Tissue and cell-specific transcriptomes in cotton reveal the subtleties of gene regulation underlying the diversity of plant secondary cell walls
- 2.3k Downloads
Knowledge of plant secondary cell wall (SCW) regulation and deposition is mainly based on the Arabidopsis model of a ‘typical’ lignocellulosic SCW. However, SCWs in other plants can vary from this. The SCW of mature cotton seed fibres is highly cellulosic and lacks lignification whereas xylem SCWs are lignocellulosic. We used cotton as a model to study different SCWs and the expression of the genes involved in their formation via RNA deep sequencing and chemical analysis of stem and seed fibre.
Transcriptome comparisons from cotton xylem and pith as well as from a developmental series of seed fibres revealed tissue-specific and developmentally regulated expression of several NAC transcription factors some of which are likely to be important as top tier regulators of SCW formation in xylem and/or seed fibre. A so far undescribed hierarchy was identified between the top tier NAC transcription factors SND1-like and NST1/2 in cotton. Key SCW MYB transcription factors, homologs of Arabidopsis MYB46/83, were practically absent in cotton stem xylem. Lack of expression of other lignin-specific MYBs in seed fibre relative to xylem could account for the lack of lignin deposition in seed fibre. Expression of a MYB103 homolog correlated with temporal expression of SCW CesAs and cellulose synthesis in seed fibres. FLAs were highly expressed and may be important structural components of seed fibre SCWs. Finally, we made the unexpected observation that cell walls in the pith of cotton stems contained lignin and had a higher S:G ratio than in xylem, despite that tissue’s lacking many of the gene transcripts normally associated with lignin biosynthesis.
Our study in cotton confirmed some features of the currently accepted gene regulatory cascade for ‘typical’ plant SCWs, but also revealed substantial differences, especially with key downstream NACs and MYBs. The lignocellulosic SCW of cotton xylem appears to be achieved differently from that in Arabidopsis. Pith cell walls in cotton stems are compositionally very different from that reported for other plant species, including Arabidopsis. The current definition of a ‘typical’ primary or secondary cell wall might not be applicable to all cell types in all plant species.
KeywordsSecondary cell wall Primary cell wall Transcription factor Cotton Gossypium hirsutum Cellulose synthase Lignin Syringyl Guaiacyl p-Hydroxyphenyl
Development of secondary cell walls (SCWs) was a key event during the evolution of land plants. The thick, specialised SCWs of xylem vessels and xylem fibres of angiosperms and tracheids of gymnosperms allow trees to grow to more than 100 m in height. In many species, pollen release and fertilisation, as well as seed dispersal are facilitated by specialised SCWs. Plants can also deposit localised ‘SCW-like’ structures in response to pathogen attack, or in specialised cells types such as transfer cells. Domesticated cottons produce long seed fibres that have extremely thick SCWs and hence are traded globally in textile and biomaterial markets. Our most comprehensive understanding on how SCWs are made has mainly come from detailed work over the last decade on xylem vessels and fibres in Arabidopsis thaliana [1, 2, 3] and other plants such as poplar , rice , grasses , spruce . The SCWs in these plants are generally considered as ‘typical’ SCWs: they are composed of approximately equal measures of cellulose, lignins and hemicelluloses, and some proteins, but the specific composition can be variable between tissues and species. SCWs are deposited between the cell’s external primary cell wall (PCW) and the plasma membrane, often once cell expansion has ceased, and are usually orders of magnitude thicker than the PCW.
The current view of the regulation of lignocellulosic SCW deposition is that a cascade of SCW-specific NAC (for NAM, ATAF1/2, and CUC2) and MYB (myeloblastosis) transcription factors (TFs) regulate downstream TFs such as other NACs, MYBs, and KNATs (knotted-like from A rabidopsis t haliana) and the SCW biosynthetic genes encoding, for example, cellulose synthases (CesAs), lignin-related enzymes, enzymes for hemicellulose synthesis, and other cell wall structural components [1, 2, 3].
Different pairs of NACs act in a top tier of these lignocellulosic SCW regulatory cascades in particular Arabidopsis cell types [8, 9]. VND6/VND7 (for vascular related NAC domain) have been shown to activate the SCW program of xylem vessels [9, 10], SND1/NST1 (for secondary wall-associated NAC domain protein1/NAC secondary thickening promoting factor1) that of xylem fibres [11, 12, 13] and anthers , whereas NST1/NST2 control the SCW program of the anther endothecium . Orthologues of these genes have been identified in species such as poplar , rice , and maize . The non-vascular moss Physcomitrella patens has VND-related NACs with roles for thick cell wall formation . Some SCW NACs such as VNI1 (for VND-Interacting 1) and VNI2 can have repressor functions .
MYBs appear to play key roles as transcriptional activators in the middle tier of the SCW regulatory cascade both in angiosperms and gymnosperms. In Arabidopsis, the critical TFs are the partially redundant MYB46 and MYB83 [19, 20, 21] and this appears to also be the case in tree species [22, 23], rice, and maize . The top tier NACs and MYB46/83 coordinate the expression of downstream TFs that may be activators or repressors and include XND1, SND2, SND3, KNAT7, MYB103, MYB20, MYB42, MYB43, MYB52, MYB54, MYB69, and MYB85 ; reviewed by [1, 9].
