ORIGINAL ARTICLE

From shoot to leaf: step-wise shifts in meristem and KNOX1activity correlate with the evolution of a unifoliate body planin Gesneriaceae

Kanae Nishii1,2 & Bing-Hong Huang3 & Chun-Neng Wang4 & Michael Möller1

Received: 6 March 2016 /Accepted: 24 November 2016 /Published online: 8 December 2016# Springer-Verlag Berlin Heidelberg 2016

Abstract Typical dicots possess equal-sized cotyledons andleaf-bearing shoots topped with a shoot apical meristem(SAM), the source of lateral organs, and where KNOX1 ho-meobox genes act as key regulators. NewWorld Gesneriaceaeshow typical cotyledons, whereas Old World Gesneriaceaeshow anisocotyly, the unequal post-germination growth ofcotyledons, and include unifoliate (one-leaf) plants. One-leafplants show an extremely reduced body plan: the adult above-ground photosynthetic tissue consisting of a single cotyledon,a macrocotyledon enlarged by the basal meristem (BM), butlacking a SAM. To investigate the origin and evolution of theBM and one-leaf plants, the meristem activity and KNOX1SHOOTMERISTEMLESS (STM) expression in cotyledonsand leaves were systematically studied by RT-PCR and in situhybridization across the family Gesneriaceae, Jovellana inCalceolariaceae (sister family to Gesneriaceae), andAntirrhinum in Plantaginaceae, all families of orderLamiales (asterids), in comparison to Arabidopsis

(Brassicales, rosids). In all examined Lamiales samples, un-like Arabidopsis, BM activity accompanied by STM expres-sion was found in both cotyledons in early stages. Foliageleaves of Gesneriaceae and Jovellana also showed the corre-lation of BM and STM expression. An extension of BM ac-tivity was found following a phylogenetic trajectory towardsone-leaf plants where it is active throughout the lifetime of themacrocotyledon. Our results suggest that KNOX1 involve-ment in early cotyledon expansion originated early on in thediversification of Lamiales and is proposed as the prerequisitefor the evolution of vegetative diversity in Gesneriaceae. Step-wise morphological shifts, driven by transfers of meristematicactivity, as evidenced by shifts in KNOX1 expression, may beone mechanism by which morphological diversity evolves inplants.

Keywords KNOX1 . Streptocarpus . Gesneriaceae .

Cotyledon . Leaf . Meristem

Introduction

Typical dicots possess equal-sized cotyledons and leaf-bearing shoots topped with a shoot apical meristem (SAM),though many variations exist. New World Gesneriaceae (sub-family Gesnerioideae), in most cases, possess an above-ground morphology with a typical shoot system with theSAM developing between two equal cotyledons (isocotyly)and produce an ‘ordinary’ shoot structure with decussate leafarrangement (Fig. 1). On the other hand, the seedling mor-phology in Old World Gesner iaceae (subfamilyDidymocarpoideae) is distinct from that of New WorldGesneriaceae, because of anisocotyly, their unequal-sizedpost-germination development of the cotyledons (Burtt1963; Fig. 1d–g). In addition, there are variable shoot systems

Communicated by Sureshkumar Balasubramanian

Electronic supplementary material The online version of this article(doi:10.1007/s00427-016-0568-x) contains supplementary material,which is available to authorized users.

* Michael Möllerm.moeller@rbge.ac.uk

1 Royal Botanic Garden Edinburgh, 20A Inverleith Row,Edinburgh, Scotland EH3 5LR, UK

2 Tokyo Gakugei University, 4-1-1 Nukuikitamachi, Koganei,Tokyo 184-8501, Japan

3 National Taiwan Normal University, No. 88, Sec. 4, Ting ChowRoad, Taipei 11677, Taiwan

4 National Taiwan University, No.1, Sec. 4, Roosevelt Road,Taipei 10617, Taiwan

Dev Genes Evol (2017) 227:41–60DOI 10.1007/s00427-016-0568-x

http://dx.doi.org/10.1007/s00427-016-0568-xhttp://crossmark.crossref.org/dialog/?doi=10.1007/s00427-016-0568-x&domain=pdf

42 Dev Genes Evol (2017) 227:41–60

recognized in Old World Gesneriaceae. The majority show atypical shoot system with decussate leaf arrangement, butsome show leafy organs without a typical SAM.

In Old World Gesneriaceae, one cotyledon, themacrocotyledon, continues to grow after germination to al-most resemble a foliage leaf, while the other, themicrocotyledon, withers away eventually (Fig. 2a, b). Thedegree of the macrocotyledon growth is different betweenevolutionary lineages of Old World Gesneriaceae. Incaulescent ones, such as Streptocarpus glandulosissimus, theanisocotyly phase, while short, delays the development of theSAM but eventually forms a typical leaf-bearing shoot(Fig. 2c). On the other hand, in acaulescent Streptocarpus,the macrocotyledon growth is greatly extended, sometimesfor years, and they can reach >1 m in length, but no SAM isformed (Fig. 2d, e; Jong 1970; Hilliard and Burtt 1971). Thus,acaulescent Streptocarpus retain a ‘leafy organ’ originatedfirst from the macrocotyledon: the macrocotyledon enlargesdue to the activity of a basal meristem located at the proximalend of the lamina and the petiolode meristem, a diffuse mer-istematic region of the petiole, near the base of the lamina,extending the petiole and midrib to form a leafy organ termed‘phyllomorph’, with the macrocotyledon forming the ‘cotyle-donary phyllomorph’ (Fig. 2d, e; Jong 1970; Jong and Burtt1975). In rosulate acaulescent species, additionalphyllomorphs are formed in an irregular arrangement fromthe groove meristem located on the adaxial side of the petiolenear the base of the lamina (Figs. 1f and 2d). In unifoliateacaulescent species, no further phyllomorphs are formed,and the cotyledonary phyllomorph is the only above-groundorgan of the plant, producing inflorescences at the base of thelamina (Figs. 1g and 2e; Jong 1970; Jong and Burtt 1975;Imaichi et al. 2000).

The leaf ontogeny in Gesneriaceae has previously beenstudied in detail only in Old World species. It was particularlywell studied in acaulescent Streptocarpus (e.g. Jong and Burtt

1975; Harrison et al. 2005; Mantegazza et al. 2007) and uni-foliate Streptocarpus or Monophyllaea species (e.g. Tsukaya1997; Imaichi et al. 2000; Nishii et al. 2004; Ayano et al.2005). Only the caulescent S. glandulosissimus with typicalshoots has been studied in detail, and an area at the proximalend of the leaf with an extended meristematic activity wasfound, similar to the basal meristem in phyllomorphs of acau-lescent Streptocarpus species (Nishii et al. 2010). Nothing isknown in this respect from New World Gesneriaceae that donot exhibit anisocotyly, and thus, the critical questions to ad-dress are when the basal meristem has evolved in the evolu-tion of Gesneriaceae, and where, in the cotyledon or leaves, itwas first established. This is particularly interesting with viewto the evolution of plant meristems leading to unifoliate one-leaf plants, whether it is caused by an evolutionarily rapidtruncation of the shoot or represents a gradual transition inshoot morphology (Hilliard and Burtt 1971; Cronk andMöller 1997; Möller and Cronk 2001).

To this end, we selected species representing growth formsof major lineages across the phylogeny of Gesneriaceae(Möller et al. 2009, 2011; Möller and Clark 2013; Fig. 1,Online Resource 1) and determined the position of cell divi-sion activity in the cotyledon and leaf. Furthermore, the ex-pression of class 1 KNOX homeobox (KNOX1) genes wasevaluated since it was shown in the previous studies that inboth cotyledons, leaves, and phyllomorphs of Streptocarpus,KNOX1 has been found linked to the areas of meristematicactivity (Mantegazza et al. 2009; Nishii et al. 2010), and it iswell established that KNOX1 genes function as regulators ofundifferentiated tissues and maintainers of meristematic prop-erties in the SAM of angiosperms (Vollbrecht et al. 1991;Sm i t h e t a l . 1 9 9 5 ; L o n g e t a l . 1 9 9 6 ) , w i t hSHOOTMERISTEMLESS (STM) being a major player in di-cots (Long et al. 1996). Thus, we isolated STM homologuesfrom our study plants and determined their expression do-mains and employed them as the developmental genetic mark-er for meristem identity. To place the findings in a larger evo-lutionary context, we compared the Gesneriaceae species tooutgroup taxa across the core eudicots (i.e. asterids: Lamiales:Jovellana punctata , Antirrhinum majus) includingArabidopsis thaliana (Brassicales, rosids; Online Resource 1).

