The main developmentally and genetically-based hypotheses and theories on the origin of angiosperms and the flower are Asama's Growth Retardation Theory (Asama 1982, 1985), Meyen's gamoheterotophic hypothesis (Meyen 1986, Meyen 1988), Frohlich and Parker's Mostly-male Theory (Frohlich and Parker 2000), Becker and Theißen's out-of-male and out-of-female hypotheses (Becker and Theißen 2003), Baum and Hileman's model of cone and floral development from hermaphroditic strobili (2006), and modifications of the latter two ideas on bisexual strobilus development suggested by Theißen and Melzer (2007).

Another interpretation of the origin of angiosperms from bisexual cone axes, based upon molecular divergence in LEAFY (LFY) genes about 130 million years ago, was published by Albert et al. (2002) who link the elimination of one LFY gene paralog "... to the sudden appearance of flowers in the fossil record" (abstract, Albert et al. 2002).

"Our theory may also be tested, without the vagaries of fossil discovery, in modern plants. Of the early-acting genes that control flower development, more should have close homologs (or orthologs, if gene trees are sufficiently resolved to demonstrate orthology) active in male gymnosperm reproductive structures rather than in female structures... In the coming years, topology-violating transformations that are of evolutionary importance may become recognizable using evidence from genes that control development."

The above quotation is from page 167 of M. W. Frohlich and D. S. Parker (2000), The mostly male theory of flower evolutionary origins: from genes to fossils (MMT), Systematic Botany 25(2): 155-170.

Frohlich suggests in two updates of MMT (2002, 2003) that early Mesozoic Corystospermales might be one of the possible ancestors of flowering plants. Frohlich (2003) placed an interesting idea on the table, namely that potential pollinators of gymnosperms were attracted by ectopic ovules displayed on upper (abaxial) leaf surfaces.

For example, none of the work on Permian gigantopterids published to date convincingly demonstrates physical connections between reproductive structures, stems, roots, and leaves. Whole plant morphologies are incompletely known for this group of Paleozoic seed plants (pages 758-763, T. N. Taylor et al. 2009).

I now state a coevolutionary hypothesis on the origin of flowering plants and certain holometabolous insect antagonists:

Coevolution between phytophagous insect antagonists and Carboniferous, Permian, and Triassic seed plant hosts at the level of their respective developmental tool kits with focus on selective forces that drive the logic of transcriptional regulation is proposed to explain the origin of angiosperms and certain clades of holometabolous insects.

I advance Takhtajan's ideas on flowering plant origins (1969, 1976) from 21st century research perspectives of evolutionary-developmental (evo-devo) biology, melded with rethinking of the paleontology of detached, fossilized remains of overlooked Paleozoic seed plant organs that might potentially be derived from massive, developmentally plastic, shoot apical meristems (SAMs) of gigantopteroid shrubs, vines, or small trees.

During the hypoxic late Devonian-early Carboniferous ice-house (DeCARB), invertebrate Hox proteins, hemocyanin respiratory enzymes, and certain nuclear receptors of plecopteran (stonefly) ancestors diverged, leading to novel multimeric proteins in several hemimetabolous and holometabolous clades of late Paleozoic insects.

Candidates for the early insect developmental tool kit may also include certain homeotic selector genes of the Hox complex such as homologs and paralogs of abdominal A (abd-A), Abdominal B (Abd-B), Hox3, proboscipedia (pb), sex combs reduced (Scr), and Ultrabithorax (Ubx), the compartment-selector gene engraled; also including the proteins these genes encode such as Ubx and engraled, the ecdysone hormone signaling machinery, hemocyanins, hexamerins, and nuclear receptor proteins.

At the same time that homeotic genes and transcription factors were diverging in early Carboniferous insect clades, key vascular plant developmental genes including AP2, ARF, Class I and II TCP, Class III HD-Zip, GRAS, KNOX, LEAFY, MIKC-type MADS-box, PIN, TIR, and YABBY, and their modular transcription factors, were evolving into molecular novelties in Paleozoic seed plant lineages.

Coevolution of insect helix-turn-helix homeodomain proteins, seed plant homeotic helix-turn-helix LEAFY (LFY) proteins, other phytophagous insect- and seed plant host developmental cis-regulatory modules, and their respective developmental tool kits, potentially occurred during the DeCARB and/or end-Guadalupian carbon cycle event (GuCCE), with local hypoxia and temperature extremes as underlying and potentially powerful selective forces.

Global oxygen levels declined to 13% during the course of millions of years of the DeCARB. Resident chewing, crawling, ovipositing, piercing, and sucking insects probably became dependent on host seed plant shrubs and trees for food, oxygen gas; and shelter from predators, cold temperatures, and ultraviolet radiation. Phytophagous insect associates of Permo-Carboniferous vegetation potentially coevolved with their oxygen-generating plant hosts.

The GuCCE and associated global hypercapnia reinforced the already close association between Permian plants and protocoleopterans, weevils, and other hemi- and holometabolous insects, ostensibly established after the DeCARB. Resident phytophagous insect associates of developmentally plastic Permian seed plant shrubs including populations and species of Vojnovskyales and gigantopteroid seed plants, allowed stressed hemimetabolous insect mutualists to survive hot, hypoxic environments.

Permo-Triassic seed plants were probably confined to small isolated populations indigenous to fragmented biomes and fitness landscapes resulting from the deleterious effects of the end-Permian extinction (EPE). Surviving insects of the EPE and Triassic-Jurassic boundary carbon cycle event (TrCCE) fed, reproduced, and might have secreted biologically active molecules while trampling the surfaces of shoot apical meristems (SAMs) and accessory bisexual cone axes.

Signaling molecules such as jasmonates and volicitins originating in the bodies of phytophagous insect associates of Permo-Triassic shrubs and trees, when applied to growing points of the host plant with mechanical force might have affected biosynthesis of certain phytohormones including gibberellic acid (GA). Transduction of biologically active insect signaling molecules potentially stimulated biosynthesis of GA, activating, modifying, and upregulating homeotic selector and meristem identity genes affecting developmental switches in expanding bisexual cone axes.

Phytoecdysones originating in tissues of Carboniferous and Permian seed plants could have signalled the hemocyanin respiratory and hexamerin storage protein manufacturing machinery of eggs and instars of phytophagous insect antagonists profoundly affecting their development and body size.

Based on a trail of biochemical clues gleaned from studies of vegetative growth and SAM organization in extant model lignophytes and monilophytes, when combined with insight gained from molecular phylogenetic analyses of class III homeodomain leucine zipper (Class III HD-Zip) genes and KNOTTED1 homeodomain proteins at least two (possibly three or more) evolutionary-developmental (evo-devo) programs might have existed in early diverging Devonian progymnosperm and/or seed plant populations.

Many of the supposed homologies of leaves, sporophylls, and ectopic structures including reproductive axes and ovular integuments could be incorrect based on the possibility that a fundamental dichotomy in the evolution of SAM development emerged in Devonian vascular plant stock.

I propose that modern models of shoot apical meristem (SAM) trafficking of mobile cis-acting homeodomain transcription factors and certain promoters, enhancers, messenger RNAs, and phytohormones may be used to simplify morphological phylogenetic analysis of progymnosperms and seed plants, or to conduct certain cladistically-based experiments.

If this proposal is correct then traditional ideas on the homology of leaf and sporophyll characters, which are gene expression products of transcription factors downstream of the SAM patterning developmental tool kit, should be reconsidered.

Bisexual cone axes consisting of inner (terminal) Phasmatocycas bridwellii megasporophylls and an outer ring of microsporophylls subtended by a spirally-arranged perianth of tepals and colored andropetals potentially existed in populations of poorly understood Paleozoic seed plants described as gigantopterids and Vojnovskyales, groups omitted from most phylogenetic analyses.

Paleozoic megasporophylls described by Axsmith et al. (2003) as Phasmatocycas bridwellii were probably detached female pieces of much larger 300 million-year-old flower-like hermaphroditic organs. Further, no "assembly" (page 816, J. A. Doyle 2008) of perianth parts, microsporophylls, and megasporophylls was needed during early stages of angiosperm evolution prior to the end-Permian extinction.

I hypothesize that insect-mediated natural hybridization among Vojnovskyales and gigantopteroid seed plants might have been a method through which MIKC-type MADS-box genetic novelties were recruited into populations of ancestral angiosperms during the Permo-Carboniferous approximately 300 million years ago.

Enormous proanthostrobili known only from fragments and detached megasporophylls were probably modifications of developmentally plastic bisexual cone axes representing divergent clades of several Permian gymnosperms that survived the end-Permian apocalypse.

Carpel, floral, and ovular transcriptional regulation in extant angiosperm model organisms does not preclude derivation of evo-devo models that explain curling, inrolling, and fusion in 260- to 300 million-year-old ovule-bearing gigantopteroid Phasmatocycas bridwellii leaves to form carpels, ovaries, and pistils.

Neoteny to include condensation of hypothetical gigantopteroid proanthostrobili is the most simple evo-devo process to explain the origin of reproductive organs in Mesozoic crown group angiosperms and extant basal Amborellanae, Austrobaileyanae, Nymphaeanae, and Magnolianae from a putative 160 million year old ghost lineage.

Reproductive traits of certain developmentally plastic Paleozoic seed plants including early Mesozoic flowering plants are ostensibly intertwined with 200 to 300 million year-old gene duplications and subsequent molecular evolution of cis-acting floral protein quartets essential to meristem, organ, and reproductive module identity.

At the same time that gene duplications led to the evolution of floral quartets in Paleozoic shrub lifeboats, certain phytophagous insect associates underwent a second burst of molecular evolution of hexamerin storage proteins. Molecular evolution comprising the second burst of hexamerin diversification led to Coleoptera (bees), Diptera (flies), Hymenoptera (bees and wasps), and certain Lepidoptera (moths).

Moulting in developmentally plastic insect mutualists of Permian seed plant shrub- and tree-like lifeboats was probably affected by phytoecdysone hormones manufactured by the host seed plant. Signaling of respiratory protein genes of the insect genome by the plant, and mechanical and secretory activities of insect antagonists leading to signaling of host plant homeotic genes in SAMs, were likely hallmarks of Permo-Carboniferous and Permo-Triassic coevolutionary compartments.

Enveloped and surrounded by an acidified hypoxic desert, shrub lifeboats were venues where reciprocal, simultaneous, and specialized communication between invertebrate and plant took place at the molecular level. Insect hexamerins might have been hypoxia-inducible moulting storage proteins of the developmental tool kit of the larger-bodied insects of Carboniferous and Permian time. Insect bodies became smaller in plant-dwelling populations that survived the EPE and TrCCE.

Innovative traits of Permo-Triassic seed plants possibly included double fertilization, encapsulation of female gametophytes inside hardened ovules, folding of megasporophylls into carpels, and development of staminodes and stamens from laminar microsporophylls, but critical fossil evidence is needed. Populations of Permo-Triassic seed plants were ostensibly subjected to the vagaries of genetic drift, gene flow, meiotic drive, mutation, and natural selection of phenotypes resulting from developmental recombination, genetic accommodation, and gene duplications potentially leading to evolutionary change and possible angiospermization.

Permo-Triassic shrubs and trees ostensibly developed anatomical and biochemical defenses against both invertebrate and vertebrate herbivores possibly ameliorating mechanical damage to SAMs and accessory fertile meristems. Dependence of insect colonies on seed plant shrubs and trees might have been reinforced by paleoclimatic changes triggered by global catastrophe.

Selection pressures in populations of Carboniferous and Permo-Triassic seed plant populations were probably much greater than believed. An herbaceous origin of flowering plants cannot be explained by mutualism and coevolution of insects and seed-bearing shrub hosts alone. A Late Jurassic aquatic origin for flowering plants is incongruent with likely deleterious effects of sulfuric acid poisoning of lakes, shorelines, and wetlands, caused by basalt outpouring from the Central Atlantic Magmatic Province (CAMP), which is associated with the Triassic-Jurassic boundary carbon cycle event (TrCCE).

Assuming coevolution of phytophagous coleopterans with angiosperms, various proposals on rapid diversification of flowering plants during the Albian Age of the Gallic Epoch of the early Cretaceous Period are inconsistent with molecular-based phylogenies of Coleoptera that suggest a Triassic origin of certain non-chrysomelid beetle lineages. Concomitantly, existence of a 160 million-year-old angiosperm ghost lineage is likely. Therefore, extant basal flowering plants such as Amborella trichopoda are probably derived species.

Further, the origin of flowering plants and their coevolving insect hosts is potentially traceable to hybridization between unrelated Vojnovskyales and gigantopteroid seed plants about 300 million years ago. When supported by future phylogenetic analyses a monophyletic origin of flowering plants becomes less likely.

When taking into account the cyclic nature of angiospermization, flowering plants as traditionally defined, are a loose amalgam of parallel evolutionary lines traceable to surviving geographically disparate early Triassic remnants of already divergent gigantopteroid Permian seed plant lineages.

I now consider and discuss evidence drawn from studies of insect- and plant anatomy and development, ecology, evolution, physiology, genetics, and phylogenetics, which is critical in understanding the origin of flowering plants.

Throughout these peer-reviewed essays, which are subject to periodic revision, I adopt the integrative biologic approach. Adjustments and refinements are necessary to take into account ongoing review and study of the considerably large body of scientific evidence available to me from nearby research libraries at the University of California, Berkeley.

Anatomic and Developmental Considerations:

The following chapter of the essay considers evidence drawn from the research perspectives of insect- and plant anatomy and development.

The image to the right is the passive insect trapping flowering plant, Darlingtonia californica (Sarraceniaceae, Nepenthales, Dilleniidae), photographed by the author at a seep on Eight Dollar Mountain located in the Klamath Region of western North America.

Studies of evolving allometries and body plans might help us understand a possible coevolutionary origin of angiosperms and certain clades of holometabolous phytophagous insect antagonists.

Evolutionary development (evo-devo) of plants encapsulated in programs of development is "highly dynamic in evolutionary time" involving the evolution of cis-acting promoters (page 83, Baum 1998).

The evo-devo research perspective followed throughout the essays posted on this web page may be help us to decipher more than 400 million years of seed plant evolution and the enigmatic origin of flowering plants in deep time.

