Unique Features of Tannin Cells in Fruit of Pollination Constant Non-
Keizo Yonemori, Ayako Ikegami, Shinya Kanzaki and Akira Sugiura
Graduate School of Agriculture, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan
Keywords: Diospyros kaki, tannin accumulation, Japanese persimmon
Among the four types of persimmon, the fruit of pollination constant non-
astringent (PCNA) cultivars has a different development pattern of tannin cells in
the flesh than the other three (PVNA, PNA, and PCA) types. The development of
tannin cells in PCNA types seems to cease at early stages of fruit growth, so that
tannin concentration in the flesh gradually decreases with fruit development by
dilution. In order to clarify this difference between PCNA and non-PCNA types, the
tannin cell size of 42 cultivars was investigated at fruit maturity in this study. Tannin
cells were collected after EDTA maceration of the flesh, and the areas and both
major and minor axes of 100 tannin cells were measured in each cultivar using
computer imagery. Among the 42 cultivars examined, tannin cell sizes of all 15
PCNA cultivars were much smaller than all 27 non-PCNA cultivars. Even
considering the density of tannin cells in the flesh, tannin cells of PCNA fruit
occupied much smaller volumes (less than one-fifth on average) per weight of flesh
than in non-PCNA cultivars. These results clearly show that the dilution of tannins is
the main cause of natural astringency-loss in PCNA persimmons.
Japanese persimmons are classified into 4 types depending on the relationship
between astringency, presence of seeds, and flesh colour. They are (1) pollination-
constant non-astringent (PCNA), (2) pollination-variant non-astringent (PVNA), (3)
pollination-variant astringent (PVA), and (4) pollination-constant astringent (PCA).
PCNA fruit is the most desirable for fresh consumption among these four types, since
PCNA fruit lose their astringency on the tree during fruit development and can be eaten
without any additional postharvest treatment to remove astringency.
PCNA-type fruit has qualitative differences from the other three (PVNA, PVA,
and PCA) types. The phenolic constituents of tannins are quite different between PCNA
and non-PCNA type fruits (Yonemori et al., 1983). In addition, tannins from PCNA fruit
are more stable than those from non-PCNA fruit, as revealed by sedimentation patterns of
tannins in ultracentrifugation (Yonemori and Matsushima, 1984). As a consequence,
tannins from PCNA fruit are difficult to coagulate using acetaldehyde vapour, whereas
those from non-PCNA fruit can easily be coagulated with it (Sugiura et al., 1979; Tomana
et al., 1977; Yonemori and Matsushima, 1984).
The developmental pattern of tannin cells is also quite different between PCNA
and non-PCNA fruits. The development of tannin cells in PCNA types ceases at early
stages of fruit growth, whereas it continues in non-PCNA types until late stages of fruit
development (Yonemori and Matsushima, 1985; 1987). Early cessation of tannin cell
development in PCNA fruit will result in the dilution of tannin concentration in the flesh
as fruit develops. This will be the main cause of natural astringency loss in PCNA fruit on
the tree, in contrast with astringency loss in PVNA or PVA fruit, which is induced by
coagulation of tannins from acetaldehyde produced by the seeds in the fruit (Sugiura et
al., 1979; Sugiura and Tomana, 1983).
The qualitative difference between PCNA and non-PCNA types has been
confirmed in results of breeding projects. PCNA offspring are only obtainable by crossing
among PCNA cultivars, and other combinations of the crosses, i.e. PCNA x non-PCNA or
non-PCNA x non-PCNA, have yielded only non-PCNA offspring, including PVNA, PVA,
and PCA types, but not PCNA types (Ikeda et al., 1985). There is a clear distinction in
IS on Persimmon
Ed. R. Collins
Acta Hort 601, ISHS 2003
inheritance of PCNA and non-PCNA characteristics.
In this study, we investigated tannin cell size in mature fruit from all 4 types of
cultivars to confirm differences between PCNA and non-PCNA types. We also discuss
qualitative differences in tannin cell size during natural astringency loss in PCNA fruit.
MATERIALS AND METHODS
Forty-two cultivars (15 PCNA, 7 PVNA, 6 PVA, and 14 PCA) were used for this
study (Table 1). Three to five mature fruits of most cultivars were collected at harvest
from mature trees growing at Kyoto University orchard, but the fruits of several cultivars
were sampled at harvest time at the Persimmon and Grape Research Center of the
National Institute of Fruit Tree Sciences.