Materials and methods

Plant materials

Plant and seed materials were provided by the Royal BotanicGarden Edinburgh (RBGE), Taipei Botanic Garden (Taiwan),and Cecilia Koo Conservation Center (Taiwan). Seeds of someoutgroup species were purchased from Suttons Seeds(Paignton, UK) (Online Resource 2). The selection of experi-mental species across the Gesneriaceae followed recent

Fig. 1 Experimental material selected across the Lamiales in aphylogenetic context (see Stevens 2001 onwards; Weber et al. 2013). aAntirrhinum majus , Plantaginaceae. b Jovellana punctata ,Calceolariaceae. c Corytoplectus speciosus, New World Gesneriaceae(isocotylous). d–g Old World Gesneriaceae (anisocotylous). dHenckelia anachoreta. e Streptocarpus glandulosissimus, caulescent withtypical shoot apical meristem (SAM). f, g Acaulescent Streptocarpuslacking a shoot and SAM (Jong and Burtt 1975). f Streptocarpus rexii,rosulate, forming leaves from the groove meristem arranged in an irreg-ular rosette (Jong and Burtt 1975). g Streptocarpus wendlandii, unifoliate,retaining the macrocotyledon as sole foliar organ (Hilliard and Burtt1971). Column 1 habit, column 2 schematic illustration of cotyledon/leaf shape, column 3 SEM micrographs of seedlings, column 4 imagesof foliage leaves and macrocotyledon of the unifoliate S. wendlandii,column 5 descriptions of materials, from left to right: M meristem type,L leaf arrangement and type, C cotyledon type, S systematic position ofthe study material: *New World subfamily Gesnerioideae, **Old Worldsubfamily Didymocarpoideae. co cotyledon, Mc macrocotyledon, mcmicrocotyledon

Dev Genes Evol (2017) 227:41–60 43

44 Dev Genes Evol (2017) 227:41–60

molecular phylogenetic studies (Möller et al. 2009, 2011;Online Resource 1). As outgroups for developmental studies,J. punctatawas selected belonging to Calceolariaceae, the clos-est sister family to Gesneriaceae (Schäferhoff et al. 2010;Weber et al. 2013), as well as A. majus (Plantaginaceae) as amodel plant from within Lamiales (Refulio-Rodriguez andOlmstead 2014), both with simple leaves and a pair of typicalequal-sized dicot cotyledons. Although the margin of the fo-liage leaf of J. punctata is slightly serrated, the leaf primordiaare simple (Online Resource 3a).

In Gesneriaceae, we chooseCorytoplectus speciosus from theNew World, also with simple leaves and isocotylous seedlings,and Henckelia anachoreta from the Old World Gesneriaceaewith simple leaves but anisocotylous seedlings. InStreptocarpus, we selected the caulescent S. glandulosissimusdeveloping anisocotylous seedlings and pairs of simple decussateleaves formed from a typical SAM, the perennial rosulateStreptocarpus rexii also producing unequal-sized cotyledonsbut additional leaves (phyllomorphs, henceforth termed leavesfor simplicity unless specifically given), and Streptocarpuswendlandii as a unifoliate one-leaf plant, only retaining the cot-yledonary phyllomorph as sole foliar organ, flowering andfruiting only once (monocarpic) (Hilliard and Burtt 1971). Inaddition, we studied a range of Gesneriaceae species to confirmthe result observed in the representative species (OnlineResource 2).

Cloning of STM homologues from Gesneriaceaeand J. punctata

Total RNAwas extracted using TRIzol (Invitrogen, Carlsbad,CA, USA). First-strand complementary DNA (cDNA) wassynthesized with an OligodT primer [d(T)18] andSuperScript III Reverse Transcriptase (Invitrogen). STM-KNOX1 was amplified by PCR using degenerate primersbased on regions from the KNOX1 domain to thehomeodomain conserved between STM (A. thaliana),NTH15 (Nicotiana tabacum), Let6 (Solanum lycopersicum),INA (A. majus), and SrSTM1 (S. rexii) (Online Resource 4).

The amplified DNA fragments were subcloned using thepGEM-T Easy Vector System (Promega, WI, USA) and se-quenced through the Edinburgh Genomics sequencing serviceof the University of Edinburgh (UK). The STM homologueswere identified as STM1 in Gesneriaceae (Harrison et al.2005). In Harrison et al. (2005), STM2 was also reported asan additional STM gene from Streptocarpus. However, since itwas characterized as a rather short sequence, with 19 aminoacids located in the homeodomain and had an intron in adifferent position compared to Gesneriaceae STM1 andArabidopsis STM, it is obviously quite divergent from these;thus, we excluded it from our phylogenetic study. For simplic-ity, we use the term ‘STM’ for STM-like KNOX1 genes ingeneral here (e.g. including Arabidopsis STM, AntirrhinumINA , HIRZ) and for the STM gene homologues inGesneriaceae STM1. The newly acquired gene sequenceswere deposited in GenBank and their accession numbers giv-en in Online Resource 5.

Homology analyses

The codon sequences of the STM genes isolated here werealigned with previously reported KNOX1 genes usingBioEdit v.5.0 (Hall 1999). A Bayesian inference (BI) phylo-genetic analysis was performed using MrBayes v.3.2.2(Ronquist et al. 2012). Best-fitting nucleotide substitutionmodels were determined for each codon position separatelyin MrModeltest v.2.3 (Nylander 2004). Under the Akaike in-formation criterion (Akaike 1974), GTR+R was chosen forthe 1st and 2nd codon positions, and GTR+I+G for the 3rdposition. Five million generations were run in two indepen-dent parallel runs of four MCMC chains and sampled every1,000th generation. The first 500 trees (10%) were discardedas burn-in, determined by checking for stationarity of the like-lihood values against generations. From the remaining trees,BI majority rule consensus trees and posterior probabilities(PP) were obtained. Convergence of the two runs wasassessed through the average standard deviation of split fre-quencies (0.009826). Accession numbers and codes of thesequences used in the tree are given in Online Resource 5.

Cotyledon morphometry and meristematic activity

For scanning electron microscopy (SEM) observations, sam-ples were fixed in FAA (3.7% formaldehyde, 5% acetic acid,50% ethanol, in distilled water), dehydrated in an ethanol se-ries and immersed in acetone. Following drying and ion-sputter coating, samples were observed with a LEO Supra55VP SEM (Zeiss, Welwyn Garden City, UK). Seedlings ofdifferent developmental stages were analysed: for the earlydevelopmental stage, seedlings 1 day after cotyledonunfolding (1D) were used. For later developmental stages,

Fig. 2 Schematic representation of phyllomorphic organization inStreptocarpus. a, b Seedling growth of Old World Gesneriaceae.Seedling just after germination shows equal cotyledons (a), and onecotyledon continues to grow to become a macrocotyledon (b). c–e Themacrocotyledon enlarges to resemble a foliage leaf in caulescentStreptocarpus (c) or become a cotyledonary phyllomorph in acaulescentStreptocarpus (d, e), and it is indistinguishable from the additionalphyllomorphs in rosulate Streptocarpus with cotyledonary phyllomorphand additional phyllomorphs (d) and unifoliate Streptocarpus with onlythe cotyledonary phyllomorph (e). f, g Schematic illustration of aphyllomorph with its unique meristems. Top view (f) and side view (g).ap additional phyllomorph, ar adventitious root, bm basal meristem, cocotyledon, cp cotyledonary phyllomorph, gm groove meristem, hyhypocotyl, if inflorescence, Mc macrocotyledon, mc microcotyledon,me mesocotyl, mr midrib, pd petiolode, pm petiolode meristem

Dev Genes Evol (2017) 227:41–60 45

isocotylous species with plumule (10D) or anisocotylous spe-cies with apparent macrocotyledon (20–30D) were used.

For observation of cotyledon morphometry and cell divisionactivity, seedlings of 1D, 10D, 20D, and 30D were fixed inethanol and glacial acetic acid (4:1). Tomeasure the area, imagesof cotyledons were taken with a Zeiss AxioCam MRc5 CCDcamera mounted on a Zeiss Axiophot brightfield microscope.The images were analysed with Zeiss AxioVision v.4.7 and thedata transferred to Microsoft Excel. The number of lateral veinswas counted in the same samples. To identify dividing cells incotyledons, Aniline Blue staining, for the fluorescent detectionof ß-1,3glucan, which is contained in newly formed cell walls,was conducted as described previously (Nishii et al. 2004;Kuwabara and Nagata 2006; Online Resource 6). Samples wereobserved under fluorescence on a Zeiss Axiophot microscopeusing UV excitation. We observed Aniline Blue-stained cellwalls of epidermal and subepidermal cell layers from the adaxialside, where no stomata occur in the examined species ofGesneriaceae and J. punctata. In A. majus, stomatal divisionwas observed on the upper surface of cotyledons, and thus, weonly counted the septum walls between two square cells, not inround stomata cells. The area of a cotyledon retaining AnilineBlue-stained cell walls was determined by tracing the outline ofcotyledons and the XY-axis locations of the cell walls in thecotyledons under fluorescence and analysed in Photoshop(Adobe Systems Inc.) and ImageJ (Schneider et al. 2012). Thedata were transferred to Microsoft Excel and 2D plots prepared(see Online Resource 6).

Leaf morphology and meristematic activity

To allow comparisons of the leaf growth between species, weused proportional leaf lengths (%). The maximum leaf lengthused for each species was as follows: A. majus 7.5 cm,J. punctata 11 cm, C. speciosus 12 cm, H. anachoreta12 cm, S. glandulosissimus 6 cm, S. rexii 29 cm, andS. wendlandii 30 cm. We plotted the number of veins againstthe proportional leaf length. Primary lateral veins were ob-served by eye in leaves more than 10 mm in length. Leavessmaller than 10 mm in length were fixed in ethanol/acetic acid(4:1), further cleared with chloral hydrate (Nishii et al. 2004),and observed under an Zeiss Axiophot brightfield microscope.

To observe the cell division arrest front, developing leaveswere divided into eight sections from tip to base and stainedwith Aniline Blue. Each leaf segment was assessed for meri-stematic activity. The relative position of the most distal leafsegment with meristematic activity was calculated for eachleaf. The highest value in a category was taken as indicatorfor the location of the arrest front. The leaves at the differentdevelopmental stages were analysed (see Online Resource7a–f), and the data at five representative stages (

specific primer pairs designed here were used (OnlineResource 4). 18S ribosomal RNA (rRNA) was amplified asinternal controls. RT-qPCR was carried out with the KAPASYBR FAST qPCR kit (Kapa Biosystems, Woburn, MA,USA), using a Bio-Rad CFX real-time PCR machine (Bio-Rad, Hercules, CA, USA). The melting curve was analysedfor each experiment individually for each primer set. The ef-ficiency of the primer pairs and the obtained threshold cycle(Ct) values were analysed in REST (Pfaffl et al. 2002). Allexperiments were conducted in triplicates. Relative STM ex-pression levels and the standard error in cotyledons were cal-culated using 18S rRNA as reference gene and whole seed-lings as controls.