"The idea is that plants have a plastic and modular developmental system such that simple changes in regulatory genes need not lead to inviability but can generate novel, potentially favored phenotypes."

"Ontogeny in land plants can be viewed as a complex, partly hierarchical, series of developmental processes, which together with their underlying genetic controls, provide the raw material for morphological innovation. Ontogeny is thus the creative force behind botanical diversification, and small modifications at the genetic level may have a disproportionate effect on plant form as their consequences cascade and multiply through development. Plant evolution occurs as variation in genetic and epigenetic developmental processes is winnowed by ecology..."

The preceding quotation is from page 161 of P. R. Crane and P. Kenrick (1997), Diverted development of reproductive organs: a source of morphological innovation in land plants, Plant Systematics and Evolution 206: 161-174.

Body plans and allometric scaling. From the research perspectives of insect- and floral biology, and paleoentomology and floral origins, allometric scaling data might be applied to understanding potentially interesting aspects of escape and radiation at the interface between development and ecology (Enquist et al. 2007).

Sean B. Carroll (1995), Rogers et al. (1997), Gellon and McGinnis (1998), and S. B. Carroll et al. (2005) review molecular evolution of homeotic genes and homeodomain transcription factors needed to understand regulation of body ground plan development in phytophagous arthropod antagonists. Molecular control over arthropod growth varies among the major clades of insects (Grimaldi and Engel 2005).

A key paper on the control of insect body size (Nijhout 2003) outlines the molecular mechanisms (involving cis-acting transcription factors and hormones) and environmental controls (nutrition and temperature) behind growth and cell division in hemimetabolous and holometabolous insects.

Several neurosecretory hormones play an important part in mechanisms that regulate cell division and growth including insulin-like peptides (Drosophila insulin-like proteins [DILPs] and bombyxins), chitenase-derived imaginal disk factor proteins, the steroid hormone ecdysone, local autocrine and paracrine transcription factors, and brain neurosecretory prothoracicotropic hormone (PTTH) (Nijhout 2003).

Cessation of growth in holometabolous insects leading to a new moulting cycle is triggered by PTTH that initiates the ecdysone growth regulatory cascade. Prothoracicotropic hormone and/or ecdysone secretion in Holometabola is negatively controlled by juvenile hormone (JH) (Truman and Riddiford 2002). Once JH circulating in the hemolymph is destroyed by juvenile hormone esterases, then PTTH secretion resumes under circadian (22-24 hour) photoperiodic control (Nijhout 2003). On the other hand, certain hemimetabolous bugs (Hemiptera) possess abdominal stretch receptors that activate secretion of PTTH (Nijhout 2003).

Arthropod body allometry is intertwined with development of larval and imaginal disc tissues (Stern and Emlen 1999, Shingleton et al. 2008) and the evo-devo of insect polyphenism (Emlen and Nijhout 2000).

A review of neotenous development in termites is available (Korb and Hartfelder 2008). Isoptera (termites) are hemimetabolous insects (Grimaldi and Engel 2005).

Insect developmental tool kit. The insect developmental tool kit is comprised of certain homeotic selector genes (including Hox genes), zygotic (gap- maternal-, and pair-rule-) genes, field-specific selector genes, compartment selector genes, cell-type-specific selector genes, and segment polarity genes; and the transcription factors they encode. In addition, the insect developmental tool kit is comprised of controlling factors behind the cessation of insect growth including bioactive PTTH, JH, juvenile hormone esterases, and ecdysone steroids (Truman and Riddiford 2002, Nijhout 2003, S. B. Carroll et al. 2005).

Thummel and Chory (2002) point to a possible coevolutionary connection between the 20E-ecdysone/cytochrome P450 biosynthetic machinery of insect antagonists and seed plant hosts. Structurally similar to bioactive plant brassinosteroids, 20E-ecdysone induces a cascade of transcription factor biosynthesis important in the regulation of insect development (Truman and Riddiford 2002, De Loof 2008).

Juvenile hormone (JH) biosynthesized in the corpora allata of the insect brain is a sesquiterpenoid epoxide methyl ester (Hartfelder 2000). Interestingly, many naturally-occurring plant sesquiterpene esters and lactones are bioactive and exhibit insecticidal properties. Juvenile hormone and its homologs are integral in vitellogenesis (Hartfelder 2000), regulation of moult cycles (Truman and Riddiford 2002), and caste development and behavior in social Hymenoptera (Guidugli et al. 2005).

Were bioactive brassinolides and sesquiterpenes manufactured by Paleozoic seed plants used as chemical warfare agents to affect growth, development, and behaviour of herbivorous insects?

Several developmental gene families, transcription factors, and enzymes involved in hormone signaling cascades are known in invertebrates based in part, on experimental studies of the Drosophila model arthropod (S. B. Carroll et al. 2005, Hittinger and S. B. Carroll 2008, among others). Key elements of the Drosophila genetic tool kit are:

Achaete-Scute [AS-C] complex (four-gene complexes encoding transcription factors critical for bristle patterning and neural cell development)

apterous [ap] (compartment-selector gene encoding Apterous protein needed for subdivision of imaginal discs into dorsal and ventral compartments)

bicoid [bcd] (maternal effect gene necessary for anterior embryo development)

caudal and nanos (maternal effect genes needed for posterior development of the embryo)

decapentaplegic [dpp], short-gastrulation [sog], snail [sna], twist [twi], and zerknüllt [zen] (maternal effect genes involved in subdivision of cells in the dorsiventral axis of larvae)

distal-less [dll] (field-specific selector gene necessary for limb [antennae, genitalia, mouthparts, legs] development)

engraled [en] (compartment-selector gene encoding distinct homeodomain transcription factors which determine the posterior identity of embryos and wings).

even-skipped, fushi-tarazu, hairy, and paired (pair-rule genes which act in the periodicity of double-segments)

eyeless [ey] (encodes Ey protein, a DNA transcription factor which regulates other field-specific genes)

giant, huckebein, hunchback, knirps, Krüppel, and tailless (gap genes that regulate larval segmentation)

hedgehog and wingless (segment polarity genes which organize the orientation and patterns of cells within the anteroposterior axis of each larval segment)

Hox genes (clusters of eight-linked homeotic selector genes absolutely required for embryo, larval, and adult development)

Runx genes (clusters of homeotic selector genes that regulate development)

scalloped [sd] and vestigial [vg] (field-specific selector genes required for haltere and wing development)

tinman (field-specific selector gene that controls development of the insect heart)

Evolution of the Hox complex. Molecular diversification of the Hox gene complex over the course of 600 million years of metazoan evolution is analogous to the 400 million-year old molecular evolution of MIKC-type MADS-box genes and related cis-acting transcription factors of land plants (Theißen et al. 2000, Theißen 2001, and Theißen and Becker 2004).

Evolution of the Hox complex probably involved small gene duplications (SGDs), whole genome duplications (WGDs), divergence of homeodomains, disintegration of the Hox cluster at breakpoints, and rapid changes in the nucleotide sequence of homeodomains (S. B. Carroll et al. 2005, Negre et al. 2005). Further, changes in the arthropod homeodomain and evolution of new protein motifs led to new Hox developmental functions in certain insect lineages (S. B. Carroll et al. 2005).

The homeotic selector genes (S. B. Carroll et al. 2005, among others) of the Drosophila Hox complex are:

abdominal A [abd-A] (selects the developmental fate of the front-middle parts of the abdomen)

Abdominal B [Abd-B] (selects the developmental fate of the rear-middle part of the abdomen)

antennapedia [ANTb] (affects development of the middle legs and middle part of the thorax)

deformed [Dfd] (affects development of the upper part of the head)

labial [lab] (selects the developmental fate of the labium of the head)

proboscipedia [pb] (affects development of the lower half of the head)

sex combs reduced [Scr] (selects the developmental fate of the most posterior region of the head, labial segment, front legs, prothorax, and front and anterior portions of the thorax) (Rogers et al. 1997)

ultrabithorax [Ubx] (selects the developmental fate of the rear legs and thorax and front of the abdomen)

Ancestral arthropods possess two additional homeotic selector genes of the Hox cluster that together comprise the HOM-C, ten-gene complex (see discussion in Negre et al. 2005). These additional genes are:

fushi tarazu (ftz), and

Hox3

Genomic analyses suggest that derived winged insects lost functional copies of ftz and Hox3 through disintegration of the HOM-C complex (Negre et al. 2005). Duplication of the Hox3 gene of ancestral Cyclorrhaphan flies gave rise to two maternal effect genes, bicoid (bcd) and zerknüllt (zen) (Stauber et al. 2002). Based upon this study it is important to include Hox3 as part of the ancestral diverging insect developmental tool kit.

Early divergent arthropod developmental tool kit. Possible candidates for the early divergent insect developmental tool kit might include certain homeotic selector genes of the Hox complex such as homologs and paralogs of abdominal A (abd-A), Abdominal B (Abd-B), Hox3, proboscipedia (pb), sex combs reduced (Scr) (Rogers et al. 1997), and Ultrabithorax (Ubx), and the field-specific selector gene necessary for limb development in Drosophila (Diptera) known as distal-less (dll) (S. B. Carroll et al. 2005).

The importance of Ubx protein encoded by the Ubx gene in the early divergent insect developmental tool kit cannot be neglected in the present analysis since significant changes in the carboxy-terminal (C-terminal) region (Galant and Carroll 2002) and serine/threonine phosphorylation sites (Ronshaugen et al. 2002) are probably behind many insect body plan novelties seen in the paleontologic record of the past 400 million years of arthropod and crustacean evolution (Pavlopoulos and Averof 2002).

I also add hexamerin moulting storage proteins which are related to hemocyanin respiratory enzymes (Burmester et al. 1998, Burmester 2001, Burmester et al. 2006, Burmester and Hankein 2007), juvenile hormone esterases, vitellogenin genes and yolk proteins (Isoe and Hagedorn 2007), pheromone chemoreceptors (Robertson and Wanner 2006), and certain nuclear receptor proteins (Bonneton et al. 2003, Bonneton et al. 2008) including ultraspiracle, and ecdysone-inducible transcription factors to the list of developmental tools in the ancestral, diverging insect kit.

Respiratory enzymes (hemocyanins) and moulting storage proteins (hexamerins) are key elements of the early divergent arthropod developmental tool kit that tie-in with the evolution of insect legs and wings from bilaterian gills. Ancient insect wings probably functioned as respiratory organs. Wings, halteres, arachnid spinnerets, and insect legs are all organs that develop from limb fields of cells where Ubx expression is prevalent (S. B. Carroll et al. 2005).

Hexamerins are also implicated as silencers of JH signaling in neotenous castes of hemimetabolous termites (X. Zhou et al. 2006, X. Zhou et al. 2007) and caste polyphenism in holometabolous wasps (J. H. Hunt et al. 2007).

Several insect systematists studying beetle (Coleoptera) evolution are employing some of the aforementioned genes and proteins of the insect development tool kit in their phylogenetic analyses (Gómez-Zurita and Galián 2005). Gómez-Zurita and Galián (2005) discuss the utility of molecular phylogenetic characters appearing in the entomological literature in a review paper, which is organized along the lines of Floyd and Bowman (2007) for land plants (see section below).

Molecular model systems used as tools in beetle genomic research and phylogenetic studies include proteins central to development (JH esterases), diapause proteins, heat shock proteins, ultraspiracle (an ecdysone nuclear receptor protein), cuticle proteins, hexamerins, genes encoding vitellogenin, and apolipophorins, among others (see review by Gómez-Zurita and Galián 2005).

Understanding the nature and timing of early molecular diversification of homeotic selector genes, developmental proteins, nuclear receptor proteins, and cis-acting transcription factors of both invertebrate antagonists and vascular plant hosts might be a critical first step in understanding a coevolutionary origin of angiosperms.

The place and time to begin a molecular phylogenetic analysis is the late Devonian-early Carboniferous hypoxic icehouse (DeCARB)

Land plant developmental tool kit. Understanding the land plant developmental tool kit and gene regulation from a deep time research perspective ties-in with models of cone and floral organization, cell geometry and regulation of growth from shoot apical meristems (SAMs), paleobiology of homeodomain transcription factor trafficking, phyllotaxis, leaf development, and morphogenesis of fertile organs.

Evolutionary development (evo-devo) of early land plants was probably intertwined with regulatory changes in polycomb repressive 2 gene complexes and other stem cell factors as evidenced from studies of the extant model bryophyte Physcomitrella (Okano et al. 2009).

Floyd and Bowman (2007) are the first workers to estimate the developmental tool kit of early land plants including Paleozoic seed plant homeotic genes potentially important in the later evolution and diversification of angiosperms and origin of the first flowers from bisexual cone axes sensu Baum and Hileman (2006).

The work by Floyd and Bowman (2007) focuses on a molecular phylogenetic analysis of Chara (a green alga), Physcomitrella (a moss), Selaginella (a lycophyte), Arabidopsis (angiosperm malvid), Antirrhinum (angiosperm asterid), Oryza (angiosperm monocot), Populus (angiosperm fabid), Picea (gymnosperm conifer), and Pinus (among others).

Is unique architecture and branch patterning in progymnosperms traceable to differentiation and growth of cells from meristematic tissue fields other than axillary buds?

Were certain seed plant morphologies exploited by respiring insect eggs, larvae, instars, and adults?

Can we identify plant adaptations that helped shield insect eggs and instars from hypoxia, temperature extremes, and ultraviolet (UV) radiation in paleoenvironments?

What are the kinds of miniature seed-plant fitness landscapes exploited by insects?

Did the architecture of some Paleozoic seed plants provide critical, oxygenated habitat for phytophagous insect and tetrapod associates?

Was the development and form of pollen and seed cones (strobili) affected by wind currents in Triassic fitness landscapes?

Are some seed plant morphologies, such as ovule placement (adaxial or abaxial) on the surfaces of large, leathery leaves a consequence of the mechanostimulatory effects of grazing tetrapods on expanding SAMs of seed plant monopodia?

A review of plant homeobox genes and homeodomain proteins offers additional insight into critical elements of the land plant developmental tool kit (Mukherjee et al. 2009).

