al. (2008) A TCP domain transcription factor controls flower type 15. Zhang W, Kramer EM, Davis CC (2009) CYCLOIDEA2 and the origin and maintenance specification along the radial axis of the Gerbera (Asteraceae) inflorescence. Proc Natl of floral zygomorphy in Malpighiaceae. Botany & Mycology 2009: Annual Meeting of Acad Sci USA 105:9117–9122. the Botanical Society of America in Snowbird, Utah, USA. Available at http://2009. 37. Cubas P, Vincent C (1999) An epigenetic mutation responsible for natural variation in botanyconference.org/engine/search/index.php?func=detail&aid=546. floral symmetry. Nature 401:157. 16. Wurdack KJ, Davis CC (2009) Malpighiales phylogenetics: gaining ground on one of 38. Kim M, et al. (2008) Regulatory genes control a key morphological and ecological trait the most recalcitrant clades in the angiosperm tree of life. Am J Bot 96:1551–1570. transferred between species. Science 322:1116–1119. 17. Davis CC, Bell CD, Mathews S, Donoghue MJ (2002) Laurasian migration explains 39. Maddison DR, Maddison WP (2003) MacClade 4: analysis of phylogeny and character Gondwanan disjunctions: evidence from Malpighiaceae. Proc Natl Acad Sci USA 99: evolution (Sinauer, Sunderland, MA), version 4.0.6. 6833–6837. 40. Abascal F, Zardoya R, Posada D (2005) ProtTest: selection of best-fit models of protein 18. Davis CC, Fritsch PW, Bell CD, Mathews S (2004) High-latitude tertiary migrations of an evolution. Bioinformatics 21:2104–2105. exclusively tropical clade: evidence from Malpighiaceae. Int J Plant Sci 165:S107–S21. 41. Stamatakis A, Ludwig T, Meier H (2005) RAxML-III: a fast program for maximum 19. Davis CC, Anderson WR, Donoghue MJ (2001) Phylogeny of Malpighiaceae: evidence likelihood-based inference of large phylogenetic trees. Bioinformatics 21:456–463. from chloroplast ndhF and trnL-F nucleotide sequences. Am J Bot 88:1830–1846. 42. Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogenetic 20. Michener CD (2000) The bees of the world (Johns Hopkins Univ Press, Baltimore). trees. Bioinformatics 17:754–755. 21. Davis CC (2002) Madagasikaria (Malpighiaceae): a new genus from Madagascar with 43. Kramer EM, Di Stilio VS, Schluter PM (2003) Complex patterns of gene duplication in implications for floral evolution in Malpighiaceae. Am J Bot 89:699–706. the APETALA3 and PISTILLATA lineages of the Ranunculaceae. Int J Plant Sci 164:1–11.

Zhang et al. PNAS | April 6, 2010 | vol. 107 | no. 14 | 6393 Supporting Information