Gene expression analyses by in situ hybridization

STMDNA fragments were amplified from cDNA preparedby PCR using gene-specific primers. The regions used forprobes were 14-A∼113-Q for J. punctata, 34-R∼120-S forC. speciosus, and S-83∼F-194 for S. glandulosissimus, ofdeduced amino acid sequences. The amplified sequenceswere cloned into P-GEM-T Easy Vector (Promega) andinsertion confirmed by sequencing as described above.The plasmids with the species-specific inserts were usedas templates. Digoxygenin-labelled (DIG) RNA probesfor STM genes were generated using an in vitro transcrip-tion kit (Roche Diagnostics GmbH, Mannheim, Germany)according to the manufacturer’s protocol. Sense tran-scripts from the same regions were used as negative con-trols. Hybridization and immunological detection was per-formed as described previously (Mantegazza et al. 2007).Purple staining was interpreted as positive signals.Occasionally, brown∼beige colouration was observed inthe sections as background staining and was characteristicof the Gesneriaceae and J. punctata material. Images of insitu hybridization were taken with an Olympus BX51brightfield microscope (Olympus, Tokyo, Japan).

Results

Isolation of STM genes

We isolated STM-KNOX1 homeobox genes from nineGesneriaceae in addition to SrSTM1 of S. rexii (Harrisonet al. 2005) and three outgroup species: partial STM sequenceswere obtained from the New World C. speciosus, Sinningiaschiffneri , Sinningia speciosa, and the Old WorldH. anachoreta, Loxostigma griffithii, and Rhynchoglossumgardneri. STM was also isolated from the outgroup samplesJ. punctata, Calceolaria angustifolia, and Calceolariaascendens (all Calceolariaceae) (Fig. 3). The deduced aminoacid sequences of the isolated sequences were aligned and the

codon sequences used to build a BI tree with those KNOX1genes previously characterized frommodel plants [A. thalianaSTM, N. tabacum (tobacco) NTH, S. lycopersicum (tomato)Let6, and A. majus (snapdragon) AmINA and AmHIRZ] re-trieved from GenBank. The Gesneriaceae and outgroup se-quences were found homologous to the previously reportedSTM genes (PP = 1.0; Fig. 4). Furthermore, all three intronsfound in A. thaliana STM (protein ID AAF70849,25193.25402, 26031.26269, 26975.27230, 27691.28119 inBAC F2401, Arabidopsis thaliana Genome Center,University of Pennsylvania, USA) were identified in theSTM genes isolated here. The intron positions were fully con-served between the study species and A. thaliana STM (Fig. 3)and S. rexii STM1 (Harrison et al. 2005).

Cotyledonary basal meristem observed in both isocotylousand anisocotylous seedlings in the early post-germinationstage

Both cotyledons in a seedling developed identical in sizeover the observed time period after cotyledon unfolding inthe isocotylous species A. majus, Digitalis purpurea (bothPlantag inaceae) , J . punctata (Calceolar iaceae) ,C. speciosus, Seemannia purpurascens, S. speciosa, andGesneria cuneifolia (all New World Gesneriaceae)(Table 1, Fig. 1 column 3, Fig. 5 column 1). On the otherhand, in the Old World Gesneriaceae species examined,H . a n a c h o re t a , Mi c ro c h i r i t a l a v a n d u l a c e a ,S. glandulosissimus, S. rexii, and S. wendlandii allshowed seedlings with an unequal cotyledon developmentfrom 10D onwards (Table 1, Fig. 5 column 1), and theformation of eglandular hairs in the proximal region of themacrocotyledons, a sign of lamina extension (see Nishiiet al. 2004), was also observed (Online Resource 8d–g).Initially, the number of lateral veins in cotyledons in-creased during their growth in both isocotylous andan i soco ty lous spec i e s (Tab le 2 ) . Howeve r, i nanisocotylous species, the number of lateral veins wasstatistically significantly higher in the larger cotyledon at10D and later, with the only exception of H. anachoreta,although it showed strong anisocotyly early on at 10D(Tables 1 and 2).

Cell division detected by Aniline Blue followed thepattern of cotyledon area expansion. Both, isocotylousand anisocotylous species showed cell division activityat 1D, and at 5D for A. majus, mostly observed in theproximal region of both cotyledons (Fig. 5 column 2,Online Resource 9). However, at 10D, cell division activ-ity in most isocotylous species had ceased, whereas thelarger cotyledon of the anisocotylous species continued toshow cell division activity at 10D and 20D with the centreof activity near the base of the macrocotyledon (Fig. 5column 3, Table 3, Online Resource 9).

Dev Genes Evol (2017) 227:41–60 47

STM expression was detected in the cotyledon in linewith the cell division activity. In all isocotylous speciesSTM expression was observed in 1D, but not in olderisocotylous seedlings (Fig. 5 column 4, Online Resource9). In the anisocotylous seedlings of Old WorldGesneriaceae species examined in this study, STM wasexpressed in both cotyledons at 1D, but later detected onlyin the macrocotyledon (Fig. 5 column 4). The RT-qPCRresults further confirmed the early STM expression in cot-yledons of the isocotylous New World Gesneriaceae

C. speciosus just after germination, but when cell divisionhad ceased at 20D, STM was found down-regulated(Fig. 6a). RNA in situ hybridization showed that in thisspecies, STM was localized in the proximal region of bothcotyledons at 1D (Fig. 7a), as well as in the SAM (Fig. 7b),where cell division was observed and the number of celllayers was increased (Online Resource 10). In A. majus,the expression of both INA and HIRZ was detected in cot-yledons just after germination at 1D by RT-qPCR, but wasdown-regulated in the cotyledons at 20D (Fig. 6b).

Fig. 3 Conservation and homology of isolated STM genes. Alignment ofdeduced amino acid sequences of STM genes. Extent of the conserveddomains, KNOX1, KNOX2, ELK, and HOMEODOMAIN indicated by

bold lines. Positions ofArabidopsis STM andGesneriaceae STM1 intronsindicated by grey arrowheads

48 Dev Genes Evol (2017) 227:41–60

Leaf basal meristem is common to Calceolariaceaeand Gesneriaceae, but extended in Gesneriaceae

The growth patterns of foliage leaves, the phyllomorph ofS. rexii and the cotyledonary phyllomorph (macrocotyledon)of the unifoliate S. wendlandii as equivalent, were examined.In A. majus, as reported, cell division ceases after 10% of thefinal leaf length has been attained, and STM gene expressionwas never observed (see Nath et al. 2003). The arrest front ofcell division for J. punctata leaves showed a more rapid ter-mination of division at 20–30% leaf length compared to theGesner iaceae C. speciosus , H. anachore ta , andS. glandulosissimus with 50–60% leaf length. Streptocarpusrexii and S. wendlandii showed the longest cell division activ-ity, up to 90–100% leaf length (Fig. 8 column 1).

The cell expansion rates were more desynchronized inleaves of Old World Gesneriaceae compared to those ofNew World Gesneriaceae and the outgroup species: the size

of epidermal cells in the proximal position (base) of the leafdifferentiated from those of the middle and tip regions at about50% leaf length in J. punctata, 10% in C. speciosus, and lessthan 5% in Old World species of H. anachoreta andS. glandulosissimus (Online Resource 7o–r).

The number of lateral veins in the outgroup A. majusreached its maximum before 10% leaf length was attained(Fig. 8 column 2). On the other hand, in J. punctata, thenumber of lateral veins gradually increased up to 40–50% leaflength. In New World Gesneriaceae, they increased duringleaf growth slightly longer than J. punctata, about 50–60%final leaf length. The Old World Gesneriaceae showed a sig-nificantly extended period (>60% leaf length) of lateral veinincrease. Additional samples of NewWorld (Besleria labiosa,Nautilocalyx lynchii, S. schiffneri, S. purpurascens, Gesneriapedicellaris × pedunculosa) and Old World Gesneriaceae(L. griffithii, Paraboea swinhoei, Oreocharis xiangguiensis)confirmed these patterns (Online Resource 7g–n). In the

Fig. 4 Bayesian inference tree ofKNOX1 genes. Posteriorprobabilities are given next to thebranches. STM genes fromGesneriaceae, Calceolariaceae,and the STM-KNOX1 genesAtSTM (Arabidopsis thaliana),AmINA, AmHIRZ (Antirrhinummajus), NTH15 (Nicotianatabacum), and Let6 (Solanumlycopersicum) form a singlehighly supported clade (posteriorprobability = 1.0), indicated by avertical box

Dev Genes Evol (2017) 227:41–60 49

phyllomorphic S. rexii and S. wendlandii, the lateral vein in-crease was observed during the entire period of leaf growth(Fig. 8 column 2).

STM expression during leaf growth was not observed inA. majus (Golz et al. 2002). Surprisingly, STM expressionwas detected in J. punctata and all Gesneriaceae species

Table 1 Seedling development of the study species: development of cotyledon area (mm2) over time, 1 day (1D), 10 days (10D), and 20 days (20D)after cotyledon unfolding

1 D 10D 20D

Larger co. Smaller co. Larger co. Smaller co. Larger co. Smaller co.