Many developmental gene families and transcription factors have been identified in land plants (Langdale 2008, Mukherjee et al. 2009). These genes, and their transcription factors, together with phytohormone biosynthetic and regulatory factors, homeotic selector genes, and certain microRNAs, make-up the developmental tool kit of land plants including lignophytes. The main developmental and regulatory gene tools of focus are:

ARF (auxin response factor gene family)

ARP (KNOX regulating genes ASYMMETRIC LEAVES1, ROUGHSHEATH2, PHANTASTICA [ARP])

AS2 (LOB family genes that function in lateral organ formation in SAMs)

Class III HD-Zip (a family of SAM maintenance genes and upstream controllers of leaf development)

Class I KNOX (essential genes to meristem maintenance and SAM formation affected by auxins and gibberellins)

Class I TCP (leaf sculpting genes and cell division controllers)

Class II TCP (genes that determine leaf shape, floral symmetry, and flower patterns)

CUC (important genes in embryo and flower development)

euANT (regulating genes involved in lateral organ development)

euAP2 (SAM maintenance and timing of flowering including APETALA 1 [AP1]and APETALA 2 [AP2])

GRAS (a family of genes necessary for radial patterning of roots and shoots and axillary meristem development [HAIRY MERISTEM, LATERAL SUPRESSOR, SCARECROW], and in mediation of gibberellic acid responses during flowering and development of trichomes [DELLA genes])

KANADI (GARP family transcriptional regulators antagonistic to Class III HD-Zip genes important to leaf and integument development)

LEAFY [LFY] and FLORICAULA [FLO] (controllers of the downstream expression of several genes including MIKC-type MADS-box genes)

MIKC-type MADS-box genes (developmental genes that control organogenesis in seed plants)

PIN (auxin related genes needed in land plant architecture)

TIR (auxin related genes)

UFO (stem cell maintenance gene)

WOX Class genes including WUSCHEL (essential genes for stem cell maintenance in organizing centers of SAMs)

YABBY (seed plant specific family of genes that promote differentiation of abaxial tissues and leaf growth)

Based on a molecular phylogenetic analysis by Floyd and Bowman (2007) the ancestral land plant homeotic gene tools of our focus to deduce the timing of the origin of the first flower should be AP2, ARF, Class I and II TCP, Class III HD-Zip, GRAS, KNOX, LEAFY, MIKC-type MADS-box, PIN, TIR, and YABBY genes.

Cell geometry of shoot apical meristems (SAMs). Much progress has been made on understanding developmental biochemistry and hormonal controls behind the anatomy and cell geometry of the shoot apical meristems (SAMs) and accessory fertile meristems of a number of diverse flowering plants and gymnosperms (Glover 2007, Busov et al. 2008). The best studied example is the diminutive mouse-ear cress (Arabidopsis thaliana [Brassicaceae, Brassicales, Rosanae]), an extant, highly derived annual species of malvid mustards (Kwiatkowska 2006).

Floral meristem identity and integrator genes of massive shoot apical meristems of the oil palm (Elaeis guineensis [Arecaceae, Arecales, Lilianae]) have also been studied (Adam et al. 2007).

Biomechanics is a fertile area of research in botany (Niklas 1992, Niklas 2000, Rowe and Speck 2005, Niklas et al. 2006, Read and Stokes 2006, Masselter et al. 2009).

Can the experiments drawn from principles of geometry and engineering answer some of the basic questions posed by Coen et al. (2004)?

"Whatever the detailed mechanism for aligning cell axes with gradients [of morphogen flow], these considerations indicate that morphogens play two distinct roles in the generation of shape. The first is a regionalizing role, involved in elaborating differences between regions [of meristems], such as levels of gene activity, growth rate, or anisotropy. The second is a polarizing role involved in specifying directions of cell axes. Morphogens with a regionalizing role often act by influencing transcription factors, which in turn can affect parameters such as growth rate and anisotropy. Cells respond only to the overall concentration of these morphogens, not to the direction of their gradient. Morphogens with a polarizing role, influence the orientation of cell axes and can determine the principal directions of growth. Cells respond to the highest slope of the morphogen gradient or to the direction of morphogen flow. Such responses [of plant cells] are unlikely to be mediated by transcription factors, as the directional information would be lost through conversion to a gene expression level."

The preceding quote is from E. Coen, A.-G. Rolland-Lagan, M. Mathews, A. Bangham, and P. Prusinkiewicz (2004), The genetics of geometry, Proceedings of the National Academy of Sciences 101(14): 4728-4735. The phrases in brackets [] are mine.

Certain anatomical and enzymatic tools of the land plant hormone signaling machinery align cell axes and control morphogen flow. Some of these affect SAM development and phyllotaxis in both morphospace and time resulting in evolutionary change.

"Evolutionary changes in developmental tissue geometry, in our terminology, are the result of heterotopy."

The preceding sentence is quoted from M. L. Zelditch and W. L. Fink (1996), Heterochrony and heterotopy: stability and innovation in the evolution of form. Paleobiology 22(2): 241-254.

Shoot apical meristem (SAM) organization. The shoot apical meristem (SAM), stem cell geometry, and fertile SAM phyllotaxis figure prominently in several evo-devo hypotheses on the origin of cones and flowers (Theißen Saedler 2001, Becker and Theißen 2003, Baum and Hileman 2006, Theißen and Melzer 2007).

Reviews of SAM development in model flowering plants are available (Bäurle and Laux 2003, Aida and Tasaka 2006 [two papers], Carraro et al. 2006), which tie together aspects of tool kit genes and the transcription factors they encode (previous sections) with phyllotaxis (next section).

The SAM of angiosperms and certain other seed plants is layered consisting of the tunica (surface layer) and tissues within (corpus). This character has been employed in many seed plant phylogenetic analyses.

Development of the model malvid Arabidopsis thaliana SAM operates three ways (Bäurle and Laux 2003):

local maintenance of stem cells

amplification of stem cell daughters

initiation of organ primordia

The essential gene WUSCHEL and the homeodomain transcription factor this gene encodes operates in the organizing center of the eudicot SAM (Mayer et al. 1998). At the periphery of the central zone, partitioned pre-founder cells and founder cells determine how lateral organs grow and the position of reproductive modules.

Partitioning of cells of the central SAM in model malvids leads to discernable boundary layers: the meristem-organ primordium (M-O) boundary and a region between developing organs known as the O-O boundary (Aida and Tasaka 2006).

Morphogenesis at the boundary cells between developing organ primordia and the SAM daughter cells is regulated by several classes of genes. Further, plasmodesmata of the boundary cells of the SAM potentially regulate flow (traffic) of small proteins and other transcription factors (see review by Aida and Tasaka 2006). Physiology of homeodomain transcription factor trafficking is discussed in a later section.

At every step of the way, the cells of the developing and expanding SAM exhibit discreet gene expression patterns. Further, early events in transcriptional regulation and post-transcriptional control determine the morphology of lateral organs of the seed plant shoot (see review by Carraro et al. 2006, among others).

Shoot apical meristem maintenance and patterning genes and transcription factors of the land plant developmental tool kit including WUSCHEL (Vásquez-Lobo et al. 2007, Nardmann et al. 2009) may affect downstream phyllotactic patterns.

Did early whole genome duplications early on in the evolution of vascular plants and progymnosperms affect the expression of WUSCHEL (WUS) genes and their transcription factors resulting in fundamental changes in the way SAMs were organized in diverging seed plant lineages?

Phyllotaxis. Phyllotactic patterns of extant flowering plants are of ongoing interest to paleobotanists and plant systematics who seek to understand the origin of the angiosperm flower from an evo-devo research perspective (J. A. Doyle and Endress 2000, Endress and J. A. Doyle 2007). Floral phyllotaxis is distinct from vegetative phyllotaxis.

Extant basal angiosperms including the Magnoliales exhibit both spiral and whorled floral phyllotactic patterns (J. A. Doyle and Endress 2000, Ronse De Craene et al. 2003, Staedler et al. 2007). When floral phyllotactic characters in basal angiosperms are subjected to phylogenetic analyses, reversals from spiral to whorled and whorled to spiral become apparent (Ronse De Craene et al. 2003, Endress and J. A. Doyle 2007).

Phyllotaxis is reviewed by Kuhlemeier (2007). The most recent Arabidopsis model suggests that in addition to PIN proteins, the family of AUX1 LAX auxin influx carrier enzymes stabilize spiral phyllotaxis (Bainbridge et al. 2008).

An earlier model of developmental phyllotaxis and initiation of procambial strands leading to vasculature in the SAMs of extant flowering plants implicates auxins, auxin efflux carrier protein (ATPIN1), mRNA transcripts such as ATHB8-GUS, and other auxin related developmental genes (R. S. Smith et al. 2006). Development of phyllotactic geometric patterns and vasculature in SAMs of some modern flowering plant species may be traced to a common genetic and physiologic mechanism (Dengler 2006).

Were the possibly conserved mechanisms of SAM differentiation that underlie geometric patterning in sporophytes operating the same way in long extinct progymnosperm and seed plant species?

According to Floyd and Bowman (2007) the answer to the second question is "yes." Orthologues of PIN and ARF, auxin related genes and transcription factors needed in land plant architecture and phyllotaxis, were part of the ancestral land plant developmental tool kit according to Floyd and Bowman (2007).

Floyd and Bowman (2007) identified several "PIN-like" genes from expressed sequence tag (EST) studies of moss gametophytes belonging to Physcomitrella. Five PIN genes were identified in sporophytes of extant Selaginella, and eight additional PIN genes were found in Arabidopsis (Floyd and Bowman 2007).

"If the hypothesis that increasing complexity of auxin metabolism, transport, and perception was related to the increased sporophytic complexity accompanying land plant evolution, then we might expect these particular gene families to be less complex in the moss and lycophyte lineages than in angiosperms, as indeed appears to be the case."

The preceding quote is from page 22 of Sandra K. Floyd and John L. Bowman (2007), The ancestral developmental tool kit of land plants, International Journal of Plant Sciences 168(1): 1-35.

Did auxin exist in Paleozoic vascular plants? Yes, as evidenced by paleobotanical data (Rothwell and Lev-Yadun 2005, Rothwell et al. 2008).

Macromolecular secretions (microRNAs, bioactive peptides, and other small-sized molecules) when applied to the surfaces of SAMs by chewing, crawling, ovipositing, piercing, and sucking resident, phytophagous insects, could potentially signal the host plant genome and affect phyllotaxis and other developmentally plastic traits but experimental studies are needed to support this hypothesis.

Leaf development. Evolutionary development of leaves is a critical body of evidence necessary to decipher paleophysiology and the origin of reproductive organs unique to flowering plants: carpels and stamens.

Knowledge of long extinct whole plants and paleopopulations often exists only as detached compressed or permineralized foliar organs traditionally described by paleobotanists as morphotype species. An unappreciated paleontologic problem is that detached perianth parts and vegetative leaves having different venation patterns might have belonged to the same mother plant.

Several books and review papers place leaf development of vascular plants including angiosperms in a paleontologic context including Melville (1969), J. A. Doyle (1978), Melville (1983), (J. A. Doyle and Donoghue 1986, 1987), Stewart and Rothwell (1993), Trivett and Pigg (1996), Sinha (1999), Beerling et al. (2001), Roth-Nebelsick et al. (2001), Boyce and Knoll (2002), Boyce (2005), Feild and Arens (2005), Harrison et al. (2005), J. A. Doyle (2006), Beerling and Fleming (2007), Brodribb et al. (2007), Sanders et al. (2007), and Boyce et al. (2009).

According to a review by Boyce and Knoll (page 74, 2002), " ... the gigantopterids bore the most complex and morphologically distinctive leaves of any plants in the Paleozoic ... "

The Permian gigantopterid plant Delnortea abbottiae possesses leaves with four orders of veins, and certain cells of the leaf hypodermis are sclerotic (pages 1423 and 1424, Fig. 36 on page 1420, Mamay et al. 1988), which is anatomically transitional to the two means of water distribution in leaves described by Brodribb et al. (2007).

The image to the left is a piece of rock slab containing the largest known Delnortea abbottiae leaf compression collected by the author and colleagues (USNM 387473, Fig. 10, page 1413, Mamay et al. 1988). The piece of fossil leaf was photographed by the writer on the day it was pulled from the rock layer in the Lower Permian (Leonardian) Cathedral Mountain Formation, Del Norte Mountains of southwestern North America in 1982.

Was the paleophysiology of gigantopterid plants as expressed by the presence of two anatomical innovations to maximize leaf hydraulics somehow linked with the success of these plants in arid Permian paleoenvironments?

"Just as the early steps of marginal meristem evolution were constrained by the scope of developmental possibility, the ultimate range of possible leaf morphologies must have been constrained by available developmental mechanisms."

This quote is from page 78 of C. K. Boyce and A. H. Knoll (2002), Evolution of developmental potential and the multiple independent origins of leaves in Paleozoic vascular plants. Paleobiology 28(1): 70-100.

Land plants possess two leaf hydraulic strategies as expressed by their anatomy and morphology (Brodribb et al. 2007). According to Brodribb et al. (pages 1894 and 1895, 2007), the first hydraulic method is to branch the system of veins to deliver water to the site of transpiration (this is the method employed by many flowering plants). The second method seen in some members of the gymnosperm clade is to modify hydraulics by lignifying certain cells of the leaf mesophyll thus channeling water to the stomates.

Can these same leaf traits be found in later Mesozoic seed plants?

Ongoing research on the anatomy and cell biology of leaf development from meristematic cells of SAMs, nuclei, and genetic tool kits is incrementally reviewed by Sinha (1999), Roth-Nebelsick et al. (2001), Bharathan et al. (2002), Boyce and Knoll (2002), Engström et al. (2004), Fleming (2005), Piazza et al. (2005), Aida and Tasaka (2006), Brodribb et al. (2007), and Champagne et al. (2007).

Development of leaves depends on the activity of Class III HD-ZIP genes (Floyd and Bowman 2007), and a KNOX/ARP genetically-based, modular switch mechanism (Beerling and Fleming 2007). At least two and probably several land plant evo-devo developmental programs existed in deep time. Several classes of homeotic genes, homeodomain proteins, and other transcription factors are probably involved.

Whether the KNOX genetic switch is on or off depends on the activity of ARP protein repressors (see review by Piazza et al. 2005), and may lead to the development of different leaf forms e.g. pinnate, palmate, or compound (Beerling and Fleming 2007).