Zhang et al. 10.1073/pnas.0910155107 SI Materials and Methods sequence variation across a diverse set of taxa. One set of nested Morphology-Based Character State Reconstruction of Floral degenerate primers (i.e., the forward primer located in the TCP Symmetry. We used maximum likelihood (ML) character state domain, 5′-GCIMGIAARTTYTTYGAYYTKCAA, and the re- reconstruction as implemented in Mesquite version 2.6 (1) to verse primer in the R domain, 5′-GCYCKYGCYCTIGCY- infer the evolution of floral symmetry in Malpighiaceae and its YTHKCYCTWGA), was selected to amplify a 350- to 400-bp closest relatives, Elatinaceae and Centroplacaceae. Floral mor- fragment of the CYC2-like genes from Malpighiaceae and its sister phology for 353 species was scored as either zygomorphic or clades. Gel-purified PCR products were cleaned and cloned. Fifty actinomorphic and treated as unordered and optimized onto the to 200 colonies were initially screened by restriction site analysis, rate-smoothed topology for the family (2, 3). Absolute di- and at least five clones were sequenced for each variant identified vergence time estimates were ascertained using the methods using this initial screen. Our criteria to further distinguish these described by Davis et al. (4, 5) and included nearly one quarter variants included the degree of nucleotide variability and the of all species of Malpighiaceae, plus numerous outgroup species presence of unique indels. Sequence variants containing no indels of Elantinaceace and Centroplacaceae. These data were ana- and differing by less than 5% sequence similarity were treated as lyzed using the Mk1 model (6) with rate parameters estimated alleles. The coding regions of CYC2A and CYC2B that fall be- directly from the data. tween the TCP and R domains are 176 ± 14 and 184 ± 5nu- cleotides long, respectively. CYC1-andCYC3-like genes, which Isolation of CYCLOIDEA Homologues in Malpighiales–Oxalidaceae– were rarely amplified with CYC2, were distinguished by using Celastrales. We used 3′ RACE to obtain all CYC homologues, in- phylogenetic analysis and excluded from subsequent consideration. cluding CYC1, CYC2,andCYC3, across the closely related clades Malpighiales, Oxalidales, and Celastrales (7). Fresh floral tissue Nucleotide Sequence Analyses. Our analyses in the main text focused was collected over a broad range of developmental stages (i.e., buds on amino acid sequence analyses. Analyses using nucleotide from multiple developmental stages from <1 mm to mature size, sequence data with third codon positions excluded, yielded a and open flowers) from four species of Malpighiaceae (Byrsonima topology nearly identical to the amino acid sequence data. The lucida DC., Janusia guaranitica A. Juss., Malpighia coccigera L., and parameters of the best-fit model for our nucleotide data were Tristellateia australasiae A. Rich.), two additional species of Mal- estimated using MODELTEST 3.06 (9). The Akaike Information pighiales [Hypericum perforatum L. (Hypericaceae) and Euphorbia Criterion (10) recommended a general time reversible model with milii Des Moul. (Euphorbiaceae)], one species of Oxalidales [Oxalis added parameters for invariable sites and a Γ distribution (“GTR herrerae R. Knuth (Oxalidaceae)], and one species of Celastrales +I+Γ”). One hundred ML bootstrap replicates were conducted [Euonymus alatus (Thunb.) Siebold (Celastraceae)]. with the optimal model of sequence evolution. Bayesian analyses Total RNAs were purified using the Concert Plant RNA were also conducted using this model. Reagent (Invitrogen). Single-stranded cDNA was synthesized from 5 μg of total RNA using SuperScript II reverse transcriptase Southern Hybridization. Ten micrograms of genomic DNA was (Invitrogen) by priming with the oligonucleotide 5′-CCGGATC digested from Bergia texana, Byrsonima crassifolia, J. guaranitica, CTCTAGAGCGGCCGC(T)17. This poly-T primer was then used and T. australasiae with restriction enzymes (i.e., HindIII, EcoRI, with primer 5′-AARGAYMGICAYAGYAARAT (modified and HindIII plus EcoRI), fractionated on 0.8% agarose gels, and from ref. 8) for PCR. PCR products were cleaned with the blotted onto a positively charged nylon membrane (GE Health- QIAquick PCR purification kit (Qiagen) and used as the template care BioSciences). In addition, for Janusia, we ran lanes contain- in a secondary PCR with the degenerate forward primer, 5′- ing CYC1, CYC2,andCYC3 plasmid DNA as controls to test GCIAGRAARTTYTTYGAYYTICARGAYATG, and the poly- probe efficiency and specificity. T primer described earlier. PCRs were performed in 100 μLof A fragment containing the 3′ end of the TCP domain and the buffer [60 mM Tris-SO4, pH 8.9; 18 mM (NH4)2SO4;2.5mM variable region between the TCP and R domains was used as a MgSO4]containing50pmol5′ primer, 10 pmol poly-T primer, 25 template to synthesize probes for detecting CYC2-like genes (Fig. pmol of dNTPs, and 3.75 units of PlatinumTaq polymerase (In- S3B). For J. guaranitica, a mixture of JgCYC2A (CYC2A) and vitrogen). The PCR amplification began with a 12-min activation JgCYC2B-3 (CYC2B) sequences in equal molar concentration was 32 step at 95 °C, followed by 37 cycles of a 1-min incubation step at used as a template to synthesize probes with P-dCTP (Perkin- 95 °C, a 30-s annealing step at temperatures ranging from 40 °C to Elmer) using the Prime-It II kit (Stratagene). This gene region 60 °C, and a 1-min extension at 72 °C. The reaction was termi- exhibits no more than 31% pair-wise sequence similarity among nated with a 10-min incubation at 72 °C. Gel-purified secondary the CYC1/2/3 paralogues, which ensures probe specificity. The PCR products were cloned using the pGEM-T easy vector system hybridization was carried out in hybridization solution [900 mM • × (Promega) and XL1-Blue competent cells (Stratagene). Fifty to NaCl; 60 mM NaH2PO4 H2O; 6 mM EDTA; 5 Denhart solution 200 colonies for each taxon were screened by restriction site (from 50×; Amresco); 1% SDS; 10 μg/mL sheared salmon sperm analysis and sequencing using the protocols established at Func- DNA; pH 7.4] at 65 °C for 18 h. The membranes were then washed • tional Biosciences. The identity of CYC-like genes was determined at low stringency (900 mM NaCl; 60 mM NaH2PO4 H2O; 6 mM by the presence of the highly conserved TCP domain. Phyloge- EDTA; 1% SDS; pH 7.4) at 65 °C and exposed for 90 h to phosphor netic analyses including numerous TCP genes from previously imaging (GE Healthcare Bio-Sciences) to detect for the presence published studies confirmed the identity of CYC homologues. of all CYC2 homologues. There are four bands for J. guaranitica in the EcoRI digest of Isolation of CYC2-Like Genes from Malpighiaceae–Elatinaceae– the CYC2 probe (Fig. S3A). This result is identical to our cloning Centroplacaceae–Oxalidaceae. Based on the CYC-like gene align- experiments, which also identified four copies of CYC2. The ment, we designed degenerate primers to amplify the CYC2-like CYC1 and CYC3 plasmid controls gave very faint signals ex- genes from genomic DNA. Our broad taxonomic sampling en- perimentally, demonstrating that the CYC2 probes are lineage sured that these degenerate primers covered a broad range of specific. The number of bands in the EcoRI digest therefore

