The malic and tartaric acid contents of grape berries display different accumulation pat- terns throughout the double sigmoidal growth cycle of the fruit (Fig. 2. J). Immature grape berries appear as highly specialised in the storage of unsalted organic acids when compared to other plant organs. Large vacuoles are present in the peri carp as soon as 2 days after anthesis (Harris et at., 1968), while cell division is very intense. There is a nine-fold increase of the mean volume of the peri carp cells between anthesis and the onset of ripening (Ojeda et at., 1999), due to intense vacuolar enlargement based on the storage of malic, tartaric and citric acid as major osmotica. Tartaric acid is more rapidly accumulated shortly after anthesis, while cell division still occurs at a significant rate.
However, malic acid becomes two to three - fold higher during the first growth period before ripening is triggered (i.e. Steffan et aI., 1975). Unsalted free acids (0.45 ± 0.1 Eq) are one order of magnitude more concentrated than cations at this stage, and it is possible that this difference is even more marked at the cellular level since potassium is preferen- tially located in some subepidennal cells near the vascular bundles (Storey, 1987). The pH of the juice is lower than the dissociation constants of both acids and lies between 2.3 and 2.7 throughout the whole growth period, before veraison. A noticeable exception to the-
Fresh Weight ~-- -;;;-
0.60 Sugars / ãE .,
,... l.0 ~~
8 ., u - Malic acid ãE 0
u ãE .. 0
<1l .. Tartaric acid .&>0
0.40 o~
U 0 0 gu
Z 0 ~o
< .. ~6
0 .. "- 0.5
0:: .;- • V)
0 !:!:. 0.20 ~o:: . <
t.L.O ::>
:. V)
0 20 40 60 80 100 120
Ti me (days)
Figure 2.1. Typical changes in fresh weight, sugars, malic and tartric acid content during grape berry development.
GRAPE BERRY ACIDITY 37 se general rules is Gora chirine, a five fold less acidic cultivar of V vinifera, which compensates such impairment in the osmotic potential of acids by the precocious storage of 0.2 M glucose before ripening (Diakou et aI., 1997).
As in many fruits, the malic and citric acid content per berry diminishes very rapidly once sugar accumulation is triggered (Fig. 2.1). Malate catabolism is specifically associ- ated with vascular tissues, creating a diffusion gradient in the surrounding tissues (Coombe, 1992). Although it is not a general rule, the water content and cellular volume of many cultivars roughly double while 1M hexoses are accumulated during ripening.
The resulting dilution largely explains the decrease in tartaric acid concentration during ripening and the level of this acid is almost constant when expressed on a per berry ba- sis. Berries fed with labelled tartaric acid emit a negligible amount of 14C02, unlike ob- servations made using labelled malate or glucose (Steffan et aI., 1975). Thus, tartrate is often more concentrated than malate once berries have reached maturity. In contrast with both organic acids, the storage of potassium persists during ripening (i.e. Coombe, 1992).
The rate of malate breakdown and the final pH at harvest depend on the cultivar, root- stock, nutritional status and temperature, while the effect of environmental factors on tartrate content is far less obvious. Conditions leading to vigorous vines and high yields generally result in a higher malate concentration at harvest (Champagnol, 1984). Increas- ing K t nutritional status strongly increases the malate content of mature berries (Fig. 2.2, recalculated from Delas et at. , 1989); however, this correlation becomes visible only during ripening, as if K+ acted as a protector against malate catabolism (Hale, 1977).
Cluster grown at 20DC exhibited 2 to 3 times more malic acid in their berries than clusters
..-, 140 0 140 ,...,
0- Tar = 147 -1.08K + r2=0.33 0 0 0-
~ E 120 • 0 120 ~
'-' Q) . 0 '-'
• Q)
~ . ... • •
~
~ 100 • " !I!. • 100
:,s ... ! ....
E-<
80 • 80
0 0 & •
60 0 • 60
30 35 40 45 50 55 60
K + (mM)
Figure 2.2. Relationship between potassium, malic and tartaric acid contents in grape berries (from Delas et al., 1989).
38 N. TERRIER and C. ROMIEU
grown at 30°C (Kliewer and Lider, 1968). Nevertheless, the variations depending on environmental factors are low when compared to that triggered by berry development.