Abaxial-adaxial leaf organ development in angiosperms begins in the shoot apical meristem: the adaxial side of leaf primordia faces toward the SAM while the abaxial surface is orientated away from the SAM. The genetic basis of cell polarity in Arabidopsis SAMs is reviewed by Engström et al. (2004) and Carraro et al. (2006).

Three classes of transcription factors establish adaxial-abaxial leaf organ identity in Arabidopsis: Class III HD-Zip (PHABULOSA, PHAVOLUTA, REVOLUTA), KANADI, and YABBY proteins (Floyd et al. 2006, Nole-Wilson and Krizek 2006). Class III HD-Zip transcription factors operate on the adaxial side of organ bulges in the SAM. KANADI and YABBY proteins are active in the abaxial domain (reviewed by Carraro et al. 2006).

James Doyle discusses YABBY within the context of ovule tissue layers and attachment point (abaxial or adaxial) to carpels and leaves in Amborella, a basalmost angiosperm (J. A. Doyle 2006).

AINTEGUMENTA (ANT) gene expression in certain extant flowering plants determines which cells of the SAM "leave the meristem to form lateral organs" (page 983, Nole-Wilson and Krizek 2006). Further, ANT and APETALA2 are transcription factors of focus in phylogenetic analyses of seed plants (Floyd and Bowman 2007) and the origin of angiosperms (S. Kim et al. 2006).

Fertile leaves of seed plants are termed sporophylls. The two kinds of sporophylls potentially derived from the ancient land plant developmental tool kit of SAMs are: megasporophylls (ovule-bearing leaves) and microsporophylls (microsporangium bearing leaves).

Sporophylls and their fertile regions may be viewed as kinds of reproductive modules (see next section). Based on studies of MIKC-type MADS-box genes (Zahn et al. 2005) and MADS-box B protein heterodimers (Hernández-Hernández et al. 2007), a whole genome duplication (WGD) occurring around the time of the angiosperm-gymnosperm split 300 million years ago, was a critical macroevolutionary step in the origin of the first flowers from bisexual cone axes (Baum and Hileman 2006).

Reproductive modules. Modular development as it pertains to seed plant reproduction may be viewed at the cellular or organ level. Aggregations of cells (tissue fields and compartments) and whole organs are also potentially modular from an evo-devo research perspective.

Paleontologic evidence of ovule development in bennettitalean seed plants suggests early innovation in the production of these reproductive modules (Rothwell and Stockey 2002).

Cellular and anatomic details of sexual reproduction in extinct seed plant groups are sometimes "accessible" contrary to statements in the literature (page 145, Williams 2009) but require tedious preparation of rare permineralizations (e.g. Nishida et al. 2004, Rothwell et al. 2009).

The image to the right is the permineralized nucellus (the megasporangium wall of a fossilized ovule) of Eoantha zerikhinii (from a gnetalean "pre-flower") from the Baisian Assemblage, early Cretaceous Period, of Transbaikalia in central Asia. The fossilized ovule contains a female gametophyte with a pollen chamber filled with Ephedrites-type pollen (this picture is from an original image provided by Professor Valentin Krassilov, posted here with his permission).

Several key papers and reviews are devoted to the study of reproductive structures in seed plants as an approach to gain a better understanding of evolution of gymnosperms and the origin of flowering plants (Friedman 1992 [two papers], Crepet and Nixon 1996, Crane and Kenrick 1997, Labandeira 2000, Eckardt 2002, Lord and Russell 2002, Friedman et al. 2004, Friedman and Williams 2004, Williams and Friedman 2004, Friedman et al. 2008, Friedman and Ryerson 2009, Williams 2009).

Friedman and Ryerson (2009) in a recent review of megagametogenesis in angiospermous seed plants argue that the female gametophyte of flowering plants consists of a modular quartet of cells.

Curious molecular and developmental similarities of angiosperm reproductive modules with animal germ lines (Dickinson and Grant-Downton 2009) require further study.

Reproductive modules are specific organs, tissue fields, or cells derived from a SAM or cone axis. Evolutionary development and genetics of the female reproductive tract is reviewed by B. C. W. Crawford and Yanofsky (2008). The bulleted list below includes some key references:

archegonia (haploid tissue receptacles in certain gymnosperms that contain the egg and accessory cells)

carpel (an inrolled [conduplicately folded] ovule-bearing leaf) (Scutt et al. 2006, Vialette-Guiraud and Scutt 2009)

central cell (female haploid cell of the embryo sac involved in double fertilization) (Bemer et al. 2008, Berger et al. 2008, R. J. Scott et al. 2008)

eggs or egg nuclei (female haploid cells or nuclei, respectively) (Berger et al. 2008, R. J. Scott et al. 2008)

embryo (an often tiny [usually dormant] diploid daughter plant [complete with a basal cell, suspensor, and growing point] that develops from the fertilized egg, usually housed within a mature ovule or ripened seed) (Tobe et al. 2000, Forbis et al. 2002, Holloway and Friedman 2008)

endosperm (diploid, triploid, or polyploid nutritive cells or nuclei usually housed inside of ovules and embryo sacs) (Baroux et al. 2002, Raghavan 2003, Rudall et al. 2009)

interseminal scale (modified female organs of bennettitaleans possibly used to house toxins needed as defense against potential herbivores)

megasporangia (multicellular sacs that produce megaspore mother cells and megaspores)

megaspore mother cells (female diploid cells where meiosis occurs)

megaspores (female haploid products of meiosis)

megagametophytes (cellular or acellular haploid tissues derived from cell- and/or nuclear divisions in megaspores [termed modular quartets of cells in flowering plants]) (Yadegari and Drews 2004, Tobe et al. 2007, Blanvillain et al. 2008, Friedman et al. 2008, Madrid and Friedman 2008, Friedman and Ryerson 2009, Madrid and Freidman 2009)

microgametophytes (Russell 1991, Rudall and Bateman 2007, Blanvillain et al. 2008, Borg et al. 2009)

microsporangia (diploid multicellular sacs that produce diploid microspore mother cells, haploid microspores, and haploid pollen) (Russell 1991, Furness et al. 2002)

microspore mother cells (male diploid meiotic cells) (Russell 1991, Furness et al. 2002)

microspores (male haploid products of meiosis) (Russell 1991, Furness et al. 2002, Nadot et al. 2008)

ovary (an organ of the mother plant consisting of one or more carpels; contains one to several ovules)

ovule (a complex organ of the mother seed plant that contains sterile diploid protective tissues termed integuments [often hardened by lignified cell walls], a vestigial megaspore wall [nucellus], megaspores; and at later stages of development: megagametophytes, other nutritive tissues in some seed plants [endosperm], archegonia [in certain seed plants], and eggs [to include accessory cells and synergids in some seed plants]) (Skinner et al. 2004, Scutt et al. 2006, Colombo et al. 2008, Kelley et al. 2009, Skinner and Gasser 2009)

polar nuclei (haploid nuclei of the embryo sac) (Berger et al. 2008)

pollen grain (an endosporic microgametophyte that consists of a hardened microspore containing haploid male nuclei and/cells) (Borg et al. 2009, Lora et al. 2009)

pollen tube (a living tubular cell derived from the male [micro] gametophyte, which is lined with callose in extant flowering plants) (Russell 1991, Williams 2008, Cai and Cresti 2009, Dresselhaus and Márton 2009, Williams 2009)

seed (a ripened ovule that contains an embryo)

sperm or sperm nuclei (male haploid cells [sometimes motile] or nuclei, respectively) (Berger et al. 2008)

stamen (a complex, highly modified male branch or grouping of fused male [micro] sporophylls that contains one or more microsporangia [and pollen when ripe]) (Alves-Ferreira et al. 2007)

staminode (a sterile, often leaf-like module derived from the male [micro] sporophyll or stamen) (Walker-Larsen and Harder 2000)

stigma (the pollen receptive surface of a carpel)

style (an organ located midway between the ovary and stigma through which the pollen tube may penetrate) (J. P. Alvarez et al. 2009)

suspensor (an often linear array of diploid cells [often containing endoreduplicated nuclei] of the embryo)

synergids (specialized cells of the embryo sac that interact with the pollen tube during double fertilization) (Sandaklie-Nikolova et al. 2007)

Sporophylls. Fertile leaves (sporophylls) of heterosporous seed plants are of two kinds: 1) megasporophylls (these are leaves that bear ovules later maturing into seeds, a morphological condition termed phyllospermy) and, 2) microsporophylls (microsporangium-bearing, polleniferous leaves). The sporophylls of some Permo-Triassic shrubs were flattened and photosynthetic.

Sporophylls in early diverging populations of Paleozoic seed plants were often clustered together on spur shoots known as simple cones and proanthostrobili. Further, prevailing models of cone and floral organization of sporophylls (Theißen Saedler 2001, Becker and Theißen 2003, Baum and Hileman 2006, Theißen and Melzer 2007) posit that the last common ancestor of seed plants possessed a hermaphroditic (bisexual) strobilus (Specht and Bartlett 2009, Rudall and Bateman 2010).

Clues on whether or not hypoxia is a selective force come from ecologic, physiologic, and rearing studies of extant insects (Gorr et al. 2006). The Holometabola exhibit complete metamorphosis of the fertilized egg to a wingless larval stage, a quiescent pupal stage, culminating with reproductively mature adults (page 331, Grimaldi and Engel 2005).

Hypoxia in extant terrestrial environments is defined as an oxygen concentration below ca. 20% (Hoback and Stanley 2001). Adaptations to hypoxia in holometabolous insects are specific to developmental stages. According to the review by Hoback and Stanley (2001), several kinds of extant insects are adapted to hypoxic environments including Coleoptera (beetles and weevils), Collembola (springtails), Diptera (flies), Isoptera (termites), Lepidoptera (butterflies and moths), and Orthoptera (grasshoppers and katydids).

Tenebrio molitor beetles exhibit 80% larval mortality, decreased instar body mass, and an increased number of moults when reared in an atmosphere of 10% oxygen in a nitrogen (Greenberg and Ar 1996). Aquatic insects are generally more susceptible to hypoxia than terrestrial insects (Hoback and Stanley 2001). Hoback and Stanley (page 540, 2001) conclude "... that the developmental plasticity of insects facilitates morphological changes in respiratory systems in chronically hypoxic situations."

Choice of respiratory enzymes (hemocyanins) and moulting storage proteins (hexamerins) as some of the right insect traits to study ties-in with the evolution of insect legs and wings from bilaterian gills. Ancient insect wings probably functioned as respiratory organs. Wings, halteres, arachnid spinnerets, and insect legs are all organs that develop from limb fields of cells where Ubx expression is prevalent (S. B. Carroll et al. 2005).

Escape and radiation. The "escape and radiation" concept was developed by J. N. Thompson in Chapter 7 of his book on coevolution to recapitulate Ehrlich and Raven's (1964) suggestion that reciprocal selection may produce "higher-level patterns of diversification within phylogenetically conserved interacting lineages" (page 150, J. N. Thompson 2005).

Since Ehrlich and Raven's classic paper, many colleagues have enthusiastically supported proposals on a coevolutionary origin and/or diversification of flowering plants based on co-radiations between specific groups of animals and seed plant hosts of the Cretaceous Period (Crepet and Niklas 2009).

Aspects of coevolution of Mesozoic arthropods and seed plants which have a bearing on the origin and diversity of angiosperms were proposed by Takhtajan (1969), Raven (1977), Thien et al. (1985), Farrell (1998), Labandeira (1998), Grimaldi (1999), Wilf et al. (2000), and S. Hu et al. (2008).

Some of these studies however, focus on possible coevolution of arthropods and the first angiosperms during the Jurassic or Cretaceous periods of the Mesozoic Era, not the Carboniferous or Permian periods of the Paleozoic Era.

There are problems associated with Cretaceous co-radiations of angiosperms and insects brought to light by phylogenetics. Presumed co-radiations are often out-of-sync, suggesting that evolution of clades of late Mesozoic phytophagous ants, bees, beetles, butterflies, flies, and moths is independent of the explosive origin and spread of eudicot orders and families.

Molecular phylogenies of chrysomelid leaf beetles suggest a younger age of the group than previously realized (Gómez-Zurita and Galián 2005, Gómez-Zurita et al. 2007). These findings contradict earlier proposals on a coevolutionary origin of angiosperms during the Cretaceous Period (Farrell 1998, Grimaldi 1999, Wilf et al. 2000).

John N. Thompson's "escape and radiation" concept (Chapter 7, 2005) might be applied to a couple of intractable or difficult questions in population paleoecology:

Was there a potential for genetic release of insect-induced molecular effects on the seed plant developmental machinery attributable to shifts of arthropod antagonists away from the plant host to other oxygen-enriched shrub lifeboat habitats and lignophyte hosts?

What specific kinds of paleontologic and molecular phylogenetic data may be used to shed light on the paleobiology of insect-seed plant coevolutionary compartments?

To my knowledge, the notion of genetic release in a host plant-insect antagonist coevolutionary relationship at the level of transcriptional regulation is unsupported by scientific data gleaned from extant populations.

Ecological studies and phylogenetic analyses of extant species groups with divergent floral characters point toward reciprocal selection and recurring pollinator shifts in the evolutionary history of certain parallel lineages (see review by Fenster et al. 2004).

The recent literature contains several key articles on this subject (P. Wilson et al. 2007, S. D. Smith et al. 2008, J. D. Thomson and P. Wilson 2008, Muchhala and J. D. Thomson 2009, Pauw et al. 2009).

I may revisit these coevolutionary studies of extant model systems within the context of the geographical mosaic theory of coevolution (J. N. Thompson 2005) and possible deep-time seed plant interactions with phytophagous insects, at a later date.

Coevolution of seed plants and tetrapods. Weishampel and Jianu (2000), in an extensive review of dinosaurian herbivory, cast doubt on the coevolutionary-based idea that ornithischian tetrapods had an effect on the rapid radiation of flowering plants during the Cretaceous Period. Dinosaurians probably did not play a key role in the contrived Cretaceous Terrestrial Revolution (G. T. Lloyd et al. 2008).

According to Weishampel and Jianu (page 137, 2000), three geologic intervals of diversification of euornithopod transverse-grinders occurred during the Cretaceous Period "with the middle peak arguably associated with the origin of angiosperms." However, when ghost lineages are added to the cladistic analyses "the middle peak disappears" (page 137, Weishampel and Jianu 2000).