Soluble tannin content of the fruit was determined by measuring phenol content of
the flesh with triplicates using the Folin-Denis method (Swain and Hillis, 1959) after 5g
flesh of each fruit was homogenized with 80 % methanol, calculated as (+)-catechin
equivalents. For size determination of parenchyma and tannin cells, specimen blocks were
taken from the equatorial portion of the fruit and fixed immediately with 2.5%
glutaraldehyde containing 0.2% tannic acid. After washing with water, the blocks were
macerated at 45˚C for 5h by oscillating at 90 times/min in a 0.05M EDTA solution
adjusted to pH 10.0, according to Letham (1960). Tannin cells were separated from
parenchyma cells by decanting several times, and were put onto a slide glass for light
microscopic observations. For size determination of parenchyma cells, the flesh tissue
macerated in the EDTA solution was used for the observation without decanting.
Images of parenchyma and tannin cells were recorded on videotape through a
CCD camera equipped with an inverted light microscope (Olympus IMT-2). Then, the
area of 100 parenchyma and tannin cells were measured from the images of videotape in
each cultivar by a Macintosh computer (Power Macintosh 7500/100) using the public
domain NIH Image v.1.61 software.
When the tannin cells were measured by a computer, the length of both major and
minor axes of every 100 tannin cells were calculated using the software and the volume of
each tannin cell was calculated assuming the tannin cell to be oval. In addition, the
density of tannin cell per unit weight (1g) of flesh was measured with triplicates in each
cultivar by counting the number of tannin cells in 0.5 ml of the solution of macerated 1g
of flesh tissue after filling the solution of 1g of flesh tissue up to 100ml. Thereafter, the
average volume occupied by tannin cells per unit weight (1g) was calculated by
multiplication of the average volume of tannin cells to the average number of tannin cells
for each cultivar.
RESULTS AND DISCUSSION
The size of parenchyma cells varied from 14.0 x 10
of ‘Kakiyamagaki’ to
38.4 x 10
of ‘Hiratanenashi’ (Fig. 1). There was no clear distinction between PCNA-
type and non-PCNA-type cultivars. It differed depending on the cultivar. However, when
the tannin cell size of each cultivar was plotted against its soluble tannin content (Fig. 2),
the tannin cell size from PCNA cultivars was revealed to be very small, centering in a
narrow area with less astringency.
All PCNA cultivars except one are thought to have developed in Japan (Yamada,
1993; Yamada et al., 1994). The exception is the cultivar ‘Luo-tian-tian-shi’, which
originated in China (Wang, 1982; Yamada et al., 1993). Phylogenetic studies on the
relationship among PCNA cultivars using amplified fragment length polymorphism
(AFLP) analysis (Kanzaki et al., 2000a) revealed that ‘Luo-tian-tian-shi’ was distantly
related to the PCNA cultivars of Japanese origin, whereas all PCNA cultivars of Japanese
origin showed a close relationship in their phylogenetic tree. Furthermore, the mechanism
of the loss of astringency in ‘Luo-tian-tian-shi’ is reported to be different from PCNA
cultivars of Japanese origin (Kanzaki et al., 2000b). Despite these findings that ‘Luo-tian-
tian-shi’ is relatively different from PCNA cultivars of Japanese origin, it is included in
the same group with regard to tannin cell size in the present study.
In contrast, non-PCNA cultivars exhibited wide variations of both tannin cell size
and soluble tannin content (Fig. 2). The size of tannin cells of non-PCNA-type cultivars
was spread across a range from 36.7 x 10
to 11.2 x 10
, although the cell size of
non-PCNA cultivars is always much bigger than that of PCNA cultivars. The tannin
content of non-PCNA cultivars was also scattered across a wide range among cultivars.
Even in PVNA cultivars, soluble tannin content was scattered, depending on the number
of seeds in the fruit. Since PVNA cvs. Ama-hyakume, Anzai, Chokenji, and Johren had
several seeds in the fruit, they mostly lost soluble tannins at harvest time, whereas the
other three PVNA cultivars (cvs. Kurokuma, Ohniwa, and Shogatsu) still showed quite
high tannin contents due to fewer seeds in the fruit sampled in this study. It is well known
that loss of astringency in PVNA cultivars depends on the existence of seeds in the fruit
(Ito, 1971; Kajiura and Blumenfeld, 1989; Kitagawa and Glucina, 1984; Yonemori et al.,
2000) because the coagulation of tannins is caused by acetaldehyde evolved from seeds
during the growing season (Sugiura et al., 1979; Sugiura and Tomana, 1983). If the fruit
of PVNA cultivars does not have seeds, tannins are not coagulated and the fruit remains
astringent until harvest time. With other astringent types (PVA and PCA), tannin content
was scattered across a relatively wide range from 0.56 to 2.36 %, and tannin cell size was
also scattered across a wide range. However, a distinct difference was clear between
PCNA and non-PCNA cultivars in regard to tannin cell size.
