Retinoic acid-independent expression of Meis2 during autopod patterning in the developing bat and mouse limb
© Mason et al.; licensee BioMed Central. 2015
Received: 31 December 2014
Accepted: 4 February 2015
Published: 14 March 2015
The bat has strikingly divergent forelimbs (long digits supporting wing membranes) and hindlimbs (short, typically free digits) due to the distinct requirements of both aerial and terrestrial locomotion. During embryonic development, the morphology of the bat forelimb deviates dramatically from the mouse and chick, offering an alternative paradigm for identifying genes that play an important role in limb patterning.
Using transcriptome analysis of developing Natal long-fingered bat (Miniopterus natalensis) fore- and hindlimbs, we demonstrate that the transcription factor Meis2 has a significantly higher expression in bat forelimb autopods compared to hindlimbs. Validation by reverse transcriptase and quantitative polymerase chain reaction (RT-qPCR) and whole mount in situ hybridisation shows that Meis2, conventionally known as a marker of the early proximal limb bud, is upregulated in the bat forelimb autopod from CS16. Meis2 expression is localised to the expanding interdigital webbing and the membranes linking the wing to the hindlimb and tail. In mice, Meis2 is also expressed in the interdigital region prior to tissue regression. This interdigital Meis2 expression is not activated by retinoic acid (RA) signalling as it is present in the retained interdigital tissue of Rdh10 trex/trex mice, which lack RA. Additionally, genes encoding RA-synthesising enzymes, Rdh10 and Aldh1a2, and the RA nuclear receptor Rarβ are robustly expressed in bat fore- and hindlimb interdigital tissues indicating that the mechanism that retains interdigital tissue in bats also occurs independently of RA signalling.
Mammalian interdigital Meis2 expression, and upregulation in the interdigital webbing of bat wings, suggests an important role for Meis2 in autopod development. Interdigital Meis2 expression is RA-independent, and retention of interdigital webbing in bat wings is not due to the suppression of RA-induced cell death. Rather, RA signalling may play a role in the thinning (rather than complete loss) of the interdigital tissue in the bat forelimb, while Meis2 may interact with other factors during both bat and mouse autopod development to maintain a pool of interdigital cells that contribute to digit patterning and growth.
KeywordsMeis2 Retinoic acid signalling Limb development Bat wing Evo-devo Interdigital webbing
Understanding the function of genes involved in patterning the tetrapod limb has traditionally been based on leads from human congenital abnormalities, studies using mouse genetics and the experimental manipulation of chicken embryos. This has led to the identification of key genes and pathways that pattern the outgrowing limb, including the three amino-acid loop extension (TALE) homeobox transcription factors, Meis homeobox 1 (Meis1) and Meis homeobox 2 (Meis2), that are expressed in the proximal limb bud [1,2].
One paradigm for understanding how limb outgrowth and patterning is regulated is the ‘two-signal model’ based on pharmacological and transplantation studies in the chick. This model proposes that the proximal-distal (P-D) axis is specified by opposing proximal and distal factors, namely retinoic acid (RA) and fibroblast growth factor (FGF) signalling, that activate or repress Meis1/2 expression, respectively [3-5]. In this model, Meis2 expression is restricted to the proximal portion of the outgrowing limb bud, where it plays a role in specifying the proximal limb identity and is used as a marker of the stylopod [2,5-7]. Overexpression/ectopic distal expression of Meis2 in the developing chick limb results in distal limb defects that include limb axis proximalisation, reductions in the length of distal skeletal elements and the persistence of interdigital webbing [1,2]. Ectopic distal expression of Meis1 in the developing mouse limb leads to a comparable proximalisation of the distal elements, a reduction of the ulna, a delay in distal element ossification and syndactyly . Ectopic expression of Meis2 throughout the hindlimb (HL) bud, and the posterior half of the forelimb (FL) bud, leads to delayed expression of homeobox A13 (Hoxa13) in the distal limb bud (a marker of the autopod region) and persistent distal expression of homeobox A11 (Hoxa11; a marker of the zeugopod region) . These data have been used to build a model of limb bud outgrowth where Meis1 and Meis2 play a pivotal role in patterning the proximal limb.
Two distinct thresholds of RA signalling are proposed to delimit the three limb regions, with the higher mediating the stylopod to zeugopod transition and the lower (together with a timing mechanism specified by specific histone acetylation signals) mediating the zeugopod to autopod transition . In this model, RA is synthesised in the flank where retinol dehydrogenase 10 (Rdh10) and retinaldehyde dehydrogenase 2 (Aldh1a2/Raldh2) are expressed [9-12]. Diffusion of RA from the flank through the nascent limb bud forms a gradient, reinforced by degradation of RA through the action of a distally expressed cytochrome P450 RA-degrading enzyme encoded by Cyp26b1 [3,5]. Although genetic studies in mice have established that FGF signalling from the apical ectodermal ridge (AER) is required for P-D patterning and the distal repression of Meis1/2 , several genetic studies that abrogate RA synthesis cast doubt on the requirement for endogenous RA signalling in P-D patterning or for expression of Meis1/2 in the proximal limb [14-16].
Genetic studies support a role for RA signalling later in limb development, maintaining the interdigital tissue of the autopod in an undifferentiated state and mediating cell death events [15,17,18], as well as playing a role in tissue specification events at the digit-interdigit junction during digit development . RA is synthesised in the interdigital region through expression of Rdh10 and Raldh2 , while RA is degraded in the digit region through Cyp26b1 expression . Together, these genes regionalise the developing autopod into RA-depleted regions (the condensing digits) and RA-rich regions (undifferentiated interdigital tissue). The interdigital mesenchyme is maintained in mice that are deficient in the RA-signalling pathway (Rdh10 trex/trex , Raldh2 −/− , RARβ −/− /RARγ −/− and RARβ +/− /RARγ −/− mice) suggesting that RA synthesis and subsequent signalling is necessary for the regression of interdigital tissues [15,17-19].
