Comparative transcriptome analysis coupled to X-ray CT reveals sucrose supply and growth velocity as major determinants of potato tuber starch biosynthesis

Abstract

Background: Even though the process of potato tuber starch biosynthesis is well understood, mechanisms regulating biosynthesis are still unclear. Transcriptome analysis provides valuable information as to how genes are regulated. Therefore, this work aimed at investigating transcriptional regulation of starch biosynthetic genes in leaves and tubers of potato plants under various conditions. More specifically we looked at gene expression diurnally in leaves and tubers, during tuber induction and in tubers growing at different velocities. To determine velocity of potato tuber growth a new method based on X-ray Computed Tomography (X-ray CT) was established.

Results: Comparative transcriptome analysis between leaves and tubers revealed striking similarities with the same genes being differentially expressed in both tissues. In tubers, oscillation of granule bound starch synthase (GBSS) expression) was observed which could be linked to sucrose supply from source leaves. X-ray CT was used to determine time-dependent changes in tuber volume and the growth velocity was calculated. Although there is not a linear correlation between growth velocity and expression of starch biosynthetic genes, there are significant differences between growing and non-growing tubers. Co-expression analysis was used to identify transcription factors positively correlating with starch biosynthetic genes possibly regulating starch biosynthesis.

Conclusion: Most starch biosynthetic enzymes are encoded by gene families. Co-expression analysis revealed that the same members of these gene families are co-regulated in leaves and tubers. This suggests that regulation of transitory and storage starch biosynthesis in leaves and tubers, respectively, is surprisingly similar. X-ray CT can be used to monitor growth and development of belowground organs and allows to link tuber growth to changes in gene expression. Comparative transcriptome analysis provides a useful tool to identify transcription factors possibly involved in the regulation of starch biosynthesis.

Figure 1

Proposed pathway of starch metabolism in photosynthetic and non-photosynthetic tissue.

Figure 2

Diurnal starch and sucrose content of leaves over a sixteen hour period. A) Starch and B) sucrose content. Error bars indicate standard deviation (n = 3).

Figure 3

Diurnal expression of genes known to be involved in starch biosynthesis. A) Import of glucose 6-phosphate and ATP into the plastid and the conversion thereof to ADP-Glucose. GPT1 (light blue), GPT2 (dark blue), PGM (orange), NTT1 (brown), AGPase LS (light green), AGPase SS (dark green). B) Starch synthases and branching enzymes. SSII (dark grey), SSIII (light grey), GBSS (blue), SSIV (black), SBEA (light purple), SBEB (dark purple), ISA1 (dark red) and ISA2 (pink). C) Photosynthetic and Calvin cycle related genes. Plastocyanin (pink solid), CAB ( pink dotted), Rubisco (blue dotted), TPT (blue solid). D) Sucrose cleavage and phosphorylation. CW-Inv (olive green), FK (blue), Susy (gold), HK1 (dark yellow) and HK2 (green) Values are the mean of two replicates.

Figure 4

Relative expression of GPT and NTT in epidermal and whole leaf tissue. A) GPT and B) NTT. Error bars represent standard deviation (n = 3).

Figure 5

Expression of genes known to be involved in starch biosynthesis during tuber induction. A) Import of glucose 6-phosphate and ATP into the plastid and the conversion thereof to ADP-Glucose. GPT1 (light blue), GPT2 (dark blue), PGM (orange), NTT1 (brown), AGPase LS (light green), AGPase SS (dark green). B) Starch synthases and branching enzymes. SSII (dark grey), SSIII (light grey), GBSS (blue), SSIV (black), SBEA (light purple), SBEB (dark purple), ISA1 (dark red) and ISA2 (pink). C) Sucrose cleavage and phosphorylation. Cw-Inv (Black), FK (blue), Susy (gold), HK1 (Dark yellow) and HK2 (green). Values are the mean of two replicates.

Figure 6

Quantitative real time PCR confirmation of microarray results. A-C) Relative expression of GPT2, Susy4 and GBSS diurnally in leaves. D-F) Relative expression of GPT2, Susy4 and GBSS in tubers. Error bars represent standard deviation (n = 3).

Figure 7

Diurnal expression of genes known to be involved in starch degradation. GWD (black), DPE1 (green), DPE2 (light blue), MEX1 (red), Alpha amylase (grey), Isoamylase 3 (dark blue), PCT-BMY (brown) and SEX4 (orange). Values are the mean of two replicates.

Figure 8

GBSS relative expression and stolon sucrose content at different time-points of the day. A) Diurnal expression of GBSS in tubers. B) GBSS expression at the 24 hour time-point from plant grown in light/dark cycle and from plants kept in twenty four hours of darkness. C) Stolon sucrose content at different time-points of the day (dark grey bars) and at the same time-points from plant kept in darkness from 0 hours onward (dark grey bars). Two values were not determined (n.d.). Error bars represent standard deviation (n = 3-7).

Figure 9

Schematic scheme illustrating how tuber volume is measured using X-ray CT. Potato plants are scanned with X-ray in a chamber containing an X-ray beam and a two dimensional detector. After projecting X-ray images of the potato plant on the detector at different angles, the projections are reconstructed in silico to create a three dimensional image. From this image the tuber volume can be calculated.

Figure 10

Potato tuber segmentation and volume calculation. A) Two dimensional X-ray images illustrating that potato tubers can be distinguished from the surrounding soil. The histogram shows that segmentation is possible. The red bar indicates the grey level threshold selected. B) Linear regression of X-ray calculated and real volume measurements confirming the accuracy of X-ray CT calculated volumes.

Figure 11

Estimated growth velocity of tubers in cubic centimetre volume increase per day. Tubers chosen for microarray hybridisation are marked by arrow heads

Figure 12

Relative expression of starch biosynthetic genes in tubers growing at different velocities. GPT1 (pink), GPT2 (yellow), ISA2 (orange), AGPase LS (purple), AGPase SS (brown) and Susy4 (green). Values are the mean of two replicates.

Figure 13

Comparative analysis of transcription profiles. A) Venn diagramm showing 1662 features differentially expressed under all conditions selected B) K means clustering showing 5 clusters (A-E) with three clusters having the desired pattern of expression for starch biosynthetic genes (B-D). C) Functional assignment of features present in clusters B-D).

Figure 14

Expression profiles of transcription factors possibly regulating starch biosynthesis. A) Diurnally in leaves, B) during tuber induction and C) growing and non-growing tubers. Micro.8007 (AT2g40820), orange, Micro.7865.c1 (AT3G16280), blue, Micro.5635.c1 (AT4G37750), black, Micro.5579.c2 (AT4G34590), light blue, Micro.15471.c1 (AT4G00870), brown and Micro.4326.c1 (AT4G32730), red.