Non-redundant functions of two proline dehydrogenase isoforms in Arabidopsis

Abstract

Background: Proline (Pro) accumulation is a widespread response of prokaryotic and eukaryotic cells subjected to osmotic stress or dehydration. When the cells are released from stress, Pro is degraded to glutamate by Pro-dehydrogenase (ProDH) and Pyrroline-5-carboxylate dehydrogenase (P5CDH), which are both mitochondrial enzymes in eukaryotes. While P5CDH is a single copy gene in Arabidopsis, two ProDH genes have been identified in the genome. Until now, only ProDH1 (At3g30775) had been functionally characterised.

Results: We demonstrate vasculature specific expression of the Arabidopsis ProDH2 gene (At5g38710) as well as enzymatic activity and mitochondrial localisation of the encoded protein. Expression levels of ProDH2 are generally low, but increased in senescent leaves and in the abscission zone of floral organs. While sucrose represses ProDH2 expression, Pro and NaCl were identified as inducers. Endogenous ProDH2 expression was not able to overcome Pro sensitivity of ProDH1 mutants, but overexpression of a GFP-tagged form of ProDH2 enabled the utilisation of Pro as single nitrogen source for growth. Amongst two intronic insertion mutants, one was identified as a null allele, whereas the other still produced native ProDH2 transcripts.

Conclusions: Arabidopsis possesses two functional ProDHs, which have non-redundant, although partially overlapping physiological functions. The two ProDH isoforms differ with respect to spatial, developmental and environmental regulation of expression. While ProDH1 appears to be the dominant isoform under most conditions and in most tissues, ProDH2 was specifically upregulated during salt stress, when ProDH1 was repressed. The characterisation of ProDH2 as a functional gene requires a careful re-analysis of mutants with a deletion of ProDH1, which were so far considered to be devoid of ProDH activity. We hypothesise that ProDH2 plays an important role in Pro homeostasis in the vasculature, especially under stress conditions that promote Pro accumulation.

Figure 1

Proline metabolism in Arabidopsis. Schematic illustration of the current knowledge on Pro metabolism and its intracellular distribution. Black arrows indicate metabolic fluxes, enzymes are given in bold blue letters, for abbreviations see the main text. The molecular identity of mitochondrial and plastidic Pro and glutamate transporters is currently not known. Alternating localisation between cytosol and plastids was demonstrated for P5CS1 and P5CS2, the localisation of P5CR requires further detailed analysis. Export or leakage of GSA/P5C from mitochondria was postulated but experimental evidence is lacking.

Figure 2

ProDH2 expression complements a yeast Δput1 mutant. Yeast wildtype or Δput1 mutant with various ProDH expression constructs were grown for 4 d on minimal medium with 2% (w/v) galactose (Gal) as inducing carbon source and 10 mM Pro (upper plate) or 10 mM urea (lower plate) as the sole nitrogen source. Expression of ScPut1 confers wildtype growth to the Δput1 mutant. Native AtProDH pre-proteins do not complement the Pro utilisation deficiency of the Δput1 mutant. Replacement of the predicted Arabidopsis mTPs by the mTP of the yeast Sdh1 gene confers functional expression and the capability to metabolise external Pro as N-source (mTP-AtProDH1 and mTP-AtProDH2).

Figure 3

Tissue distribution of ProDH2 expression. A-C: ProDH2-promoter activity visualised in transgenic plants carrying a ProDH2-promoter-Gus fusion construct. Tissues were stained for 6 h in 1 mM X-Gluc and destained in 80% (v/v) EtOH. A: Two mature leaves of a 6-week-old plant. B: Three-week-old seedling showing enhanced staining with progressing leaf age. C: Inflorescence and young siliques showing GUS activity in the abscission zone between the pistils and the floral stalk. D: Northern blot analysis of transcript levels of Pro degradation genes in various tissues and two different cell cultures. The picture of the EtBr stained gel is shown to demonstrate RNA integrity and equal loading.

Figure 4

ProDH2 is induced by Pro and repressed by sucrose. Three-week-old wildtype (Col-8) or ProDH1-knockout mutant (pdh1-1) seedlings from sterile culture were transferred for 6 h to liquid half-strength MS medium with the indicated supplements prior to RNA extraction and northern blot analysis. The same membrane was hybridised consecutively with specific probes for ProDH2, P5CS1 and ProDH1. During the detection of P5CS1, no remainders of the ProDH2 specific signal were observed. The histogram above the northern blot panel indicates the intensity of the ProDH2-specific signal normalised to the level of Col-8 in medium with 30 mM sucrose.

Figure 5

p5cdh-2 mutants are more sensitive to external Pro supply than pdh1 mutants. A: Col-8 and Ler wildtypes, pdh1-1, pdh1-4, pdh2-1 and p5cdh-2 seedlings were cultivated for 2 weeks on MS medium with 5 mM Pro as the sole source of nitrogen and 15 mM sucrose as carbon source, buffered to pH 5.8 with 5 mM MES. B: Hypocotyl elongation of Col-8 wildtype, pdh1-1, pdh1-4 and p5cdh-2 seedlings cultivated on MS medium supplemented with 30 mM sucrose and varying concentrations of Pro. Error bars indicate SD of ≥ 10 seedlings per genotype. The experiments were repeated with similar results.

Figure 6

Knockout lines for ProDH1 and ProDH2 and T-DNA excision during splicing. A: Graphic representation of the exon-intron structure of the ProDH1 and ProDH2 genes with the position of T-DNA insertions in the analysed mutant lines (see material and methods section). Arrows with letters indicate the binding sites of primers used for PCR analyses in B and C. B: PCR on genomic DNA with two primer pairs, one for RbohD (At5g47910 as internal control, fragment size 1272 bp) and one ProDH1 or ProDH2 specific primer pair. Absence of ProDH specific PCR products demonstrates that all mutant plants were homozygous for the respective T-DNA insertion. C: RT-PCR analysis of ProDH1 and ProDH2 expression in the mutant lines demonstrates the presence of native transcripts in pdh1-4 and pdh2-2. PCR reactions with cloned cDNAs and genomic DNA (leftmost three lanes) demonstrate specificity of the PCR products. cDNAs for C and genomic DNA for B were obtained from the same samples, a slight contamination of the RNA samples with genomic DNA caused the additional amplification of intron-containing PCR products in the RT-PCRs.

Figure 7

Subcellular localisation of ProDH2. False colour images of protoplasts isolated from Arabidopsis plants stably transformed with a 35S:ProDH2:GFP fusion construct and stained with MitoTracker Orange. A: GFP fluorescence depicted in green; B: MitoTracker fluorescence depicted in red; C: Overlay of A and B demonstrating co-localisation of ProDH2-GFP with mitochondria. D: Overlay of a transmitted light picture with chlorophyll autofluorescence.

Figure 8

ProDH-GFP expression rescues the pdh1-1 mutant. Wildtype Col-8, pdh1-1 and pdh1-1 transformed with 35S:ProDH1:GFP or 35S:ProDH2:GFP were cultivated for 3 weeks on MS-N medium with 30 mM sucrose and 5 mM Pro as the only source of nitrogen. The transformants are segregating T₂ populations, therefore not all seedlings are able to utilise Pro as nitrogen source.