The genes encoding enzymes of SCW biosynthesis have been investigated in most detail in Arabidopsis, but also in woody plants [2, 25, 26, 27, 28, 29, 30]. Of the ten known Arabidopsis CesAs, AtCesA4, AtCesA7, and AtCesA8 are considered important for SCW synthesis, whereas AtCesA1, AtCesA3, and AtCesA6 are essential for PCWs. Homologs of the SCW CesAs are known in many other species, such as cotton [31, 32], rice , poplar , and Brachypodium . Other proteins/enzymes involved in cellulose biosynthesis, structure, and deposition, including in SCWs [25, 26], include COBL4 (Cobra-like 4), CTL1 (chitinase-like protein 1), CTL2 (chitinase-like protein 2), TED6 (tracheary element differentiation-related 6), POM2/CSI (pom pom 2/cellulose synthase-interactive protein1), KOR (KORRIGAN), certain RLKs (receptor-like kinases) such as HERK1 (Hercules1) , FLA11 and FLA12 (fasciclin-like arabinogalactan proteins 11, 12) , FRA1 (fragile fibre 1) , and SuS (sucrose synthase) .
Lignins are composed of syringyl (S), guaiacyl (G), and/or p-hydroxyphenyl (H) units derived from the phenylpropanoid pathway [40, 41]. Lignin composition varies across cell types and genera, but the enzymes of the general and lignin-specific phenylpropanoid pathway are well conserved across many species and include phenylalanine ammonia lyase (PAL), cinnamate-4-hydroxylase (C4H), 4-hydroxycinnamate CoA ligase (4CL), hydroxycinnamoyl transferase (HCT), coumarate 3-hydroxylase (C3H), caffeoyl-CoA O-methyltransferase (CCoAOMT), ferulate 5-hydroxylase (F5H), caffeic acid O-methyltransferase (COMT), cinnamoyl-CoA reductase (CCR), and cinnamyl alcohol dehydrogenase (CAD) [42, 43]. Monolignols are oxidized in the cell wall by laccases or peroxidases and form lignin in a spontaneous process [40, 43, 44]. Transcriptional activators of lignin synthesis include MYBs such as AtMYB58, AtMYB63, and PtrMYB28  and other TFs. On the other hand, AtMYB4-, AtMYB7-, and AtMYB32-related TFs from different species have been shown to specifically repress lignin biosynthetic genes [46, 47, 48].
‘Typical’ SCWs also contain hemicelluloses , of which xylans such as glucuronoxylan (major dicot form) and glucuronoarabinoxylan (major gymnosperm and monocot form) are major components. Some genes encoding enzymes for xylan synthesis have been proposed .
Cotton seed fibres are long single cells surrounded by a PCW that elongates from the seed coat to reach a few centimetres in length before filling with a thick SCW. They are an exceptional example of a ‘highly cellulosic’ SCW being composed of ~94% cellulose [50, 51, 52] and with almost no lignin [53, 54]. Its SCW is so different to lignocellulosic SCWs that it is not surprising that the genes involved in its formation, or at least their expression, are different [55, 56]. On the other hand, the SCWs in cotton stem xylem are more likely to be similar to those in other plants. In a novel approach, that enables understanding of how the same genome can translate to entirely different cell walls, we have used cotton as a model to study global gene expression differences in a range of tissues and cells depositing compositionally different cell walls to understand how those compositions may be regulated at a transcriptional level. We also use state of the art NMR and biochemical analyses as well as microimaging to determine the compositional differences amongst xylem, pith, and seed fibre cells within the cotton plant. We discovered that cotton tissues contain a number of ‘atypical’ cell wall types that are regulated through modified hierarchical gene cascades that have diverged from those found in other plants.
Plant growth and tissue harvest
Gossypium hirsutum Coker 315–11 plants were grown in temperature controlled glasshouse conditions (31 °C 18 h day-time; 26 °C 6 h night-time) under natural summer daylight in Canberra, Australia. Plants were grown in pots containing soil and Osmocote fertiliser for approximately 10 weeks. Xylem and pith samples were hand-dissected from internodes 5, 6, and 7 (distance from cotyledons) from stem and flash-frozen in liquid nitrogen and stored at −80 °C; three biological replicates were harvested. Hand sections were checked using microscopy to ensure there was no cross-contamination from other tissues (Additional file 1). Seeds and attached seed fibres were harvested at 7 DPA (days post anthesis), pooled 14, 15, 16 DPA (average 15 DPA), and 25 DPA, from three biological replicates (2 bolls each), flash-frozen in liquid nitrogen, and stored at −80 °C. Seed fibres were separated from the seeds using tweezers and hand-ground to a fine powder using a mortar and pestle in liquid nitrogen. Appropriate time points for seed fibre harvest were identified based on in-house chemical analyses of a developmental series of seed fibre (Pettolino et al., submitted) and .
Freshly harvested stem segments were fixed in 70% (v/v) ethanol and cross-sectioned to approximately 120 μm. Sections were stained with toluidine blue for 1 min and washed twice with water. Sections were viewed and photographed under a Leica DMR upright fluorescence microscope using brightfield settings.
Cell wall polysaccharide composition from monosaccharide-linkage analysis
Ground seed fibre, pith, and xylem tissue was extracted with successive washes of 70% (v/v) ethanol (three times), chloroform and methanol (1:1), methanol and then acetone before drying to prepare an alcohol-insoluble residue (AIR) enriched in cell walls. AIR was de-starched using porcine pancreatic α-amylase (Sigma A3176), then carboxyl reduced (to assist in the determination of uronic acids and their methyl esters) before derivatisation by methylation, hydrolysis, reduction and acetylation according to Pettolino, Walsh . Deduced monosaccharide linkages, as determined by GC-MS, were grouped according to most likely polysaccharide assignments for summation of mol% of those derivatives to give estimates of the relative proportion of each polysaccharide in the sample .
