PROLINE ACCUMULA nON IN GRAPE BERRIES 95 Vvp5cs encodes an 82.6 kDa protein with homology to P5CS sequences from other organisms, for example, 79% amino acid identity with A. thaliana ATP5CS-l and 76%
amino acid identity with V. aconitifolia P5CS (Hu et aI., 1992, Savoure e/ aI., 1995). It also shares significant homology with the Escherichia coli genes encoding GK and GPR, confirming that it encodes both domains of the bi-functional plant enzyme. Vvoat en- codes a 48 kDa protein with homology to OATs from several organisms (Stines, 1999), for example 72% amino acid identity with A. thaliana OAT (Roosens et aI. , 1998).
Insertion of the cDNAs into the pET14-b vector (Novagen) enabled production of recombinant VVP5CS and VVOA T proteins in E. coli (Stines, 1999; Stines et aI., 1999).
Whilst most of the recombinant protein produced was insoluble, in both cases the recovery of soluble protein could be enhanced by incubation of the E. coli cultures at 16°C rather than 37°C (Fig. 4.3). The recombinant proteins were purified by affinity chromatography and SDS-P AGE and used to generate polycJonal antibodies in rabbits.
Southern hybridisation analyses indicated ã that both the P5CS and OAT encoding genes are present as single copies in the V. vinifera genome. Studies of gene expression by Northern and Western blot hybridisation demonstrated that Vvp5cs is expressed in grape berries as well as in a range of other grapevine tissues (Fig. 4.4) (Stines, 1999).
A B
kDa Unind Ind kDa
94 .... 94
67 67
43 43
30
30 20
20 14
Figure 4.3. A) Expression of recombinant VVP5CS in E. coli grown and induced at 37°C resulted in predominantly insoluble VVP5CS protein which was purified by affinity chromatography and SDS-PAGE, then used to produce polyc\onal antibodies. Unind (protein from uninduced cultures), lnd (protein from induced cultures). B) Soluble recombinant VVP5CS was purified from E. coli cells induced and grown at 16 °C. This preparation could be used to determine the activity of the VVP5CS enzyme. Source: Stines (1999).
96 R. V AN HEESWIJCK et ai.
A PU SK SE FB FL YL ML RO kOa
B PU SK SE FB FL YL ML RO WP5CS~ _-ã ~ .... --
94 68 43 30
94 68
Figure 4.4. Relative levels of VVP5CS in a range of V vinifera cv Chardonnay tissues. A) Protein was extracted from pulp (PU), skin (SK) and seeds (SE) of mature berries, flower buds (FB), flowers (FL), young leaves (YL), mature leaves (ML) and roots (RO) and subjected to SDS- PAGE. B) Western blot hybridisation was performed on a replica gel using anti-VVP5CS antibodies. Source: Stines (1999).
Expression of Vvoat could not be detected by Northern blot hybridisation indicating that it is not very highly expressed in the tissues examined. Subsequent use of RT-PCR however enabled detection of Vvoat transcripts in berries, as well as in flowers, leaves and roots (Fig. 4.5) (Stines, 1999). This was confirmed by Western blot hybridisation which further demonstrated that in cv Chardonnay, there was little, if any, VVOAT protein present in berry skin.
OA T activity was readily detected in prepared grape berry extracts using buffers speci-
FB YL FL RO BE
WOAT~ ... - - 210 b
Figure 4.5. RT-PCR analyses enabled detection of Vvoat mRNA in a range of V vinifera cv Chardonnay tissues. Total RNA used in the analyses was extracted from flower buds (FB), young leaves (YL), flowers (FL), roots (RO) and mature berries (BE). (-) indicates the control reaction (no RNA template). Source: Stines ( 1999).
PROLINE ACCUMULA nON IN GRAPE BERRlES 97 fically designed for tissues containing high levels of phenolic compounds (Ford and H0j, 1998; Stines, 1999). The assay is based on the production ofP5C and its reaction with ninhydrin under hot acidic conditions (Kim et at., 1994). A red pigment is formed, the absorbance of which can be measured at 510 nrn. Unfortunately, the direct measurement of VVP5CS activity in grape berry extracts has not yet been possible, presumably due to technical problems similar to those experienced in other studies on plant P5CS enzymes (Kishor et aI., 1995; Zhang et al., 1995). The activity ofthe soluble recombinant P5CS was however readily assayed according to the method of Garcia-Rios et al. (1997) where the A TP and glutamate dependent conversion of NADPH to NADP is measured by a decrease in absorbance at 340 nrn. Assays of recombinant VVP5CS demonstrated that whilst it is subject to feedback inhibition by proline, it is relatively insensitive to this form of regulation, requiring 25 mM proline to achieve 50% inhibition in the presence of 5mM glutamate (Stines et aI., 1999). This suggests that VVP5CS can retain considerable activity even in mature grape berries where proline concentrations only rarely reach such concentrations.
Together with the cloning and gene expression results described above, it is clear that the capacity for de novo proline synthesis via both the glutamate and ornithine pathways exists in grape berries. What needed to be determined, however, is whether the regula- tion of these pathways and the "cross-talk" between them affects the spatial and temporal patterns of proline accumulation in grape berries. The information and tools gained from the isolation and characterisation of the Vvp5cs and Vvoat cDNAs and recombinant pro- teins provided an excellent basis for investigating these questions.