Interdigital tissue is also naturally retained during development in the autopods of several vertebrates, a feature that possibly evolved as an adaptive trait to facilitate locomotion and predation. The bat wing is a clear example of this, containing asymmetrically elongated digits (II to V) that support an expansive wing membrane used in powered flight. In contrast, the bat’s hindlimb digits are short, clawed and (in most species) free, being adapted for crawling and roosting . This unique contrast in autopod morphology provides a powerful comparative system that can be used to inform current developmental models that are established using conventional systems such as the mouse and the chick [22,23].
Based on leads from a transcriptome analysis of developing bat autopods, we present data showing that Meis2 is upregulated in the interdigital tissue of mammalian autopods. Furthermore, we find that RA signalling and Meis2 expression are not coupled in the bat or in the mouse autopod.
Sample collection, ethics and preparation
Fold change data for Meis2 transcripts over both the 5′-Meis2 and the 3′-Meis2 region
Bat (CS17) vs mouse (E13.5)
FL vs HL
9.9 (8.4 to 11.7)
2 (1.5 to 3.3)
6.1 (5.4 to 12.6)
287.1 (56.5 to 829.9)
0.5 (0.5 to 0.7)
3.0 (2.3 to 3.9)
8.4 (7.7 to 8.5)
7.9 (5.8 to 8.4)
1.5 (1.2 to 2.3)
1 (0.6 to 1.2)
1.4 (1.2 to 2.6)
8.6 (8.3 to 12.3)
0.9 (0.6 to 1.2)
1.9 (1.6 to 2.6)
5.9 (5.5 to 6.0)
4.9 (2.2 to 5.1)
Autopods from the FLs and HLs of bat embryos and the FLs of E13.5 mouse ICR embryos (UCT strain 1, University of Cape Town Medical School, Animal Ethics Committee application number: 006/040) were dissected off while immersed in RNAlater® (Sigma-Aldrich, St Louis, MO, US). This area was distinguished as the dorso-ventrally flattened tissue on the distal portion of the limb and was dissected along the point of constriction of the presumptive wrist or ankle, excluding the tissue associated with the stylopod and the zeugopod as well as that associated with the proto- and plagiopatagium. This autopod tissue is subsequently referred to as either the FL or the HL. Autopod RNA was extracted using the RNeasy® Lipid Tissue Mini Kit (QIAGEN, Hilden, DE) and checked for quality by the Centre for Proteomic and Genomic Research (CPGR, Cape Town, ZA) using the Agilent 2100 Bioanalyser (Agilent Technologies), in addition to running denaturing agarose gels and quantified using the Nanodrop 1000 (Thermo Scientific, Pittsburgh PA, US).
In situ hybridisation experiments were performed using several mouse strains. Meis2 in situ hybridisation experiments were performed on wildtype NIMR:Parkes mice. RA-signalling pathway in situ hybridisation experiments were performed on wildtype C57BL/6 (UCT strain 3) mice (UCT Medical School Animal Ethics number 012/052). The comparison of Meis2 expression in wildtype (Black Swiss) compared to Rdh10 trex/trex mutant mice, a strain that lacks RA signalling in limbs and exhibits retained interdigital webbing in the hindlimb [9,15], conformed to the regulatory standards adopted by the Animal Research Committee at the Sanford-Burnham Medical Research Institute.
Input RNA samples (0.5 μg) were amplified and labelled using the Amino Allyl MessageAmp™ II Cy3 aRNA Amplification Kit (Ambion, Austin, TX, US). Each labelled amplified RNA (aRNA) sample (250 ng) was hybridised to OpArray™ Mouse v4.0 slides and processed according to the manufacturer’s instructions (OPERON Biotechnologies, Inc., Huntsville, AL, US). Slides were scanned using a Genepix® 4000A scanner (Axon Instruments Inc., Molecular Devices, Sunnyvale, CA, US). Data were captured, background signal calculated and spot quality checked using GenePix™ Pro ver. 6.0 microarray analysis software (Amersham Biosciences, Piscataway, NJ, US). Background correction, normalisation and subsequent pre-processing steps were performed in R ver. 2.6.0 (R Development Core Team). Poor quality features were removed, missing data were imputed, batch artefacts were corrected for and variance filtering was performed. Filtered and pre-processed data were analysed for differential expression using the DEDS package (ver. 1.12). Genes were ranked according to their cumulative q value and determined to be differentially expressed if fold change (FC) > 2 (interspecies analysis) or FC > 1.5 (intraspecies analysis) and q < 0.01. Single-channel normalised data were used to graph array signal of developmental genes of interest. Data were imported into Microsoft Excel (Microsoft Office 2007), and biological replicate averages were compared using the moderated t-statistic with Benjamini and Hochberg False Discovery Rate correction. Protocols and data were deposited in the NCBI Gene Expression Omnibus [26,27] and can be accessed through the GEO Series accession number GSE51042 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE51042).