Cell wall lignin content was determined as acetyl bromide lignin essentially as described  using alkali lignin from Aldrich (370959) as a standard. AIR was de-starched as above and protein depleted by successive washes with phosphate buffer, Triton X-100, sodium chloride, water, and acetone according to Moreira-Vilar, Siqueira-Soares .
RNA extraction, cDNA synthesis, and quantitative real-time PCR (qPCR)
RNA was extracted from xylem, pith, and seed fibre samples using a Qiagen RNeasy plant miniKit following the manufacturer’s instructions, with the exception that the RLT buffer was modified to include dithiothreitol (0.31% w/v; Boehringer), proteinase K (0.33% w/v; Amresco K525), polyvinylpyrrolidone-40 (2% w/v; Sigma PVP-40) and that the RLT-incubation step was performed for approximately 5 min at 40 °C to increase RNA yield and quality as assessed on a NanoDrop spectrophotometer (Thermo Scientific). Rnase-free DNase (Qiagen) treatment was performed on-column as recommended. Quality of RNA intended for sequencing was confirmed on a BioAnalyzer 2100 (Agilent Inc.).
cDNA synthesis was performed on 700 ng RNA using Superscript III (Invitrogen) following the manufacturer’s instructions and an oligo(dT)22V primer. qPCR was performed with four technical replicates using diluted cDNA, specified primer pairs (Additional file 2), Fast SYBR Green Master Mix (LifeTechnologies), and a 7900HT Fast Real-Time PCR System (Applied Biosystems) using comparative quantitation analysis against an internal reference gene (Gh_D_13g1487).
Individual sequencing libraries were prepared at the ACRF Biomolecular Resource Facility, The Australian National University, from total RNA from stem xylem and pith and seed fibre using in-house protocols. Each library had unique barcode adaptors and were pooled before sequencing across 5 lanes using Illumina HiSeq2000 according to the manufacturer’s protocol. All reads were aligned to the G. raimondii reference genome [61, 62] using TopHat  (version 2.0.9, default parameters). Aligned BAM files were further processed using Cufflinks  (version 2.1.1, default parameters) and Cuffmerge  was used to produce a non-redundant set of transcripts. Read counts were then generated for each gene in each sample using the intersectBed program  (BedTools version 2.16.2) and custom Perl scripts by using annotated gene locations. Unaligned reads were not further analysed.
Differential expression analysis was performed using edgeR  (version 2.6.12) between samples following the authors’ recommendations; read counts per gene were normalised using the trimmed-mean of M-values (TMM) method .
Gene set enrichment analysis was performed using G:Profiler [68, 69]. Benjamini-Hochberg  adjusted p value (false-discovery rate) thresholds of 0.05 were used to evaluate the overall number of differentially expressed genes and a p-value of <1e−6 to identify the more highly significant differentially regulated genes.
Results and discussion
Cotton plants have compositionally ‘atypical’ primary and secondary cell walls
We determined the cell wall composition, in particular the lignin and polysaccharide content, of stem xylem and pith (considered to be SCW and PCW tissue types, respectively) as well as seed fibres (a developmental series from PCW, transition to SCW and SCW deposition stages) to understand the variety of cell wall types present in cotton. Previous work had suggested that SCWs of mature seed fibres  and those of cotton stem pith  may be lignified, so we measured acetyl bromide lignin (ABL) in the cell walls from xylem, pith, and mature seed fibre and characterised the lignins by NMR (Fig. 1, Additional file 3). Stem xylem and pith from the same stem segments both contained high levels of lignin. Mature seed fibre, on the other hand, contained only low amounts of measurable ABL. NMR of mature and immature (25 DPA) seed fibre could not detect signals for guaiacyl nor syringyl lignin but could detect two peaks in aromatic regions that are consistent with the existence of p-hydroxphenyl units. If lignin does exist in cotton seed fibre it has an unusual composition in that it is not composed of the typical “G” and “S” subunits found in ‘normal’ dicot lignins. NMR indicated that the lignin in the pith tissue had a substantially higher S:G ratio (~70:30) than that in the xylem (~40:60). The higher S:G ratio of pith lignin was also reflected in its lignin structure as indicated by the larger peak of Aβ(S) than of Aβ(G) in the aliphatic region of the spectra.