Zhang et al. www.pnas.org/cgi/content/short/0910155107 1of10 reflects the approximate CYC2 copy number. In the HindIII and Notes on the Floral Orientation of Bergia texana and Bhesa double digests, we expected more than four bands as a result of paniculata. B. texana has a single dorsal petal with respect to the presence of a restriction site within the probed region. the stem, which is uncommon relative to most angiosperms, Giventheability of our probe todetect lineage specific CYC2gene which have one ventral petal (11, 12). Two dorsal and one copies, we conducted Southern hybridizations only for CYC2 on the ventral sepal are glandular, and two lateral sepals are eglandular remaining taxa (i.e., B. texana, B. crassifolia, and T. australasiae). (Fig. 3 in the main text). Moreover, the two dorsal sepals tightly BtCYC2-1 and BtCYC2-2 of B. texana, BcCYC2A and BcCYC2B clasp the stem, so it is relatively easy to determine that there is a from B. crassifolia,andTaCYC2A of T. australasiae were mixed in single dorsal petal in B. texana. In Centroplacaceae, the ori- equal molar concentrations and used as a template to synthesize our entation is more difficult to interpret. The flowers of B. pan- 32 fi P-labeled probes. Our CYC2-speci c probes revealed two bands in iculata are sessile, and three tightly congested flowers are B. crassifolia, one band in T. australasiae, and six bands in B. texana commonly borne in an inflorescence. More than one kind of (see EcoRI digest in Fig. S4), which was identical to the CYC2 copy floral orientation was observed, in which case these flowers may number inferred by PCR and cloning. twist during development.

Floral Organ Arrangement in New and Old World Malpighiaceae. New Character State Reconstruction of CYC2-Like Gene Expression. We World Malpighiaceae, e.g., B. crassifolia and J. guaranitica, pos- fl used ML character state optimization as implemented in Mes- sess a dorsal ag/banner petal that is always innermost in bud. quite 2.6 (1) to reconstruct the evolution of CYC2-like gene The remaining four petals can be arranged in one of two ways expression. We used the single ML topology inferred from CYC2 that form mirror images of each other (Fig. S5A). These enan- nucleotide sequences to reconstruct the pattern of CYC2 gene tiomorphic flowers can be found on the same inflorescence within a single species. The Old World species T. australasiae has expression under the general Mk1 model (6) with the rate pa- a similar petal aestivation (Fig. S5B; also see main text). The rameter estimated from the data. The character states of gene innermost petal of T. australasiae is homologous to the New expression were coded as follows: none, uniform, broad differ- World dorsal flag/banner petal and can also initiate on the left or ential, and narrow differential. Ancestral character states were right side of the dorsoventral plane of symmetry (Fig. S5B). We reconstructed assuming that all transition states are unordered. used the relative positions of these floral organs to sample ho- Although we used genetic distance as an approximate measure mologous tissue types from these New and Old World species, of the “opportunity for selection” (13), we also conducted our and took great care to sample only those Tristellateia flowers for analysis with the topology calibrated for absolute divergence RT-PCR in which the innermost petal was located on the left time estimates (4, 5, 14). Those results are very similar to those side of the dorsoventral plane of symmetry. presented here and do not affect our conclusions.

1. Maddison WP, Maddison DR (2009) Mesquite: a modular system for evolutionary 8. Howarth DG, Donoghue MJ (2005) Duplications in CYC-like genes from Dipsacales analysis, version 2.6. Available at http://mesquiteproject.org/mesquite/mesquite.html. correlate with floral form. Int J Plant Sci 166:357–370. Accessed June 25, 2009. 9. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. 2. Zhang W, Kramer EM, Davis CC (2009) Exploring the developmental genetic basis of Bioinformatics 14:817–818. independent reversals in floral symmetry in Malpighiaceae. Botany & Mycology 2009: 10. Akaike H (1973) Petrov BN, Csaki F, eds (1973) , ed (1973) Information theory and an Annual Meeting of the Botanical Society of America in Snowbird, Utah, USA.Available extension of the maximum likelihood principle. Second International Symposium in at http://2009.botanyconference.org/engine/search/index.php?func=detail&aid=851. Information Theory, eds Petrov BN, Csaki F (Akademiai Kiado, Budapest), pp 267–281. 3. Zhang W, Kramer EM, Davis CC (2009) CYCLOIDEA2 and the origin andmaintenance 11. Donoghue MJ, Ree RH (2000) Homoplasy and developmental constraint: A model and of floral zygomorphy in Malpighiaceae. Botany & Mycology 2009: Annual Meeting of an example from plants. Am Zool 40:759–769. the Botanical Society of America in Snowbird, Utah, USA. Available at http://2009. 12. Eichler AW (1875) Blüthendiagramme construirt und erläutert. 2 theile in 1 (W. botanyconference.org/engine/search/index.php?func=detail&aid=546. Engelmann, Leipzig). 4. Davis CC, Bell CD, Mathews S, Donoghue MJ (2002) Laurasian migration explains 13. Pagel M (1997) Inferring evolutionary processes from phylogenies. Zool Scr 26: Gondwanan disjunctions: Evidence from Malpighiaceae. Proc Natl Acad Sci USA 99: 331–348. 6833–6837. 14. Davis CC, Fritsch PW, Bell CD, Mathews S (2004) High-latitude tertiary migrations of an 5. Davis CC, et al. (2005) Explosive radiation of Malpighiales supports a mid-Cretaceous exclusively tropical clade: Evidence from Malpighiaceae. Int J Plant Sci 165:S107–S121. origin of modern tropical rain forests. Am Nat 165:E36–E65. 15. Davis CC, Chase MW (2004) Elatinaceae are sister to Malpighiaceae; Peridiscaceae 6. Lewis PO (2001) A likelihood approach to estimating phylogeny from discrete belong to Saxifragales. Amer J Bot 91:262–273. morphological character data. Syst Biol 50:913–925. 16. Davis CC, Anderson WR, Donoghue MJ (2001) Phylogeny of Malpighiaceae: Evidence 7. Stevens PF (2001 onwards) Angiosperm Phylogeny Website; version 9, June 2008. from chloroplast ndhF and trnL-F nucleotide sequences. Amer J Bot 88:1830–1846. Available at http://www.mobot.org/MOBOT/research/APweb/. Accessed January 5, 17. Wurdack KJ, Davis CC (2009) Malpighiales phylogenetics: gaining ground on one of 2010. the most recalcitrant clades in the angiosperm tree of life. Am J Bot 96:1551–1570.