Analysis of Miniopterus schreibersii RNA-seq dataset
mRNA-seq datasets (Accession GSE50699) from embryonic autopods of the common bent-wing bat (Miniopterus schreibersii) were downloaded from the Gene Expression Omnibus (GEO). For each gene of interest, the longest mouse (GRCmm38) coding transcript sequence (cDNA) was obtained (Ensemble Genome Browser, release 75) and used to perform a BLASTn on the RNA-seq bat sequence dataset. M. schreibersii sequences with high similarity were noted [see Additional file 1: Table S2], and the normalised read counts for these were extracted. Where there was more than one high-similarity sequence for a gene of interest, the sum of the normalised read counts was taken.
5′ and 3′ RACE and high-fidelity PCR
5′ and 3′ Rapid Amplification of cDNA Ends (RACE)-ready complementary DNA (cDNA) was synthesised from the total RNA extracted from the FLs and heads of CS17 M. natalensis and E13.5 Mus musculus embryos using the SMARTer™ cDNA Amplification kit (Clontech, Mountain View, CA, US), using oligo(dT) to prime first strand synthesis. Various gene-specific primers (GSPs) and nested gene-specific primers (NGSPs) were designed to generate full-length 5′ and 3′ RACE cDNA transcripts from the Meis2 locus [Additional file 1: Table S3 and Additional file 2: Figure S1] in combination with Clontech universal primers, using PCR conditions summarised in Additional file 1: Table S4 and Additional file 3. Additional details of PCR conditions, cloning and sequencing can be found in Additional file 3. A Meis2 overlap region, between the genome-encoded first adenine-rich region and the start of the coding region [see Additional file 2: Figure S1], was amplified from bat (CS17) and mouse (E13.5) embryonic head cDNA as described in Additional file 3.
Relative qPCR analysis
RT-qPCR was performed following the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) conventions . Primers were designed to amplify the transcript regions that were specifically hybridised by the corresponding mouse OPERON probes [see Additional file 1: Table S5]. Experiments were performed using Amino allyl aRNA samples that were reverse transcribed into cDNA using Superscript III Reverse transcriptase (Roche Molecular Diagnostics, Pleasanton, CA, US) and second-round primers (Ambion, Austin, TX, USA). Relative qPCR was performed using the Sensimix SYBR Kit (BIOLINE, London, GB). A portion of each cDNA sample was pooled and serially diluted by a factor of two to generate five standard curve samples for each run. Experiments were performed on the Rotogene 6000 (QIAGEN). Standard curves and E values were generated using the Rotogene 6000 real time rotary analyser software (Ver. 1.7, QIAGEN). Standard curves were plotted for each experiment and the E values [see Additional file 1: Table S5] and Ct values were imported into Microsoft Office Excel 2010 (Microsoft Corporation) and analysed. Outliers that could be attributed to a failed amplification were discarded and the technical repeats averaged. Data were analysed using the efficiency correction method : data were calibrated to the average dilution in the serial dilution standard curve and normalised to the reference gene, Tbpl1. After calibration, data were scaled to the bat CS15 HL. Statistical analyses were performed in IBM® SPSS Statistics ver. 22 (IBM Corporation, Armonk, NY, US).
Whole mount in situ hybridisation (WISH)
Bat-specific whole mount in situ hybridisation (WISH) probe templates were generated from bat FL pooled cDNA (CS15 to CS18). Primers for Meis2 WISH probes were designed to include the regions that were specifically hybridised by the corresponding mouse OPERON probes. Primer and probe template information for all bat WISH experiments can be found in the supplementary information [see Additional file 1: Table S6]. PCR products were purified (Promega, Fitchberg, WI, US) and cloned into the pGEM®-T Easy vector (Promega) and sequenced (Central Analytical Facilities, Stellenbosch, WC, ZA) [see Additional file 1: Table S6]. A mouse 5′-Meis2 probe was generated from RIKEN FANTOM™ mouse clone A830011L22 (GenBank: AK043601.1) (Source Bioscience Lifesciences, Cambridge, GB). Information for mouse probes of RA-signalling-associated genes can be found in previously published work: mRdh10 , mAldh1a2 , mCyp26b1  and mRarβ . Plasmids were linearized using the appropriate restriction enzymes, and DIG-labelled probes were generated using in vitro transcription with T3, T7 or SP6 polymerase as required (Roche Molecular Diagnostics). Bat embryos were halved along their rostral-caudal axis prior to WISH to conserve samples and enable direct comparisons between genes from the same developmental stage. WISH was performed as previously described  with the following modifications: proteinase K digestion times were as recommended by , overnight hybridisation temperature set at 70°C and detection performed using nitro-blue tetrazolium and 5-bromo-4-chloro-3′-indolyphosphate solution (NBT-BCIP).
Data deposition footnote
NCBI Gene Expression Omnibus GEO Series accession number GSE51042 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE51042). Genbank Accession numbers for all sequences in this paper are given in Additional file 1: Table S7 and S9 to S12.
Cross-species microarray analysis reveals a set of genes that are robustly differentially expressed between the bat fore- and hindlimb, including Meis2 and Hoxd11
In contrast to the interspecies analysis, far fewer genes were DE in the intraspecies comparison (Figure 1B (i) and (ii)). Only 15 genes were significantly DE (FC > 1.5, q < 0.01) when comparing the bat FL to the HL at CS17 and only six at CS16 (Figure 1B (iii), Additional file 1: Table S7). Two genes were upregulated at both developmental stages and were significantly DE across all comparisons. The most significantly DE gene corresponded to an unannotated probe (M400017713) that mapped upstream of the mouse Meis2 gene locus, while the second probe mapped to the homeobox D11 (Hoxd11) locus. Here, we explore the role of Meis2 in autopod formation.
