Amyloplasts that sediment in protonemata of the moss Ceratodon purpureus are nonrandomly distributed in microgravity

Little is known about whether or how plant cells regulate the position of heavy organelles that sediment toward gravity. Dark-grown protonemata of the moss Ceratodon purpureus displays a complex plastid zonation in that only some amyloplasts sediment along the length of the tip cell. If gravity is the major force determining the position of amyloplasts that sediment, then these plastids should be randomly distributed in space. Instead, amyloplasts were clustered in the subapical region in microgravity. Cells rotated on a clinostat on earth had a roughly similar non-random plastid distribution. Subapical clusters were also found in ground controls that were inverted and kept stationary, but the distribution profile differed considerably due to amyloplast sedimentation. These findings indicate the existence of as yet unknown endogenous forces and mechanisms that influence amyloplast position and that are normally masked in stationary cells grown on earth. It is hypothesized that a microtubule-based mechanism normally compensates for g-induced drag while still allowing for regulated amyloplast sedimentation.     

Cytoplasmic streaming affects gravity-induced amyloplast sedimentation in maize coleoptiles

Living maize (Zea mays L.) coleoptile cells were observed using a horizontal microscope to determine the interaction between cytoplasmic streaming and gravity-induced amyloplast sedimentation. Sedimentation is heavily influenced by streaming which may (1) hasten or slow the velocity of amyloplast movement and (2) displace the plastid laterally or even upwards before or after sedimentation. Amyloplasts may move through transvacuolar strands or through the peripheral cytoplasm which may be divided into fine cytoplasmic strands of much smaller diameter than the plastids. The results indicate that streaming may contribute to the dynamics of graviperception by influencing amyloplast movement.     

Plastid sedimentation kinetics in roots of wild-type and starch-deficient mutants of Arabidopsis.

Sedimentation and movement of plastids in columella cells of the root cap were measured in seedlings of wild-type, a reduced starch mutant, and a starchless mutant of Arabidopsis. To assay for sedimentation, we used both linear measurements and the change of angle from the cell center as indices in vertical and reoriented plants with the aid of computer-assisted image analysis. Seedlings were fixed at short periods after reorientation, and plastid sedimentation correlated with starch content in the three strains of Arabidopsis. Amyloplasts of wild-type seedlings showed the greatest sedimentation, whereas plastids of the starchless mutant showed no significant sedimentation in the vertically grown and reoriented seedlings. Because previous research has shown that a full complement of starch is needed for full gravitropic sensitivity, this study correlates increased sensitivity with plastid sedimentation. However, although plastid sedimentation contributed to gravisensitivity, it was not required, because the gravitropic starchless mutant had plastids that did not sediment. This is the first study, to our knowledge, to measure plastid sedimentation in Arabidopsis roots after reorientation of seedlings. Taken together, the results of this study are consistent with the classic plastid-based and protoplast-based models of graviperception and suggest that multiple systems of perception exist in plant cells.     

Microtubule distribution in gravitropic protonemata of the mossCeratodon

Summary Tip cells of dark-grown protonemata of the mossCeratodon purpureus are negatively gravitropic (grow upward). They possess a unique longitudinal zonation: (1) a tip group of amylochloroplasts in the apical dome, (2) a plastid-free zone, (3) a zone of significant plastid sedimentation, and (4) a zone of mostly non-sedimenting plastids. Immunofluorescence of vertical cells showed microtubules distributed throughout the cytoplasm in a mostly axial orientation extending through all zones. Optical sectioning revealed a close spatial association between microtubules and plastids. A majority (two thirds) of protonemata gravistimulated for >20 min had a higher density of microtubules near the lower flank compared to the upper flank in the plastid-free zone. This apparent enrichment of microtubules occurred just proximal to sedimented plastids and near the part of the tip that presumably elongates more to produce curvature. Fewer than 5% of gravistimulated protonemata had an enrichment in microtubules near the upper flank, whereas 14% of vertical protonemata were enriched near one of the side walls. Oryzalin and amiprophos-methyl (APM) disrupted microtubules, gravitropism, and normal tip growth and zonation, but did not prevent plastid sedimentation. We hypothesize that a microtubule redistribution plays a role in gravitropism in this protonema. This appears to be the first report of an effect of gravity on microtubule distribution in plants.Abbreviations APM amiprophos-methyl - DIC differential interference contrast - DMSO dimethyl sulfoxide - EGTA ethylene glycolbis-(-amino-ethylether) N,N,N',N'-tetraacetic acid - FITC fluorescein isothiocyanate - GS gravitropic stimulus - MT microtubule - PIPES piperazine-N,N'-bis-2-ethanesulfonic acid     