"Whatever coevolutionary interactions may have existed between flowering plants (angiosperms) and ornithischian dinosaurs, they are not apparent at the onset of the evolutionary radiation of angiosperms, as least as far as diversity patterns are concerned."

The preceding quotation is from page 139 of Weishampel and Jianu (2000), Chapter 5, Plant-eaters and ghost lineages: dinosaurian herbivory revisited. Pp. 123-143 In: H.-D. Sues (ed.), Evolution of Herbivory in Terrestrial Vertebrates: Perspectives from the Fossil Record, Cambridge: Cambridge University Press, 256 pp.

Tiffney (2004), in an extensive review of vertebrate dispersal of seed plants, concludes that tight coevolution between animal dispersers and plant is rare. Biotic dispersal of seed plant propagules probably arose through "transfer of function," according to Tiffney (page 6, 2004). Further, he suggests that "the Mesozoic dynamic was dominated by reptile-gymnosperm interactions to which the angiosperms were newcomers" (page 20, Tiffney 2004).

Fitness landscapes. Niklas (1997) expands and modifies Wright's fitness landscape concept in a review of the evolution of early land plants.

How do we apply Sewall Wright's adaptive (fitness) landscape concept toward a better understanding of evolutionary ecology of biome fragments, paleohabitats, seed plant shrubs, trees, mycorrhizal fungi, vertebrate herbivores, and resident phytophagous insects of Pangaea?

Shrub lifeboats. Pockets of air trapped by sheathing, leathery photosynthetic leaves of a monopodium and other crevices in organs of arborescent and herbaceous seed plant species might have had higher amounts of oxygen gas than the surrounding atmosphere of Devonian, Carboniferous, Permian, and Triassic time.

I name the individual seed plant shrubs and small trees of hypoxic late Devonian-early Carboniferous, Permian, and Triassic fitness (adaptive) landscapes that house herbivorous insect antagonists in oxygenated air spaces and crevices, shrub lifeboats.

A higher partial oxygen pressure within the air spaces between host plant leaves could potentially protect resident insect mutualists from atmospheric disruption and hypoxic conditions attributable the late Devonian-early Carboniferous ice-house, GuCCE, EPE, TrCCE, and BaCCE. Trapped oxygen gas of air pockets in the host shrub might have been critical for the survival and metamorphosis of instars (fertilized eggs, larvae, pupae).

Optimal thermogenic microhabitats cast from food-bearing, sheathing leaves and woody tissues could have sustained a resident insect and tetrapod population during times of temperature extremes, hypoxia, and high ultraviolet radiation levels.

The diagram on the left side of the page illustrates a hypothetical seed plant shrub lifeboat. The monopodium with its shoot apical meristem (SAM) and large, leathery sheathing leaves is shown from an overhead view. The resident insect colony of eggs and instars (or larvae, pupae, and adults) shown as red-colored objects occupies the oxygenated air spaces that surround the growing tip and hollows formed by clasping leaves of the shrub.

Invoking coevolutionary nomenclature drawn from studies of extant biomes, a shrub lifeboat might also be viewed as a compartment housing antagonists (herbivores and predators) in a locally hot, hypercapnic, hypoxic, Permo-Triassic fitness landscape.

Did perennial seed plant shrubs furnish respiring insects with cavities and air spaces having a higher partial pressure of oxygen than the surrounding hypoxic oxygen desert?

What happened to complex food webs, plant communities and habitats of early Triassic time following the disruptive and extirpatory forces of the end-Permian extinction (EPE)?

Some early Paleozoic forests of conifers, ferns, lycophytes, and sphenopsids were stratified and contained complex food webs involving decomposers, primary producers, mycorrhizal fungi, and populations of invertebrates and amphibians (Z.-Q. Wang 2000).

Were elevated carbon dioxide levels (hypercapnia) in the atmosphere of late Permian time, when combined with low oxygen levels (hypoxia), a selective force in animal populations, including terrestrial arthropod invertebrates and vertebrate tetrapods?

Knoll et al. (2007) review the possible physiological effects of hypercapnia on animals and plants in connection with the end-Permian global apocalypse and biotic recovery of early Triassic aquatic and terrestrial landscapes. Geochemical studies suggest that late Permian and early Triassic air of Pangaean shorelines had oxygen levels comparable to the highest summits (>5000 meters) of most modern mountain ranges (Huey and Ward 2005).

Based on studies of fossils, paleosols, and geochemistry it was possible that during millions of years of Paleozoic global climate changes, complex networks of generalist and antagonist animal and plant species broke down and/or were disrupted (Retallack et al. 2003, Retallack et al. 2005).

Hypoxic ice-house late Devonian-early Carboniferous fitness landscapes. The ice-house terrestrial paleoenvironments of the late Devonian and early Carboniferous Periods (DeCARB) might have been challenging to the respiratory systems of invertebrates including ancient plecopteran insects (Labandeira 2006). The DeCARB was also a time of seed plant innovation (Rothwell et al. 1996).

Early arthropods exhibited an explosive radiation of molecular novelties of the hemocyanin class of gas-binding respiratory enzymes during the hypoxic DeCARB. At least seven clades of hemimetabolous and holometabolous insects diverged from the thawing ice-house as evidenced by molecular clock estimates of hexamerin moulting storage proteins (Hagner-Holler et al. 2007).

Was the burst of evolution of insect hexamerin proteins from hemocyanins somehow linked to low temperature and low oxygen levels in late Devonian-early Carboniferous terrestrial paleoenvironments?

Hypoxic and hypercapnic hot-house Permo-Triassic fitness landscapes. High carbon dioxide levels in the atmosphere and water column preceding and immediately following the end-Permian extinction (EPE) might have had profound effects on the physiology of animals (Knoll et al. 2007). Several unanswered questions remain:

How did terrestrial invertebrates and vertebrates cope with high carbon dioxide content and low levels of oxygen in the global atmosphere (<12%) following the end-Guadalupian carbon cycle event (GuCCE) and end-Permian extinction (EPE)?

Was hypoxia greater at higher altitudes in, for example, the Hercynian Mountains?

What was the effect of oxygen depletion (hypoxia) and elevated carbon dioxide (hypercapnia), methane, and hydrogen sulfide levels on animals and plants indigenous to aquatic and terrestrial environments?

According to Andy Knoll et al. (page 302, 2007), "... hypercapnia can reduce respiratory capacity, exacerbating the physiological effects of environmental oxygen limitation."

"It is possible, therefore, that the high CO₂ [Silurian] concentrations acted as an environmental 'barrier' that prevented the origination and spread of large megaphylls. Once that environmental barrier was removed, ecological processes, especially competition between species for light, soil and water nutrients, would have driven both the co-evolution of the root and shoot and the well-documented increase in plant height (Beerling and Berner 2004)."

The previous statement is quoted from page 10 of David J. Beerling and Andrew J. Fleming (2007), Zimmermann's telomb theory of megaphyll leaf evolution: a molecular and cellular critique, Current Opinion in Plant Biology 10: 4-12.

Did Paleozoic seed plant shrubs and trees (or even herbs) furnish respiring insect larvae with cavities and air spaces having a higher partial pressure of oxygen than the surrounding hypoxic landscape? Could plant organs, tissues, and waxy cuticles shield insect eggs and instars from harmful ultraviolet solar radiation?

The end-Permian extinction marking the beginning of the Triassic Period of the Mesozoic Era also had devastating effects on growing substrates including extirpation of mycorrhizal fungi and certain soil bacteria. Global terrestrial and aquatic environments became acidified by 600,000 years of global volcanism and more than a million years of precipitation containing dissolved sulfur dioxide (SO₂) outgassed from magma underlying the Siberian Traps, a name applied by geologists to early Triassic epicontinental lava lakes.

The art board below shows a community of potentially coevolving animal antagonists that inhabit shrubs and trees of a biome fragment of an hypoxic late Permian fitness landscape based on descriptions by DiMichele and Hook (1992). A biome fragment near a water source in an otherwise arid landscape is depicted with overstory conifers and understory scrub, which is punctuated by dicynodont burrows. Herbivorous dicynodonts were probably the most common vertebrate animals of such late Permian landscapes (DiMichele and Hook 1992).

The vertebrate tetrapod population resided with the oxygen generating vegetation consisting of seed plant shrubs and coniferous trees, and an understory of bryophytes, ferns, horsetails, lycopsids, and seed ferns. Seeds rained on to the floor of the biome forming a seed bank sensu Harper (1977), and bisaccate pollen was transported to other shrubs and trees. Insect colonies and free-living invertebrates (antagonistic generalists and predators) are shown in dark red.

Free-living insects and certain generalists were probably only able to crawl or fly short distances from the oxygenated crevices of one shrub lifeboat to another. The surrounding hypoxic desert might have been lethal to most respiring animals. The increasingly small body size of Triassic invertebrates and larger lung cavities of certain tetrapodial vertebrates might have been adaptations of these animals to withstand oxygen deprivation.

The preceding diagram emphasizes one possible fate (of many) of a biome fragment immediately following the EPE. In this case, the biome fragment is dominated by antagonistic networks of interacting plants and animals compartmentalized around oxygen-generating vegetation including shrub- and tree-like lifeboats. Several questions may be posed:

What were adaptations of land plants available for exploitation by resident insect colonies?

Were these invertebrate colonies subject to predation by insectivorous amphibians and reptiles?

How did the activities of dicynodonts, therapsids, and archosaurian dinosaurs (among other vertebrates) affect the evolution, diversification, and dispersal of seed plants and their angiospermization?

Paleoecological processes in Paleozoic animals and plants were probably not homologous to living systems. Inextricably linked to paleosuccession in Permo-Triassic terrestrial ecosystems following disturbance was the feeding behavior of herbivorous and insectivorous amphibians and early reptiles including archosaurians, dicynodonts, and therapsids, among other tetrapods (Tiffney 1992).

Hypocapnic and hyperoxic Cretaceous fitness landscapes.  Impacts of hyperoxic oxygen levels in the atmosphere of Permo-Carboniferous time are reviewed by Beerling and Berner (2000) and Scott and Glasspool (2006). McElwain et al. (2005) in a review chapter on the effects of atmospheric carbon dioxide on angiosperm origin, radiation, and paleodiversity conclude:

"It is likely, therefore, that a long-term trend of CO₂ decline through the Cretaceous period may have played an extremely important role in the diversification of angiosperms ..."

The previous phrase is quoted from page 158 of J. C. McElwain, K. J. Willis, and R. Lupia, 2005, Chapter 7. Cretaceous CO₂ decline and the radiation and diversification of angiosperms. Pp. 133-165 In: J. R. Ehleringer, T. E. Cerling, and M.-D. Dearing, A History of Atmospheric CO₂ and its Effects on Plants, Animals, and Ecosystems. New York: Springer, 530 pp.

Cretaceous angiosperm diversification might have been triggered by abiotic effects of carbon dioxide decline (hypocapnia) on the physiology of flowering plants as McElwain et al. suggest. However, a concomitant rise in atmospheric oxygen content probably also had physiologic effects on insect antagonists of early Aptian angiosperms.

Coevolutionary arms race. The scientific literature on the coevolutionary arms race ("Darwin's race" title, Pauw et al. 2009) is vast and beyond the scope of this essay. Key references include published work by I. T. Baldwin (1998, among others) and Cornell et al. (2003). High points of this line of scientific inquiry are stated in this section and in the next chapter.

A coevolutionary arms race at the molecular level between populations of coevolving flowering plant hosts and insect antagonists is supported by real-time experimental studies. Further, evolutionary-developmental studies using biochemically altered probes, genetically-engineered transcription factors, and microRNAs demonstrate the lability of flowers and insect bodies.

Insects are unable to manufacture cholesterol ring systems required for the biosynthesis of 20E-ecdysone, an important steroid in endocrine signaling, but must obtain precursor molecules from digested products of animal and plant tissues, which are then assembled by cytochrome P₄₅₀ enzymes (Truman and Riddiford 2002).

Certain enzymes of lepidopteran insect guts are able to isomerize biosynthetic precursors of oxylipin phytohormones (Dabrowska et al. 2009). Experimental evidence on the efficacy of arthropod digestive enzymes to alter the molecular structure of ingested phytohormone precursors may have implications toward an understanding of coevolutionary arms race dynamics.

Plants manufacture many small polypeptides that affect reproductive development (BRICK, CLAVATA3-like/embryo-surrounding [ESR], CLE, cyclotide mini-proteins, and Dresselhaus Proteins) or operate in defense and wounding responses (systemins), or in molecular mimicry (PNP). Some of these possess insulin-like (e.g. leginsulin) or neuropeptidal (e.g. plant atrial natiuretic peptides) properties that affect insect development and physiology (Germain et al. 2006).

Does the inherent lability of floral and insect organs inferred from laboratory studies have a molecular genetic basis applicable at the population level to coevolving plants and animals?

"Coevolution is a well characterized process that takes place at all biological levels, from ecosystems to molecules."

The preceding quotation is from page 934 of Juan et al. (2008), High-confidence prediction of global interactomes based on genome-wide coevolutionary networks, Proceedings of the National Academy of Sciences 105(3): 934-939.

Understanding the selective forces behind molecular coevolution of insect and seed plant DNA-binding transcription factors and cis-regulatory modules in deep-time i.e. hundreds of millions of years, might be an important first-step in unraveling the enigmatic origin of angiosperms, evolutionary development of the first flower, and adaptive radiation of certain clades of coevolving Holometabola.

The next two chapters consider scientific evidence drawn from plant physiology and genetics critical in understanding the Paleozoic beginnings and Mesozoic adaptive radiation of the first angiosperms.

Physiologic Considerations:

This chapter considers evidence from the physiology of interacting plants and insects from a different research perspective. A discussion of phytohormones and defense substances and insect developmental and respiratory proteins, hormones, and neurotransmitters is necessary to understand possible molecular coevolution of insect and seed plant developmental tool kits, animal-plant signaling, and the ongoing coevolutionary arms race (Becerra et al. 2009, Pauw et al. 2009).

Some of the topics covered here are discussed by the contributing authors to chapters in a book edited by Hemsley and Poole (2004).