Cell wall composition of cotton stem xylem, stem pith, and seed fibres
1.0 ± 0.1
4.0 ± 0.6
19.7 ± 5.0
13.7 ± 0.5
0.7 ± 0.4
0.6 ± 0.1
0.9 ± 0.02
4.3 ± 1.3
4.8 ± 0.2
0.4 ± 0.1
0.5 ± 0.2
0.7 ± 0.1
15.1 ± 1.8
8.6 ± 0.1
0.2 ± 0.1
Type I AG
0.3 ± 0.02
0.5 ± 0.1
2.9 ± 0.3
2.9 ± 0.1
0.1 ± 0.02
Type II AG
0.6 ± 0.1
0.4 ± 0.1
8.2 ± 0.2
8.2 ± 0.01
0.9 ± 0.01
27.6 ± 4.3
19.8 ± 7.1
2.9 ± 1.1
1.7 ± 0.1
0.6 ± 0.1
0.1 ± 0.001
0.1 ± 0.05
2.1 ± 0.5
12.2 ± 0.9
8.7 ± 2.5
3.4 ± 0.5
2.8 ± 0.1
5.6 ± 0.3
5.6 ± 0.4
1.4 ± 0.02
4.8 ± 0.6
4.1 ± 0.7
9.5 ± 1.4
10.6 ± 0.7
1.4 ± 0.5
61.0 ± 5.9
66.7 ± 7.3
23.4 ± 0.1
29.2 ± 0.5
85.6 ± 3.4
6.4 ± 0.2
2.6 ± 0.5
42.5 ± 3.5
51.0 ± 5.7
49.0 ± 5.7
42.5 ± 3.5
20.0 ± 0.00
Compositional differences between stem tissues and seed fibres are reflected in their tissue-specific transcriptomes
Differential expression comparisons were performed amongst the transcriptomes of the different tissues and cells to identify genes important for cotton SCW biogenesis (Fig. 2d). Approximately 12,500 genes were differentially expressed (at an adjusted p-value of 0.05) between the SF25 and stem xylem samples (Fig. 2d). Comparing SCW and PCW tissue types, approximately 7000 genes were differentially expressed between SF25 and SF07, and approximately 4400 genes between stem xylem and pith samples (Fig. 2d). Comparison of SF15 with SF07 or SF25 revealed about 3000 and 5000 genes differentially expressed at the transition to seed fibre SCW synthesis, respectively (Fig. 2d).
We examined expression levels (as FKPM – fragments per kilobase of million reads mapped) in more detail for the following differentially expressed gene classes: NACs, MYBs, WRKYs, auxin-related TFs, KNATs, BEL1-likes, bHLH TFs, cellulose synthase-related, phenylpropanoid pathway-related, and FLAs. Some but not all of these classes contain members that are considered key components of the current ‘typical’ SCW gene regulatory network.
Homologs of top tier NAC transcriptional regulators of secondary cell wall deposition show tissue-specific expression in cotton
In Arabidopsis the second or middle tier NAC, SND2, is considered to act downstream of SND1/NST1 in the xylem fibre SCW network [8, 21] and this appears to be similar in cotton seed fibre SCW development. The cotton SND2/3-related transcripts were observed in 25 DPA, but not in 7 DPA or 15 DPA seed fibre, whereas SND1-like expression was already induced in 15 DPA seed fibre (Fig. 3). One of those SND2-related NAC genes, Gorai009G166300, showed predominant expression in 25 DPA fibres and had the highest transcript level of any of the NAC-related genes across the seed fibre series. In Arabidopsis, SND2 is able to directly activate the CesA8 promoter  and some have suggested it specifically controls cellulose, mannan, and xylan biosynthesis, as well as lignin polymerization, but not monolignol synthesis , and thus Gorai009G166300 may have a similar role in cotton fibres (excluding lignin polymerization as that does not occur in the fibre SCW).
In Arabidopsis, VND-related TFs regulate SCW deposition in xylem vessels and are considered to be top tier TFs together with SND1 and NST1/2 [8, 9]. In our samples, VND-related genes showed expression mainly in xylem, but not much in seed fibre (Fig. 3). SND1- and NST1/2-related TFs are therefore likely to be top tier SCW NACs in seed fibre SCW biogenesis, whereas they may be accompanied by VND-related TFs in cotton stem xylem.
There was no consistent expression pattern for the VNI2/1-related genes in our samples (Fig. 3). However, two of the genes (Gorai005G195300 and Gorai012G125500) showed highest expression in xylem and pith, indicating that they have a specific function in stem. In contrast, Gorai007G065300 showed increasing transcript levels in seed fibre from 7 to 25 DPA, with lower expression in the stem samples. It has been suggested that VNIs can function as repressors of NAC function by interacting with other NAC TFs  and this may also be the case in cotton seed fibre.
In Arabidopsis and poplar, XND1 is thought to act as a repressor of xylem SCWs . In our cotton samples we observed moderate levels of expression of XND1-related genes in xylem, but they were completely absent from seed fibre (Fig. 3).
Members of several NAC groups, that have not previously featured in SCW regulation, were specifically up-regulated to sometimes high levels in cotton xylem compared to pith or seed fibres, and these include ANAC036/61- (Gorai001G122800, 006G113000), ANAC071/96- (Gorai003G180600, 008G227800) and ATAF1/2-related NACs (Gorai012G037600) (Fig. 3). These NACs are good candidates for future studies of cotton xylem SCW development. We also found some NAC-related TFs that were expressed highly in just pith compared to xylem or seed fibres, i.e., ANAC074 (Gorai001G254300, Gorai004G221300, Gorai007G147300) and CUC3/NAC1-related (Gorai009G354900) NACs. In contrast, few of the identified NAC-related genes showed particularly high expression in elongating or transition stage seed fibres. This indicates that NAC TFs, with the exception of the SND1-related genes, do not play a major role at those developmental stages.
Our data are in general agreement with the current model of SCW regulation in Arabidopsis as SND1-, NST1/2-, and VND-related NACs are all present in cotton cells undergoing SCW deposition and are amongst the early genes to be induced. However, a further level of detail might have to be added to the current model as our seed fibre developmental series strongly suggests that SND1-like TFs act earlier than NST1/2-related TFs in the SCW regulatory network. Additionally, many so far uncharacterised members of the NAC TF family displayed xylem-specific expression, suggesting possible roles for these genes in the lignocellulosic SCW gene regulatory network different to the highly cellulosic network found in seed fibre and these warrant further investigation.