Antirrhinum majus0.88 ± 0.06

(4)

0.82 ± 0.04

(4)

11.18 ± 0.72

(5)

10.03 ± 0.56

(5)

12.28 ± 0.72

(5)

11.62 ± 0.59

(5)

Plan – isocotyl P = 0.42 P = 0.24 P = 0.50

Digitalis purpurea0.84 ± 0.07

(5)

0.82 ± 0.07

(5)

2.68 ± 0.08

(5)

2.48 ± 0.10

(5)

2.99 ± 0.13

(5)

2.71 ± 0.11

(5)

Plan – isocotyl P = 0.88 P = 0.17 P = 0.14

Jovellana punctata0.29 ± 0.03

(5)

0.28 ± 0.02

(5)

1.17 ± 0.23

(3)

1.12 ± 0.21

(3)

1.35 ± 0.08

(3)

1.23 ± 0.06

(3)

Cal – isocotyl P = 0.67 P = 0.90 P = 0.33

Corytoplectus speciosus0.22 ± 0.01

(5)

0.22 ± 0.01

(5)

0.75 ± 0.05

(5)

0.72 ± 0.04

(5)

1.08 ± 0.03

(5)

1.01 ± 0.05

(5)

NW – isocotyl P = 0.64 P = 0.63 P = 0.27

Seemannia purpurascens0.23 ± 0.04

(5)

0.21 ± 0.04

(5)

1.83 ± 0.10

(5)

1.75 ± 0.11

(5)

1.86 ± 0.10

(5)

1.74 ± 0.10

(5)

NW – isocotyl P = 0.72 P = 0.63 P = 0.66

Sinningia speciosa0.41 ± 0.06

(5)

0.40 ± 0.07

(5)

3.52 ± 0.35

(5)

3.32 ± 0.31

(5)

4.22 ± 0.37

(5)

4.05 ± 0.35

(5)

NW – isocotyl P = 0.89 P = 0.67 P = 0.75

Gesneria cuneifolia0.21 ± 0.02

(5)

0.20 ± 0.02

(5)

1.32 ± 0.04

(5)

1.26 ± 0.03

(5)

1.83 ± 0.05

(5)

1.74 ± 0.08

(5)

NW – isocotyl P = 0.78 P = 0.29 P = 0.38

Henckelia anachoreta0.16 ± 0.03

(4)

0.15 ± 0.03

(4)

0.61 ± 0.08

(9)

0.35 ± 0.03

(9)

2.54 ± 0.1

(5)

0.78 ± 0.02

(5)

OW – anisocotyl P = 0.86 P = 0.007** P = 1.64E-5***

Microchirita lavandulacea0.27 ± 0.05

(5)

0.24 ± 0.06

(5)

5.56 ± 0.06

(5)

2.56 ± 0.11

(5)

20.85 ± 0.75

(5)

3.39 ± 0.33

(5)

OW – anisocotyl P = 0.74 P = 7.55E-9*** P = 2.44E-8***

Streptocarpus glandulosissimus0.18 ± 0.02

(5)

0.17 ± 0.02

(5)

0.52 ± 0.03

(5)

0.35 ± 0.02

(5)

0.62 ± 0.03

(5)

0.38 ± 0.01

(5)

OW – anisocotyl P = 0.65 P < 0.01 (0.0018)** P < 0.01 (0.0004)**

Streptocarpus rexii0.12 ± 0.01

(5)

0.12 ± 0.01

(5)

0.40 ± 0.03

(5)

0.29 ± 0.01

(5)

0.51 ± 0.04

(5)

0.33 ± 0.01

(5)

OW – anisocotyl P = 0.74 P = 0.006** P = 0.002**

Streptocarpus wendlandii0.08 ± 0.01

(5)

0.08 ± 0.01

(5)

0.50 ± 0.05

(6)

0.38 ± 0.02

(6)

0.84 ± 0.06

(5)

0.41 ± 0.03

(5)

OW – anisocotyl P = 0.75 P = 0.04* P = 0.0003***Means ± standard errors of N replicates (in brackets). P values calculated by one-way ANOVA between the area of the larger cotyledon and the smallercotyledon. Significant comparisons shaded in grey

Plan Plantaginaceae, Cal Calceolariaceae, NW New World Gesneriaceae, OW Old World Gesneriaceae

*Significant at P = 0.05; **significant at P = 0.01; ***significant at P = 0.001

50 Dev Genes Evol (2017) 227:41–60

examined in this study, although to different extents (Fig. 8column 3, Online Resource 11). In J. punctata, STM wasexpressed in leaves up to 10% final size. In situ hybridizationof STM also showed a signal in leaf primordia of J. punctata(Online Resource 12b). The NewWorld Gesneriaceae speciesC. speciosus (Fig. 8) and S. schiffneri (Online Resource 11a)and the OldWorld Gesneriaceae, S. glandulosissimus (Fig. 8),and L. griffithii (Online Resource 11b) showed STM expres-sion up to 20–30% leaf size. Phyllomorphic Streptocarpusspecies showed the longest periods of STM expression in theirleaves, with S. rexii up to 50–60% and S. wendlandii through-out its vegetative growth phase. When leaves were dividedinto proximal and distal halves, STM expression was detectedonly in the proximal region of the lamina but not in the distalregion in all species analysed (Fig. 8 column 4). In the NewWorld C. speciosus, in situ hybridization of STM also showedthat the gene was localized in the proximal region of the lam-ina (Fig. 7c), similar to the leaf of the Old WorldS. glandulosissimus (Online Resource 13b). In both cases,STM expression was restricted to the basal region of the lam-ina where the basal meristem is located, seen as protrusions onthe midrib in transverse sections (Fig. 7c, Online Resource13b).

Discussion

We studied the vegetative development of strategically sam-pled members of the morphologically diverse familyGesneriaceae and outgroup samples, to illustrate changes inthe meristem activity leading to one-leaf plants in an evolu-tionary setting, by following cell division and meristemKNOX1-STM gene expression patterns. It should be noted thateven though STM mutants lack a SAM and may superficiallyresemble unifoliates (Barton and Poethig 1993), there is evi-dence from genetic studies of unifoliate vs. rosulate speciesdemonstrating that STM is not the causative gene forunifoliateness in Streptocarpus (Harrison et al. 2005).Instead, in the present study, we used STM as locator for mer-istems (Long et al. 1996; Mantegazza et al. 2007, 2009). Ourcomparative studies revealed striking differences in the loca-tion and extent of meristematic activities that allowed to pin-point key evolutionary events in the rise of cotyledonary one-leaf plants.

Comparisons to model plants

While our outgroup sampling was dictated by the availabil-ity of thoroughly studied systems, trends could beascertained of the characteristics of non-gesneriads with aview to cotyledon and leaf development. Cell division incotyledons of A. thaliana (rosids; Brassicales) is only ob-served in marginal regions of the palisade mesophyll, but

throughout the cotyledons in tobacco and petunia, bothSolanales (asterids), an order closely related to Lamiales towhich family Gesneriaceae belongs, though in both casesdivision ceases before the cotyledons unfold (Fridlenderet al. 1996; Stoynova-Bakalova et al. 2004). Within theLamiales, we observed a further trend of intensificationand localization of cell division towards the base of bothcotyledons in line with their relatedness, from Antirrhinumover Jovellana to the New World then Old WorldGesneriaceae (Fig. 5, Online Resource 9). A strong temporaland spatial disjunction occurred at the evolutionary split be-tween the New and Old World Gesneriaceae, whereby theOld World Gesneriaceae acquired anisocotyly, based on theextended basal meristem activity of the macrocotyledon(Fig. 5). These lineage-specific patterns were confirmed ina larger sampling of the Plantaginaceae and Gesneriaceae(Tables 1, 2, and 3) and seem to be conserved within eachgroup. Additionally, whereas in A. thaliana (Stoynova-Bakalova et al. 2004) and tobacco (Avery 1933) cotyledonsdevelop through anticlinal cell divisions, Gesneriaceae suchas Streptocarpus (Imaichi et al. 2000), Monophyllaea(Tsukaya 1997; Ayano et al. 2005), and C. speciosus(Online Resource 10) display both anticlinal and periclinaldivisions, adding depth to the cotyledonary tissue, a pre-requisite for the stability of larger organs.

In foliage leaves, cell division in A. thaliana occurs onlyvery early on and across the emerging primordia but sharplydeclines towards the base soon afterwards (Donnelly et al.1999). The Lamiales samples here followed a similar pattern,but at greatly different rates between the species analysed(Fig. 8, Online Resource 7), basically following the trend ofcell division of the cotyledons, intensifying from an earlyceasing activity in A. majus (Nath et al. 2003), overJ. punctata with a slightly extended period and theGesneriaceae with a greatly extended period of cell division,in parallel with a more rapid shift of the cell cycle arrest fronttowards the lamina base (Fig. 8, Online Resource 7). Theunifoliate S. wendlandii showed the swiftest focus of cell di-visions and the most extended activity across its entire leafdevelopment. The pattern of lateral vein formation, an indirectmeasure of lamina extension (Nishii et al. 2010), supportedthe observed extended cell division activity in the leaves of theexamined Gesneriaceae species and J. punctata (Fig. 8,Online Resource 7).

The rate of cell differentiation between the proximal, mid-dle, and distal leaf regions, illustrated by the distribution ofcell sizes over time, showed a gradual increase from the prox-imal towards the distal region in A. majus (Nath et al. 2003),J. punctata, and the isocotylous New World Gesneriaceae.However, a distinct size shift across the lamina was observedfor anisocotylous Old World Gesneriaceae, indicating that thedifferentiation between the basal and remaining lamina tissuewas rapid (Online Resource 7o–r).