Insect development is apparently controlled by cis-acting transcription factors (S. B. Carroll et al. 2005) and hormone signals acting in regulatory networks and cascades (Bonneton et al. 2008). Further, structural similarities exist in the trihelical bundle of certain helix-turn-helix cis-regulatory homeodomain proteins of the model insect Drosophila melanogaster and angiosperm malvid Arabidopsis thaliana.

The image to the right is a 20 × 20 cm polyamid thin layer chromatogram representing successive gel filtration column fractions from a mixture of floral flavonoids (x-axis), partitioned in an aqueous solvent mixture (y-axis) by TLC. Development of the thin layer chromatogram and treatment with Natural Products Reagent A reveals a separation of flavonol mono-, di-, and triglycosides (viewed under long-wave ultraviolet light and captured on film). The yellow-orange spot to the right is a standard sample of quercetin-3-O-glucoside, a flavonol monoglycoside.

Natural products, phytohormones, and signaling. This section briefly discusses the high points of a great body of literature in plant physiology, which is devoted to bioactive secondary plant products. Certain natural products are natural insecticides and phytohormones.

According to Floyd and Bowman's phylogenetic analyses (2007) the auxin, ethylene, and gibberellin metabolic pathways are highly conserved and involve several ancient land plant enzyme systems including APETALA2/auxin response proteins (AP2/ARFs), DELLAs (transcription factors of the GRAS family), and ethylene response factors (ERFs).

Did Paleozoic seed plants manufacture bioactive plant natural products and phytohormones and signal moulting insect instars residing in oxygenated crevices and leaf bases of massive SAMs?

Likewise, did phytophagous insect antagonists of Permo-Carboniferous and Permo-Triassic seed plants signal the homeotic genetic machinery of host plants affecting SAM development and morphology?

An important clue comes from naturally occurring but fossilized triterpenoids known as oleanone triterpanes (oleananes) and class 1c components of fossilized resins (amber).

Some flavonoids are antioxidants that inhibit auxin efflux carriers. Many flavonoids are implicated in hormone signaling, plant defense, and regulation of gene expression (Koes et al. 2005, Owens et al. 2008).

Biosynthesis of chalcone synthetase is affected by the hormone gibberellic acid (GA) in some flowering plants. Chalcone synthetase manufactures many classes of flavonoids involved as antioxidants, bee-violet attractants in flowers such as Rudbeckia hirta, and as ultraviolet radiation shields.

Steroids are well-known natural products manufactured by plants for use in defense and signaling (Truman and Riddiford 2002).

Auxin down-regulates Class I KNOX genes in developing SAMs of extant flowering plant experimental systems permitting GA activity in leaf primordia and promoting growth (Weiss and Ori 2007). Membrane proteins known as efflux and influx carriers control auxin movement from cell to cell in the SAM.

Further, auxin response factor (ARF) genes encode cis-acting ARF proteins that bind to auxin responsive genes (Busov et al. 2008, Delker et al. 2008, R. S. Smith 2008). Auxin response factors figure prominently in the regulation of leaf morphology and belong to the ancestral developmental tool kit of land plants (Floyd and Bowman 2007).

Interestingly, the MIKC-type MADS-box transcription factor SEPALLATA3 interacts with ARF and other elements of several hormone signaling cascades in Arabidopsis (Kaufmann et al. 2009).

Clues on the possible existence of polar auxin flow in woody stems have been uncovered in comparative anatomical studies of a 375 million-year-old fossilized progymnosperm known as Archaeopteris and modern conifers (Rothwell and Lev-Yadun 2005). The fossil record reveals that auxin regulation of secondary growth was a conserved physiologic process (Rothwell et al. 2008).

Based on the phylogenetic analysis of Class I KNOX genes (Floyd and Bowman 2007), paleobotanical evidence of polar auxin flow in Devonian plant fossils (Rothwell and Lev-Yadun 2005), and deep conservation in terpenoid enzymology (S. Lee and Chappell 2008) needed to biosynthesize diterpenes including GA, it is not unreasonable to add auxins and gibberellins to the regulatory tool kit of early diverging land plants.

Brassinosteroids (BRs) also known as brassinolides, are structurally similar to 20S-ecdysone, a potent insect steroid moulting hormone. Brassinolides are also implicated in endogenous ethylene production (J. Li and Chory 1999, Thummel and Chory 2002, Zullo and Adam 2002).

Cytokinins are modified nucleotide phytohormones that induce the expression of WUSCHEL genes in CLAVATA pathways in developing SAMs (Gordon et al. 2009), among other profound regulatory effects on plant growth and development.

Ethylene is a gaseous hormone with profound effects on plant growth, development, and responses to biotic stress including herbivory. A family of five membrane-bound receptor proteins perceives ethylene molecules from the external environment. The ethylene signaling cascade crosstalks with the biosynthetic machinery of most other phytohormones including abscisic acid (ABA), auxins, brassinosteroids (BAs), and gibberellins (Y. F. Chen et al. 2005).

For example, ethylene adversely affects gibberellic acid (GA) by activating DELLA repressing proteins. In turn, DELLAs repress the floral integrator genes LEAFY (LFY) and SOC1 resulting in delayed flowering in Arabidopsis (Achard et al. 2007). Ethylene and gibberellic acid (GA) biosynthesis in plants is enhanced by insect neurotransmitters such as catecholamines (Schultz 2002).

Gibberellins are key diterpenoid hormones of the developmental tool kit of extant plants (Weiss and Ori 2007). The effects of nuclear regulatory growth repressing DELLA proteins that belong to the GRAS family of transcriptional regulators are relieved by gibberellic acid (GA) via the ubiquitin 26S proteasome biosynthetic pathway (Yu et al. 2004, Axtell and Bartel 2005, Lu et al. 2005, Jiang and Fu 2007, Weiss and Ori 2007, Gao et al. 2008, Hartweck 2008).

Studies of Arabidopsis provide evidence that GA controls the development and morphology of trichomes on leaves, flowers, and shoots by de-repressing DELLAs (Gan et al. 2007). Whether the mechanism of GA action on the negative effects of nuclear regulatory growth repressing DELLA proteins on glandular hair and trichome development in other seed plants is the same as Arabidopsis is not known. Trichomes and glandular hairs play an important role in defense against herbivores.

DELLAs restrain plant growth and development under adverse environmental conditions in extant flowering plants, and repress the accumulation of reactive oxygen species (Achard et al. 2008).

Did these nuclear repressing proteins play a role in the diversification and adaptive radiation of flowering plants in fluctuating paleoclimates?

The ubiquitin 26S proteasome pathway is employed by animals and plants to control levels of- and degrade regulatory proteins such as gibberellin sensitive plant DELLAs and certain ethylene regulatory enzymes (Y. F. Chen et al. 2005).

Several other plant hormones including abscisic acid, auxin, cytokinins, and ethylene, interact with GA; some of these interactions are negative and reciprocal. For example, when rice is grown under hypoxic conditions, culm and internode elongation requires both ethylene and GA. KNOX1 proteins induce the production of cytokinins in SAMs that in turn inhibit the biosynthesis of GA (Weiss and Ori 2007).

Jasmonic acid (JA) and related esters belong to the oxylipin family of molecules (Wasternack 2007, Chico et al. 2008). Jasmonates are hormones implicated in both specific and conserved plant morphogenic effects (I. T. Baldwin 1998, Biondi et al. 2001, Balbi and Devoto 2008, Katsir et al. 2008).

Specific effects include the ability of oxylipins to reprogram expression of jasmonate-inducible genes (Pauwels et al. 2009). Jasmonates, biosynthesized as a response to mechanostimulation (wounding) of tissues by herbivores (I. T. Baldwin 1998), affect development of stamens and pollen (Devoto and Turner 2003).

Nitric oxide (NO) is a gas with profound mediating effects on phytohormone cross-talk, development of flowers and pollen tubes, defense, and signaling (Perazzolli et al. 2006, Besson-Bard et al. 2008 [two papers], Palavan-Unsal and Arisan 2009). Nitric oxide activates the enzyme protein kinase G (PKG) in mammals and both molecules are important to survival of insects in artificial hypoxic environments (Wingrove and O'Farrell 1999, Hoback and Stanley 2001).

Interestingly, NO also binds to hemoglobins, a class of oxygen-binding enzymes found in plants (Perazzolli et al. 2006), and vertebrates and invertebrates including certain species of the insect orders Coleoptera, Diptera, Hemiptera, Hymenoptera, and Lepidoptera (Van Holde et al. 2001, Burmester and Hankein 2007). Further, both NO and PKG enhance the survival of Drosophila larvae when deprived of oxygen in the laboratory (Wingrove and O'Farrell 1999).

Were Paleozoic insect antagonists attracted to potential emissions of nitric oxide of host plants where NO possibly played a role in ameliorating the detrimental effects of local hypoxia on egg and larval development?

Carbon dioxide is also a gaseous signaling molecule, which is perceived by the olfactory receptors of blood-sucking, meat-eating, and herbivorous insects. At certain concentrations carbon dioxide (CO₂) becomes an odor to Manduca moths (Goyret et al. 2008).

How did hypercapnic paleoenvironments following the end-Guadalupian extinction (GuCCE), end-Permian apocalypse (EPE), Triassic-Jurassic carbon cycle event (TrCCE), and end-Barremian biogeochemical perturbation (BaCCE) affect molecular evolution of insect olfactory receptor proteins and coevolution of stem group angiosperms and their pollinators?

Salicylic acid (SA) is a well-known natural signaling molecule that a play key role in activating genes for defense and integrating WRKY transcription factors during plant responses to herbivory and wounding (Wasternack 2007, Balbi and Devoto 2008, Blanco et al. 2009).

Strigolactones are carotenoid-derived hormones implicated in negative repression of axillary bud growth (Dun et al. 2009). Potential biological activity of strigolactones in insect antagonist and host plant systems should be researched.

Signaling by herbivory and wounding. The mechanism of action of hormone signals in SAMs is an active area of research in plant physiology. Through a process termed signal transduction biochemical signals that originate in the insect exoskeleton manipulate and control the genetic machinery needed for orderly (or disorderly) growth and expansion of shoots and other organs of the plant (Pandey et al. 2008).

Molecular studies on the herbivorous sphingid moth, Manduca sexta by Ian Baldwin and students, demonstrate several molecular interactions between the insect and the wild tobacco plant host Nicotiana attenuata (Solanales, Asteranae) that potentially affect the ecology and development of both organisms. The activities of the specialized insect herbivore induces "... a dramatic ethylene release, a jasmonate burst, and a suppression of nicotine accumulation ..." in simulations of herbivory on the host plant (abstract, Winz and I. T. Baldwin 2001). Further, plant mRNAs accumulate "... in response to insect-derived cues" (title, Schittko et al. 2001).

Gall-forming insects inject plant cells with modified transfer RNAs (tRNAs), which are in fact, cytokinins. Injected cytokinins biosynthesized by phloem-feeding insects cause rapid differentiation and disorderly growth in plant tissues resulting in gall formation (Schultz 2002).

Do hormones secreted by parasitic invertebrates or insect instars and larvae upregulate or repress homeotic genes in host seed plant SAMs?

Small CSE polypeptides biosynthesized in the esophageal glands of parasitic nematodes when injected through feeding stylet into plant cells, affect development of floral meristems (Germain et al. 2006). Interestingly, the elusive flowering hormone florigen is a small protein that is transported through the phloem of angiosperms (Shalit et al. 2009).

Plant light receptors. Phytochrome, an enzymatic light receptor implicated in several physiological processes in plants, including flowering, affects regulation of MIKC-type MADS-box genes. Issue Number 12 of Volume 58 (2007) of the Journal of Experimental Botany is devoted to photomorphogenesis.

Further discussion of photoreceptors in plants is beyond the scope of this essay although phytochrome genes figure prominently in early attempts to unravel angiosperm phylogeny (Mathews and M. J. Donoghue 1999).

The review by Mathews et al. (2010) presents new data and valuable insight on the timing of major events in flowering plant evolution based on amino acid sequences of the phytochrome enzyme.

Insect hexamerins: molecular evolution of hemocyanins. As the Neoproterozoic snowball Earth slowly warmed during the Cambrian and Ordovician periods global oxygen concentration increased. Photosynthesis in the chloroplasts of aquatic algae, protozooans, and land plants was the source of oxygen gas (Berner 1999, Berner and Kothavala 2001).

Hemocyanins (Hc) of invertebrate circulatory systems evolved during the early Paleozoic as a class of gas-binding, copper-containing enzymes needed in respiration (Burmester and Hankein 2007, Pick et al. 2009). Iron-containing hemoglobins (Hg) of animals also evolved and diversified during this time (Burmester et al. 1998, Van Holde et al. 2001, Burmester 2004).

Molecular evolution of copperless insect storage hexamerins (Hx) from the respiratory hemocyanin class (Hc) of invertebrate copper-containing enzymes may be linked with the late Devonian/early Carboniferous ice-house cooling of the Earth, accompanied by a drop in oxygen levels to 13% (Berner and Kothavala 2001, Ward et al. 2006). Protein storage macromolecules of the hemocyanin family of enzymes are termed hexamerins (Burmester 2001, Moreira et al. 2004).

Thorsten Burmester and coworkers (1998) describe the molecular evolution of hexamerin hemolymph protein subunits from the hemocyanin family of enzymes in two important reviews (Burmester 2001, Burmester et al. 2006). Hexamerins are moulting storage proteins composed of several polypeptide chains (multimers). Each enzyme subunit has a molecular weight of about 80,000.

Hemocyanins, including hexamerins are critical to developing and respiring entognathous hexapods and true (ectognathous) insects, being necessary for instar development and successive moults (Hagner-Holler et al. 2004, Pick et al. 2009). Larvae digest hexamerin storage proteins to release certain amino acids required for development (Moreira et al. 2004).

Hexamerins are also implicated as silencers of JH signaling in neotenous castes of hemimetabolous termites (X. Zhou et al. 2006, X. Zhou et al. 2007) and caste polyphenism in holometabolous wasps (J. H. Hunt et al. 2007).

Moulting food storage proteins, homeotic selector genes including Hox genes, zygotic genes, field-specific selector genes, compartment selector genes, cell-type-specific selector genes, segment polarity genes, and controlling factors behind the cessation of insect growth (brain neurosecretory prothoracicotropic hormone [PTTH], juvenile hormone [JH], and ecdysone steroids) comprise key elements of the developmental tool kit of insects (Truman and Riddiford 2002, S. B. Carroll et al. 2005).

