MOLECULAR BIOLOGY OF SUGAR AND ANTHOCYANIN ACCUMULATION IN GRAPE BERRIES
3.3. Grape anthocyanin gene expression
3.3.1. Anthocyanin gene expression in grapevine seedlings
The first grape cDNA clones coding for anthocyanin structural gene were isolated by Sparvoli et al. (1994). They used heterologous probes from maize and snapdragon to screen a cDNA library constructed using total RNA extracted from 14-day-old grape seedlings grown in continuous light for 48 hours. Full-length cDNA clones that showed homology to chs, chi, j3h, dfr and ldox genes were obtained and partial clones of pal and ufgt were also isolated (Sparvoli et aI., 1994). In order to study the pattern of induction of these genes during anthocyanin synthesis in grape tissues, Sparvoli et al. (1994) ex- posed etiolated grapevine seedlings to continuous light, which induces anthocyanin pro- duction. It was found that pal was constitutively expressed in seedlings, whereas the mRNA levels of chs, chi, j3h, djr, ldox and ufgt were dramatically increased upon expo- sure to light (Fig. 1.6). The data suggests that exposing seedlings to light can induce the anthocyanin pathway in grapevine and that genes are coordinated, like the anthocyanin genes in maize where all the genes from chs down are upregulated together. Dark grown seedlings also showed a low basal level expression of these six genes which was thought to be involved in the biosynthesis of other flavonoids (Sparvoli et al., 1994).
3.3.2. Anthocyanin gene expression in berry skins during development
When northern analysis of the grapevine anthocyanin genes was carried out on Shiraz berry
PAL
CHS
CHI
F3H
DFR
LDOX
UFGT
SUGARS AND ANTHOCY ANINS IN GRAPE BERRIES 19
D L L L L
o (; 12 24 48
Figure 1.6. Northern analysis of anthocyanin gene expression in grapevine seedlings. Seedlings were grown in darkness (D) and then exposed to light (L) for 6, 12,24 and 48 hours. Total RNA extracted from the seedlings was northern blotted and probed with the grapevine cDNA clones encoding pal, chs, chi, f3h, dfr, ldox and ufgt. Reproduced from Plant Molecular Biology, Sparvoli et al. "Cloning and molecular analysis of structural genes involved in flavonoid and stilbene biosynthesis in grape (V vinifera L.)", 24: 743-755: 1994.
20 P.K. BOSS and C. DAVIES
skin samples taken throughout development the results suggested that the grape anthocya- nin pathway is controlled in a manner different from that seen for maize, petunia and snap- dragon (Boss et aI., 1996a). In grape berry skins during development there were two phases of anthocyanin pathway gene expression (Fig. 1.7). Early in development all the genes except ufgt were expressed followed by a reduction in the expression of all the genes during the lag phase of grape berry development (Fig. 1.7). Then, after veraison, there was a co-ordinate induction of all the genes from the pathway including ufgt and this coincided with the accumulation of anthocyanins in the skin. These results suggest that ufgt is under a different regulatory regime than the genes from the rest of the anthocyanin pathway. It appears that ufgt plays an important role in the accumulation of anthocyanins in grape berry skins. However, the gene encoding the putative dehydratase that is thought to cata- lyse the reaction between ldox and ufgt (Fig. 1.5; Heller and Forkmann, 1988) has not been cloned and could be controlled in a manner similar to ufgt. Genes involved in transporting the anthocyanins into the vacuole, whose functions are required after ufgt in the anthocya- nin pathway, may also be controlled in the same manner as ufgt. Therefore, it is perhaps more correct to state that the control of anthocyanin biosynthesis in grapevines occurs be- yond the ldox step of the anthocyanin biosynthesis pathway.
The link between ufgt expression and anthocyanin biosynthesis in berry skins has also been demonstrated by retarding the ripening process through the application of the synthetic auxin-like compound BTOA to grape berries (Davies et aI., 1997). When expression of ufgt was analysed in the treated and untreated samples, it was found that ufgt was only expressed in those berries producing anthocyanins. In the BTOA-treated samples the onset of UFGT expression was delayed by four weeks, which paralleled the delay in other ripening parame- ters caused by the treatment (Davies et ai., 1997). Ford et ai. (1998) were able to produce recombinant ufgt protein and raise antibodies to this protein. It was demonstrated that the presence of this protein correlated strongly with the quantitative ability of the berries to glu- cosylate anthocyanidins but not with the ability to glucosylate quercetin (Ford et ai., 1998).
