to BioTechniques free email alert service to receive content updates.
Separation of integument and nucellar tissues from cotton ovules (Gossypium hirsutum L.) for both high- and low-throughput molecular applications
CSIRO Plant Industry, Canberra, ACT, Australia
BioTechniques, Vol. 54, No. 1, January 2013, pp. 44–46
Full Text (PDF)
Supplementary Material
Protocol (.pdf)

Here we present a quick and low-cost method to separate the different layers of tissue from the ovules and young seeds of cotton (Gossypium hirsutum L.) for use in high- and low-throughput molecular applications. This method is performed at room temperature using standard laboratory equipment and does not require embedding of the samples, time-consuming fixation, or micro-sectioning procedures. We show that the three main tissues can be efficiently separated from isolated ovules collected on the day of anthesis. RNA and genomic DNA extracted from tissues separated by this method are of good quality and suitable for a variety of molecular applications to study the early stages of cotton seed and fiber development.

Cotton (Gossypium hirsutum L.) is the most economically important crop used in the textile industry worldwide because of the exceptional characteristics of its seed fibers. Moreover, the ability of these single-celled seed trichomes to become highly elongated and their exceptional biochemical compositions make cotton seed fibers an ideal model for studies of plant cell elongation and cell wall biogenesis (1). Seed fiber initials are epidermal cells that bubble out and grow from the outer layer of the ovule (i.e., outer-integument, OI) on the day of anthesis ± 1 day, suggesting that this tissue is more important than the underlying tissues (i.e., inner-integument, II, and nucellus, N) in the study of fiber development (2, 3). As such, analysis of the OI separated from the rest of the ovule and young seed is indispensable to elucidate the molecular mechanisms involved in the early stages of fiber development.

Until now, transcriptome and proteome studies published about cotton seed fiber initiation and development used the whole ovule when studying events occurring on or before the fibers can be mechanically removed from the seed (4, 5), thus pooling all tissues together and leading to possible misinterpretation of the results. Very few studies have used laser capture microdissection to isolate initial and pavement cells or fibers and inner integument tissues from early stage cotton seeds, probably because this technology requires access to complex equipment and provides very low recovery rates of macromolecules (6-8). Only one research article reported the isolation of the OI from ovules collected on the day of anthesis and analyzed it for gene expression levels (6); it clearly demonstrated that this tissue is instrumental in providing genes directly associated with fiber initiation and development (9-12). Unfortunately the method used to dissect the ovules was not described and the underlying tissues were not investigated (6). Here we present a quick and low-cost method to separate the different type of tissues from the ovules and young seeds of cotton (Gossypium hirsutum L.) in order to provide RNA, genomic DNA (gDNA) and potentially protein that can be used in low- and high- throughput molecular applications.

To demonstrate our method we show here that the three main tissues of the ovule can be efficiently separated from isolated ovules collected on the day of anthesis, but we were also able to work with -4 days post anthesis (dpa) ovules up to +4 dpa seeds (data not shown). A detailed working protocol including preparation of the working environment, ovule isolation and sequential pictures of the ovule dissection is available.

Method summary

Here we present a quick and low-cost method to separate the three main tissues from ovules and young seeds of cotton (Gossypium hirsutum L.) in order to provide RNA, genomic DNA, and potentially protein that can be used in low- and high-throughput molecular applications. Our method should enable studies on biological tissues relevant to the initiation and development of fibers as well as early seed development during the first hours after fertilization

Ovule microdissection is performed under a microscope allowing about 16× magnification (Leica Microsystems Pty., North Ryde, NSW, Australia). A single ovule is gently picked up with the forceps and placed onto a dry portion of a sterile Petri dish lid. In order to separate the OI from one ovule, hold the ovule in place with forceps in one hand and with the other hand hold a size 11 scalpel. The ovule is then positioned crest side up as in Figure 1A and 1B, and the forceps tips are located centrally to the ovule touching the ovule's left and right side. Next, a shallow incision is made in the OI along the crest running the blade from the chalazal end (CE) to the micropyle end (ME) (Figure 1A). Finally the OI is peeled away by slipping the scalpel tip under the OI at the incision site and proceeding to gently prise it apart from the II. Once an obvious flap of OI is free from the underlying II, the OI can be completely removed by (i) sliding the forceps tips under the OI on both sides of the incision by following along the surface of the II and pinching the forceps closed underneath pushing out the II with the N inside, or (ii) by continuing to peel the OI away from the II using the forceps to roll the ovule as you use the scalpel blade to separate the two layers in a scraping action. When the OI is separated, it is placed into a dedicated 250μL droplet of RNAlater (Ambion, Austin, TX, USA) to protect the integrity of RNA and gDNA. The same method is used to separate the II from the N; the tissues are collected into their dedicated droplets of RNAlater and transferred to an Eppendorf tube containing 500μL of RNAlater. Figure 1C shows a dissected dpa(0) ovule. In our hands, all ovules from one ovary (~15–25) can be dissected to obtain all three tissues in 40–50 min. When collecting OI only, the time is reduced to 15–20 min. Storage of the microdissected tissues are according to the RNAlater manufacturer's instructions.