Trans-acting chromatin proteins and cis-regulatory transcription factors of eukaryotic genomes respond to oxygen levels in controlled environments. For example, lowering oxygen concentration in controlled environments affects levels of nonubiquitinated proteins and activates hypoxia-responsive genomic DNA elements (HREs) (Flück et al. 2007). Further, "hypoxia-regulated gene expression very likely reflects an ancient adaptive trait that was established in primitive eukaryotes in response to alterations in levels of oxygen" (page 529, Flück et al. 2007).

The below graphic is a crude plot of changing oxygen levels in Earth's atmosphere over geologic time based on GEOCARB III (Berner and Kothavala 2001), and other work (Beerling and Berner 2002, Huey and Ward 2005, Ward et al. 2006). I superimposed an evolutionary tree showing cladogenesis of insect hexamerins from invertebrate hemocyanins, which is based on the most recent molecular clock estimates by Hagner-Holler et al. (2007) and morphological data. In addition, I plotted three of the four insect herbivore expansion phases outlined by Labandeira (2006) on the graph.

The preceding plot is redrawn from Figure 6 on page 1072 of Hagner-Holler et al. (2007), Diversity of stonefly hexamerins and implication for the evolution of insect storage proteins, Insect Biochemistry and Molecular Biology 37: 1064-1074, with additional data added from GEOCARB III (Berner and Kothavala 2001), GEOCARBSULF (Berner 2006), and Labandeira (2006).

Three bursts of molecular evolution of insect hexamerins are apparent from the Hagner-Holler et al. data (2007). The first of the hexamerin molecular evolutionary radiations corresponds with the late Devonian/early Carboniferous hypoxic ice-house (DeCARB) Earth from 360 to 320 M.Y.A. (Hagner-Holler et al. 2007), and the initiation of Insect Herbivore Expansion Phase 2 (Labandeira 2006). During the hypoxic interval of the DeCARB seven clades of hexamerin storage proteins diverge from primitive Plecoptera (stoneflies) leading to many derived hemimetabolous and holometabolous insect orders (Hagner-Holler et al. 2007).

A second burst of insect hexamerin evolution occurs prior to the end-Guadalupian extinction and a decline in atmospheric oxygen levels from 31% to 12% (Berner and Kothavala 2001, Ward et al. 2006). During this interval, hexamerins diverged into clades leading to the Coleoptera (beetles and weevils), Hymenoptera (bees and wasps), and Diptera (flies), and certain Lepidoptera (butterflies and moths). This interval of hexamerin diversification corresponds with Insect Herbivore Expansion Phase 2 (Labandeira 2006).

The point of divergence between the hexamerin class belonging to beetles and those Hxs belonging to bees and wasps coincides with the second burst of hexamerin evolution.

Was the second explosion of molecular evolution of insect hexamerin protein subunits driven by developing hypoxia following the end-Guadalupian extinction (GuCCE) of the late Permian, or did diversification occur before the GuCCE as the data suggest?

Dipteran hexamerins diverged about the same time as oxygen levels in Earth's atmosphere declined to 12%. Interestingly, the third burst of hexamerin evolution leading to blattarians (cockroaches) and Isoptera (termites) occurs at a point in geologic time coincident with the Triassic-Jurassic carbon cycle event (TrCCE) (Beerling and Berner 2002).

Despite the rough estimate of molecular divergence times of insect hexamerins by Hagner-Holler et al. (2007) plotted against GEOCARB III, future phylogenetic inference could employ other cladistic algorithms and be fine-tuned with more data on insect hexamerins, possibly including a detailed study of hemoglobin evolution of beetles and wasps (Hgs).

A more detailed synthesis of invertebrate gene duplications and later molecular evolution of hexamerins and vitellogenins in populations of coevolving insect antagonists and Permo-Triassic seed plants is possible. Finely-tuned molecular phylogenetic and isotope studies are needed to implicate global hypoxia as an underlying selective force in the evolution of insect hexamerins. Insect hemoglobins (Hgs) remain relatively unstudied in this context. Several questions come to mind:

Were Paleozoic changes in atmospheric oxygen concentration the selective force driving the molecular evolution of gas-binding hemocyanin respiratory enzymes and moulting storage proteins of Permian insects?

Are hexamerins hypoxia-inducible storage proteins of insects?

What was the role of nitric oxide (NO) in mediating developmental and physiologic effects of hypoxia on Paleozoic insects living with plants?

Did certain steroid hormones manufactured by Permo-Triassic seed plant host shrubs act like ecdysones while sending signals to the genetic machinery of developing eggs, instars of phytophagous insects inhabiting shrub lifeboats?

At the same time that hexamerin molecules were evolving in insect populations and colonies of shrub lifeboats, trampling and secreting insect antagonists might have sent signals to the genetic machinery of SAMs of host plants. Signals originating in the bodies of phytophagous insects had potential far-reaching effects on the biochemical machinery of differentiating cells in SAMs of certain developmentally plastic Permo-Triassic seed plants.

One example of a far-reaching effect might be signaling of the homeotic gene biosynthetic machinery in the nuclei of cells of bisexual cone axes of seed plants sensu Becker and Theißen (2003), Baum and Hileman (2006), and Theißen and Melzer (2007).

Insect sensing and hormone signaling. Arthropods manufacture several classes of hormones that regulate development and, together with anatomically modified appendages they facilitate communication among individuals, within castes of social insects, or possibly with plant hosts.

Some of the hormones produced by insects which are active in signaling and sensing include ecdysone steroids, epoxide methyl esters of sesquiterpenoids, peptides and polypeptides, and pheromones (Robertson and Wanner 2006).

Insect olfactory and gustatory chemoreceptors (Abdel-latief 2007) and innervated bristles are of potential interest with respect to plant to insect, insect to insect, and insect to plant communication by biomolecules and thigmo.

Interestingly, from the research perspective of possible neotenous coevolution of insect antagonists and seed plant hosts, Nicholas Strausfeld (2009) reviews the deeply conserved evo-devo of arthropod mushroom bodies in the biochemical and phylogenetic context. Mushroom bodies are paedomorphic neuropils of insects and firebrats that function in control of behavior, and in olfactory perception (Strausfeld et al. 1998).

Olfactory organs and chemoreceptors (Ray et al. 2008) allow certain insects to sense minute concentrations of pheromones, nitric oxide (NO), carbon dioxide, and other volatiles (VOCs) in the air. Beetles exhibit many adaptations to smell and taste and this is reflected in considerable anatomical, biochemical, and morphological diversification in sensillae of antennae, legs, mouthparts, and wings (Abdel-latief 2007).

Physiology and molecular biology of cone development and flowering. Since the pioneering work of Coen and Meyerowitz (1991) tremendous advances have been made in understanding the physiology and molecular biology of flowering within an evolutionary-developmental (evo-devo) context (Endress 2006, Irish 2006, Glover 2007, Whitney and Glover 2007, Hileman and Irish 2009, P. S. Soltis et al. 2009, Specht and Bartlett 2009).

Regulation of flowering is reviewed in Volume 44 of Advances in Botanical Research, Developmental Genetics of the Flower (D. E. Soltis et al., eds. 2006), and in a later book by Glover (2007). A chapter (Baum and Hileman 2006) in a third book presents a developmental genetic model of the origin of the flower.

Specht and Bartlett (2009) and Rudall and Bateman (2010) review the biology and physiology of floral origins.

Induction of flowering from vegetative SAMs and changes in accessory meristem identity occurs four different ways (Glover 2007):

induction of flowering by perception of photoperiodism

meristem determination and flower formation by DELLA de-repression, the ubiquitin 26S proteasome pathway, and gibberellic acid (GA)

autonomous flowering

vernalization

FLORICAULA (FLO) AND LEAFY (LFY) genes and their DNA-binding HTH proteins are main integrators of the biosynthetic and developmental pathway that leads to organ identity and formation of reproductive modules in the fertile SAMs of seed plants (reviewed by Becker et al. 2000, Irish 2006, Jack 2004, Theißen and Kaufmann 2006, Theißen and Melzer 2007), but UFO, WOX5, and WUSCHEL (WUS) might also be involved (Vásquez-Lobo et al. 2007, Nardmann et al. 2009).

Development of fertile organs (bract and cone scale complexes, carpels, perianth parts, sporophylls) following the integration and induction of MIKC-type MADS-box genes in bisexual strobili may be divided into four sequential steps:

Response of the SAM to environmental (light, temperature, touch) and endogenous/exogenous biochemical signals (hormones produced by the insect and the plant, polypeptides, and microRNAs) that switches the meristem from vegetative to reproductive

Integration of signals leading to activation of SAM identity genes

Meristem identity genes activate reproductive organ identity genes in specific regions of the expanding SAM

Reproductive identity genes activate organ building genes of the growing SAM

However, ferns and fern allies, carry out development of fertile structures differently (Hasebe et al. 1998, Himi et al. 2001). FLORICAULA (FLO) AND LEAFY (LFY) genes of the fern model organism Ceratopteris "do not directly induce MADS-box genes" (page 1205, Shindo et al. 2001).

Induction of seed plant reproductive structures in the four whorls of monopodial SAMs is under genetic control (Irish 2006). While the biochemistry and development of expanding cells of SAMs and accessory fertile meristems is just barely understood (and only in a few extant seed plant species), at least two working models on the molecular mechanism of floral control and evolution, either the ABCDE model (Ditta et al. 2004) or floral quartet model (Theißen 2001) have been proposed.

Evidence of the involvement of floral quartets in floral development is supported by in vitro experiments using B and E-class homeotic proteins (Melzer and Theißen 2009, Melzer et al. 2009).

LEAFY and the biochemical logic behind FLORICAULA (FLO), LEAFY (LFY), and NEEDLY (NLY) DNA-binding HTH proteins figure prominently in papers that explore the use of evo-devo approaches in understanding gymnosperm phylogeny, the origin of angiosperms, and floral origins (Frohlich and Parker 2000, Shindo et al. 2001, Albert et al. 2002, Frohlich 2003, Baum and Hileman 2006, Frohlich 2006, Theißen and Melzer 2007, Vásquez-Lobo et al. 2007, Rudall et al. 2009).

Vásquez-Lobo et al. (2007), in an elegant study of FLO/LFY orthologs in Picea (Coniferales), Podocarpus (Podocarpales), and Taxus (Taxales), while employing ultrasensitive radioactive probes and histochemical methods, clearly demonstrate in situ expression of FLO/LFY gene transcripts in all the organs of female fertile SAMs. These findings, when taken together with similar studies of other gymnosperms including Gnetum parvifolium [Gnetales] (Shindo et al. 2001) cast doubt on some of the proposals outlined in MMT.

The ancestral function of LEAFY, its paralog, NEEDLY, and various orthologs; and the putative arthropod and seed plant biological activity of the DNA-binding HTH proteins they encode remain as enigmatic as ever (see the later subsection on the LEAFY [LFY] switch and insect engraled protein).

DELLA de-repression and GA. What are DELLA nuclear regulatory proteins? The DELLA subfamily of nuclear repressing proteins belongs to the larger family of plant-specific GRAS proteins (Bolle 2004). Further, the DELLA subfamily of proteins is gibberellin sensitive and functions in holding-back plant growth and development in cells of floral meristems in extant angiosperms (Jiang and Fu 2007).

Hao Yu et al. (2004) suggest a close relationship between gibberellins (GAs) and other indeterminate morphogens and expression of homeotic selector genes. Yu et al. state:

"Taken together, the work presented here suggests that GA promotes normal development of floral organs partly by up-regulating the expression of floral homeotic genes AP3, PI, and AG. GA achieves this effect by suppressing the function of two DELLA proteins, RGA and RGL2."

The preceding phrase is quoted from page 7832 of H. Yu, T. Ito, Y. Zhao, J. Peng, P. Kumar, and E. M. Meyerowitz (2004), Floral homeotic genes are targets of gibberellin signaling in flower development, Proceedings of the National Academy of Sciences 101(20): 7827-7832.

Further studies of Arabidopsis suggest that DELLA proteins mediate the GA-dependent development of floral organs however, the detailed interactions between DELLAs and the floral meristem organ identity and integrator genes they repress remain incompletely understood (Hou et al. 2008).

It might be important to understand the details of DELLA repression of floral homeotic genes (and relief of repression by GA) in other experimental seed plant systems (and in a molecular phylogenetic context), as floral regulators are probably a key part of the puzzle behind the origin of the angiosperm flower (Theißen and Melzer 2007).

A possible relationship exists between the plant gibberellin nuclear receptor GIBBERELLIN INSENSITIVE DWARF (GID1), DELLAs, and hormone specific lipases (HSLs) (Itoh et al. 2008, Murase et al. 2008).

Do certain insect nuclear receptor proteins and HSLs bind to plant GA?

Are gibberellins manufactured by host plants bioactive in developing instars and larvae of insect antagonists?

The LEAFY (LFY) switch and insect engraled protein. The LEAFY (LFY) gene encodes a diffusible 47 kilodalton-sized transcription factor which is unique to plants (Sessions et al. 2000, X. Wu et al. 2003). LEAFY (LFY) protein regulates patterning of cells in developing shoot apical meristems (SAMs) of the model flowering plant, Arabidopsis thaliana (Maizel et al. 2005, Moyroud et al. 2009).

LEAFY (LFY) regulates gibberellic acid (GA) mediated transition to flowering and downstream expression of MIKC-type MADS-box genes that determine floral organ identity (H. Yu et al. 2004, Theißen and Melzer 2007, Hou et al. 2008, among others). The paralog of LFY in gymnosperms is NEEDLY (NLY) (Mouradov et al. 1998).

Engraled (en), also spelled "engrailed" in the literature, is an insect compartment selector gene that encodes a homeodomain transcription factor that determines the posterior identity of embryos and wings of the Drosophila model arthropod. The Apterous (ap) compartment selector gene encodes Apterous protein, which is another transcription factor involved in subdivision of imaginal discs into dorsal and ventral compartments of developing fruit flies (S. B. Carroll et al. 2005).