Dev Genes Evol (2017) 227:41–60 51

52 Dev Genes Evol (2017) 227:41–60

STM is involved in lateral organ growth in Gesneriaceae

In A. thaliana and maize, KNOX1 genes mark the SAM butare excluded from the cotyledons during embryogenesis andafter germination (Smith et al. 1995; Long et al. 1996). It wassurprising to find STM expression in cotyledons early on (1D)for all Lamiales samples analysed here (Fig. 5 column 4,Online Resource 9f). The STM expression closely followedchanges in cell division patterns in the cotyledons over time,being down-regulated in cotyledons in the isocotylousGesneriaceae C. speciosus and S. speciosa, in J. punctata,

Table 2 Seedling development of the study species: development of lateral vein bifurcation per cotyledon

1 D 10D 20D

Larger co. Smaller co. Larger co. Smaller co. Larger co. Smaller co.

Antirrhinum majus 4.6 ± 0.3 (4) 4.2 ± 0.2 (4) 6.8 ± 0.6 (5) 5.8 ± 0.3 (5) 6.6 ± 0.3 (5) 6.4 ± 0.4 (5)

Plan – isocotyl P = 0.54 P = 0.31 P = 0.68

Digitalis purpurea 5.0 ± 1.2 (3) 5.0 ± 0.8 (3) 7.3 ± 0.5 (3) 5.7 ± 0.3 (3) 8.0 ± 0.8 (3) 6.7 ± 0.7 (3)

Plan – isocotyl P = 1 P = 0.09 P = 0.37

Jovellana punctata 0.0 ± 0.0 (7) 0.0 ± 0.0 (7) 1.3 ± 0.4 (3) 1.3 ± 0.4 (3) 2.0 ± 0.0 (5) 2.0 ± 0.0 (5)

Cal – isocotyl a P = 0.72 a

Corytoplectus speciosus 0.0 ± 0.0 (7) 0.0 ± 0.0 (7) 0.0 ± 0.0 (5) 0.0 ± 0.0 (5) 0.6 ± 0.4 (5) 0.4 ± 0.2 (5)

NW – isocotyl a a P = 0.66

Seemannia purpurascens 0.4 ± 0.2 (3) 0.2 ± 0.2 (3) 1.8 ± 0.2 (5) 1.6 ± 0.2 (5) 2.0 ± 0.0 (5) 1.8 ± 0.3 (5)

NW – isocotyl P = 0.37 P = 0.54 P = 0.61

Sinningia speciosa 2.0 ± 0.0 (5) 2.0 ± 0.0 (5) 2.2 ± 0.2 (6) 2.2 ± 0.2 (6) n.e. n.e.

NW – isocotyl a P = 1

Gesneria cuneifolia 0.0 ± 0.0 (6) 0.0 ± 0.0 (6) 0.6 ± 0.2 (5) 0.4 ± 0.2 (5) 2.0 ± 0.3 (5) 1.4 ± 0.3 (5)

NW – isocotyl a P = 0.58 P = 0.17

Henckelia anachoreta 0.0 ± 0.0 (4) 0.0 ± 0.0 (4) 0.0 ± 0.0 (4) 0.0 ± 0.0 (4) 0.7 ± 0.5 (3) 0.0 ± 0.0 (3)

OW – anisocotyl a a P = 0.37

Microchirita lavandulacea 0.0 ± 0.0 (5) 0.0 ± 0.0 (5) 7.2 ± 0.2 (5) 2.6 ± 0.2 (5) 10.8 ± 0.6 (5) 2.2 ± 0.6 (5)

OW – anisocotyl a P = 4.89E-7*** P = 6.2E-6***

Streptocarpus glandulosissimus 1.0 ± 0.4 (5) 1.0 ± 0.4 (5) 2.2 ± 0.2 (5) 1.4 ± 0.4 (5) 3.6 ± 0.2 (5) 2.2 ± 0.2 (5)

OW – anisocotyl P = 1 P = 0.11 P = 0.002*

Streptocarpus rexii 0.0 ± 0.0 (3) 0.0 ± 0.0 (3) 1.3 ± 0.3 (3) 0.3 ± 0.3 (3) 1.8 ± 0.2 (4) 0.3 ± 0.2 (4)

OW – anisocotyl a P = 0.10 P = 0.005*

Streptocarpus wendlandii 0.0 ± 0.0 (5) 0.0 ± 0.0 (5) 0.4 ± 0.2 (7) 0.0 ± 0.0 (7) 1.6 ± 0.5 (4) 0.0 ± 0.0 (4)

OW – anisocotyl a P = 0.14 P = 0.013*Means ± standard errors of N replicates (in brackets). P values calculated by one-way ANOVA between the area of the larger cotyledon and the smallercotyledon. Significant comparisons shaded in grey

n.e. not evaluated, Plan Plantaginaceae, Cal Calceolariaceae, NW New World Gesneriaceae, OW Old World Gesneriaceae

*Significant at P = 0.05; **significant at P = 0.01; ***significant at P = 0.001aNo variance

Fig. 5 Meristem behaviour associated with STM expression contributesto cotyledon development. a–g As in Fig. 1. Column 1 Cotyledon areawithin seedlings equal at seedling stage 1 day after unfolding (1D). OnlyOld World Gesneriaceae (d–g) show unequal cotyledon growth at 10–30D. The diagonal line indicates values for a 1:1 ratio of cotyledon size.Columns 2–3 meristem activity detected by Aniline Blue staining in theproximal region of both cotyledons 1D or 5D in all species (column 2)and only in the macrocotyledon 20D for anisocotylous Gesneriaceae(column 3). Column 4 RT-PCR expression patterns of STM. STM expres-sion observed in cotyledons of all species 1D, but not in the 20D cotyle-dons of isocotylous species (arrowheads), but observed in themacrocotyledon at 20D. Arrowheads indicate samples without STM ex-pression. co cotyledon, w whole seedling, Mc macrocotyledon

Dev Genes Evol (2017) 227:41–60 53

Table 3 Seedling development of the study species: cell division activity in developing cotyledons

1 D 10D 20D

Larger co. Smaller co. Larger co. Smaller co. Larger co. Smaller co.

Antirrhinum majusPlan – isocotyl

100%

(4/4)

100%

(4/4)

20%

(1/5)

20%

(1/5)

0%

(0/5)

0%

(0/5)

Digitalis purpureaPlan – isocotyl

100%

(3/3)

100%

(3/3)

0%

(0/3)

0%

(0/3)

0%

(0/3)

0%

(0/3)

Jovellana punctateCal – isocotyl

100%

(7/7)

100%

(7/7)

0%

(0/3)

0%

(0/3)

0%

(0/5)

0%

(0/5)

Corytoplectus speciosusNW – isocotyl

100%

(7/7)

100%

(7/7)

0%

(0/5)

0%

(0/5)

0%

(0/5)

0%

(0/5)

Seemannia purpurascensNW – isocotyl

100%

(3/3)

100%

(3/3)

0%

(0/5)

0%

(0/5)

0%

(0/5)

0%

(0/5)

Sinningia speciosaNW – isocotyl

100%

(5/5)

100%

(5/5)

0%

(0/6)

0%

(0/6)

n.e. n.e.

Gesneria cuneifoliaNW – isocotyl

100%

(6/6)

67%

(4/6 )

0%

(0/5)

0%

(0/5)

0%

(0/5)

0%

(0/5)

Henckelia anachoretaOW – anisocotyl

100%

(4/4)

100%

(4/4)

100%

(4/4)

25%

(1/4)

100%

(3/3)

0%

(0/3)

Microchirita lavandulaceaOW – anisocotyl

100%

(5/5)

60%

(3/5)

100%

(5/5)

0%

(0/5)

100%

(5/5)

0%

(5/5)

Streptocarpus glandulosissimusOW – anisocotyl

100%

(5/5)

100%

(5/5)

100%

(5/5)

0%

(5/5)

100%

(5/5)

0%

(5/5)

Streptocarpus rexiiOW – anisocotyl

100%

(3/3)

100%

(3/3)

100%

(3/3)

0%

(0/3)

100%

(4/4)

0%

(0/4)

Streptocarpus wendlandiiOW – anisocotyl

100%

(5/5)

100%

(5/5)

100%

(7/7)

29%

(2/7)

100%

(4/4)

0%

(0/4)

Assessed by Aniline Blue staining of cell walls in epidermal and subepidermal cells on the adaxial side of cotyledons. Numbers are percentages ofcotyledons with divisions. The actual number of seedlings with divisions and the total number observed are given in brackets. Shaded cells indicate thepresence of cell divisions

n.e. not evaluated, Plan Plantaginaceae, Cal Calceolariaceae, NW New World Gesneriaceae, OW Old World Gesneriaceae

Fig. 6 Isocotylous species showKNOX1 expression in cotyledonsat an early growth stage. a, b STMRT-qPCR results. a Corytoplectusspeciosus. b Antirrhinum majus.CsSTM1, AmINA, and AmHIRZwere expressed in cotyledon at theseedling stage 1 day afterunfolding (1D), but were sup-pressed in older seedlings (20D).Relative expression levelscalculated against wholeseedlings (±SE bars, n = 3)

54 Dev Genes Evol (2017) 227:41–60

and in A. majus, while being continuously expressed in themacrocotyledon of anisocotylous Gesneriaceae (Fig. 5column 4) (Mantegazza et al. 2007, 2009). It was interestingto find that STM was localized in the proximal region of bothcotyledons in the isocotylous C. speciosus in the early stagesof germination (Fig. 7a), a pattern similar to the anisocotylousS. rexii just after germination (Mantegazza et al. 2007). Thismay indicate that the early cotyledonary expression of STMmay be shared among the Lamiales.