The microarray probe M400017713 corresponds to the 5′ UTR of the Meis2 mRNA transcript
BLASTn analysis of the sequence corresponding to M400017713 had a 100% match to a RIKEN mouse clone AK043601, annotated as a cDNA from a RIKEN, full-length enriched, 10-day neonate cortex library. This small (472bp) cDNA clone ended in a polyadenylated tract, and mapped upstream of the mouse Meis2 locus in the sense direction. The cDNA was predicted to not encode a protein and we consequently refer to it as a long non-coding Meis transcript (lncMeis2).
Meis2 transcripts are upregulated in developing bat wings
The microarray signal from the 5′-Meis2 probe was compared alongside that from the 3′-Meis2 probe, in addition to the signal from the Meis1 and Meis3 probes (Figure 2A). The signal from the 5′-Meis2 probe was higher in the FL of both the CS17 (over sixfold) and CS16 (twofold) bat (Table 1) compared to their HLs and was significantly higher (over ninefold) in the CS17 FL compared to that of the mouse (Table 1, Figure 2A). The 3′-Meis2 probe was also higher in CS17 FL as compared to the HL (1.4-fold) and the mouse FL (1.5-fold), but these fold changes were modest with the probe having a lower array signal overall. Both mouse probes have high similarity to the bat genomic sequence (96% and 98%, respectively). The array signals from the Meis1 and Meis3 probes were low and not significantly different between limb types (Figure 2A). These data correspond with recently published RNA-seq data in the closely related bat species M. schreibersii [see Additional file 6: Figure S3] [34,35]. Our analysis of the M. schreibersii dataset showed that Meis2 was the most highly expressed of the three Meis genes. Its expression in tissue taken from the anterior and posterior regions of the early limb bud (CS14) corresponds well to the known expression pattern of Meis2 at early stages. In the pooled CS15-17 bat limbs, Meis1 and Meis3 have a very low read count in both limb types. In contrast to this, Meis2 transcripts are abundant in the FL interdigital tissue (digits II to V), while expression is low in FL digit I, FL digits II to V and HL samples [see Additional file 6: Figure S3] .
RT-qPCR validation confirms the upregulation of Meis2 transcripts in autopods of developing bat forelimbs
Expression of Meis2 occurs in the interdigits of the mouse autopod and the interdigital webbing of the bat forelimb but is absent from the bat hindlimb
The early expression of Meis2 in the CS15 bat is comparable to that of the E11.5 mouse, with expression confined to the proximal region of the limb bud (Figure 4C, D). The first evidence of expression of Meis2 in the bat autopod is in the CS16 FL where a strong domain of expression can be seen in the interdigital region between digits IV and V (Figure 4G). A less intense region of staining is apparent in the interdigital region between digits III and IV, and no expression is evident in interdigital regions I to II or II to III at this stage (Figure 4G). Interestingly, Meis2 is also strongly expressed in the presumptive plagiopatagium and uropatagium from CS15 onwards (red arrows in Figure 4G, H, K, L, O, P). At later stages, Meis2 expression is evident in all interdigital tissues of the FL, with a strong proximal bias (Figure 4K, O). The ‘comma-shaped’ domain of expression observed in the mouse HL is not found in the HL of any bat embryos (Figure 4H and Additional file 9: Figure S6). Although Meis2 expression is undetectable in CS15 to CS17 HLs, we could detect a faint line interdigital expression in CS18 HL autopods (Figures 4P).
WISH using the 5′- and 3′-Meis2 probes gave similar results, except in mouse HLs where only the 3′-probe detected the aforementioned ‘comma-shaped’ domain [see Additional file 9: Figure S6 and Additional file 10: Figure S7]. This could be due to variation in the sensitivity of the different-sized probes or a difference among the 5′ UTR of Meis2 transcripts for this domain.
Meis2 transcripts are expressed in the absence of RA signalling in the interdigital tissue of Rdh10 mutants
Our discovery of Meis2 expression in the interdigital tissue of the developing mouse and bat gave us the opportunity to test the dependence of its expression on RA signalling and examine the role of RA signalling in shaping the bat FL autopod with its retained interdigital membranes.
Retention of interdigital webbing in the bat forelimb is not a consequence of diminished RA signalling
In mice, interdigital RA signalling plays a role in apoptotic events that result in freely separated digits [15,19,37]. We used a combination of our microarray expression dataset and in situ hybridisation studies to investigate whether retention of the interdigital membrane in the bat FL autopod was linked to downregulation of RA signalling in comparison to the bat HL and mouse FL.
An analysis of the microarray data shows that there is no significant difference in the abundance of mRNA transcripts for the key RA-signalling genes between bat FL and HL autopods [see Additional file 10: Figure S7]. The Rdh10, Aldh1a2 and Cyp26b1 genes, which encode enzymes that perform the entirety of RA synthesis and degradation in the mouse limb [9,11,15,20], have signals that are similar among all limb types. Rarβ plays a predominant role in limb RA signal transduction in the interdigital region and also showed very little signal differences across limb types [see Additional file 10: Figure S7A].
In this paper, we describe a novel, RA-independent expression domain for the transcription factor Meis2 in interdigital tissue during mouse and bat embryonic limb development. The expression of Meis1/2 in the proximal limb during early limb bud outgrowth has been described in several vertebrates including zebrafish, urodele, chick and mouse embryos [1,2,38-44]. Consistent with these reports, we show that early Meis2 expression in the proximal limb mesenchyme of Natal long-fingered bat (M. natalensis) embryos appears to be conserved, resembling that of the mouse. Additionally, we show that Meis2 has a second phase of expression in the autopod later in limb development. Meis2 expression is robustly upregulated in M. natalensis CS16 and CS17 FL autopods compared to HL autopods. Furthermore, Meis2 transcripts are expressed in the interdigital tissue of mouse FL and HL autopods, prior to tissue regression.