2.2. Organic acid metabolic pathways in grape berries 2.2.1. Organic acid synthesis
The pathway of malate synthesis in grapes does not differ from the classical one in plants; malate is formed by ~ carboxylation of PEP, according to the general equation C6Hl206 +2C02 ~ 2 malate + 4 H+ (Fig. 2.3). Omitting the last step of vacuolar storage, this pathway permits a net conversion of hexose to malate without any changes in the redox state of the NAD pool, the energy charge of the adenylate pool, and the Pi concen- tration (Latzko and Kelly, 1983; Lance and Rustin, 1984). This pathway is strongly retroinhibited by protons and malate as reaction products, due to the regulatory proper- ties of the key enzyme PEPcase (Diakou, 1999). Of the CO2 formed by respiration, 75%
is recaptured by photosynthesis in immature berries exposed to the light, so that the compensation point is never reached in green berries (Koch and Alleweldt, 1978). It is therefore widely accepted that most organic acids are formed at the expense of imported sucrose. In this respect, clusters grown in the dark are c.a. 20% smaller than those ex- posed before veraison, and photosynthesis would contribute to less than 1 % of the dif- ference in dry weight (Dokoozlian and Kliewer, 1996). Reports on the intensity of dark CO2 fixation by excised immature berries were somewhat contradictory. Some meas- urements have suggested that it predominates over respiration (Lakso and Kliewer, 1978), yet Koch and Alleweldt (1978) could not confirm this despite using the same methods.
The biosynthetic pathways of tartaric acid formation, which is not common in plants, has remained a matter of debate (RUffner, 1982a). Recent results showed that virtually all tartaric acid found in berries is synthesised from glucose via ascorbic acid (Saito,
1994; Loewus, 1999).
2.2.2. The induction a/malate respiration during ripening
Respiration measurements have been conducted for more than a century on grape berries (Peynaud and Ribereau-Gayon, 1971). The grape berry is classified as a non-climacteric fruit since it does not display a significant increase in respiration during ripening. It was well established that the Qr is significantly above I during the malate breakdown period (Saulnier-Blache and Bruzeau, 1967; Harris et aI., 1971; Koch and Alleweldt, 1978).
This fundamental result suggested to many authors that the initiation of malate break- down was the consequence of its oxidation in place of imported sugars (Peynaud and Ribereau-Gayon, 1971). Experiments with labelled substrate confirmed increased malate oxidation during ripening. However, pulses with labelled sugar, malate and CO2 also unambiguously showed that the malate/sugar interconversion pathways remain re- versible in berries at both developmental stages (Hardy, 1968; Steffan et aI., 1975).
Moreover, the in vivo activity of PEPcase assayed with 14C02 increases during ripening,
GRAPE BERRY ACIDITY
Cytoplasm
TP PPi
M itochondria
, ,
Vac uo le
Organic acid channel
\if ,
ME Malate
AOA i
P EPCarbOXY-/
kinase ! t£Pca,~
neoglucogenesi \ lyCOIysis ugar
Figure 2.3. Metabolism of malate in grape berry cells.
39
and the relative labelling of sugars and malate is the opposite of their net accumulation throughout the developmental cycle (Table 2.1 recalculated from Meynhardt, 1963).
This classical result suggested that the nature of the respiratory substrate shifts according to
40 N. TERRIER and C. ROMIEU
the relative size of sugar and malate metabolic pools, as governed by changes in vacuo- lar compartmentation. Similar conclusions could be drawn from the results of enzy- mological studies, which showed a lack of correlation between PEPcase activity and malate accumulation, or between ME and PEPcarboxykinase activities and acid decrease (Hawker, 1969; Lakso and Kliewer, 1975; RUffner and Kliewer, 1975). These enzy- mological results were confirmed during the last decade using different enzyme extrac- tion procedures, immunological and molecular biology approaches (Guttierez-Granda and Morrisson, 1992; Taureilles-Saurel et aI., 1995; Franke and Adams, 1995; Diakou, 1999). It was therefore proposed that compartmentation control and known regulatory properties, such as differential enzyme activity, feedback inhibition and energy charge could explain the fluctuations in malic acid concentration during berry ripening (RUffner, 1982b).