In foliage leaves, absence of KNOX1 expression is wide-spread in plants and was first reported for maize, A. thaliana,and A. majus (Smith et al. 1995; Long et al. 1996; Golz et al.2002). Only occasionally KNOX1 is expressed outside theSAM, in plants with compound leaves or complex leafprimordia (Hareven et al. 1996; Bharathan et al. 2002; Hayand Tsiantis 2006). Streptocarpus, however, have simpleleaves and simple leaf primordia (Mantegazza et al. 2007;Nishii et al. 2004; Imaichi et al. 2007) but still expressKNOX1 there (Harrison et al. 2005; Nishii et al. 2010;Mantegazza et al. 2007). Intriguingly, we found STM expres-sion also in the foliage leaf of J. punctata and all Gesneriaceaesamples studied including those of the New World (Fig. 8,Online Resource 11). All these plants have simple leavesand simple leaf primordia (Fig. 1, Online Resource 3a–c).

Thus, in Lamiales, at least the Gesneriaceae plus their closestfamily the Calceolariaceae do not follow the correlation be-tween compound leaf and leaf KNOX1 expression.

It appears that the leaf meristematic activity becomes morefocused and extends in duration in Gesneriaceae, since the STMexpression was strongly focussed in the proximal region of thelamina of the developing leaf in Gesneriaceae. This basal meri-stem activity appears to expand ‘step-wise’ in Gesneriaceae andis seemingly ‘indeterminate’ in unifoliate Streptocarpus.Intriguingly, rosulate Streptocarpus showed a shorter durationof STM expression (compare Fig. 8 column 3 f vs. g), suggestingthat the size of phyllomorphs is determined during their growth,perhaps interacting with previously or newly formed ones.

The basal meristem

The Oxford Dictionary of Plant Sciences (Allaby 2012) de-fines a meristem as ‘a group of plant cells that are capable ofdividing indefinitely and whose main function is the produc-tion of new growth’. In this sense, the areas at the base of theleaves of anisocotylous Gesneriaceae that show a high densityof dividing cells and KNOX1 expression could be seen as anintercalary meristem (Esau 1953), since it gives rise to similarcells or to cells that differentiate to produce diverse specialized

Fig. 7 STM in situ expression patterns in the isocotylous Corytoplectusspeciosus. a Longitudinal section of cotyledons just after germination atthe seedling stage 1 day after unfolding (1D). b Shoot apical meristembetween cotyledons, 10D. c–d Transverse section through the proximal

region of a foliage leaf. a–c Sections were hybridized with antisenseprobe (a–c) or sense negative control (d). Arrows indicate the purple-coloured in situ signals. Scale bars = 100 μm

Dev Genes Evol (2017) 227:41–60 55

Fig. 8 Foliage leaf development associated with meristem behaviour andSTM expression. a–g As in Fig. 1. Column 1 progression of cell divisionarrest front in segments of leaves detected by Aniline Blue staining.Column 2 number of lateral veins plotted against leaf length peaks earlyin Antirrhinum majus and increased in Jovellana punctata andGesneriaceae species until later stages. The line indicates the trend ofincrease of lateral veins. Column 3 RT-PCR revealed STM expression in

leaf primordia in J. punctata and developing leaves in Gesneriaceae spe-cies, as well as an extended presence in the acaulescent Streptocarpusrexii and S. wendlandii. Column 4 RT-PCR shows STM expression in theproximal region of the lamina (pL) but not in the distal region (dL).Arrowheads indicate samples without STM expression. w whole shoots,pL proximal region of lamina, dL distal region of lamina, *Nath et al.2003, **Golz et al. 2002

56 Dev Genes Evol (2017) 227:41–60

tissues (i.e. epidermis, parenchyma, and vasculature includingmain lateral veins), and has indeed been first termed ‘basalmeristem’ by Jong (1970).

This basal meristem could be seen as an evolution from therelatively scattered and transiently active meristemoids inA. thaliana (Donnelly et al. 1999) withoutKNOX1 expression,through intermediate steps where the correlation between celldivision andKNOX1 is initially relatively disjunct in an organ,as in the leaf of J. punctata, with a wider and more diffuse celldivision pattern than STM expression, to Streptocarpus wherethe area of meristematic activity becomes highly defined andthe STM expression is more in line with this area of cell divi-sion, perhaps to keep cells there in an indeterminate state toform a basal region of pluripotent cells (Figs. 5 and 8;Mantegazza et al. 2007). This is supported by previous workin Streptocarpus: STM and WUSCHEL, both regulating themeristem identity and expressed in the SAM in A. thaliana(Laux et al. 1996; Long et al. 1996), are also expressed in thebasal meristem of S. rexii (Mantegazza et al. 2007, 2009).Further, YABBY, expressed in lateral organs and correlatedwith the abaxial cell fate of the lamina, is excluded from theSAM of A. thaliana and is also excluded from the basal mer-istem in S. rexii (Siegfried et al. 1999; Tononi et al. 2010).

It is also known that meristems are fine-regulated by planthormones. For example, cytokinins (CK) and gibberellins(GA) regulate functioning of the SAM through KNOX1 genesin A. thaliana. Here, high CK and low GA levels maintainindeterminacy of the SAM, whereas lateral organs show high

GA that prevent meristem activity there (Hay et al. 2002;Jasinski et al. 2005; Yanai et al. 2005). Intriguingly, the basalmeristem in Streptocarpus is also affected by GA and CK.Here, GA application suppresses basal meristem activity andthus the formation of a macrocotyledon, while CK applicationinduces basal meristem activity in both cotyledons to formtwo macrocotyledons (Rosenblum and Basile 1984; Nishiiet al. 2004, 2012; Mantegazza et al. 2007, 2009). The genesinvolved in GA synthesis and degradation have recently beenisolated and studied in S. rexii, and the GA synthesis gene,SrGA20ox, found expressed in the lamina of the cotyledons,and the GA degradation gene, SrGA2ox, in the basal meri-stem. This would result in low levels of GA in the basal mer-istem and high levels in the remaining lamina, similar to theSAM and leaves in A. thaliana, respectively (Nishii et al.2014).

All these features support the view that the basal meristemis a highly evolved and organized structure that can beregarded as a meristem with a region of undifferentiated cellsrather than a lamina with transient unorganized cell division.

Consequences of the step-wise evolution of the basalmeristem

We report here a case of progressive temporal extensions andspatial shifts of cell division activity from the SAM to cotyle-dons and leaves, which can be seen as developmental steppingstones in the evolution of the one-leaf plant. In this instance, it

Fig. 9 Hypothesised model of step-wise meristem evolution towards theone-leaf plant in Gesneriaceae. Meristem activities associated withKNOX1 expression. a In SAM in Arabidopsis. b Additionally in cotyle-dons. c Extended into foliage leaf primordia, but disjunct from later lam-ina expansion. d Expanded into foliage leaf, associated with lamina ex-pansion, facultative anisophylly (indicated as stippled outline). e

Evolution of anisocotyly. f Loss of shoot and SAM. g, h Shift of SAMfunction to leaf. g Rosulate with groove meristem forming additionalleaves. h Unifoliate with indefinite KNOX1 expression in themacrocotyledon. Black meristem associated with KNOX1, light grey cot-yledon, grey foliage leaf and stem

Dev Genes Evol (2017) 227:41–60 57

is not solely the result of a single macromutation as ponderedon previously (Cronk and Möller 1997), but originates froman intricate series of small changes in meristem behaviour(Fig. 9). It involves an expansion of STM expression fromthe SAM to cotyledons in the Lamiales lineage (Fig. 9b).STM further extended its expression domain outside theSAM in leaf pr imordia in the Gesneriaceae andCalceolariaceae (Fig. 9c). Both events are without immediate-ly apparent developmental consequences, but represent essen-tial precursory steps for the consecutive evolutionary innova-tions. These include facultative anisophylly, the unequal sizeof leaves in a pair, in some Gesneriaceae (Fig. 9d, OnlineResource 3d, e), perhaps to avoid self-shading, e.g. inColumnea among New World genera (Online Resource 3d),Cyrtandra and Loxostigma among Old World genera (OnlineResource 3e), and anisocotyly only in the Old WorldGesneriaceae (Fig. 9e), perhaps a prerequisite to expand pho-tosynthetic area rapidly after germination that allows theplants to grow in the dark forest they often inhabit (Burtt1970). The presence of species showing anisophylly acrossall Gesneriaceae and the absence of anisocotyly in NewWorld Gesneriaceae suggest that the genetic cascade and con-trols required for asymmetric lateral organ growth seeminglyevolved first in foliage leaves (Fig. 9d). Without this pre-acquired lateral dominance, one-leaf plants would requirethe acquisition of several interconnected traits simultaneouslyto be able to survive (Fig. 9e–h). Whether rosulates, basicallySAM-lacking acaulescents that form additional leaves from agroove meristem (Jong 1970; Jong and Burtt 1975), arose firstfrom caulescents or unifoliates is still unresolved (Möller andCronk 2001; Nishii et al. 2015).