This is important because it demonstrates that this gene encodes a UDP-gIucose: anthocya- nidin 3-0-glucosyltransferase and not a flavonoid 3-0-glucosyltransferase and therefore should only be present when anthocyanins are being synthesised.
Why are the early genes of the pathway expressed before veraison? It is possible that these flavopoid pathway genes are expressed in tissues not producing flavonoids. How- ever, it is more likely that other flavonoid compounds are being produced in the flower and early berry samples. Aurones, flavones, flavonols, isoflavonoids and proanthocyanid- ins are all produced as the result of biosynthetic pathways branching from the anthocya- nin pathway as is shown in Figure 1.5. Flavonols are essential for pollen viability in maize and petunia (Mo et aI., 1992) and are known to accumulate in the ovaries ofpetu- nia flowers (Koes et aI., 1990) and are perhaps being synthesised in grape flowers as well.
Proanthocyanidins are known to be present in young berries (Boss, 1998) and the produc- tion of these compounds requires the expression of the anthocyanin genes up to and in- cluding 4fT. The puzzle that remains is the observation that ldox was expressed in the flower and early berry samples. The product of this enzyme is presumed to be an unstable intermediate in the conversion of leucoanthocyanidins into anthocyanidins and the reason
A
B
SUGARS AND ANTHOCY ANINS IN GRAPE BERRIES 21
s ..
I
'!;,.. -:
H-• r x" ...
U ... ..
~ .
,. J
..
~ t,
. . . . . t4I " t~ "
- - fIo-"'oI
FL 2 4 6 8 10121416
PAL
CHS CHI F3H
DFR
LDOX - ....
UFGT
Figure 1.7. Total anthocyanin accumulation and northern analysis of anthocyanin gene expression during the development of grape berry skins. A) Changes in total anthocyan ins per gram fresh weight of berry skin (top) and changes in total soluble solids in the berry juice, measured as °Brix (bottom) during the development and ripening of Shiraz grape ber- ries. The vertical, dotted line represents verai- son. B) Northern blots of total RNA from grape flowers (FL) and grape berry skin sam- ples taken at fortnightly intervals throughout development, probed with grape cDNA clones for pal, chs, chi, j3h, dfr, ldox and ufgt. The numbers indicate weeks postflowering at which the RNA was extracted from berry skins. Reproduced from Plant Physiology, Boss et al. "Analysis of the expression of an- thocyanin pathway genes in developing V.
vinifera L. cv Shiraz grape berries and the im- plications for pathway regulation", Ill : 1059- 1066: J 996, with permission from the Ameri- can Society of Plant Physiologists.
for the presence of its mRNA in these samples is unknown.
The coordinate induction of the anthocyanin pathway genes after veraison suggests that the pathway in grapevines is tightly controlled by regulatory genes, as has been ob- served in other plant species. The two patterns of expression seen in berry skins pre- and post-veraison (i.e. with and without ufgt) result in the tissue either producing antho- cyanins or not. The pathway appears to be controlled beyond the step catalysed by LDOX, later in the pathway than seen in maize, petunia and snapdragon.
3.3.3. Anthocyanin gene expression in red and white grapes
White grape varieties are thought to have arisen from red- or black-skinned varieties (Slinkard and Singleton, 1984) and it is quite likely that the white grape has arisen in de-
22 P.K. BOSS and C. DAVIES
pendently a number of times from different red or black varieties. The mutations could be different in each case and to investigate the possible nature of these mutations, ex- pression of the anthocyanin pathway genes was compared in a number of white and red or black varieties (Boss et aI., 1996b; 1996c). Transcripts of all of the anthocyanin struc- tural genes below ehs were detected in the red or black varieties studied (Fig. 1.8; Boss et aI., 1996c). Most of the white grape varieties tested showed moderate levels of ex- pression of all the genes except ufgt, which could not be detected in any of the white berry skin samples (Fig. 1.8). However, Muscat Gordo and Sultana were unusual in that all the genes were expressed at very low or non-detectable levels.
The consistent difference between the white and red or black grapes was that the ex- pression of ufgt was not detected in any of the white grapes (Fig. 1.8; Boss et aI., 1996c).