The structure of the LFY-C protein of Arabidopsis thaliana consists of a helix-turn-helix (HTH) DNA-binding motif with uncanny similarities to the active motif of Drosophila engraled (en) homeodomain protein, insect Tc3A transposase, and arthropod Hin recombinase enzymes (Hamès et al. 2008).

Now that Hamès et al. (2008) deduced the 3-dimensional crystal structure of LFY-C protein from Arabidopsis, it is possible to formulate explicit hypotheses on the origin of angiosperms and the origin of the first flower involving molecular coevolution of seed plant LEAFY (LFY) and insect cis-acting regulatory modules and homeodomain proteins including engraled (Hittinger and S. B. Carroll 2008).

"Our study reveals that the LFY master regulator, which determines flower meristem fate and controls the expression of floral organ identity genes, shares structural similarity with other HTH proteins, indicating that this universal DNA-binding motif has also been adopted in plants to trigger major developmental switches."

The next chapter of the essay considers evidence from evolutionary-developmental genetics and genomics needed to better understand the origin of the first flower and the origin of angiosperms.

Genetic Considerations:

In this chapter I discuss molecular evolution of transcription factors and cis-regulatory modules, co-option and exaptation, developmental recombination and reprogramming, genetic accommodation, homeotic selector genes, horizontal transfer of genes (HGT), chloroplast capture, gene duplications, and duplication of whole genomes (WGDs) including polyploidy.

The image on the right is a stained preparation of the chromosomes of Joinvillea plicata (Joinvilleaceae, Poales, Lilianae), a species of grass-like monocots with a fully developed perianth. The chromosomes in this image were fixed at the moment of metaphase I of meiosis in a microspore mother cell, × 2000. In the living Joinvillea plant, meiotic chromosomes would be identifiable in two places: inside the immature megasporangium of the ovule and within young microsporangia of very young stamens.

In addition, I summarize key research advances in our understanding of the possible effects of signals on whole genomes, homeotic selector genes, gene transcripts such as microRNAs (miRNAs and siRNAs), and cis-regulatory sequences that bind transcription factors.

"A central concept that is emerging from comparative studies of developmental genes and their interactions is the co-option of these for different functional roles as evolution proceeds."

This quotation is from page 761 of W. Arthur (2002), The emerging conceptual framework of evolutionary developmental biology. Nature 415: 757-764.

Could the application of mechanical force and secretion of phytohormone signals by chewing, crawling, feeding, ovipositing, and sucking insect mutualists on elongated (or shortened) cone axis SAMs affect movements of homeodomain proteins and trigger developmental switches that determine downstream reproductive organ identity and formation?

"I argue that the origin of species differences can be explained, and the synthesis of Darwinism with genetics can be improved, by invoking two concepts: developmental recombination and genetic accommodation. Developmental recombination, or developmental reorganization of the ancestral phenotype, explains where new variants come from: they come from the preexisting phenotype, which is developmentally plastic and therefore subject to reorganization to produce novel variants when stimulated to do so by new inputs from the genome or the environment. Genetic accommodation, or genetic change in the regulation or form of a novel trait, is the process by which new developmental variants become established within populations and species because of genetic evolution by selection on phenotypic variation when it has a genetic component."

The preceding passage is quoted from page 6544 of M. J. West-Eberhard (2005), Developmental plasticity and the origin of species differences, Proceedings of the National Academy of Sciences 102(Supplement 1): 6543-6549.

These and many other questions are in need of answers:

What is the role of developmental recombination and reprogramming, and genetic accommodation, as underlying processes that lead to co-option, modularity, dissociation, and diverted development of seed plant characters involved in angiospermization?

Were the genes, morphogens, and regulatory factors having to do with mechanoperception and shoot development linked or the same in extinct Permo-Triassic seed plants?

Are phytophagous insects and their secretions a source of signals perceived by the developing plant SAM (including accessory reproductive meristems) and sent to nuclei of cells where they upregulate (or repress) floral integrator and meristem identity genes?

Does mRNA and homeodomain protein trafficking in meristematic tissues of model ophioglossoid monilophytes occur differently than seed plants?

Do certain insect nuclear receptor proteins and HSLs bind to plant GA?

Are microRNAs passed on to male and female reproductive modules (pollen and ovules)?

Can we identify potentially interfertile sources of reproductive modules, and model the flow of pollen and rain of seeds in pre-angiospermous seed plant populations following the end-Permian extinction (EPE) and Triassic-Jurassic boundary carbon cycle event (TrCCE)?

What were the effects of epigenesis, including input of foreign genetic material, on developmental plasticity and phenotypic variation in populations of Permo-Triassic seed plants and the first angiosperms?

How has the horizontal transfer of genes in paleopopulations affected the origin and evolution of flowering plants?

Do phytophagous insects send biochemical signals via the host plant cytoskeleton and microtubules to the nucleus thereby upregulating SAM maintenance- and floral integrator and meristem identity genes?

Homeotic selector genes. Homeotic selector genes are known from metazoans, fungi, and plants. These genes are key regulators in insect and land plant developmental tool kits being important in plant organ or body segment identity (S. B. Carroll et al. 2005, Floyd and Bowman 2007). Homeotic selector genes of insects include Hox genes, zygotic (gap- maternal-, and pair-rule-) genes, field-specific selector genes, compartment selector genes, cell-type-specific selector genes, and segment polarity genes (S. B. Carroll et al. 2005).

Homeotic selector genes of plants include shoot apical meristem (SAM) maintenance genes and floral meristem organ identity and integrator genes that encode MIKC-type MADS-box transcription factors.

Developmental proteins encoded by homeotic selector genes of plants have been the subject of active research by biochemists, plant cell biologists, and molecular systematists for nearly two decades (Coen and Meyerowitz 1991, Theißen et al. 1996, Kramer et al. 1998, Becker et al. 2000, Cubas et al. 2001, Stuurman et al. 2002, Becker and Theißen 2003, Irish 2003, S. Tucker 2003, Yamada et al. 2003, Kaufmann et al. 2009, Melzer et al. 2009, among others).

More recent studies having to do with the origin of the flower, perianth parts, bisexual cone axes, ovules and carpels include Buzgo et al. (2004), Jack (2004), S. Kim et al. (2004), P. Zhang et al. (2004), Adam et al. (2005), Buzgo et al. (2005), Irish and Litt (2005), Kaufmann et al. (2005), S. Kim et al. (2005 [two papers]), Zahn et al. (2005 [two papers]), Baum and Hileman (2006), Cronk (2006), De Bodt et al. (2006), Doerner (2006), Endress (2006), Feng et al. (2006), Irish (2006), S. Kim et al. (2006), Scutt et al. (2006), P. S. Soltis et al. (2006), Kramer et al. (2007), and Theißen and Melzer (2007), among others.

Developmental cis-regulatory modules. Non-enzyme coding DNA sequences involved in the control of gene expression are termed cis-regulatory modules (Zinzen et al. 2009, Borok et al. 2010).

Clusters of regulatory DNA spaced at irregular intervals on the chromosomes of diverse eukaryotes sometimes contain enhancer and promoter cis-regulatory sequences that bind transcription factors (Borok et al. 2010). Promoters are cis-regulatory sequences that affect transcription. Some promoters work together with additional factors termed modular enhancers, which are cis-regulatory elements [CREs] (Rebeiz et al. 2009, Borok et al. 2010).

Modular enhancer CREs of certain insects are subject to mutation. Further mutations in CREs within Drosophila populations potentially lead to adaptive morphological changes with "large phenotypic effect" (abstract, Rebeiz et al. 2009).

"Plant evolution may often entail changes to the promoter/enhancer elements that cause changes in the expression patterns of regulatory genes."

The preceding quotation is from page 59 of D. A. Baum and M. J. Donoghue (2002), Transformation of function, heterotopy, and the evolution of plant development. Pp. 52-69 In: Q. C. B. Cronk, R. Bateman, and J. Hawkins (eds.), Developmental Genetics and Plant Evolution, The Systematics Association Special Volume Series 65. London: Taylor and Francis, 543 pp.

Transcription factors are proteins that bind to cis-regulatory sequences (Wray et al. 2003).

It is increasingly clear that cis-regulatory MADS-box domain (see section below) transcription factors interact even with their own promoters. Further, "MADS-box proteins seem to function mostly as subunits of larger protein complexes" (page 34, Rijpkema et al. 2007).

Understanding the selective forces behind molecular coevolution of insect and seed plant developmental cis-acting regulatory modules in deep-time may be an important first-step in unraveling the enigmatic origin of angiosperms, evolutionary development of the first flower, and adaptive radiation of certain clades of coevolving Holometabola.

Evolution of cis-regulatory modules in relation to animal and plant body plan patterning, development of organs of insects and seed plants, and natural selection is reviewed by Rodríques-Trelles et al. (2003), Wray (2003), Wray et al. (2003), Theißen and Becker (2004), S. B. Carroll (2005), De Bodt et al. (2006), Wittkopp (2006), Hoekstra and Coyne (2007), Wray (2007), and S. B. Carroll (2008).

The diagram below, which is redrawn from Carroll (2005), shows how the cis-regulatory molecular evolution and duplication of genes in insects or plants may potentially increase gene function while minimizing pleiotropy. The blue lines are segments of coding and non-coding DNA. Intron pairs are gray-colored rectangles. Orange circles are cis-regulatory promoters. Finally, asterisks denote mutations.

The first line of "A" illustrates a progenitor gene with its introns and promoter. The second two lines of "A" shows a gene duplication. Potential mutations of the cis-regulatory promoter and gene intron are denoted by asterisks. "B" illustrates expansion of cis-regulatory elements. "C" depicts the genesis of a novel intron and potential splicing sites that may lead to biosynthesis of alternative protein forms without the deleterious effects of mutations.

The above graphic is redrawn from Figure 1 of Sean B. Carroll, 2005, Evolution at two levels: on genes and form, PLoS Biology 3(7): 1159-1166.

A typical promoter consists of fewer than 50 binding sites for less than 15 different transcription factors. Transcription factor binding sites are about a dozen nucleotide base pairs wide, spaced irregularly along a segment of chromatin. Some promoters are highly conserved while others are not.

Shoot apical meristem (SAM) maintenance genes. Function and patterning in the shoot apical meristem (SAM) requires several gene families and their cis-acting transcription factors. The main shoot apical meristem identity genes are Class III HD-Zip, KANADI, CLASS 1 KNOX (KNOTTED1-like), and ARP genes (Ori et al. 2000, Reiser et al. 2000, Hake et al. 2004, Di Giacomo et al. 2008).

Class III homeodomain-leucine zipper (Class III HD-Zip) proteins, which are encoded by Class III HD-Zip genes, include AtHB8, CNA, PHB, PHV, and REV (see reviews by Floyd et al. 2006 and Prigge and S. E. Clark 2006). Class III HD-Zip genes and the enzymes they encode have been isolated from extant flowering plants, gymnosperms, ferns, lycophytes, bryophytes, and charophytes including most of the model photosynthetic organisms used in evo-devo research. All were probably key players in the land plant developmental tool kit of cis-acting genes, transcription factors, and homeodomain proteins (Floyd and Bowman 2007).

The diagram below is a phylogeny of Class III HD-Zip genes redrawn from Floyd et al. (2006). Cha = Chara (Charophyta), Phy = Physcomitrella (Bryophyta), Pha = Phaeoceros (Anthocerotophyta), Mar = Marchantia (Hepatophyta), Psi = Psilotum (Psilophyta), Cer = Ceratopteris (Pteridophyta), Pse = Pseudotsuga (Coniferophyta), Gin = Ginkgo (Ginkgophyta), Tax = Taxus (Coniferophyta), Ory = Oryza (angiosperm), Zin = Zinnia (angiosperm).

Class III homeodomain-leucine zipper (Class III HD-Zip) genes are abbreviated on the diagram (e.g. CNA = CORONA/AtHB15, PHB = PHABULOSA, PHV = PHAVOLUTA, REV = REVOLUTA). The CNA, PHB, PHV, and REV cis-regulatory modules are principal components of the plant SAM, which are also needed for initiation of axillary SAMs in vascular plants including seed plants (see Floyd et al. 2006 for references).

Many of the supposed homologies of leaves, sporophylls, and ectopic structures including reproductive axes and ovular integuments could be incorrect based on the possibility that a fundamental dichotomy in SAM evo-devo emerged in Devonian lignophyte populations.

Interestingly, lateral branches of the Devonian progymnosperm, Archaeopteris (Archaeopteridales) may arise from tissue fields resembling leaf primordia, and not axillary buds (page 481, T. N. Taylor et al. 2009).

Did early-diverging Devonian progymnosperms sprout branches without axillary SAMs and buds?

Three of the Class III HD-Zip genes (PHB, PHV, and REV) are needed by the plant for development of leaves and the placement of organs on upper (adaxial) leaf surfaces (Floyd et al. 2006). Additional functions of Class III HD-Zip genes are known (Floyd and Bowman 2007).

Colored bars on the cladogram denote landmark developmental events in vascular plant evolution. Yellow = vascularization of the microphyll of lycophytes, Orange = evo-devo of the sporophyte SAM, Blue = organization of the stele of the vascular plant stem, Light Green = site of lateral organ initiation, Dark Green = evo-devo of adaxial domains.

Phylogenetic analyses suggest that duplications in floral meristem identity genes took place prior to the origin of angiosperms more than 260 million years ago (Bowe et al. 2000, Bowers et al. 2003, S. Kim et al. 2004). Further, mathematical modeling studies of Arabidopsis, the highly researched extant model malvid plant, reveal that diversification of genes involved in signal transduction, SAM development, and the regulation of transcription are attributable to three WGDs during the last 350 million years (Maere et al. 2005).

Similarly, a phylogenetic analysis of KNOX1 genes by Guillet-Claude et al. (page 2242, 2004) "suggests that three duplication events (D1, D2, and D3) occurred between the split of the lineage leading to conifers from that leading to angiosperms, around 300 m.y.a. ..." and the Cretaceous divergence of pines and spruces.

Paleopolyploidy. A great body of data has been assembled on paleopolyploidy in plants (Bennett and Leitch 2005, J. J. Doyle et al. 2008). Evidence from an exhaustive genomic study of the cultivated grape overwhelmingly supports the existence of paleohexaploidy (Jaillon et al. 2007).