The clear cut distinction between PCNA and non-PCNA types was confirmed with
the volume occupied by tannin cells per unit weight of flesh (Fig. 3), where PCNA fruit
clearly showed a much smaller volume occupied by tannin cells than in non-PCNA
cultivars, with cv. ‘Yamato-gosho’ showing a slightly larger volume among PCNA types.
Smaller volumes occupied by tannin cells per unit weight in PCNA fruit are closely
related to the natural loss of astringency in PCNA cultivars, which is caused by gradual
dilution of tannins due to the cessation of tannin cell development at an early stage of
fruit enlargement (Yonemori and Matsushima, 1985; 1987). The main factor inducing the
loss of astringency in PCNA types is the small size of tannin cells. Even considering that
‘Luo-tian-tian-shi’ has a different mechanism of astringency loss (Kanzaki et al., 2000b),
it showed the same tendency of small volume occupied by tannin cells per unit weight of
flesh, similar to PCNA cultivars of Japanese origin. Tannin cell size and area are
qualitatively different between non-PCNA and PCNA types, including ‘Luo-tian-tian-shi’.
As previously mentioned, qualitative differences were reported between PCNA
and non-PCNA types regarding loss of astringency. Qualitative differences between
chemical properties of tannins (Yonemori et al, 1983; Yonemori and Matsushima, 1984)
causes a distinction of the coagulation reaction of tannins between them; tannins from
PCNA fruit are difficult to coagulate using acetaldehyde, whereas tannins from non-
PCNA fruit are easily coagulated with it (Sugiura et al., 1979; Tomana et al., 1977;
Yonemori and Matsushima, 1984). Due to this difference of coagulation reaction of
tannins between PCNA and non-PCNA types, Sugiura (1984) proposed a new
classification of Japanese persimmon in which cultivars are grouped into volatile-
independent groups (VIG) and volatile dependent groups (VDG), corresponding to PCNA
and non-PCNA types respectively. The inheritance of natural astringency loss in PCNA
fruit is also reported to be qualitative. The trait in PCNA types appears to be homozygous
recessive, so that PCNA offspring are only obtainable in F1 progeny when the crosses
were made among PCNA cultivars or selections (Ikeda et al., 1985). Natural astringency
loss is reported to be regulated by at least two loci on different chromosomes (Kanzaki et
al., 2000c). PCNA cultivars seem clearly different to non-PCNA cultivars with regard to
chemical properties of tannins and their inheritance. However, this distinction is not clear
when considering the Chinese PCNA cultivar ‘Luo-tian-tian-shi’, because tannins from
‘Luo-tian-tian-shi’ are easily coagulated using acetaldehyde and the cross between ‘Luo-
tian-tian-shi’ and a PCNA cultivar ‘Taishu’ yielded both PCNA and non-PCNA offspring
in the F1 progeny (unpublished data).
Consequently, the clear distinction between PCNA and non-PCNA types should be
based on the difference in tannin cell size as revealed in this study. An absolutely
necessary factor to be a PCNA type is that the size of tannin cells is small enough and that
tannin concentration is diluted during fruit enlargement so as to create a non-astringent
taste at harvest. Based on these criteria, all PCNA cultivars, including ‘Luo-tian-tian-shi’,
can be clearly separated from non-PCNA cultivars.
We thank Dr. Yamada for providing fruit materials of some cultivars for this study
at Persimmon and Grape Research Center, National Institute of Fruit Tree Sciences,
Akitsu, Hiroshima, Japan.
Ikeda, I., Yamada, M., Kurihara, A. Nishida, T. 1985. Inheritance of astringency in
Japanese persimmon. J. Japan. Soc. Hort. Sci. 54: 39-45. (in Japanese with English
Ito, S. 1971. The persimmon, p. 281-301. In: A.C. Hume (ed.). The biochemistry of fruits
and their products. Vol. 2. Academic press, New York.
Kajiura, I. and Blumenfeld, A. 1989. Diospyros kaki. p.298-306. In: A.H. Halevy (ed.).
CRC handbook of flowering. Vol.VI. CRC Press, Baca Raton, FL.