The importance of a series of micromutational steps alteringthe spatial-temporal pattern of cell division and KNOX1 expres-sion has been indicated here in the evolution of the one-leafplant. There are examples of gradual shifts of homeobox (Hox)gene expression in animals and their effects on body plan evo-lution (e.g. paired fins in vertebrate; Tanaka and Onimaru 2012;snake body plan; Guerreiro et al. 2013), where flexible cis-ele-ments play important roles. As such, each mutation has under-gone natural selection processes and has proven its adaptivemerits and thus added to the chances of the success of a novelevolutionary trajectory (e.g. the one-leaf plants here) that at firstseems unlikely to have any great evolutionary significance.However, there exist >50 species with unifoliate habit, just underone third of all species in the genus Streptocarpus (Nishii et al.2015), and this habit has apparently evolved several times inde-pendently in Africa and Madagascar (Möller and Cronk 2001;Nishii et al. 2015), a testimony of their evolutionary potential.Perhaps the combination of anisocotyly, allowing photosynthet-ic tissues to be added swiftly, and unifoliateness, allowing axil-lary inflorescences to be formed without delay, enables theplants to survive in unfavourable conditions where resourcesare scarce. Strong support comes from other species such as

Streptocarpus nobilis or other Gesneriaceae such asMicrochirita hamosa, Microchirita micromusa, andRhynchoglossum obliquum, caulescent species that form unifo-liates under adverse conditions (Lawrence 1943; Burtt 1970).Detailed studies of these intermediates would shed more lighton the alternative developmental and genetic pathways that al-lows plants to adapt with extreme reductions in form to adverseconditions.

Acknowledgements We thank M. Gibby for the support and helpfulcomments on this work, as well as P. Hollingsworth, T. Pennington, andC. Kidner for helpful comments. We thank J. Preston at the University ofVermont (USA) for critical comments on an earlier version of the manu-script. The work was supported by the Royal Botanic Garden Edinburgh(RBGE, UK) and the Sibbald Trust (project 2012#9) at the RBGE. KNwas supported by the Top100-University scheme of the National TaiwanUniversity (NTU, Taiwan, Grant Number 10R40044) and the JapanSociety of Promotion of Science (JSPS KAKENHI Grant Number15K18593). We thank S.-T. Jeng and T.-P. Lin for supporting KN’s stayat NTU; A. Iwamoto and H. Iida for supporting KN’s stay at TokyoGakugei University; K.-J. Tang and Y.-Y. Gao (Techcomm, NTU) fortechnical support at NTU; and M. Hart, F. Christie, R. Holland, and L.Forrest for technical support at RBGE. We thank S. Barber, S. Scott, andC. Morter for the growing research materials. RBGE is supported by theRural and Environment Science and Analytical Services Division(RESAS) in the Scottish Government.

References

Akaike H (1974) A new look at the statistical model identification. IEEETrans Auto Con 19:716–723. doi:10.1109/TAC.1974.1100705

Allaby M (2012) Oxford dictionary of plant sciences. Oxford UniversityPress, Oxford

Avery GS (1933) Structure and germination of tobacco seed and the de-velopmental anatomy of the seedling plant. Amer J Bot 20:309–327

Ayano M, Imaichi R, Kato M (2005) Developmental morphology of theAsian one-leaf plant,Monophyllaea glabra (Gesneriaceae) with em-phasis on inflorescence morphology. J Pl Res 118:99–109.doi:10.1007/s10265-005-0195-5

BartonMK, Poethig RS (1993) Formation of the shoot apical meristem inArabidopsis thaliana: an analysis of development in the wild typeand in the shoot meristemless mutant. Development 119:823–831.doi:10.1016/0168-9525(94)90143-0

Bharathan G, Goliber TE, Moore C, Kessler S, Pham T, Sinha N (2002)Homologies in leaf form inferred from KNOXI gene expressionduring development. Science 296:1858–1860. doi:10.1126/science.1070343

Burtt BL (1963) Studies in the Gesneriaceae of the Old World, XXIV:tentative keys to the tribes and genera. Notes Roy Bot GardEdinburgh 24:205–220

Burtt BL (1970) Studies in the Gesneriaceae of the Old World XXXI:some aspects of functional evolution. Notes Roy Bot GardEdinburgh 30:1–9

Cronk QCB, Möller M (1997) Strange morphogenesis—organ determi-nation in Monophyllaea. Trends Pl Sci 2:327–328. doi:10.1016/S1360-1385(97)84614-6

Donnelly PM, Bonetta D, Tsukaya H, Dengler RE, Dengler NG (1999)Cell cycling and cell enlargement in developing leaves ofArabidopsis. Dev Biol 215:407–419. doi:10.1006/dbio.1999.9443

Esau K (1953) Plant Anatomy. Wiley, New York

58 Dev Genes Evol (2017) 227:41–60

http://dx.doi.org/10.1109/TAC.1974.1100705http://dx.doi.org/10.1007/s10265-005-0195-5http://dx.doi.org/10.1016/0168-9525(94)90143-0http://dx.doi.org/10.1126/science.1070343http://dx.doi.org/10.1126/science.1070343http://dx.doi.org/10.1016/S1360-1385(97)84614-6http://dx.doi.org/10.1016/S1360-1385(97)84614-6http://dx.doi.org/10.1006/dbio.1999.9443

Fridlender M, Lev-Yadun S, Baburek I, Angelis K, Levy AA (1996) Celldivisions in cotyledons after germination: localization, time courseand utilization for a mutagenesis assay. Planta 199:307–313.doi:10.1007/BF00196573

Golz JF, Keck EJ, Hudson A (2002) Spontaneous mutations in KNOXgenes give rise to a novel floral structure in Antirrhinum. Curr Biol12:515–522. doi:10.1016/S0960-9822(02)00721-2

Guerreiro I, Nunes A, Woltering J, Casaca A, Nóvoa A, Vinagre T,Hunter M, Duboule D, Mallo M (2013) Role of a polymorphismin a Hox/Pax-responsive enhancer in the evolution of the vertebratespine. Proc Natl Acad Sci U S A 110:10682–10686. doi:10.1073/pnas.1300592110

Hall TA (1999) BioEdit: a user-friendly biological sequence alignmenteditor and analysis program for Windows 95/98/NT. Nucleic AcidsSymp Ser 41:95–98. doi:10.1021/bk-1999-0734.ch008

Hareven D, Gutfinger T, Parnis A, Eshed Y, Lifschitz E (1996) Themakingof a compound leaf: genetic manipulation of leaf architecture in to-mato. Cell 84:735–744. doi:10.1016/S0092-8674(00)81051-X

Harrison J, Möller M, Langdale J, Cronk QCB, Hudson A (2005) Therole of KNOX genes in the evolution of morphological novelty inStreptocarpus. Plant Cell 17:430–443. doi:10.1105/tpc.104.028936

Hay A, Kaur H, Phillips A, Hedden P, Hake S (2002) The gibberellinpathway mediates KNOTTED1-type homeobox function in plantswith different body plans. Curr Biol 12:1557–1565. doi:10.1016/S0960-9822(02)01125-9

Hay A, Tsiantis M (2006) The genetic basis for differences in leaf formbetween Arabidopsis thaliana and its wild relative Cardaminehirsuta. Nature Genet 38:942–947. doi:10.1038/ng1835

Hilliard OM, Burtt BL (1971) Streptocarpus. An African plant study.Natal University Press, Pietermaritzburg

Imaichi R, Nagumo S, Kato M (2000) Ontogenetic anatomy ofStreptocarpus grandis (Gesneriaceae) with implications for evolutionof monophylly. Ann Bot 86:37–46. doi:10.1006/anbo.2000.1155

Imaichi R, Omura-ShimadateM, AyanoM,KatoM (2007) Developmentalmorphology of the caulescent species Streptocarpus pallidiflorus(Gesneriaceae), with implications for evolution of monophylly. Int JPl Sci 168:251–260. doi:10.1086/510410

Jasinski S, Piazza P, Craft J, Hay A, Woolley L, Rieu I, Phillips A,Hedden P, Tsiantis M (2005) KNOX action in Arabidopsis is medi-ated by coordinate regulation of cytokinin and gibberellin activities.Curr Biol 6:1560–1565. doi:10.1016/j.cub.2005.07.023

Jong K (1970) Developmental aspects of vegetative morphology ofStreptocarpus. Dissertation, University of Edinburgh

Jong K, Burtt BL (1975) The evolution of morphological novelty exem-plified in the growth patterns of some Gesneriaceae. New Phytol 75:297–311. doi:10.1111/j.1469-8137.1975.tb01400.x

Kuwabara A, Nagata T (2006) Cellular basis of developmental plasticityobserved in heterophyllous leaf formation of Ludwigia arcuata(Onagraceae). Planta 224:761–770. doi:10.1007/s00425-006-0258-4

Lawrence WJC (1943) Photoperiodism in Streptocarpus. Gardeners’Chronicle 3:113–156

Laux T, Mayer KF, Berger J, Jürgens G (1996) The WUSCHEL gene isrequired for shoot and floral meristem integrity in Arabidopsis.Development 122:87–96

Long JA, Moan EI, Medford JI, Barton MK (1996) A member of theKNOTTED class of homeodomain proteins encoded by the STMgene of Arabidopsis. Nature 379:66–69. doi:10.1038/379066a0

Mantegazza R, Möller M, Harrison CJ, Fior S, De Luca C, Spada A(2007) Anisocotyly and meristem initiation in an unorthodox plant,Streptocarpus rexii (Gesneriaceae). Planta 225:653–663.doi:10.1007/s00425-006-0389-7

Mantegazza R, Tononi P, Möller M, Spada A (2009) WUS and STMhomologs are linked to the expression of lateral dominance in theacaulescent Streptocarpus rexii (Gesneriaceae). Planta 230:529–542. doi:10.1007/s00425-009-0965-8

Möller M, Clark JL (2013) The state of molecular phylogenetic work inthe family Gesneriaceae: a review. Selbyana 31:95–125

Möller M, Cronk QCB (2001) Evolution of morphological novelty: aphylogenetic analysis of growth patterns in Streptocarpus(Gesneriaceae). Evolution 55:918–929. doi:10.1111/j.0014-3820.2001.tb00609.x