Important ‘MYB master switches’ are missing in some cotton tissues with secondary cell walls
Transcripts of genes encoding putative MYB42/85 homologs were almost exclusively present in xylem, in agreement with their proposed function in Arabidopsis in activating transcription of genes in the lignin-specific part of the phenylpropanoid pathway . The absence of notable levels of MYB42/85-related gene transcripts and presumably also protein in cotton seed fibre provides at least a partial explanation for the low lignin content observed for these cells. One of the MYB42/85-related genes (Gorai006G064000) was expressed at moderate levels in pith. However, whether this is the reason for the surprisingly high lignin content of cotton pith cell walls requires further analysis. AtMYB58/63 have also been found to positively regulate lignin deposition in Arabidopsis . However, the expression levels of two putative MYB58/63-related genes (Gorai004G269400 and Gorai002G106300) were almost negligible in the lignin-containing xylem and absent from the other samples, including pith, indicating that MYB58/63 homologs do not play a major role in lignification of cell walls in cotton. We did identify five genes encoding putative homologs of AtMYB4/7/32, known repressors of lignin biosynthetic genes [46, 47, 48], that had moderate transcript levels in xylem and/or seed fibre, but no notable amounts in pith. The relatively high transcript level of the AtMYB4/7/32-related Gorai008G35700 in 25 DPA fibre could be another explanation for the absence of notable lignin deposition in seed fibres. On the other hand, transcript levels of MYB4/7/32-related genes in pith and xylem were not negatively correlated with the lignin level in the walls of those cells, which argues for a different function for these MYBs in cotton stems.
We identified only one gene with detectable expression that potentially encodes a MYB46/83 homolog (Gorai013G251100), with its transcript mainly present in the 25 DPA sample. No notable amounts of transcript were detected for any of the other genes that potentially encode MYB46/83 homologs (Gorai006G129100, Gorai001G138800, Gorai004G172700, Gorai004G037300). Arabidopsis AtMYB46/83 and homologs in several other plant species can activate the entire suite of SCW biosynthetic programs . The very low level of expression in cotton stem xylem with its more ‘typical’ SCW, suggests that there are ways of regulating composition of these cell walls other than those in Arabidopsis.
For many of the genes in the group we have classed as “other MYBs” we found tissue-specific expression patterns, suggesting specific functions in different cell types. The three genes encoding putative AtMYB48 homologs, for example, had the highest transcript levels for any MYBs across all samples. However, their expression was entirely absent from seed fibres, while only one of the genes was expressed in pith (Gorai001G107900). To our knowledge, AtMYB48 has not been previously associated with SCW formation, but our results strongly suggest at least a potential xylem-specific function for the cotton homologs.
Another player in the current ‘typical’ SCW regulatory network that is apparently missing in cotton is MYB75. Arabidopsis AtMYB75 is a positive regulator of anthocyanin synthesis and together with AtKNAT7 represses SCW synthesis . The cotton genome does not appear to encode a protein with significant similarity to AtMYB75, but it does encode at least 3 KNAT7 homologs (Additional file 8, C) two of which have their highest expression in 25 DPA seed fibres.
WRKY transcription factors are involved in stem, but not seed fibre SCW development
Other classes of TFs with differentially expressed genes across the comparisons were auxin/IAA-related, BEL1-like, and bHLH TFs (Additional files 6 and 8). We identified 21 auxin/IAA-related TF genes (Additional file 8, A), with many of them having high transcript levels in xylem and/or pith. Two genes, one IAA14/SLR- and one IAA30-related gene (Gorai006G246000 and Gorai007G117700, respectively) showed seed fibre-specific expression: Gorai006G246000 showed moderate expression at 7 DPA and highest transcript level at 15 DPA, whereas Gorai007G117700 was exclusively expressed at 15 DPA. These two TFs are potential candidates as activators of SCW synthesis at the seed fibre transition stage, but further experiments are required to test their function. In Arabidopsis, IAA14/SLR has been shown to negatively regulate lateral root initiation  and IAA30 has been shown to be specifically expressed in the quiescent centre of the root apical meristem , but their specific mode of action is still unclear. BEL1-like TFs were highly expressed in xylem and pith, but with no clear predominance in either tissue and some of the genes were also expressed in seed fibre, especially at 25 DPA (Additional file 8, B). Consequently, there is no clear bias in expression of these TFs in cells with PCW or SCWs. bHLH-related genes had very little expression in seed fibre compared to stem (Additional file 8, D). Two exceptions were Gorai002G016500 and Gorai009G176000 which were much more highly expressed than in xylem suggesting a seed fibre SCW-specific role. Cotton xylem had two bHLH genes, Gorai011G292100 and Gorai013G242800, with high levels compared to any of the other samples, suggesting that these may play a specific role in lignocellulosic SCW synthesis.
The cotton seed fibre is a ‘cellulose synthesis machine’
The cellulose synthase-like genes (CSLs) identified in this study include CSLA, CSLC, CSLD, CSLE, and CSLG. In contrast to the CesAs, the CSL transcripts were generally low in abundance in all samples. AtCSLA2 and AtCSLA9 reportedly act in the synthesis of glucomannan in Arabidopsis stems , but it seems that amongst the homologs detected here, at least one CSLA2 gene is early seed fibre-specific. The CSLA9 homologs were most highly expressed in stem xylem, consistent with their known role in Arabidopsis stems. There is evidence for the involvement of CSLCs in xylogucan biosynthesis  and we saw expression of CSLC homologs in 7 and 15 DPA seed fibre and the stem xylem samples, consistent with the presence of xyloglucans in these tissues (Table 1). CSLDs, including CSLD3, have been suggested to be involved in mannan biosynthesis . We saw expression of CSLD3-related genes at reasonably high levels in stem tissue, but lower levels in fibre, except for Gorai003G052200, which was high-moderately expressed across the tissue samples examined.