6. Vvp5cs GENE EXPRESSION DURING GRAPE BERRY DEVELOPMENT
As reviewed above, previous studies of proline accumulation in response to abiotic stress have demonstrated that P5CS is the key regulatory enzyme of the glutamate pathway of proline biosynthesis and that enhanced proline biosynthesis occurs primarily due to tran- scriptional activation of the P5CS gene. Attention has recently turned to elucidating the role of P5CS in proline accumulation in developing plant organs. In kiwi fruit (Actinidia deliciosa), the glutamate pathway of proline biosynthesis and transcriptional activation of p5cs gene expression was found to play a significant role in the accumulation of proline in buds during bud-break (Walton et al., 1997). The Atp5csl gene of A. thaliana has been shown to be strongly expressed in reproductive organs and tissues (flower buds, pollen and pistils) which accumulate proline (Savoure et at., 1995; Yoshiba et al. 1999).
In contrast, the transcript level of tompro2, one of the two p5cs genes in tomato, was found to be lowest in pollen, the tissue which contains the highest amount of proline (Fujita et at., 1998). Hence, the mechanisms regulating proline accumulation during normal plant development are still not well understood.
In order to examine the expression of Vvp5cs during grape berry development, North-
98 R. VAN HEESWIJCKef al.
em blot analysis was perfonned on RNA prepared from V vinifera cv Chardonnay fruit sampled from four weeks postflowering to maturity. The steady-state levels of Vvp5cs mRNA were observed to remain relatively constant throughout grape berry development (Stines et aI., 1999). There were no significant changes in Vvp5cs mRNA levels which would parallel the significant increase in proline accumulation late in berry development.
Thus, unlike the numerous reports on stress-induced proline accumulation in vegetative tissues, the primary basis for proline accumulation in the case of grape berry develop- ment does not appear to be a temporally programmed induction of p5cs gene expression.
Virtually all of the published infonnation regarding p5cs gene expression in plants is restricted to analyses at the mRNA level. Therefore it is not clear whether the transcrip- tional activation of p5cs under osmotic stress is also accompanied by an increase in its protein levels. A direct relationship between mRNA and protein is assumed by most re- searchers, however, post-transcriptional and post-translational regulation mechanisms are an important and sensitive means of controlling cellular metabolism. To investigate the role of post-transcriptional control of Vvp5cs in the regulation of proline synthesis and accumulation in developing grape berries, Western blot analyses were perfonned on protein extracts derived from fruit of a number of different V vinifera cultivars at differ- ent stages of development. The results for cvs Chardonnay and Gewurztraminer are illustrated in Figure 4.6. Similar results were obtained for cvs Cabernet Sauvignon and Muscat Gordo Blanco (Stines, 1999). VVP5CS protein is present throughout berry de- velopment in all cultivars and there is no apparent difference in pattern of expression between cultivars which accumulate relatively high levels of proline (Chardonnay and Cabernet Sauvignon) and cultivars which accumulate more moderate levels of proline (Gewurztraminer and Muscat Gordo Blanco). In particular, the dramatic increase in free proline observed late in berry development in cultivars such as Chardonnay and Caber- net Sauvignon is not associated with a concurrent increase in steady-state levels of VVP5CS protein. Quantitative western blot analyses enabled the amount of VVP5CS protein present in mature berries of Chardonnay and Gewurztraminer to be detennined (data not shown). A direct comparison demonstrated that there were no significant dif- ferences in the amount of VVP5CS protein which might correlate with the different lev- els of proline accumulated.
Taken together, the above results demonstrate that proline accumulation late in grape berry development is independent of changes in steady-state levels of both Vvp5cs mRNA and protein. This suggests that the mechanisms regulating proline accumulation during nonnal plant development are quite different to those operating during the abiotic stress response. Clearly there are a number of points of regulation other than simple changes in Vvp5cs mRNA and protein levels which could account for the temporal pat- tern of proline accumulation in developing grape berries. The western blot analysis of different grapevine tissues (Fig. 4.4) indicates that the skin and pulp of grape berries contain relatively high levels of P5CS protein compared with other tissues such as seeds and leaves. This suggests that the capacity for proline synthesis, in terms of enzymic protein, remains relatively high in berry tissue throughout development. Other factors which
PROLINE ACCUMULA TlON IN GRAPE BERRIES A
B
WP5CS ~ - - -
kDa 94 68 43 30 21
94
C weeks postflowering
4 8 10 12 13 14 16
o
68 WP5CS
99
14
Figure 4.6. Steady-state levels of Vvp5cs protein throughout development of V vinifera cv Gewurztraminer (A,B) and cv Chardonnay (C,D) beFfies: A,C. Proteins extracted from whole berry homogenatcs were separated by SDS-PAGE. Each lane contained an extract from equivalent amounts of berry homogenate on a fresh weight basis. B,D. Western blot hybridisation was performed on replica gels using anti-VVP5CS antibodies. Source: Adapted from Stines (1999) and Stines et al. (1999) Plant Physiology 120: 923-931 with permission from the American Society of Plant Physiologists.
might modulate the levels of proline accumulation could, for example, control P5CS substrate supply, or change the flux through metabolic pathways such as those involved in proline degradation.