Comparative analysis of the heat stable proteome of radicles of Medicago truncatula seeds during germination identifies late embryogenesis abundant proteins associated with desiccation tolerance

Abstract

A proteomic analysis was performed on the heat stable protein fraction of imbibed radicles of Medicago truncatula seeds to investigate whether proteins can be identified that are specifically linked to desiccation tolerance (DT). Radicles were compared before and after emergence (2.8 mm long) in association with the loss of DT, and after reinduction of DT by an osmotic treatment. To separate proteins induced by the osmotic treatment from those linked with DT, the comparison was extended to 5 mm long emerged radicles for which DT could no longer be reinduced, albeit that drought tolerance was increased. The abundance of 15 polypeptides was linked with DT, out of which 11 were identified as late embryogenesis abundant proteins from different groups: MtEm6 (group 1), one isoform of DHN3 (dehydrins), MtPM25 (group 5), and three members of group 3 (MP2, an isoform of PM18, and all the isoforms of SBP65). In silico analysis revealed that their expression is likely seed specific, except for DHN3. Other isoforms of DNH3 and PM18 as well as three isoforms of the dehydrin Budcar5 were associated with drought tolerance. Changes in the abundance of MtEm6 and MtPM25 in imbibed cotyledons during the loss of DT and in developing embryos during the acquisition of DT confirmed the link of these two proteins with DT. Fourier transform infrared spectroscopy revealed that the recombinant MtPM25 and MtEm6 exhibited a certain degree of order in the hydrated state, but that they became more structured by adopting alpha helices and beta sheets during drying. A model is presented in which DT-linked late embryogenesis abundant proteins might exert different protective functions at high and low hydration levels.

Figure 1.

Two-dimensional electrophoresis of the HS proteome of NG radicles excised from M. truncatula seeds that were imbibed for 16 h using 24 cm nonlinear immobilized pH gradient strips (3–10). pI and molecular mass (in kilodaltons) are noted. Numbers indicate the spots that were identified.

Figure 2.

The relation between DT and the water content of the radicles of M. truncatula at different intervals during fast drying. Germinated seeds exhibiting a protruded radicle of 2.8 and 5 mm were first incubated or not (control) in a PEG solution for 2 d then dried for different intervals. The data from two independent experiments are pooled together.

Figure 3.

Phylogenic tree of LEA proteins of M. truncatula. ClustalX was used to create an alignment of the following translated tentative consensus that was found in the Medicago database (http://www.tigr.org/tdb/tgi/mtgi). The alignment was bootstrapped (n = 1,000 replicates) to create the final tree with the bootstrap values indicated. Underlined LEA proteins indicate those that have been found associated with DT (see Table III). BudCar5 (TC 100264 homolog of Q9M603 of Medicago sativa); CapLEA-1a (TC94389 similar to 049816 of Cicer arietinum); CapLEA-1b (TC112317 similar to 049816 of C. arietinum); DHN (dehydrin; TC101013 similar to 023957 of G. max); DHN3 (or dehydrin like, TC 100921, similar to Q945Q7 pea); DHN-Cog (dehydrin cognate, TC106659 similar to Q43430 of pea); DIP (drought induced protein, TC95389 similar to Q941N0 of R. raetam); ECP31 (TC 96862 similar Q96245 of Arabidopsis); Em6 (TC96799 homolog to P93510 of R. pseudoacacia); Lea14 (TC 101891, similar to P46519 of G. max); LEA5 (TC105834 similar to O24422 of G. max), LEA D34a (TC102411 weakly similar to P09444 LEA D34 of G. hirsutum); LEA D34b (combination of TC107159 and −158, both similar to P09444 LEA D34 of G. hirsutum); LEA likea (TC108292 weakly similar to Q6Z4J9 of rice); LEA likeb (TC 95012 similar to Q8LCW6 of Arabidopsis); PM32 (mitochondrial LEA, TC101811 and TC78559 similar to Q5NJL5 of pea); MP2 (maturation polypeptide; TC95538 and TC87025 similar to Q39871 of G. max); MtPM25 (this study); PM1 (TC96465 similar to Q01417 of G. max); PM10 (TC100258, TC85220 similar to Q39801 of G. max); PM18 (or 35 kD seed maturation protein, TC96265 similar to Q9ZTY1 of G. max); PM22 (translated BQ124186 similar to Q9XER5 of G. max); PM24 (TC96521 weakly similar to Q9SEL0 of G. max); RAB21 (TC 107383 similar to P12253 of rice); SBP65 (seed biotinylated protein, TC102224 similar to Q41060 of pea); SMP LEA-4 (TC94509 similar to Q9FNW8 of G. tabacina). Whenever possible, the classification according to Dure (1993) and Cuming (1999) together with the PFam assignments are indicated. The scale bar (number of substitutions/site) corresponds to the relative branch length.