Several FLAs are likely to be key proteins in cell wall deposition and function in seed fibre
Several other genes associated with cellulose synthesis, structure, and deposition were also very highly expressed in 25 DPA seed fibre, including: COBL4-related (Gorai004G063600, Gorai007G176400), CTL2-related (Gorai006G078900, Gorai011G198500), FRA1 (Gorai002G134500), HERK1 (Gorai001G107000), and KOR1 (Gorai010G143300) (Additional file 9). Additionally, some putative xylan glycosyl transferase genes, such as IRX9-related (Gorai006G168500) and IRX10-related (Gorai005G197500), were also highly and specifically expressed in 25 DPA seed fibre.
Phenylpropanoid pathway genes are highly expressed in tissues with lignified SCWs
Curiously, the highest transcript level for one of the genes encoding a F5H homolog (Gorai.007G197700) was in 25 DPA seed fibre with no notable expression in any of the other seed fibre samples or pith and approximately ten times less in xylem, despite there being no substantial monolignol synthesis in this tissue. As mentioned earlier, F5H expression is directly activated by MYB103 in Arabidopsis . However, MYB103 has also been shown to regulate CesA8 expression , a regulatory step that most likely contributes to efficient cellulose synthesis in SF25 as discussed earlier. Similarly, we detected high levels for one CAD-related gene (Gorai003G049600) specifically in 7 and 15 DPA seed fibre. The roles of these genes in this tissues are unknown, but it is possible that both F5H and CAD enzymes, if produced, are used in metabolic pathways, such as in secondary metabolite or cuticle formation, rather than in monolignol production.
In contrast to the genes encoding enzymes of the lignin-specific phenylpropanoid pathway, there was seed fibre-specific expression of genes encoding enzymes of the flavonoid-specific pathway, with the highest transcript levels in elongating fibres at 7 and 15 DPA. We also identified at least one gene encoding each of the enzymes of the general pathway (PAL, C4H, and 4CL) that showed high transcript levels in at least one of the seed fibre samples. This suggests that these enzymes are providing the precursors for flavonol rather than for monolignol synthesis in seed fibre. Seed fibres indeed contain flavonoids and expression of flavonoid synthesis-related genes in seed fibre has been reported previously [55, 88]. The function of flavonoids in seed fibre has not been established, but it is generally accepted that flavonoids have various functions in plants including responses to abiotic and biotic stress. Suppressed development of seed fibre upon repression of F3H in cotton indicates, however, that flavonoids are necessary for seed fibre development .
Despite both having SCWs, cotton seed fibres and xylem tissues are very different in both cell wall composition and transcript profiles. Cotton seed fibre transcriptomes after about 15 DPA show that they become dedicated to the production of their main cell wall polysaccharide, cellulose, leading to their unusual highly cellulosic SCWs that are essentially devoid of any lignin at maturity. It has been suggested [55, 56] that SCW production during seed fibre development has been re-programmed during domestication and selection in modern breeding to suppress lignin biosynthesis whilst recruiting stress-response genes needed to achieve greater fibre cell length and make cotton fibres more useful for textiles, and this is evident in our transcriptome data.
We used stem pith as an example of a non-seed fibre PCW tissue, but discovered that cotton pith walls appear to be another example of an ‘atypical’ dicot wall. Histology and transcript analysis suggest they are PCWs, but compositional analysis detected significant levels of lignin and xylan normally characteristic of SCWs. Few studies have compared the chemistry of PCWs and SCWs from the same species and in the same part of the plant, so it is difficult to determine if this truly is unusual or specific to cotton.
Comparisons between different tissues within cotton have indicated that cotton SCW deposition, like that in Arabidopsis, is probably regulated by a hierarchical cascade of transcriptional activation and repression that regulates cell wall polysaccharide, lignin, and protein composition. Top tier factors like SND1/NST1 and VND, that are conserved between cotton and Arabidopsis, initiate cotton SCW development, but key MYB regulators in lower tiers are missing from seed fibres and in the other tissues are different to those prominent in Arabidopsis, explaining their ‘atypical’ cell wall compositions. Even stem xylem that contains a more ‘typical’ SCW, has a regulatory network that is different in detail to those described for other species.
The authors would like to thank Dina Yulia for her assistance with the cell wall analysis, and Ping hua He for providing RNA samples from floral and seedling tissues. CPM, HB, DL, and FAP would like to thank Cotton Breeding Australia for their support.
CPM was funded by Cotton Breeding Australia, a joint venture between Cotton Seed Distributors and CSIRO (Project No. CBA19). HB was funded by the CSIRO’s Office of the Chief Executive (OCE) Postdoctoral Fellowship program. YT and JR were funded in part by Stanford University’s Global Climate and Energy Program, and in part by the DOE Great Lakes Bioenergy Research Center (DOE BER Office of Science, DE-FC02–07ER64494).
Availability of data and materials
The dataset supporting the conclusions of this article is included within the article and its additional files.