Southern analysis showed that all the white varieties tested contained at least part of the ufgt gene (Boss, 1998). Thus the lack of ufgt gene expression in the white-berried varie- ties seems to be due to a lack of transcription of the gene and not simply an absence of ufgt in the genome. It is possible that the ufgt promoter has been mutated in some way in each of these white-skinned varieties which has resulted in a loss of ufgt gene expres- sion.
Perhaps the loss of ufgt gene expression in the white-skinned varieties is due to the absence of a regulatory gene product. In many instances, lack of production of antho- cyanin pigments is caused by mutations in regulatory genes resulting in decreased ex- pression of a number of structural genes in the pathway. A similar picture is apparent in the skins of white grapes, where there were lower levels of expression of most of the genes in the flavonoid synthesis pathway (Fig. 1.8). This suggests that alterations of regulatory genes may have occurred in these varieties resulting in decreased synthesis of a number of enzymes of the pathway, preventing accumulation of anthocyanins. The two patterns of expression shown by the white-skinned varieties suggest that there may be more than one regulatory factor involved in regulating expression of the grapevine an- thocyanin structural genes. With the pattern seen in Riesling, Semillon and Chardonnay, it could be that a regulatory factor vital for ufgt expression, which also has a qualitative effect on the rest of the pathway, is not present or active. However, Muscat Gordo and Sultana have most of the pathway repressed, which may be due to mutations in another regulatory factor which is necessary for the activation of gene expression for most of the early pathway genes as well as ufgt.
Alternatively, the reduction in expression of the early pathway genes in the white- skinned varieties may represent a feedback inhibition of these genes, when a product pro- duced further down the pathway accumulates. This has been shown to occur at an early step in the phenylpropanoid pathway where a reduction in C4H activity resulted in an accumulation of trans-cinnamic acid and a reduction in PAL activity (Blount et ai., 2000). However, the product of LDOX activity has not been identified as it is extremely unstable and thus unlikely to accumulate.
It is surprising that there were no cases where the expression of an early gene (for ex- ample ehs) alone was absent, or where all the genes were expressed, but anthocyanin syn- thesis was inhibited by a mutation in an early gene resulting in a loss of function of the resu-
CHS CHI
F3H
DFR LDOX
UFGT
SUGARS AND ANTHOCY ANINS IN GRAPE BERRIES 23
Figure 1.8. Northern analysis of anthocya- nin gene expression in the skins of various grapevine varieties. Total RNA was iso- lated from the skins of ripe berries from the varieties indicated at the top of the figure. Northern blots of the RNA were probed with grape cDNA clones for chs, chi,j3h, dfr, ldox and ufgt. Reprinted from Plant Molecular Biology, Boss et al. "Ex- pression of anthocyanin biosynthesis pathway genes in red and white grapes", 32: 565-569: 1996.
!tant enzyme. This may reflect natural selection against these mutants, as they may lack essential, flavonoid based compounds, or selection against such white-skinned grape- vines by viticulturalists, as they may be less useful for wine-making or have other unde- sirable properties, such as decreased disease resistance. Another reason for the lack of ufgt expression in white-skinned grapes could be because this gene is only induced when its substrate begins to accumulate. 1t is possible that mutations in early structural genes exist that abolish the function of the protein but not the expression of the gene. If ufgt is only induced when its substrate accumulates, its expression would not be seen in these circumstances.
Mutations in grapevines that restore anthocyanin production in the skins of white ber- ries, have also been studied with regards to anthocyanin genes (Boss et aI., 1996c). In many cases ufgt was again the only anthocyanin gene differentially expressed between red and white wine varieties. This is an interesting phenomenon as it suggests that, if white-skinned varieties have derived from pigmented varieties, it is possible to regain the anthocyanin forming capability. However, the anthocyanin profiles of these "rever- tant" varieties often Jack acylated anthocyanins, possess a high cyanid in :malvidin ratio, and have Jess total anthocyan ins (Boss et al., 1996c). As a result, their colour is gener- ally lighter (red rather than black) than that of the usual red-wine varieties. The identifi-
24 P.K. BOSS and C. DAVIES
cation of the nature of the mutations causing the loss and regain of anthocyanin synthesis in grape berries would be invaluable for devising strategies to manipulate anthocyanin biosynthesis in grapevines.