Meis2 is conventionally described in terms of its role in specifying the proximal limb, that is the stylopod [3,6,7]. The proposed dependence of Meis1 and Meis2 on RA signalling from the embryonic flank is an important component of this two-signal model of proximal-distal patterning [4,5]. However, the role of RA signalling in this model has been challenged. Genetic loss-of-function experiments show that expression of Meis1/2 expression is maintained in the limbs of Raldh2 −/−, Raldh3 −/− and Rdh10 trex/trex mouse embryos, and even in Rdh10 trex/trex embryos treated with a retinoic acid receptor (RAR) antagonist [14-16]. These mutants show no endogenous limb-field RA signalling as indicated by a retinoic acid responsive element-β-galactosidase (RARE-lacZ) transgenic reporter, shown to be sensitive to 0.25 nM RA . Although these studies show that RA signalling is required for mouse FL bud initiation, as evidenced by stunted FL buds, RA is not required for HL bud initiation or proximal expression of Meis1/2 in either the FL or HL [14,16,45]. We show that Meis2 is expressed in the interdigital membrane of Rdh10 trex/trex mice, despite an absence of RA signalling (clear from the prominent syndactyly defect and lack of RARE-lacZ activation/expression ). This observation is consistent with a model of limb patterning in which the expression of Meis2 does not require RA signalling.
A role for Meis2 in shaping autopods has not been previously considered; we do so in the context of RA signalling and interdigital apoptosis. RA signalling is known to play an important role in autopod patterning by regulating the expression of genes that activate cell death and inhibit cell differentiation in the interdigital regions, as well as controlling tissue specification events at the digit-interdigit junction [15,19,37]. Mutants that are deficient in RA signalling display syndactyly  raising the possibility that diminished RA signalling in the bat FL might explain the retention of the interdigital membrane. We show that this not the case. An analysis of mRNA expression data from microarrays (this study), as well as recently published RNA-seq datasets , did not identify any significant differences in the expression of key enzymes in the RA synthesis pathway (Rdh10, Raldh2 and Cyp26b1) in bat FL autopods compared to HL autopods. Furthermore, in situ hybridisation analysis shows that the RA-responsive gene, Rarβ, is expressed in the interdigital tissue of bat FL and HL autopods, as well as in the plagiopatagium and uropatagium, suggesting that RA signalling is active in these tissues. Therefore, a loss of RA signalling does not explain the phenotype of retained interdigital webbing phenotype in the bat FL autopod.
It is possible that the bat FL lacks the downstream response to RA signalling required for complete removal of interdigital tissue. Another possibility is that RA signalling in the bat FL is modulated, allowing the thinning but not the complete removal of the bat wing interdigital membranes. RA signalling is known to activate BMP-mediated cell death in the mouse limb interdigital region . However, this appears to be limited in the bat by the expression of Gremlin (a BMP antagonist) in the FL at CS16 and CS17, Msh homeobox 1 (Msx1) and Msh homeobox 2 (Msx2) (apoptosis markers) are still expressed . Also, expression of fibroblast growth factor 8 (Fgf8) in the interdigital mesenchyme of the bat CS16 and CS17 FL likely limits distal apoptosis and prevents the ingression of interdigital tissue adjacent to digits II to V. This hypothesis is supported by experiments showing that inhibition of FGF in the bat FL (using a bead soaked in the FGF inhibitor SU5402) alongside the addition of ectopic BMP causes increased cell death in interdigital regions . Thus, while RA-mediated apoptosis is likely occurring in the bat interdigital membranes, this process appears to be modulated by the presence of Gremlin and Fgf8 in the FL, resulting in the thinning of the interdigital tissue rather than the complete regression of the membranes.
In both the mouse and bat autopod, Meis2 may mark the regions of interdigital tissue that are retained as a proliferating pool of cells that feed digit development. In the mouse, Meis2 is predominantly expressed in the more proximal interdigital regions, which undergo thinning, but do not appear to regress [47,48]. This observation fits the differential growth model of autopod formation whereby the distal regions of the interdigital tissue undergo massive apoptosis [47-49]. These distal regions are made up of cells that have previously received FGF signals from the AER and subsequently undergo apoptosis due to a loss of this signalling [47-49]. The digit regions, over which FGF signals are maintained, proliferate, extending distally at a faster rate than that of the interdigital regions. Meis2 may play a role in the survival of the regions of interdigital tissue that are retained during this process.
This proposal is consistent with our observations in the bat. The pattern of Meis2 expression in the FL autopod (hand plate), and lack of expression in the HL autopod (footplate), is associated with digit and interdigital tissue outgrowth, rather than interdigital regression. Meis2 expression in the bat hand plate first appears at CS16 between digits IV and V. This region of the hand plate undergoes a rapid expansion at this stage of development [24,25]. In the Carollia stage 16 late (CS16L) hand plate, Meis2 expression expands to the tissue between digital rays II to V. The interdigital tissue between the thumb and digit II recedes by stage CS17, while the tissue between the digits II and V persists to form the chiropatagium (wing membrane) that extends distally as digits II to V grow out rapidly. These observations extend to other bat species. Meis2 transcripts are abundantly expressed in the interdigital tissue of FL digits II to V, but not in the HLs for stages CS15L-CS19 in M. schreibersii embryos [34,50]. Elevated expression of Meis2 has also been noted in tissue taken from FL digits and interdigital tissue (II to V) compared to FL digit I or the HL (digits I to V) at the later (fetal) stages in Myotis ricketti and Hipposideros armiger . Therefore, it appears that the expression of Meis2 coincides with the thinning and growth of the interdigital tissue into expansive webbing, a process that begins during autopod formation and continues throughout fetal development . Meis2 expression may mark and determine the fate of this region by keeping these cells in a proliferating, undifferentiated state, similar to the role of Meis1/2 in retinal progenitor cells .