Kanzaki, S., Yonemori, K., Sato, A., Yamada, M. and Sugiura, A. 2000a. AFLP Analysis
of the genetic relationships among PCNA cultivars of persimmon (Diospyros kaki
Thunb.) including ‘Luo Tian Tian Shi’ from China. J. Japan. Soc. Hort. Sci. 69: (in
Kanzaki, S., Yonemori, K., Sato, A., Yamada, M. and Sugiura, A. 2000b. Examination of
the effectiveness of RFLP analysis for selecting PCNA genotype in some persimmon
cultivars. J. Japan. Soc. Hort. Sci. 69: (in press)
Kanzaki, S., Yonemori, K., Sugiura, A., Sato, A. and Yamada, M. 2000c. Identification of
molecular markers linked to the trait of natural astringency-loss of Japanese persimmon
(Diospyros kaki Thunb.) fruit. J. Am. Soc. Hort. Sci. (submitted)
Kitagawa, H., and Glucina, P.G. 1984. Persimmon culture in New Zealand. p. 58-60.
Science Information Publishing Centre, DSIR, Wellington, New Zealand.
Letham, D.S. 1960. The separation jof plant cells with ethylenediamineteraacetic acid.
Exp. Cell. Res. 21: 353-360.
Sugiura, A. 1984. The origin of persimmon and its cultivar differentiation. p. 30-37. In: T.
Ohmura, Y. Yomogihara, H. Ito, T. Kinoshita, and J. Kadota (eds.), Recent advances in
plant breeding. Vol. 25. Keigaku Pub, Tokyo, Japan. (in Japanese)
Sugiura, A., and Tomana, T. 1983. Relationships of ethanol production by seeds of
different types of Japanese persimmons and their tannin content. HortScience 18: 319-
Sugiura, A., K. Yonemori, H. Harada, and T. Tomama. 1979. Changes of ethanol and
acetaldehyde contents in Japanese persimmon fruits and their relation to natural
deastringency. Studies from Inst. Hort. Kyoto Univ. 9: 41-47. (in Japanese with English
Swain, T., and Hillis, W.E. 1959. Phenolic constituents of Prunus domestica. I. The
quantitative analysis of phenolic constituents. J. Agr. and Food Chem. 10: 63-8.
Tomana, T., Mizutani, F. and Takahashi, Y. 1977. Studies on artificial astringency removal
by carbon dioxide on persimmon – Relationship to the cahnges in ethanol and
acetaldehyde content. Abstr. Spring Meeting of Japan. Soc. Hort. Sci. 482-483. (in
Wang, R. 1982. The origin of ‘Lou Tian Tian Shi’. Chinese Fruit Tree 2: 16-19 (in
Yamada, M. 1993. Persimmon breeding in Japan. Japan. Agr. Res. Quart. 27 (1): 33-37.
Yamada, M., Yamane, H., Sato, A., Hirakawa, N. and Wang, R. 1994. Variations in fruit
ripening time, fruit weight and solble solids content of oriental persimmon cultivars
native to Japan. J. Japan. Soc. Hort. Sci. 63:485-491.
Yamada, M., Sato, A., Yakushiji, H., Yoshinaga, K., Yamane, H. and Endo, M. 1993.
Characteristics of ‘Luo Tian Tian Shi’, a non-astringent cultivar of oriental persimmon
(Diospyros kaki Thunb.) of Chinese origin in relation to non-astringent cultivars of
Japanese origin. Bul. Fruit Tree Res. Sta. 25: 19-32. (in Japanese with English
Yonemori, K., and Matsushima. J. 1984. Chemical characteristics of tannins from non-
astringent and astringent type fruits of Japanese persimmon (Diospyros kaki) with
particular reference to ultracentrifugal behavior. J. Japan. Soc. Hort. Sci. 53: 121-126.
(in Japanese with English summary)
Yonemori, K., and Matsushima, J. 1985. Property of development of the tannin cells in
non-astringent type fruits of Japanese persimmon (Diospyros kaki) and its relationship
to natural deastringency. J. Japan. Soc. Hort. Sci. 54: 201-208. (in Japanese with
Yonemori, K. and Matsushima, J. 1987. Changes in tannin cell morphology with growth
and development of Japanese persimmon fruit. J. Am. Soc. Hort. Sci. 112: 818-821.
Yonemori, K., Matsushima, J. and Sugiura, A. 1983. Differences in tannin of non-
astringent and astringent type fruits of Japanese Persimmon (Diospyros kaki Thunb.). J.
Japan. Soc. Hort. Sci. 52: 135-144. (in Japanese with English summary)
Yonemori, K., Yamada, M. and Sugiura, A. 2000. Persimmon Genetics and Breeding. In:
J. Janick (ed.). Plant breeding reviews. Vol. 19. John Wiley & Sons, Inc., New York. (in