Zhang et al. www.pnas.org/cgi/content/short/0910155107 2of10 Centroplacus657kw Verrucularia glaucophylla394 Galphimia glauca1981 Galphimia multicaulis1603 Galphimia mexiae1584 Galphimia glandulosa1582 Centroplacaceae Galphimia mirandae1597 Galphimia_sp.3_1598 Galphimia platyphylla1577 Galphimia brasiliensis1639 Galphimia gracilis Lophanthera hammelii1673 Spachea correae Spachea elegans1675 Lophanthera lactescens Lophanthera pendula1553 Lophanthera longifolia1592 Byrsonima lucida583 Byrsonima coccolobifolia633 Byrsonima duckeana652 Byrsonima crispa748 Byrsonima basiloba634 Byrsonima macrophylla632 Byrsonima crassifolia Byrsonima morii770 Byrsonima triopterifolia641 Blepharandra heteropetala Blepharandra hypoleuca710 Blepharandra fimbriata712 Diacidia ferruginea Diacidia galphimioides1928 Coleostachys genipifolia Pterandra arborea188 Acmanthera latifolia Brachylophon curtisii281 Acridocarpus excelsus66 Acridocarpus adenophorus602 Acridocarpus smeathmannii168 Acridocarpus_sp.1184 Acridocarpus spectabilis181 Acridocarpus chevalieri170 Acridocarpus alternifolius171 Acridocarpus orientalis68 Acridocarpus natalitius50 Acridocarpus zanzibaricus177 Acridocarpus scheffleri1183 Acridocarpus macrocalyx166 Acridocarpus staudtii164 Glandonia macrocarpa47 Burdachia sphaerocarpa Mcvaughia bahiana732 Malpighiaceae Barnebya dispar686 Dinemagonum gayanum700 Dinemandra ericoides119 Ptilochaeta nudipes48 Ptilochaeta bahiensis105 Lasiocarpus_sp.724 Lasiocarpus_sp.228 Henleophytum echinatum460 Heladena multiflora Tristellateia madagascariensis Tristellateia_sp.1180 Tristellateia_sp.378 Tristellateia_sp.373 Tristellateia africana179 Tristellateia_sp.1500 Tristellateia_sp.1182 Thryallis latifolia1739 Thryallis longifolia Bunchosia macrophylla600 Bunchosia mcvaughii685 Bunchosia nitida1783 Bunchosia glandulifera651 Bunchosia odorata1784 Bunchosia swartziana1980 Bunchosia ternata680 Bunchosia veluticarpa679 Bunchosia mollis1782 Bunchosia glandulosa1775 Bunchosia armeniaca1764 Bunchosia angustifolia1983 Bunchosia pilocarpa1786 Bunchosia deflexa1771 Bunchosia decussiflora1770 Bunchosia apiculata1760 Bunchosia polystachia1788 Bunchosia glandulosa1551 Bunchosia armeniaca Echinopterys eglandulosa227 Lophopterys inpana1987 Lophopterys floribunda Hiraea fagifolia Hiraea wiedeana689 Hiraea smilacina277 Hiraea_sp.756 Mascagnia hiraea Mascagnia dipholiphylla Mascagnia stannea841 Aenigmatanthera doniana1522 Elatinaceae Mascagnia lasiandra1152 Mascagnia lasiandra847 Mascagnia lasiandra1153 Mascagnia lasiandra1151 Mascagnia stannea99 Tetrapterys acutifolia812 Tetrapterys acutifolia813 Tetrapterys sericea811 Tetrapterys sericea692 Tetrapterys arcana2095 Tetrapterys ambigua1220 Tetrapterys salicifolia619 Tetrapterys microphylla162 Tetrapterys phlomoides635 Tetrapterys schiedeana755 Tetrapterys tinifolia670 Tetrapterys goudotiana675 Tetrapterys megalantha751 Tetrapterys discolor665 Christianella surinamensis778 Christianella multiglandulosa1527 Jubelina riparia1709 Jubelina rosea137 Jubelina uleana1683 Jubelina wilburii293 Mascagnia bracteosa Flabellaria paniculata126 Mezia includens296 Mezia araujoi Mascagnia anisopetala Callaeum psilophyllum1703 Callaeum psilophyllum731 Callaeum antifebrile1697 Callaeum septentrionale Callaeum clavipetalum1715 Callaeum malpighioides1702 Callaeum nicaraguense1679 Heteropterys steyermarkii426 Heteropterys byrsonimifolia419 Heteropterys leona162 Heteropterys aureosericea475 Heteropterys catingarum415 Heteropterys lindleyana414 Heteropterys ternstroemiifolia413 Heteropterys sanctorum412 Heteropterys conformis462 Heteropterys ciliata52 Heteropterys cordifolia506 Heteropterys nordestina435 Heteropterys imperata439 Heteropterys capixaba438 Heteropterys sincorensis473 Heteropterys sericea418 Heteropterys nitida423 Heteropterys thomasii464 Heteropterys chrysophylla433 Heteropterys bahiensis468 Heteropterys macrostachya668 Heteropterys megaptera467 Heteropterys bicolor457 Clonodia racemosa838 Clonodia ovata839 Heteropterys dumetorum465 Heteropterys angustifolia742 Heteropterys rhopalifolia430 Heteropterys trichanthera428 Heteropterys rufula511 Heteropterys brachiata584 Heteropterys pteropetala629 Heteropterys leschenaultiana425 Heteropterys pauciflora422 Tricomaria usillo Dicella bracteosa1716 Dicella nucifera94 Dicella macroptera1164 Dicella macroptera1165 Dicella macroptera1166 Dicella