Figure 1.  Ovule microdissection from cotton (Gossypium hirsutum L. cv. Coker 315–11) on the day of anthesis. (Click to enlarge)

Table 1, for comparison, shows the yield and quality of RNA and gDNA isolated from the three type of dissected tissues stored in RNAlater as well as that from whole ovules (WO) of dpa(0) ovaries processed immediately and not stored in RNAlater. RNA and gDNA yields were higher for WO compared with the dissected tissues, the nucellus having the lowest yield in both cases. RNA yield for OI and II was about the same (6–14μg), while gDNA yield was about twice as high for OI (5–8μg) compared with II. RNA quality assessed by the 260/280 and 260/230 ratios was very good and similar to that for WO (i.e., ≥2 meaning “pure” RNA) including for the nucellus, although it generally had a lower 260/230 ratio ranging from 1.57 to 1.99. RNA quality has also been assessed using a Bioanalyzer and gave very good RNA integrity numbers (RIN) for all dissected samples ranging from 8 to 8.9 (Supplementary Figure 1). Quality of gDNA from dissected tissues according to the 260/280 ratio was very good and similar to WO (╛1.8, meaning “pure” DNA), but was less consistent for the nucellus (Table 1), probably as a consequence of the low yields. Only the 260/230 ratio was found to be different between WO and dissected samples, suggesting to remove the presence of contaminants potentially related to the storage or extraction method. A gel image showing the high integrity of the isolated gDNA is presented in Supplementary Figure 1. Additional cleaning steps to improve the 260/230 ratio of the gDNA were investigated using the DNeasy kit (Qiagen) or sodium acetate precipitation, but the DNeasy kit gave strong yield reductions and neither method showed significant improvements in the ratio (data not shown).

Table 1.  Yield and quality of RNA and genomic DNA from whole ovules and dissected tissues collected on the day of anthesis. (Click to enlarge)


As a proof of concept of our dissecting method we measured the transcript level of GhMYB25Like, a transcription factor predominantly expressed in epidermal tissues during early seed fiber development (11). Figure 1D shows an equal transcript level between the WO and OI, with almost no expression in II and N, validating that the II was not contaminated by OI and that the isolated tissues are suitable for molecular analysis.

In conclusion, such a method of dissecting these three tissues from ovules and young seeds will allow researchers to place emphasis on the biologically active tissues relevant to the initiation and development of fibers as well as early seed development during the first hours after fertilization.


The authors are very grateful to Liz Brill for sharing her improvement steps used in RNA extraction from cotton ovules.

Competing interests

The authors declare no competing interests.

Address correspondence to Danny Llewellyn, CSIRO Plant Industry, P.O. Box 1600, Canberra ACT 2601, Australia. Email: [email protected]

1.) Kim, H.J., and B.A. Triplett. 2001. Cotton fiber growth in planta and in vitro. Models for plant cell elongation and cell wall biogenesis. Plant Physiol. 127:1361-1366.

2.) Ramsey, J.C., and J.D. Berlin. 1976. Ultrastructure of early stages of cotton fiber differentiation. Bot. Gaz. 137:11-19.

3.) Lee, J.J., A.W. Woodward, and Z.J. Chen. 2007. Gene expression in early events in fibre development. Ann. Bot. 100:1391-1401.

4.) Wang, Q.Q., F. Liu, X.S. Chen, X.J. Ma, H.Q. Zeng, and Z.M. Yang. 2010. Transcriptome profiling of early developing cotton fiber by deep-sequencing reveals significantly differential expression of genes in a fuzzless/lintless mutant. Genomics 96:369-376.

5.) Liu, K., M. Han, C. Zhang, L. Yao, J. Sun, and T. Zhang. 2012. Comparative proteomic analysis reveals the mechanisms governing cotton fiber differentiation and initiation. J. Proteomics 75:845-856.

6.) Wu, Y., A. Machado, R.G. White, D.J. Llewellyn, and E.S. Dennis. 2006. Identification of early genes expressed during cotton fibre initiation using cDNA microarrays. Plant Cell Physiol. 47:107-127.

7.) Wu, Y., D.J. Llewellyn, R.G. White, R. Ruggerio, Y. Al-Ghazi, and E.S. Dennis. 2007. Laser Capture Microdissection and cDNA microarrays used to generate gene expression profiles of the fibre initial cells on the surface of cotton ovules. Planta 226:1475-1490.

8.) Guan, X., J.J. Lee, M. Pang, X. Shi, D.M. Stelly, and Z.J. Chen. 2011. Activation of Arabidopsis seed hair development by cotton fiber-related genes. PLoS One 6:e21301.

9.) Machado, A., Y. Wu, Y. Yang, D.J. Llewellyn, and E.S. Dennis. 2009. The MYB transcription factor GhMYB25 regulates early fibre and trichome development. Plant J. 59:52-62.

10.) Zhang, F., K. Zuo, J. Zhang, X. Liu, L. Zhang, X. Sun, and K. Tang. 2010. An L1 box binding protein, GbML1, interacts with GbMYB25 to control cotton fibre development. J. Exp. Bot. 61:3599-3613.

11.) Walford, S.A., Y. Wu, D.J. Llewellyn, and E.S. Dennis. 2011. GhMYB25-like: a key factor in early cotton fibre development. Plant J. 65:785-797.

12.) Walford, S.A., Y. Wu, D.J. Llewellyn, and E.S. Dennis. 2012. Epidermal cell differentiation in cotton mediated by the homeodomain leucine zipper gene, GhHD-1. Plant J. 71:464-478.

13.) Wu, Y., D.J. Llewellyn, and E.S. Dennis. 2002. A quick and easy method for isolating good-quality RNA from cotton (Gossypium hirsutun L.) tissues. Plant Mol. Biol. Rep. 20:213-218.