In the developing bat forelimb, there is a rapid growth and elongation of digits II to V and their associated patagia; these processes contrast greatly with what is occurring in the hindlimb where digits I to V remain shortened and become free. The expression of Meis2 transcripts in the developing bat hand plate corresponded to these processes and is consistent with an important role for Meis2 in stimulating both the outgrowth of digits II to V and the expansion of the chiropatagium in the bat. Furthermore, expression of Meis2 in developing mouse FL and HL autopods may play a similar, albeit less dramatic, part in sculpturing the tetrapod hand and foot, by maintaining a pool of proliferating interdigital cells that stimulates digit outgrowth. We have shown that these functions of Meis2 in the autopod operate independently of RA signalling.
apical ectodermal ridge
- Aldh1a2/Raldh2 :
retinaldehyde dehydrogenase 2
coding transcript sequence
Carollia stage 15
Carollia stage 16
Carollia stage 16 late
Carollia stage 17
Carollia stage 18
- Cyp26b1 :
cytochrome P450 RA-degrading enzyme
mouse embryo 13.5 days
fibroblast growth factor
- Fgf8 :
fibroblast growth factor 8
- Hoxa11 :
- Hoxa13 :
- Hoxd11 :
- M. natalensis :
- M. schreibersii :
- Meis1 :
Meis homeobox 1
- Meis2 :
Meis homeobox 2
Minimum Information for Publication of Quantitative Real-Time PCR Experiments
- Msx1 :
Msh homeobox 1
- Msx2 :
Msh homeobox 2
nitro-blue tetrazolium and 5-bromo-4-chloro-3′-indolyphosphate solution
Rapid Amplification of cDNA Ends
retinoic acid receptor
- RARE-lacZ :
retinoic acid responsive element-β-galactosidase
- RARβ :
retinoic acid receptor, beta
- Rdh10 :
retinol dehydrogenase 10
reverse transcriptase and quantitative polymerase chain reaction
three amino-acid loop extension
whole mount in situ hybridisation
We thank Robyn Verrinder for assistance in sample collection and advice concerning the R programming language, the management of De Hoop Nature Reserve for accommodating us during field trips, the Western Cape Nature Conservation Board for providing collection permits, Ania Kucharska and staff of the NIMR Animal Research Unit, Jaap Visser and staff of the UCT Medical School Animal Research Unit. DH and MKM were supported by National Research Foundation Prestigious Masters Scholarships. LC was supported by a Research Chair NRF scholarship. DH was supported by the Society for Integrative and Comparative Biology Grant-in-Aid-of-Research Award. MKM was supported by the DST NRF Doctoral Innovation Scholarship and Commonwealth Split-site Scholarship. The following grants funded the research: National Research Foundation, South Africa #78982 (NI); Department of Science & Technology South African Research Chair Initiative # 64798 (DSJ); National Research Foundation Grant #65525 (DSJ); National Institutes of Health, AR067731 (GD); Medical Research Council (MRC) grant U117560477 (ML).
- Mercader N, Leonardo E, Azpiazu N, Serrano A, Morata G, Martinez C, et al. Conserved regulation of proximodistal limb axis development by Meis1/Hth. Nature. 1999;402:425–9.View ArticlePubMedGoogle Scholar
- Capdevila J, Tsukui T, Rodriquez Esteban C, Zappavigna V, Izpisua Belmonte JC. Control of vertebrate limb outgrowth by the proximal factor Meis2 and distal antagonism of BMPs by Gremlin. Mol Cell. 1999;4:839–49.View ArticlePubMedGoogle Scholar
- Yashiro K, Zhao X, Uehara M, Yamashita K, Nishijima M, Nishino J, et al. Regulation of retinoic acid distribution is required for proximodistal patterning and outgrowth of the developing mouse limb. Dev Cell. 2004;6:411–22.View ArticlePubMedGoogle Scholar
- Rosello-Diez A, Ros MA, Torres M. Diffusible signals, not autonomous mechanisms, determine the main proximodistal limb subdivision. Science. 2011;332:1086–8.View ArticlePubMedGoogle Scholar
- Mercader N, Leonardo E, Piedra ME, Martinez AC, Ros MA, Torres M. Opposing RA and FGF signals control proximodistal vertebrate limb development through regulation of Meis genes. Development. 2000;127:3961–70.PubMedGoogle Scholar
- Zeller R, Lopez-Rios J, Zuniga A. Vertebrate limb bud development: moving towards integrative analysis of organogenesis. Nat Rev Genet. 2009;10:845–58.View ArticlePubMedGoogle Scholar
- Tabin C, Wolpert L. Rethinking the proximodistal axis of the vertebrate limb in the molecular era. Genes Dev. 2007;21:1433–42.View ArticlePubMedGoogle Scholar
- Rosello-Diez A, Arques CG, Delgado I, Giovinazzo G, Torres M. Diffusible signals and epigenetic timing cooperate in late proximo-distal limb patterning. Development. 2014;141:1534–43.View ArticlePubMedGoogle Scholar