CPM conceived the project, prepared the samples for RNAseq analysis, including histology, conceptualised the RNAseq experimental design with AC, performed the qPCR profiling, analysed data and was a major contributor in writing the manuscript. HB prepared samples for NMR and acetyl bromide lignin determination, analysed data and was a major contributor in writing the manuscript. AC conducted the RNAseq bioinformatics. EB provided samples and optimised protocols. YT and JR performed NMR analysis and contributed to writing the manuscript. ESD and DW analysed data and edited the manuscript. FAP performed monosaccharide-linkage analysis and acetyl bromide lignin determination, analysed data and was a major contributor in writing the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 3.Taylor-Teeples M, Lin L, de Lucas M, Turco G, Toal TW, Gaudinier A, et al. An Arabidopsis gene regulatory network for secondary cell wall synthesis. Nature. 2015;517:571–5.Google Scholar
- 7.Duval I, Lachance D, Giguère I, Bomal C, Morency MJ, Pelletier G, et al. Large-scale screening of transcription factor-promoter interactions in spruce reveals a transcriptional network involved in vascular development. J Exp Bot. 2014;65:2319–33.Google Scholar
- 10.Demura T, Tashiro G, Horiguchi G, Kishimoto N, Kubo M, Matsuoka N, et al. Visualization by comprehensive microarray analysis of gene expression programs during transdifferentiation of mesophyll cells into xylem cells. Proc Natl Acad Sci U S A. 2002;99:15794–9.Google Scholar
- 13.Mitsuda N, Iwase A, Yamamoto H, Yoshida M, Seki M, Shinozaki K, et al. NAC transcription factors, NST1 and NST3, are key regulators of the formation of secondary walls in woody tissues of Arabidopsis. Plant Cell. 2007;19:270–80.Google Scholar
- 17.Xu B, Ohtani M, Yamaguchi M, Toyooka K, Wakazaki M, Sato M, et al. Contribution of NAC transcription factors to plant adaptation to land. Science. 2014;343:1505–8.Google Scholar
- 18.Yamaguchi M, Ohtani M, Mitsuda N, Kubo M, Ohme-Takagi M, Fukuda H, et al. VND-INTERACTING2, a NAC domain transcription factor, negatively regulates xylem vessel formation in Arabidopsis. Plant Cell. 2010;22:1249–63.Google Scholar
- 22.Patzlaff A, McInnis S, Courtenay A, Surman C, Newman LJ, Smith C, et al. Characterisation of a pine MYB that regulates lignification. Plant J. 2003;36:743–54.Google Scholar
- 23.Goicoechea M, Lacombe E, Legay S, Mihaljevic S, Rech P, Jauneau A, et al. EgMYB2, a new transcriptional activator from Eucalyptus xylem, regulates secondary cell wall formation and lignin biosynthesis. Plant J. 2005;43:553–67.Google Scholar
- 31.Betancur L, Singh B, Rapp RA, Wendel JF, Marks MD, Roberts AW, et al. Phylogenetically distinct cellulose synthase genes support secondary wall thickening in Arabidopsis shoot trichomes and cotton fiber. J Integr Plant Biol. 2010;52:205–20.Google Scholar
- 32.Li A, Xia T, Xu W, Chen T, Li X, Fan J, et al. An integrative analysis of four CESA isoforms specific for fiber cellulose production between Gossypium hirsutum and Gossypium barbadense. Planta. 2013;237:1585–97.Google Scholar
- 34.Kumar M, Thammannagowda S, Bulone V, Chiang V, Han KH, Joshi CP, et al. An update on the nomenclature for the cellulose synthase genes in Populus. Trends Plant Sci. 2009;14:248–54.Google Scholar
- 35.Handakumbura PP, Matos DA, Osmon KS, Harrington MJ, Heo K, Kafle K, et al. Perturbation of Brachypodium distachyon CELLULOSE SYNTHASE A4 or 7 results in abnormal cell walls. BMC Plant Biol. 2013;13:131.Google Scholar
- 36.Pinard D, Mizrachi E, Hefer CA, Kersting AR, Joubert F, Douglas CJ, et al. Comparative analysis of plant carbohydrate active enzymes and their role in xylogenesis. BMC Genomics. 2015;16:402.Google Scholar
- 39.Brill E, van Thournout M, White RG, Llewellyn D, Campbell PM, Engelen S, et al. A novel isoform of sucrose synthase is targeted to the cell wall during secondary cell wall synthesis in cotton fiber. Plant Physiol. 2011;157:40–54.Google Scholar
- 44.Ralph J, Lundquist K, Brunow G, Lu F, Kim H, Schatz PF, et al. Lignins: natural polymers from oxidative coupling of 4-hydroxyphenyl- propanoids. Phytochem Rev. 2004;3:29–60.Google Scholar
- 46.Sonbol F-M, Fornalé S, Capellades M, Encina A, Touriño S, Torres J-L, et al. The maize ZmMYB42 represses the phenylpropanoid pathway and affects the cell wall structure, composition and degradability in Arabidopsis thaliana. Plant Mol Biol. 2009;70:283–96.Google Scholar
- 47.Legay S, Lacombe E, Goicoechea M, Brière C, Séguin A, Mackay J, et al. Molecular characterization of EgMYB1, a putative transcriptional repressor of the lignin biosynthetic pathway. Plant Sci. 2007;173:542–9.Google Scholar
- 48.Shen H, He X, Poovaiah CR, Wuddineh WA, Ma J, Mann DGJ, et al. Functional characterization of the switchgrass (Panicum virgatum) R2R3-MYB transcription factor PvMYB4 for improvement of lignocellulosic feedstocks. New Phytol. 2012;193:121–36.Google Scholar