Möller M, Forrest A, Wei Y-G, Weber A (2011) A molecular phyloge-netic assessment of the advanced Asiatic and Malesiandidymocarpoid Gesneriaceae with focus on non-monophyletic andmonotypic genera. Plant Syst Evol 292:223–248. doi:10.1007/s00606-010-0413-z

Möller M, Pfosser M, Jang C-G, Mayer V, Clark A, Hollingsworth ML,Barfuss MHJ, Wang Y-Z, Kiehn M, Weber A (2009) A preliminaryphylogeny of the ‘didymocarpoid Gesneriaceae’ based on three mo-lecular data sets: incongruence with available tribal classifications.Amer J Bot 96:989–1010. doi:10.3732/ajb.0800291

Nath U, Crawford BCW, Carpenter R, Coen E (2003) Genetic control ofsurface curvature. Science 299:1404–1407. doi:10.1126/science.1079354

Nishii K, Kuwabara A, Nagata T (2004) Characterization ofanisocotylous leaf formation in Streptocarpus wendlandii(Gesneriaceae): significance of plant growth regulators. Ann Bot94:457–467. doi:10.1093/aob/mch160

Nishii K, Möller M, Kidner CA, Spada A, Mantegazza R, Wang C-N,Nagata T (2010) A complex case of simple leaves: indeterminateleaves co-express ARP andKNOX1 genes. DevGenes Evol 220:25–40. doi:10.1007/s00427-010-0326-4

Nishii K,WangCN, SpadaA, Nagata T,Möller M (2012) Gibberellin as asuppressor of lateral dominance and inducer of apical growth in theunifoliate Streptocarpus wendlandii. N Z J Bot 50:267–287.doi:10.1080/0028825X.2012.671775

Nishii K, Ho MJ, Chou YW, Gabotti D, Wang CN, Spada A, Möller M(2014) GA2 and GA20-oxidase expressions are associated with themeristem position in Streptocarpus rexii (Gesneriaceae). PlantGrowth Regul 72:123–140. doi:10.1007/s10725-013-9844-1

Nishii K, Hughes M, Briggs M, Haston E, Christie F, DeVilliers MJ,Hanekom T, Roos WG, Bellstedt DU, Möller M (2015)Streptocarpus redefined to include all Afro-MalagasyGesneriaceae: molecular phylogenies prove congruent with geogra-phy and cytology and uncovers remarkable morphological homo-plasies. Taxon 64:1243–1274

Nylander JA (2004) MrModeltest, version 2. Program distributed by theauthor. Evolutionary Biology Centre. Uppsala University, Sweden

Pfaffl MW, Horgan GW, Dempfle L (2002) Relative Expression SoftwareTool (REST) for group-wise comparison and statistical analysis ofrelative expression results in real-time PCR. Nucleic Acids Res 30:E36. doi:10.1093/nar/30.9.e36

Refulio-Rodriguez NF, Olmstead RG (2014) Phylogeny of Lamiidae.Amer J Bot 101:287–299. doi:10.3732/ajb.1300394

Ronquist F, Teslenko M, Van der Mark P, Ayres D, Darling A, Höhna S,Larget B, Liu L, Suchard MA, Huelsenbeck JP (2012) MrBayes 3.2:efficient Bayesian phylogenetic inference and model choice across alarge model space. Syst Bot 61:539–542. doi:10.1093/sysbio/sys029

Rosenblum IM, Basile DV (1984) Hormonal-regulation of morphogene-sis in Streptocarpus and its relevance to evolutionary history of theGesneriaceae. Amer J Bot 71:52–64

Siegfried KR, Eshed Y, Baum SF, Otsuga D, Drews GN, Bowman JL(1999) Members of the YABBY gene family specify abaxial cell fatein Arabidopsis. Development 126:4117–4128

Schäferhoff B, Fleischmann A, Fischer E, Albach DC, Borsch T, HeublG, Müller KF (2010) Towards resolving Lamiales relationships:insights from rapidly evolving chloroplast sequences. BMC EvolBiol 10:352–371. doi:10.1186/1471-2148-10-352

Schneider CA, Rasband WS, Eliceiri KW (2012) NIH image to ImageJ:25 years of image analysis. Nat Methods 9:671–675. doi:10.1038/nmeth.2089

Dev Genes Evol (2017) 227:41–60 59

http://dx.doi.org/10.1007/BF00196573http://dx.doi.org/10.1016/S0960-9822(02)00721-2http://dx.doi.org/10.1073/pnas.1300592110http://dx.doi.org/10.1073/pnas.1300592110http://dx.doi.org/10.1021/bk-1999-0734.ch008http://dx.doi.org/10.1016/S0092-8674(00)81051-Xhttp://dx.doi.org/10.1105/tpc.104.028936http://dx.doi.org/10.1016/S0960-9822(02)01125-9http://dx.doi.org/10.1016/S0960-9822(02)01125-9http://dx.doi.org/10.1038/ng1835http://dx.doi.org/10.1006/anbo.2000.1155http://dx.doi.org/10.1086/510410http://dx.doi.org/10.1016/j.cub.2005.07.023http://dx.doi.org/10.1111/j.1469-8137.1975.tb01400.xhttp://dx.doi.org/10.1007/s00425-006-0258-4http://dx.doi.org/10.1038/379066a0http://dx.doi.org/10.1007/s00425-006-0389-7http://dx.doi.org/10.1007/s00425-009-0965-8http://dx.doi.org/10.1111/j.0014-3820.2001.tb00609.xhttp://dx.doi.org/10.1111/j.0014-3820.2001.tb00609.xhttp://dx.doi.org/10.1007/s00606-010-0413-zhttp://dx.doi.org/10.1007/s00606-010-0413-zhttp://dx.doi.org/10.3732/ajb.0800291http://dx.doi.org/10.1126/science.1079354http://dx.doi.org/10.1126/science.1079354http://dx.doi.org/10.1093/aob/mch160http://dx.doi.org/10.1007/s00427-010-0326-4http://dx.doi.org/10.1080/0028825X.2012.671775http://dx.doi.org/10.1007/s10725-013-9844-1http://dx.doi.org/10.1093/nar/30.9.e36http://dx.doi.org/10.3732/ajb.1300394http://dx.doi.org/10.1093/sysbio/sys029http://dx.doi.org/10.1186/1471-2148-10-352http://dx.doi.org/10.1038/nmeth.2089http://dx.doi.org/10.1038/nmeth.2089

Smith LG, Jackson D, Hake S (1995) Expression of Knotted1 marksshoot meristem formation during maize empryogenesis. DevGenet 16:344–348. doi:10.1002/dvg.1020160407

Stevens PF (2001 onwards) Angiosperm PhylogenyWebsite. Version 12.http://www.mobot.org/MOBOT/ research/APweb/.

Stoynova-Bakalova E, Karanov E, Petrov P, Hall MA (2004) Cell divi-sion and cell expansion in cotyledons ofArabidopsis seedlings. NewPhytol 162:471–479. doi:10.1111/j.1469-8137.2004.01031.x

Tanaka M, Onimaru K (2012) Acquisition of the paired fins: a view fromthe sequential evolution of the lateral plate mesoderm. Evolution &Development 14:412–420

Tononi P, Möller M, Bencivenga S, Spada A (2010) GRAMINIFOLIAhomolog expression in Streptocarpus rexii is associated with thebasal meristems in phyllomorphs, a morphological novelty in

Gesneriaceae. Evol & Dev 12:61–73. doi:10.1111/j.1525-142X.2009.00391.x

Tsukaya H (1997) Determination of the unequal fate of cotyledons of aone-leaf plant, Monophyllaea. Development 124:1275–1280.doi:10.1111/j.1525-142X.2012.00561.x

Vollbrecht E, Veit B, Sinha N, Hake S (1991) The developmental geneKnotted-1 is a member of a maize homeobox gene family. Nature350:241–243. doi:10.1038/350241a0

Weber A, Clark J, Möller M (2013) A new formal classification ofGesneriaceae. Selbyana 31:68–94

Yanai O, Shani E, Dolezal K, Tarkowski P, Sablowski R, Sandberg G,Samach A, Ori N (2005) Arabidopsis KNOXI proteins activate cy-tokinin biosynthesis. Curr Biol 15:1566–1571. doi:10.1016/j.cub.2005.07.060

60 Dev Genes Evol (2017) 227:41–60

http://dx.doi.org/10.1002/dvg.1020160407http://www.mobot.org/MOBOT/http://dx.doi.org/10.1111/j.1469-8137.2004.01031.xhttp://dx.doi.org/10.1111/j.1525-142X.2009.00391.xhttp://dx.doi.org/10.1111/j.1525-142X.2009.00391.xhttp://dx.doi.org/10.1111/j.1525-142X.2012.00561.xhttp://dx.doi.org/10.1038/350241a0http://dx.doi.org/10.1016/j.cub.2005.07.060http://dx.doi.org/10.1016/j.cub.2005.07.060

From...AbstractIntroductionMaterials and methodsPlant materialsCloning of STM homologues from Gesneriaceae and J.punctataHomology analysesCotyledon morphometry and meristematic activityLeaf morphology and meristematic activityGene expression analyses by RT-PCRGene expression analyses by RT-qPCRGene expression analyses by in situ hybridization

ResultsIsolation of STM genesCotyledonary basal meristem observed in both isocotylous and anisocotylous seedlings in the early post-germination stageLeaf basal meristem is common to Calceolariaceae and Gesneriaceae, but extended in Gesneriaceae

DiscussionComparisons to model plantsSTM is involved in lateral organ growth in GesneriaceaeThe basal meristemConsequences of the step-wise evolution of the basal meristem

References