julianii1707 Flabellariopsis acuminata716 Hiptage candicans799 Hiptage_sp.1126 Hiptage benghalensis268 Hiptage benghalensis1511 Hiptage detergens760 Hiptage_sp.1127 Carolus anderssonii1538 Carolus chlorocarpus1540 Mascagnia chasei Carolus sinemariensis775 Carolus sinemariensis1543 Banisteriopsis mathiasiae1148 Banisteriopsis ferruginea1853 Banisteriopsis cornifoliaKW982 Banisteriopsis cinerascens1944 Banisteriopsis sellowiana1986 Banisteriopsis nummifera1863 Banisteriopsis gardneriana1977 Banisteriopsis martiniana749 Banisteriopsis prancei1925 Banisteriopsis calcicola1850 Banisteriopsis angustifolia1848 Banisteriopsis muricata674 Banisteriopsis_aff._malifolia750 Banisteriopsis schwannioides1871 Banisteriopsis caapi669 Banisteriopsis adenopoda1198 Banisteriopsis pulchra1870 Banisteriopsis confusa1945 Banisteriopsis latifolia1948 Banisteriopsis laevifolia622 Banisteriopsis acerosa1943 Banisteriopsis argyrophylla1201 Banisteriopsis harleyi1946 Banisteriopsis paraguariensis1866 Sphedamnocarpus pruriens Sphedamnocarpus galphimiifolius708 Sphedamnocarpus galphimiifolius1517 Sphedamnocarpus angolensis1518 Sphedamnocarpus_sp.1335 Sphedamnocarpus_sp.380 Sphedamnocarpus poissoni1338 Sphedamnocarpus_sp.369 Philgamia glabrifolia1318 Philgamia hibbetioides1319 Peixotoa parviflora1669 Peixotoa reticulata1670 Peixotoa cordistipula1695 Peixotoa paludosa1668 Peixotoa glabra102 Gallardoa fischeri134 Cordobia argentea Cordobia argentea1179 Mionandra camareoides Janusia californica134 Janusia linearis Janusia janusioides1749 Janusia hexandra1720 Janusia mediterranea1728 Janusia mediterranea136 Janusia anisandra133 Janusia christianeae1994 Janusia guaranitica1514 Janusia guaranitica707 Janusia guaranitica1516 Peregrina linearifolia103 Aspicarpa pulchella86 Aspicarpa harleyi746 Aspicarpa sericea614 Aspicarpa sericea1678 Gaudichaudia krusei2076 Gaudichaudia mcvaughii130 Aspicarpa brevipes Gaudichaudia arnottiana705 Gaudichaudia cynanchoides597 Camarea axillaris455 Aspicarpa hyssopifolia2098 Gaudichaudia albida Aspicarpa hirtella Gaudichaudia hexandra2082 Stigmaphyllon paralias110 Stigmaphyllon sagraeanum729 Stigmaphyllon sagraeanum112 Stigmaphyllon sagraeanum1549 Stigmaphyllon aberrans1734 Stigmaphyllon calcaratum615 Stigmaphyllon lindenianum673 Stigmaphyllon finlayanum663 Ryssopterys tiliifolia758 Ryssopterys timoriensis694 Ryssopterys_sp.820 Ryssopterys_sp.397 Stigmaphyllon ciliatum647 Stigmaphyllon bogotense1735 Stigmaphyllon bogotense1885 Stigmaphyllon puberum111 Diplopterys pubipetala1905 Diplopterys virgultosa1908 Diplopterys nutans1902 Diplopterys cabrerana274 Banisteriopsis hypericifolia Banisteriopsis lutea730 Banisteriopsis valvata1239 Ectopopterys soejartoi125 Amorimia exotropica1534 Mascagnia rigida295 Amorimia velutina1535 Amorimia amazonica842 Amorimia pubiflora843 Amorimia kariniana1536 Mascagnia parvifolia623 Caucanthus auriculatus1175 Caucanthus auriculatus Triaspis nelsonii1090 Triaspis nelsonii1091 Triaspis glaucophylla558 Triaspis hypericoides1088 Triaspis hypericoides Triaspis hypericoides1089 Actinomorphy Triaspis niedenzuiana1176 Triaspis_sp.1176 Triaspis macrapteron1174 Digoniopterys microphylla943 Microsteira diotostigma1661 Microsteira_sp.379 Microsteira_sp.1662 Microsteira_sp.1186 Rhynchophora humbertii372 Zygomorphy Madagasikaria andersonii374 Rhynchophora phillipsonii377 Caucanthus edulis1093 Caucanthus albidus551 Aspidopterys elliptica693 Aspidopterys elliptica239 Aspidopterys_sp.1283 Aspidopterys_sp.1284 Aspidopterys tomentosa549 Malpighia setosa1569 Malpighia cnide1568 Malpighia albiflora757 Malpighia mexicana1559 Mascagnia leticiana800 Malpighia souzae1572 Malpighia emarginata Malpighia stevensii27 Malpighia romeroana1561 Malpighia_sp.720 Malpighia glabra580 Malpighia coccigera593 Malpighia urens1643 Malpighia fucata1557 Malpighia incana588 Malpighia incana1550 Mascagnia tomentosa1922 Mascagnia tomentosa2064 Mascagnia australis744 Mascagnia arenicola662 Mascagnia arenicola Mascagnia cordifolia621 Mascagnia affinis1909 Mascagnia dissimilis1911 Mascagnia lilacina1913 Mascagnia polybotrya824 Mascagnia polybotrya832 Triopterys jamaicensis Triopterys paniculata802 Mascagnia brevifolia834 Mascagnia divaricata743 Mascagnia tenuifolia1936 Mascagnia vacciniifolia754 Elatine_sp.Qiu99051 Elatine triandra521 Elatine minima524 Bergia texana514 Bergia pedicellaris769