- Sandell LL, Sanderson BW, Moiseyev G, Johnson T, Mushegian A, Young K, et al. RDH10 is essential for synthesis of embryonic retinoic acid and is required for limb, craniofacial, and organ development. Genes Dev. 2007;21:1113–24.View ArticlePubMed CentralPubMedGoogle Scholar
- Cammas L, Romand R, Fraulob V, Mura C, Dolle P. Expression of the murine retinol dehydrogenase 10 (Rdh10) gene correlates with many sites of retinoid signalling during embryogenesis and organ differentiation. Dev Dyn. 2007;236:2899–908.View ArticlePubMedGoogle Scholar
- Niederreither K, McCaffery P, Drager UC, Chambon P, Dolle P. Restricted expression and retinoic acid-induced downregulation of the retinaldehyde dehydrogenase type 2 (RALDH-2) gene during mouse development. Mech Dev. 1997;62:67–78.View ArticlePubMedGoogle Scholar
- Mic FA, Haselbeck RJ, Cuenca AE, Duester G. Novel retinoic acid generating activities in the neural tube and heart identified by conditional rescue of Raldh2 null mutant mice. Development. 2002;129:2271–82.PubMed CentralPubMedGoogle Scholar
- Mariani FV, Ahn CP, Martin GR. Genetic evidence that FGFs have an instructive role in limb proximal-distal patterning. Nature. 2008;453:401–56.View ArticlePubMed CentralPubMedGoogle Scholar
- Zhao XL, Sirbu IO, Mlc FA, Molotkova N, Molotkov A, Kumar S, et al. Retinoic acid promotes limb induction through effects on body axis extension but is unnecessary for limb patterning. Curr Biol. 2009;19:1050–7.View ArticlePubMed CentralPubMedGoogle Scholar
- Cunningham TJ, Chatzi C, Sandell LL, Trainor PA, Duester G. Rdh10 mutants deficient in limb field retinoic acid signaling exhibit normal limb patterning but display interdigital webbing. Dev Dyn. 2011;240:1142–50.View ArticlePubMed CentralPubMedGoogle Scholar
- Cunningham TJ, Zhao XL, Sandell LL, Evans SM, Trainor PA, Duester G. Antagonism between retinoic acid and fibroblast growth factor signaling during limb development. Cell Rep. 2013;3:1503–11.View ArticlePubMed CentralPubMedGoogle Scholar
- Ghyselinck NB, Dupe V, Dierich A, Messaddeq N, Garnier JM, RochetteEgly C, et al. Role of the retinoic acid receptor beta (RAR beta) during mouse development. Int J Dev Biol. 1997;41:425–47.PubMedGoogle Scholar
- Dupe V, Ghyselinck NB, Thomazy V, Nagy L, Davies PJA, Chambon P, et al. Essential roles of retinoic acid signaling in interdigital apoptosis and control of BMP-7 expression in mouse autopods. Dev Biol. 1999;208:30–43.View ArticlePubMedGoogle Scholar
- Zhao XL, Brade T, Cunningham TJ, Duester G. Retinoic acid controls expression of tissue remodeling genes Hmgn1 and Fgf18 at the digit-interdigit junction. Dev Dyn. 2010;239:665–71.View ArticlePubMed CentralPubMedGoogle Scholar
- Abu-Abed S, MacLean G, Fraulob V, Chambon P, Petkovich M, Dolle P. Differential expression of the retinoic acid-metabolizing enzymes CYP26A1 and CYP26B1 during murine organogenesis. Mech Dev. 2002;110:173–7.View ArticlePubMedGoogle Scholar
- Neuweiler G. The biology of bats. New York: Oxford University Press; 2000.Google Scholar
- Cretekos CJ, Rasweiler JJ, Behringer RR. Comparative studies on limb morphogenesis in mice and bats: a functional genetic approach towards a molecular understanding of diversity in organ formation. Reprod Fertil Dev. 2001;13:691–5.View ArticlePubMedGoogle Scholar
- Weatherbee SD, Behringer RR, Rasweiler JJ, Niswander LA. Interdigital webbing retention in bat wings illustrates genetic changes underlying amniote limb diversification. Proc Natl Acad Sci U S A. 2006;103:15103–7.View ArticlePubMed CentralPubMedGoogle Scholar
- Hockman D, Cretekos CJ, Mason MK, Behringer RR, Jacobs DS, Illing N. A second wave of sonic hedgehog expression during the development of the bat limb. Proc Natl Acad Sci U S A. 2008;105:16982–7.View ArticlePubMed CentralPubMedGoogle Scholar
- Hockman D, Mason MK, Jacobs DS, Illing N. The role of early development in mammalian limb diversification: a descriptive comparison of early limb development between the Natal long-fingered bat (Miniopterus natalensis) and the mouse (Mus musculus). Dev Dyn. 2009;238:965–79.View ArticlePubMedGoogle Scholar
- Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30:207–10.View ArticlePubMed CentralPubMedGoogle Scholar
- Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, et al. NCBI GEO: archive for functional genomics data sets–update. Nucleic Acids Res. 2013;41:D991–5.View ArticlePubMed CentralPubMedGoogle Scholar
- Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009;55:611–22.View ArticlePubMedGoogle Scholar
- Pfaffl MW. Quantification strategies in real-time PCR. In A-Z of Quantitative PCR (Editor S.A. Bustin.). La Jolla: International University Line. 2004:87–112.