- 51.Singh B, Avci U, Eichler Inwood SE, Grimson MJ, Landgraf J, Mohnen D, et al. A specialized outer layer of the primary cell wall joins elongating cotton fibers into tissue-like bundles. Plant Physiol. 2009;150:684–99.Google Scholar
- 52.Hernandez-Gomez MC, Runavot JL, Guo X, Bourot S, Benians TA, Willats WG, et al. Heteromannan and heteroxylan cell wall polysaccharides display different dynamics during the elongation and secondary cell wall deposition phases of cotton fiber cell development. Plant Cell Physiol. 2015;56:1786–97.Google Scholar
- 53.Macmillan CP, Birke H, Bedon F, Pettolino FA. Lignin deposition in cotton cells – where is the lignin? J Plant Biochem Physiol 1: e106.Google Scholar
- 54.Fan L, Shi W-J, Hu W-R, Hao X-Y, Wang D-M, Yuan H, et al. Molecular and biochemical evidence for phenylpropanoid synthesis and presence of wall-linked phenolics in cotton fibers. J Integr Plant Biol. 2009;51:626–37.Google Scholar
- 55.Tuttle JR, Nah G, Duke MV, Alexander DC, Guan X, Song Q, et al. Metabolomic and transcriptomic insights into how cotton fiber transitions to secondary wall synthesis, represses lignification, and prolongs elongation. BMC Genomics. 2015;16:1–28.Google Scholar
- 59.Moreira-Vilar FC, RdC S-S, Finger-Teixeira A, DMd O, Ferro AP, da Rocha GJ, et al. The acetyl bromide method is faster, simpler and presents best recovery of lignin in different herbaceous tissues than Klason and thioglycolic acid methods. PLoS One. 2014;9:e110000.Google Scholar
- 60.Kim H, Ralph J. Solution-state 2D NMR of ball-milled plant cell wall gels in DMSO-d6/pyridine-d5. Org Biomol Chem. 2010;8:576–91.Google Scholar
- 61.Paterson AH, Wendel JF, Gundlach H, Guo H, Jenkins J, Jin D, et al. Repeated polyploidization of Gossypium genomes and the evolution of spinnable cotton fibres. Nature. 2012;492:423–7.Google Scholar
- 62.Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res. 2012;40:D1178–D86. https://phytozome.jgi.doe.gov/pz/portal.html
- 67.Nookaew I, Papini M, Pornputtapong N, Scalcinati G, Fagerberg L, Uhlén M, et al. A comprehensive comparison of RNA-Seq-based transcriptome analysis from reads to differential gene expression and cross-comparison with microarrays: a case study in Saccharomyces cerevisiae. Nucleic Acids Res. 2012;40:10084–97.Google Scholar
- 70.Benjamini Y, Hochberg Y. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J Roy Stat Soc B Met. 1995;57:289–300.Google Scholar
- 72.Musel G, Schindler T, Bergfeld R, Ruel K, Jacquet G, Lapierre C, et al. Structure and distribution of lignin in primary and secondary cell walls of maize coleoptiles analyzed by chemical and immunological probes. Planta. 1997;201:146–59.Google Scholar
- 74.Lamesch P, Berardini TZ, Li D, Swarbreck D, Wilks C, Sasidharan R, et al. The Arabidopsis information resource (TAIR): improved gene annotation and new tools. Nucleic Acids Res. 2012;40:D1202–D10. www.arabidopsis.org
- 75.Hussey S, Mizrachi E, Spokevicius A, Bossinger G, Berger D, Myburg A. SND2, a NAC transcription factor gene, regulates genes involved in secondary cell wall development in Arabidopsis fibres and increases fibre cell area in Eucalyptus. BMC Plant Biol. 2011;11:173.CrossRefPubMedPubMedCentralGoogle Scholar
- 77.Öhman D, Demedts B, Kumar M, Gerber L, Gorzsás A, Goeminne G, et al. MYB103 is required for FERULATE-5-HYDROXYLASE expression and syringyl lignin biosynthesis in Arabidopsis stems. Plant J. 2013;73:63–76.Google Scholar
- 80.Bhargava A, Ahad A, Wang S, Mansfield S, Haughn G, Douglas C, et al. The interacting MYB75 and KNAT7 transcription factors modulate secondary cell wall deposition both in stems and seed coat in Arabidopsis. Planta. 2013;237:1199–211.Google Scholar
- 82.Ding M, Chen J, Jiang Y, Lin L, Cao Y, Wang M, et al. Genome-wide investigation and transcriptome analysis of the WRKY gene family in Gossypium. Mol Gen Genomics. 2015;290:151–71.Google Scholar
- 83.Nawy T, Lee J-Y, Colinas J, Wang JY, Thongrod SC, Malamy JE, et al. Transcriptional profile of the Arabidopsis root quiescent center. Plant Cell. 2005;17:1908–25.Google Scholar
- 84.Goubet F, Barton CJ, Mortimer JC, Yu X, Zhang Z, Miles GP, et al. Cell wall glucomannan in Arabidopsis is synthesised by CSLA glycosyltransferases, and influences the progression of embryogenesis. Plant J. 2009;60:527–38.Google Scholar
- 87.Kumar S, Kumar K, Pandey P, Rajamani V, Padmalatha KV, Dhandapani G, et al. Glycoproteome of elongating cotton fiber cells. Mol Cell Proteomics. 2013;12:3677–89.Google Scholar
- 89.Tan J, Tu L, Deng F, Hu H, Nie Y, Zhang X. A genetic and metabolic analysis revealed that cotton fiber cell development was retarded by flavonoid naringenin. Plant Physiol. 2013;162:86–95.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 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.