Fig. S1. Ancestral character state reconstruction of floral symmetry. ML analysis indicates the relative likelihood of floral symmetry at the nodes of Centroplacaceae– Elatinaceae–Malpighiaceae (actinomorphy, 0.77; zygomorphy, 0.23), Elatinaceae–Malpighaceae (actinomorphy, 0.75; zygomorphy, 0.25), and crown group Malpighiaceae (actinomorphy, 0.0; zygomorphy, 1.0). These results indicate that the common ancestor of Centroplacaceae–Elatinaceae–Malpighiaceae and Elatinaceae–Malpighiaceae are likely actinomorphic, and that zygomorphy evolved in the common ancestor of all Malpighiaceae.

Zhang et al. www.pnas.org/cgi/content/short/0910155107 3of10 Oxalis (Oh) Oxalidaceae Bhesa (Bp) Centroplacaceae Bergia (Bt) Elatinaceae Elatine (Em) Byrsonima (Bc) Diacidia (Diag) Galphimia (Gm, Gg) Verrucularria (Vg) Spachea (Se) Lophanthera (Lp, Ll) Acridocarpus (As) Lasiocarpus (Las) Ptilochaeta (Pb, Pn) Dinemagonum (Ding) Dinemandra (De) Echinopterys (Ee) Tristellateia (Ta, Taf) Malpighiaceae Henleophytum (He) Heladena (Hb) Hiptage (Hd) Flabellariopsis (Fa) Mascagnia (Mb) Flabellaria (Fp) Madagasikaria (Ma) Malpighia (Mc) Stigmaphyllon (Sp) Ryssopterys (Rt) Janusia (Jg) Banisteriopsis (Bl) Sphedamnocarpus (Sph)

Fig. S2. Phylogeny showing accepted species relationships (derived from refs. 2, 3, 15–17).

Zhang et al. www.pnas.org/cgi/content/short/0910155107 4of10 A. CYC2 Probe: JgCYC2A and JgCYC2B-3

E H E+H CYC2 CYC1 CYC3

12,000

6,000

3,000

1,600

500

B.

5’ TCP R 3’

CYC2 probe C.

HindIII (129) JgCYC2A (299) HindIII (123) HindIII (251) JgCYC2B-1 (278) HindIII (120) HindIII (239) JgCYC2B-2 (266)

JgCYC2B-3 (278)

Fig. S3. CYC2 Southern hybridization results for J. guaranitica.(A) To assess gene copy number, we developed a CYC2-specific probe and tested it against J. guaranitica. That probe preferentially identified only CYC2 and not other CYCLOIDEA homologues in Janusia, i.e., CYC1 and CYC3. The result of this test of the probe’s specificity is shown on the right side of the Janusia blot. Restriction digests using EcoRI (E), HindIII (H), and EcoRI + HindIII (E+H) are shown for Janusia.(B) The position of the probe region within CYC-like genes. (C) Restriction cut sites were inferred from sequence analysis and are indicated on the CYC2 gene copies shown at bottom. Arrows and numbers indicate molecular size markers (in base pairs).

Zhang et al. www.pnas.org/cgi/content/short/0910155107 5of10 A. B.

Byrsonima Tristellateia Bergia Byrsonima crassifolia

E H E+H E H E+H E H E+H HindIII (123) BcCYC2A (266) HindIII (126) BcCYC2B (275)

12,000 Tristellateia australasiae

HindIII (111) 6,000 TaCYC2A (260)

Bergia texana

EcoRI (248) 3,000 BtCYC2-1 (269)

BtCYC2-2 (272) BtCYC2-3 (278)

1,600 BtCYC2-4 (272) HindIII (241) BtCYC2-5 (275) EcoRI (254) (275) 1,000 BtCYC2-6

500

Fig. S4. CYC2-like gene Southern hybridization results for Byrsonima crassifolia, T. australasiae, and Bergia texana.(A) Restriction digests using EcoRI (E), HindIII (H), and EcoRI + HindIII (E+H) are shown for B. crassifolia, T. australasiae, and B. texana.(B) Restriction cut sites were determined from sequence analysis and are indicated on the CYC2 gene copies shown at bottom. Arrows and numbers indicate molecular size markers (in base pairs).

Zhang et al. www.pnas.org/cgi/content/short/0910155107 6of10 A.

dp dp

lp1 lp1 lp2 lp2

vp1 vp1

vp2 vp2

B.

