Polygalacturonase gene expression in ripe melon fruit supports a role for polygalacturonase in ripening-associated pectin disassembly

Abstract

Ripening-associated pectin disassembly in melon is characterized by a decrease in molecular mass and an increase in the solubilization of polyuronide, modifications that in other fruit have been attributed to the activity of polygalacturonase (PG). Although it has been reported that PG activity is absent during melon fruit ripening, a mechanism for PG-independent pectin disassembly has not been positively identified. Here we provide evidence that pectin disassembly in melon (Cucumis melo) may be PG mediated. Three melon cDNA clones with significant homology to other cloned PGs were isolated from the rapidly ripening cultivar Charentais (C. melo cv Reticulatus F1 Alpha) and were expressed at high levels during fruit ripening. The expression pattern correlated temporally with an increase in pectin-degrading activity and a decrease in the molecular mass of cell wall pectins, suggesting that these genes encode functional PGs. MPG1 and MPG2 were closely related to peach fruit and tomato abscission zone PGs, and MPG3 was closely related to tomato fruit PG. MPG1, the most abundant melon PG mRNA, was expressed in Aspergillus oryzae. The culture filtrate exponentially decreased the viscosity of a pectin solution and catalyzed the linear release of reducing groups, suggesting that MPG1 encodes an endo-PG with the potential to depolymerize melon fruit cell wall pectin. Because MPG1 belongs to a group of PGs divergent from the well-characterized tomato fruit PG, this supports the involvement of a second class of PGs in fruit ripening-associated pectin disassembly.

Figure 6

A, Gel-diffusion assay of culture filtrates from A. oryzae transformed with MPG1 (XMPG1) or untransformed. Aliquots were removed from cultures at 24, 34, 50, 58, and 72 h and filtered through two layers of Miracloth, and equal amounts of protein were assayed for PG activity. B, Viscometric and reducing sugar assays of XMPG1 and untransformed culture filtrates. Culture filtrates from one time point were incubated for up to 6.5 h and assayed at multiple time points for the ability to decrease the viscosity or release reducing groups of a pectin solution. •, Untransformed viscosity; ▪, XMPG1 viscosity; and ⋄, XMPG1 reducing sugar.

Figure 1

Pectin-degrading activity in high-salt protein extracts from developing melon fruit. Extracts from fruit at six developmental stages of ripening (IG, MG, and R1–R4) were assayed viscometrically using 10% esterified citrus pectin. One unit was defined as the amount of enzyme that reduced the viscosity by 1% per hour. Each value is an average ± sd of three independent measurements. Low-salt and boiled (30 min) high-salt protein extracts did not cause a significant reduction in viscosity (data not shown). gfw, Grams fresh weight.

Figure 2

Melon genomic DNA gel-blot analysis of PG. Genomic DNA (5 μg/lane) was digested with EcoRI (E), BamHI (B), or HindIII (H). The blots were probed with the MPG1, MPG2, or MPG3 full-length cDNA or the PG4, PG5, or PG6 PCR fragments, and washed at a stringency of Tm −8°C.

Figure 3

RNA-blot analysis of PG RNA in developing melon fruit and nonfruit tissues. Each lane was loaded with 2 μg of poly(A⁺) RNA isolated from fruit tissues at six stages of development (IG, MG, and R1–R4) or 1 μg of poly(A⁺) RNA from roots (R), young leaves (L), stems (S), pistils (P), anthers (A), and fruit-abscission zones (AZ). The blots were probed with the MPG1, MPG2, or MPG3 full-length cDNA or the PG4, PG5, or PG6 PCR fragments, and washed at a stringency of Tm −8°C.

Figure 4

Sequence analysis of the MPG1, MPG2, and MPG3 cDNAs and alignment of their deduced amino acid sequences. Asterisks and dots indicate identical and conserved amino acid residues, respectively, between the MPG1, MPG2, and MPG3 sequences. Arrows indicate predicted signal sequence cleavage sites. Potential Asn glycosylation sites (N-X-S/T) in each sequence are underlined.

Figure 5

Phylogram for 19 plant and 1 fungal PG cDNA and genomic clones. The phylogram was generated from the alignment of the full-length deduced amino acid sequences described in Methods and was created using the PAUP software package (Swofford et al., 1990). The PG sequences segregate into three major clades that we have designated A, B, and C.