Figure 4.

Changes in three HS proteins identified as PM25 (A and B), Em6 (C and D), and MP2 (E and F) associated with DT in germinating radicles of M. truncatula. Representative 2D gels (A, C, and E) and relative spot quantities (B, D, and F) during germination (white bars) and after a PEG treatment (black bars) in 2.8 and 5 mm long emerged radicles. Spot numbers are taken from Figure 1. pI and molecular mass (in kilodaltons) are indicated on the left gel. For MP2, the histogram corresponds to the most preponderant isoform (spot 19). Different letters shown above the bars represent significant differences after multiple comparison of the means (P < 0.05).

Figure 5.

Changes in spots identified as PM18 (A and B) and SBP65 (C and D) that are associated with DT in germinating radicles of M. truncatula. Representative 2D gels (A–C) and relative spot intensities (B and D) corresponding to PM18 (B, spots 15–18) and SBP65 (D, spots 1–5) during germination (white bars) and after a PEG treatment (black bars) in 2.8 and 5 mm long radicles. Spot number assignment is taken from Figure 1. pI and molecular mass (in kilodaltons) are indicated on the left gel. Different letters shown above the bars represent significant differences after multiple comparison of the means (P < 0.05).

Figure 6.

Changes in several spots identified as DHN3 (A and B) and Budcar5 (C and D). Representative 2D gels (A) and relative spot intensities (B) during germination (white bars) and after a PEG treatment (black bars) in 2.8 and 5 mm long emerged radicles. Spot number assignment is taken from Figure 1. pI and molecular mass (in kilodaltons) are indicated on the left gel. Different letters shown above the bars represent significant differences after multiple comparison of the means (P < 0.05).

Figure 7.

In silico expression analysis of LEA genes in M. truncatula identified in this study (see Table III). The number of ESTs present in various libraries that were found at http://www.tigr.org/tdb/tgi/mtgi and exhibited similar characteristics (organs, developing seeds, stress, etc.) were averaged and expressed as percent of detected EST × 1,000. The absence of value means that the EST was absent in the libraries. The corresponding libraries are: leaf (T10110, 4046, 4049), flower (#9D5, #ARC), nodulated root (T1617, T10109, 4047), seed (#ARD, T10493, T10494, T11127), drought (5413), biotic stress (#A8V, T11031, T1581, T10014), N deficit (#8GI, 5518, #8GF), P starvation (5415), and plant microbe interaction (T1748, 7263, T10173, T1815, T1510, T1707, T1682, #ARE, #CDE, 5520).

Figure 8.

Western-blot analysis of MtPM25 (A–D) and MtEm6 (E–H) in relation to DT: in radicles (A and E) and cotyledons (B and F) during seed imbibition; in emerged radicles (C and G) having a length of 2.8 and 5 mm, before and after the PEG treatment; and in embryos (D and H) during the acquisition of DT during seed maturation (10–40 DAP). Twenty (embryos and radicles) and 50 μg (cotyledons) of soluble proteins were hybridized with rabbit antibodies against recombinant MtPM25 and MtEm6. Independent protein extractions and immunoblots were performed in triplicates and yielded identical results. Percentage of DT and molecular marker mass (kilodaltons) are indicated. DS, Dry seeds.

Figure 9.

FTIR absorption spectra in the amide region of detagged, recombinant MtPM25 and MtEm6. Conditions of the proteins were: A, hydrated in D₂O; B, after fast drying (FD) in an air stream of 3% RH; and C, after slow drying (SD) in circulating air of 67% RH.

Figure 10.

Curve-fitting procedure illustrated for the recombinant MtEm6 after fast drying in an air stream of 3% RH. The absorption maxima of the different protein secondary structures in the amide-I band were selected on account of peak positions in the second-derivative spectrum (A). With these different components a least square iterative curve fitting was performed to fit Voigt line shapes to the original spectrum between 1,720 and 1,600 cm⁻¹ (B). Prior to curve fitting a straight base line passing through the ordinates at 1,720 and 1,600 cm⁻¹ was subtracted. Coaddition of all of the dashed peaks that were mathematically produced resulted in a fit (crosses) that resembled the original absorption spectrum (gray line).