- MacLean G, Abu-Abed S, Dolle P, Tahayato A, Chambon P, Petkovich M. Cloning of a novel retinoic-acid metabolizing cytochrome P450, Cyp26B1, and comparative expression analysis with Cyp26A1 during early murine development. Mech Dev. 2001;107:195–201.View ArticlePubMedGoogle Scholar
- Giguere V, Lyn S, Yip P, Siu CH, Amin S. Molecular-cloning of cDNA-encoding a 2nd cellular retinoic acid-binding protein. Proc Natl Acad Sci U S A. 1990;87:6233–7.View ArticlePubMed CentralPubMedGoogle Scholar
- Riddle RD, Johnson RL, Laufer E, Tabin C. Sonic-hedgehog mediates the polarizing activity of the Zpa. Cell. 1993;75:1401–16.View ArticlePubMedGoogle Scholar
- Rasweiler JJ, Cretekos CJ, Behringer RR. Whole-mount in situ hybridization of short-tailed fruit bat (Carollia perspicillata) embryos with RNA probes. Cold Spring Harb Protocols. 2009;2009:pdb.prot5164.
- Wang Z, Dai MY, Wang Y, Cooper KL, Zhu TT, Dong D, et al. Unique expression patterns of multiple key genes associated with the evolution of mammalian flight. Proc R Soc London Biol. 2014;281:20133133.View ArticleGoogle Scholar
- Miller-Butterworth CM, Eick G, Jacobs DS, Schoeman MC, Harley EH. Genetic and phenotypic differences between South African long-fingered bats, with a global miniopterine phylogeny. J Mammal. 2005;86:1121–35.View ArticleGoogle Scholar
- Mendelsohn C, Ruberte E, Lemeur M, Morrisskay G, Chambon P. Developmental analysis of the retinoic acid-inducible rar-beta-2 promoter in transgenic animals. Development. 1991;113:723–34.PubMedGoogle Scholar
- Diaz-Hernandez M, Rios-Flores A, Abarca-Buis R, Bustamante M, Chimal-Monroy J. Molecular control of interdigital cell death and cell differentiation by retinoic acid during digit development. J Dev Biol. 2014;2:138–57.View ArticleGoogle Scholar
- OuladAbdelghani M, Chazaud C, Bouillet P, Sapin V, Chambon P, Dolle P. Meis2, a novel mouse Pbx-related homeobox gene induced by retinoic acid during differentiation of P19 embryonal carcinoma cells. Dev Dyn. 1997;210:173–83.View ArticleGoogle Scholar
- Biemar F, Devos N, Martial JA, Driever W, Peers B. Cloning and expression of the TALE superclass homeobox Meis2 gene during zebrafish embryonic development. Mech Dev. 2001;109:427–31.View ArticlePubMedGoogle Scholar
- Waskiewicz AJ, Rikhof HA, Hernandez RE, Moens CB. Zebrafish Meis functions to stabilize Pbx proteins and regulate hindbrain patterning. Development. 2001;128:4139–51.PubMedGoogle Scholar
- Zerucha T, Prince VE. Cloning and developmental expression of a zebrafish Meis2 homeobox gene. Mech Dev. 2001;102:247–50.View ArticlePubMedGoogle Scholar
- Mercader N, Tanaka EM, Torres M. Proximodistal identity during vertebrate limb regeneration is regulated by Meis homeodomain proteins. Development. 2005;132:4131–42.View ArticlePubMedGoogle Scholar
- Coy SE, Borycki AG. Expression analysis of TALE family transcription factors during avian development. Dev Dyn. 2010;239:1234–45.View ArticlePubMedGoogle Scholar
- Sanchez-Guardado LO, Ferran JL, Rodriguez-Gallardo L, Puelles L, Hidalgo-Sanchez M. Meis gene expression patterns in the developing chicken inner ear. J Comp Neurol. 2011;519:125–47.View ArticlePubMedGoogle Scholar
- Kawakami Y. Redefining the role of retinoic acid in limb development. Cell Rep. 2013;3:1337–8.View ArticlePubMedGoogle Scholar
- Chimal-Monroy J, Abarca-Buis RF, Cuervo R, Diaz-Hernandez M, Bustamante M, Rios-Flores JA, et al. Molecular control of cell differentiation and programmed cell death during digit development. IUBMB Life. 2011;63:922–9.View ArticlePubMedGoogle Scholar
- Salas-Vidal E, Valencia C, Covarrubias L. Differential tissue growth and patterns of cell death in mouse limb autopod morphogenesis. Dev Dyn. 2001;220:295–306.View ArticlePubMedGoogle Scholar
- Hernandez-Martinez R, Castro-Obregon S, Covarrubias L. Progressive interdigital cell death: regulation by the antagonistic interaction between fibroblast growth factor 8 and retinoic acid. Development. 2009;136:3669–78.View ArticlePubMedGoogle Scholar
- Hernandez-Martinez R, Covarrubias L. Interdigital cell death function and regulation: new insights on an old programmed cell death model. Dev Growth Diff. 2011;53:245–58.View ArticleGoogle Scholar
- Dai M, Wang Y, Fang L, Irwin DM, Zhu T, Zhang J, et al. Differential expression of Meis2, Mab21l2 and Tbx3 during limb development associated with diversification of limb morphology in mammals. PLoS One. 2014;9:e106100.View ArticlePubMed CentralPubMedGoogle Scholar
- Wang Z, Dong D, Ru BH, Young RL, Han NJ, Guo TT, et al. Digital gene expression tag profiling of bat digits provides robust candidates contributing to wing formation. BMC Genomics. 2010;11:619.View ArticlePubMed CentralPubMedGoogle Scholar
- Heine P, Dohle E, Bumsted-O’Brien K, Engelkamp D, Schulte D. Evidence for an evolutionary conserved role of homothorax/Meis1/2 during vertebrate retina development. Development. 2008;135:805–11.View ArticlePubMedGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.