2 2 1 1

3 5 5 3

4 4

Fig. S5. Floral aestivation and enantiomorphy of Malpighiaceae. (A) New World Malpighiaceae. Petal identities are indicated as follows: dp, dorsal petal; lp1-2, lateral petals; vp1-2, ventral petals. (B) T. australasiae, an Old World Malpighiaceae species. The dorsoventral planes of floral symmetry are indicated with a colored vertical line. Petal identities are indicated as follows: 1 and 2, dorsal petals; 3 and 5, lateral petals; 4, ventral petal. The petals are not intended to be drawn proportional to scale in Tristellateia.

Zhang et al. www.pnas.org/cgi/content/short/0910155107 7of10

dp/1 dp/1 lp1/2

lp2/5 /2 lp2/5 lp1

vp1 /3 vp1/3

vp2/4 vp2/4

Tristellateia Tristellateia NW NW

Fig. S6. Hypothesized shift in the dorsoventral plane of symmetry in the Old World species T. australasiae. Floral arrangement of New World Malpighiaceae shown for comparison. The dorsoventral planes of floral symmetry of New World Malpighiaceae and Tristellateia are indicated in orange and blue, re- spectively. The arrow illustrates the hypothesized 36° rotation in Tristellateia relative to their New World ancestors. For the New World arrangement ab- breviations are as follows: dp, dorsal petal; lp1-2, lateral petals; vp1-2, ventral petals. For the Old World arrangement, abbreviations are as follows: 1 and 2, dorsal petals; 3 and 5, lateral petals; 4, ventral petal.

Centroplacaceae Elatinaceae Malpighiaceae BpCYC2 BtCYC2-4 BtCYC2-2 BtCYC2-3 BtCYC2-1 BtCYC2-6 BtCYC2-5 BcCYC2A JgCYC2A BcCYC2B JgCYC2B-1 JgCYC2B-2 JgCYC2B-3

CYC2A CYC2B

Expression patterns: none CYC2 uniform differential broad differential narrow

Fig. S7. Character state reconstruction of CYC2 gene expression. Expression patterns are treated as character states shown in different colors. Areas of pies indicate the relative degree of support for alternative ancestral character states. Gray highlighting in flower diagram indicates the spatial pattern of CYC2 expression. The most recent common ancestor of Centroplacaceae–Elatinaceae–Malpighiaceae likely exhibited the pattern of gene expression observed in Centroplacaceae (Bhesa paniculata). CYC2 expression in the most recent common ancestor of Malpighiaceae–Elatinaceae is equivocal. The most recent common ancestor of Malpighiaceae likely exhibited dif- ferential expression of CYC2 before the CYC2A/B duplication. In addition, our results support the independent evolution of a broader pattern of gene expression of JgCYC2B-3, one of the CYC2B copies in J. guaranitica, from a narrow patterned CYC2B ancestor.

Zhang et al. www.pnas.org/cgi/content/short/0910155107 8of10 Table S1. Species sampled, with collection locations, voucher information, and CYC2 identities Identity of obtained CYC2 copies