Table 2.1. Incorporation of 14C02 in malic acid and sugars in grape berries after 6 hours of incuba- tion at 25°C (recalculated from Meynhardt, 1963).
Treatment Light Dark
Maturity of berri es green red black green red black
Tota1 14C 93344 90114 30561 5377 5438 6502
Incorporated
14C Malate/sugars 0.14 0.24 1.61 0.13 0.32 11.1
2.2.3. Aerobicfermentation and malate breakdown
Many quantitative inconsistencies were found between respiration measurements, in detached berries, and the evolution of major solutes during fruit development. Several cultivars exhibited a Qr greatly exceeding the 1.33 value of malate oxidation. Moreover, such elevated values were still observed at room temperature when malate breakdown practically ceases once maturity is reached (Koch and Alleweldt, 1978). Excessive Qr values could be attributed neither to tartrate breakdown, which in practice never occured at significant rate, nor to neoglucogenesis, since its inhibition at elevated temperature was incompatible with the strong increase ofQr (Steffan et aI., 1975).
The oxidation of organic acids was studied in vitro on purified grape mitochondria (Romieu and Flanzy, 1988; Romieu et al.,1992). Substrate breakdown by the TCA cycle was cyanide sensitive and activated by ATP demand. However, the oxidation of (J..-
ketoglutarate and succinate was progressively inhibited during measurements, and the in vitro oxidation of malate alone was surprisingly low. Such inhibition disappeared when the oxaloacetate formed was removed by transamination or citrate formation in the pres- ence of exogenous pyruvate. The accumulation of an inhibitory concentration of ox- aloacetate resulted by the absence of ME in the grape mitochondrial matrix. This ab- sence was unexpected in a fruit using malate as the major respiratory substrate, since this enzyme is generally present in plant mitochondria. As a result, high malate breakdown triggered during ripening should not depend only on energy demand, but also on pyru-
GRAPE BERRY ACIDITY 41 vate formation in the cytosol, which raises the importance of the regulation of cytosolic ME by acidic pH (Fig. 2.3).
Recent results suggest that the pathway of malate breakdown and carbon entry in the TCA cycle is more complex. Excised grape berries exhibited a significant rate of aerobic fermentation (Romieu et al., 1992) that strongly increased above room temperature and played a major role in the increase of the Qr with temperature (Terrier et aI., 1995). Adh was actually induced to a considerable level during the malate breakdown period and displayed a ripening related expression pattern in strong contrast with that of ME, PEP- case and PEPcarboxykinase, which clearly appeared as housekeeping enzymes (Sauvage et aI., 1991; Terrier et al., 1995). This induction relied on the activation of a specific Adh isogene of a small multigenic family in V vinifera (Tesniere and Verries, 2000). Never- theless, aerobic fermentation, and perhaps Qr, would have been promoted upon harvest since undetached berries readily oxidised 14C-ethanol and did not accumulate this prod- uct beyond the millimolar range (Romieu, unpublished data).
PDC is active throughout the whole berry developmental cycle (Hawker, 1969). This uncommented result suggests that the PDH bypass, recently documented in plants (Ta- dege et al., 1997) could provide an additional route for carbon entry in the TCA cycle.
The induction of Adh indicated that ethanol recycling would become critical for efficient oxidation of malate in ripening berries. Cytosolic malate breakdown and ethanol forma- tion are key reactions typically involved in the prevention of lethal cytosolic acidosis in plant tissues submitted to environmental stress (Roberts et aI., 1984; Rivoal et aI., 1991).
In this respect, the marked pH dependence of ME and Adh extracted from berries (Latzko and Kelly, 1983; Molina et al., 1986) brings up the importance of the pH stat function of the reaction 2H+ + malate ~ ethanol during ripening (Fig. 2.3). This proton scavenging pathway could allow plant survival during transient malic acid decompart- mentation in excess of the oxidative capacity of the TCA cycle.
Therefore, the induction of Adh constitutes another indirect argument in favour of an impairment of acid compartmentation as the prime cause for the initiation of malate breakdown in berries.