Species Location Voucher 22A2B

Acridocarpus smeathmanni Ghana Davis 99–13 (A) – AsCYC2A AsCYC2B Guill. and Perr. Banisteriopsis latifolia Distrito Federal, Brazil Azeuedo 698 (MICH) – BlCYC2A BlCYC2B-1, (A.Juss.) B. Gates BlCYC2B-2 Bergia texana Seub. ex Walp. Butte County, Zhang, Ahart, and BtCYC2-1 –– California, US Bartholomew 84 (A) ∼ BtCYC2-6 Bhesa paniculata Arn. Negeri Sembilan, Malaysia Zhang and Boufford 160 (A) BpCYC2 –– Byrsonima crassifolia Kunth Cult. OEB, Harvard U. Matamoros and Cerda 301 – BcCYC2A BcCYC2B Diacidia galphimioides Griseb. Amazonas, Venezuela Berry et al., 5275 (MICH) – DiagCYC2A - Dinemagonum gayanum A.Juss. Chile Simpson 83–10-23–2c (MICH) – DingCYC2A DingCYC2B Dinemandra ericoides A.Juss. Chile Dillon and Teillier 5103 (MICH) – DeCYC2A DeCYC2B Echinopterys eglandulosa Sonora, Mexico Van Devender 98–178 (MICH) – EeCYC2A EeCYC2B (A. Juss.) Small Elatine minima (Nutt.) Fisch. & Gemini lake, North Voss 11739 (MICH) EmCYC2-1, – – C. A. Meyer Michigan, US EmCYC2-2 Flabellaria paniculata Cav. Tanzania Congdon 414 (K) –– FpCYC2B Flabellariopsis acuminata Tanzania Faulkner 783 (K) – FaCYC2A FaCYC2B-1, (Engl.) R. Wilczek FaCYC2B-2 Galphimia gracilis Bartl. Fairchild T.G., Florida, US FTG 79–235 (FTG) – GgCYC2A-1, GgCYC2B GgCYC2A-2 Galphimia mexiae C.E. Anderson Jalisco, Mexico Anderson and Anderson – GmCYC2A GmCYC2B 6122 (MICH) Heladena bunchosioides A. Juss. Espírito Santo, Brazil Folli 4653 (MICH) – HbCYC2A HbCYC2B Henleophytum echinatum nr. Havana, Cuba Curtiss 688 (K, NY) – HeCYC2A HeCYC2B (Griseb.) Small Hiptage detergens Craib Thailand Middleton et al., – HdCYC2A-1, HdCYC2B-1, 2095 (A, MICH) HdCYC2A-2 HdCYC2B-2 J. guaranitica A. Juss. Cult. OEB, Harvard U. Zhang 165 (A) – JgCYC2A JgCYC2B-1, JgCYC2B-2, JgCYC2B-3 Lasiocarpus sp. Mexico Anderson 13828 (MICH) – LasCYC2A LasCYC2B Lophanthera longifolia Griseb. Amazonas, Venezuela Zimmerman 27 (MICH) – LlCYC2A LlCYC2B Lophanthera pendula Ducke Amazonas, Brazil Lima and Lima 3185 (MICH) – LpCYC2A LpCYC2B Madagasikaria andersonii Madagascar Davis 20–01 (A) – MaCYC2A – C. Davis Malpighia coccigera L. UMBG UMBG 20626 (MICH) – McCYC2A McCYC2B Mascagnia bracteosa Griseb. Manaus, Brazil Anderson 13777 (MICH) – MbCYC2A MbCYC2B Oxalis herrerae R.Knuth Cult. OEB, Harvard U. Zhang 20 (A) OhCYC2-1, –– OhCYC2-2 Ptilochaeta bahiensis Turcz. Bahia, Brazil Anderson 13725 (MICH) –– PbCYC2B-1, PbCYC2B-2, PbCYC2B-3 Ptilochaeta nudipes Griseb. Jujuy, Argentina Anderson 13588 (MICH) – PnCYC2A PnCYC2B-1, PnCYC2B-2 Ryssopterys timoriensis (DC.) Cult. Bogor XVIII.F.172 (BO) – RtCYC2A RtCYC2B-1, Blume ex A. Juss. RtCYC2B-2 Spachea elegans A. Juss. Guyana Janson-Jacobs et al., – SeCYC2A SeCYC2B-1, 3907 (MICH) SeCYC2B-2 Sphedamnocarpus sp. Madagascar Phillipson et al., 4104 –– SphCYC2B (MICH, MO, P) Stigmaphyllon paralias A. Juss. Bahia, Brazil Anderson 13693 (MICH) –– SpCYC2B Tristellateia africana S. Moore Dar es Salaam, Tanzania Davis 99–25 (A) – TafCYC2A-1, – TafCYC2A-2, TafCYC2A-3 T. australasiae A. Rich. Cult. OEB, Harvard U. Zhang 163 (A) – TaCYC2A – Verrucularia glaucophylla Bahia, Brazil Amorim 3662 – VgCYC2A-1, VgCYC2B A. Juss. (CEPEC, MICH) VgCYC2A-2

Arnold Arboretum (Arn. Arb.) is in Jamaica Plain, MA. A, Arnold Herbarium, Harvard University Herbaria; BO, Herbarium Bogoriense, Bogor, West Java, Indonesia; CEPEC, Herbário Centro de Pesquisas do Cacau, Bahia, Brazil; FTG, Fairchild Tropical Botanic Garden; K, Royal Botanic Gardens, Kew, England; MICH, University of Michigan Herbarium; MO, Missouri Botanical Garden, St. Louis, Missouri; NY, New York Botanical Garden, Bronx, New York; P, Muséum National d’Histoire Naturelle, Paris. GenBank numbers are given for each copy found. The GenBank numbers are GU982187–GU982264.

Zhang et al. www.pnas.org/cgi/content/short/0910155107 9of10 Table S2. RT-PCR primer sequences used in this study Name Taxa Forward (5′ to 3′) Reverse (5′ to 3′)

BtCYC2-1 Bergia texana GGTCTTACAATTCTACTAGTGAATTACTTG GAATTCTGGAAGCAAACTTTTGTAAT BtCYC2 (-2, -4) Bergia texana CAAGAWACTYTAGGGTTTGATAAAGCAA CAAATGACTCCATYTTTGAAACTGTCC BtCYC2 (-3, -5, -6) Bergia texana CAAGAWACTYTAGGGTTTGATAAAGCAA GAATCCAMCTTTGAAACTSTTGTCC BpCYC2 Bhesa paniculata GATCTTCAAGACATTCTAGGGTTTGAC GACTCCTTTGCAAGAAGTGTACTG BcCYC2A Byrsonima crassifolia AAGATTTGTTAGGGTTTGATAGGG AGGTCTCATTTCACTATAATCAACACA BcCYC2B Byrsonima crassifolia AAGACCTTCTAGGGTTTGATAGGG TCTCCCCTCACTAGAATCAACAGT JgCYC2A J. guaranitica ACAATCTCTGGAGCTGAAAAGG ACTCACATCCTGCCTGAACC JgCYC2B-1 J. guaranitica TTCCATAYTCAAGATCCGATTTA CTCAATTGTTCTGATGATGACCT JgCYC2B-2 J. guaranitica TTCCATAYTCAAGATCCGATTTA CTTCCCATGATTTGCAGTATACTTATT JgCYC2B-3 J. guaranitica CTAACAAGCGATCAAATCGAA GATCTCAATTGTTCTGGTGATCTT TaCYC2A T. australasiae TTAGGGTTTGACAGGGCAAG GCTTAGCAAGAAGTGGGATTT

Zhang et al. www.pnas.org/cgi/content/short/0910155107 10 of 10