Gravitropism in caulonemata of the moss Pottia intermedia

The gravitropism of caulonemata of Pottia intermedia is described and compared with that of other mosses. Spore germination produces primary protonemata including caulonemata which give rise to buds that form the leafy moss plant, the gametophore. Primary caulonemata are negatively gravitropic but their growth and the number of filaments are limited in the dark. Axenic culture of gametophores results in the production of secondary caulonemata that usually arise near the leaf base. Secondary protonemata that form in the light are agravitropic. Secondary caulonemata that form when gametophores are placed in the dark for several days show strong negative gravitropism and grow well in the dark. When upright caulonemata are reorientated to the horizontal or are inverted, upward bending can be detected after 1 h and caulonemata reach the vertical within 1-2 d. Clear amyloplast sedimentation occurs 10-15 minutes after horizontal placement and before the start of upward curvature. This sedimentation takes place in a sub-apical zone. Amyloplast sedimentation also takes place along the length of upright and inverted Pottia protonemata. These results support the hypothesis that amyloplast sedimentation functions in gravitropic sensing since sedimentation occurs before gravitropism in Pottia and since the location and presence of a unique sedimentation zone is conserved in all four mosses known to gravitropic protonomata.     

Joining forces: The interface of gravitropism and plastid protein import

In flowering plants, gravity perception appears to involve the sedimentation of starch-filled plastids, called amyloplasts, within specialized cells (the statocytes) of shoots (endodermal cells) and roots (columella cells). Unfortunately, how the physical information derived from amyloplast sedimentation is converted into a biochemical signal that promotes organ gravitropic curvature remains largely unknown. Recent results suggest an involvement of the Translocon of the Outer Envelope of (Chloro) plastids (TOC) in early phases of gravity signal transduction within the statocytes. This review summarizes our current knowledge of the molecular mechanisms that govern gravity signal transduction in flowering plants and summarizes models that attempt to explain the contribution of TOC proteins in this important behavioral plant growth response to its mechanical environment.Key words: gravitropism, root, amyloplast, TOC complex, TOC132, TOC75     

A positively gravitropic mutant mirrors the wild-type protonemal response in the moss Ceratodon purpureus

Wild-type Ceratodon purpureus (Hedw.) Brid. protonemata grow up in the dark by negative gravitropism. When upright wild-type protonemata are reoriented 90°, they temporarily grow down soon after reorientation (“initial reversal”) and also prior to cytokinesis (“mitotic reversal”). A positively gravitropic mutant designated wrong-way response (wwr-1) has been isolated by screening ultraviolet light-mutagenized Ceratodon protonemata. Protonemata of wwr-1 reoriented from the vertical to the horizontal grow down with kinetics comparable to those of the wild-type. Protonemata of wwr-1 also show initial and mitotic reversals where they temporarily grow up. Thus, the direction of gravitropism, initial reversal, and mitotic reversal are coordinated though each are opposite in wwr-1 compared to the wild-type. Normal plastid zonation is still maintained in dark-grown wwr-1 apical cells, but the plastids are more numerous and plastid sedimentation is more pronounced. In addition, wwr-1 apical cells are wider and the tips greener than in the wild-type. These data suggest that a functional WWR gene product is not necessary for the establishment of some gravitropic polarity, for gravitropism, or for the coordination of the reversals. Thus, the WWR protein may normally transduce information about cell orientation. Received: 4 November 1996 / Accepted: 26 November 1996     

Disruption of the F-actin cytoskeleton limits statolith movement in Arabidopsis hypocotyls

The F-actin cytoskeleton is hypothesized to play a role in signal transduction mechanisms of gravitropism by interacting with sedimenting amyloplasts as they traverse statocytes of gravistimulated plants. Previous studies have determined that pharmacological disruption of the F-actin cytoskeleton with latrunculin B (Lat-B) causes increased gravitropism in stem-like organs and roots, and results in a more rapid settling of amyloplasts in the columella cells of Arabidopsis roots. These results suggest that the actin cytoskeleton modulates amyloplast movement and also gravitropic signal transduction. To determine the effect of F-actin disruption on amyloplast sedimentation in stem-like organs, Arabidopsis hypocotyls were treated with Lat-B and a detailed analysis of amyloplast sedimentation kinetics was performed by determining amyloplast positions in endodermal cells at various time intervals following reorientation. Confocal microscopy was used to confirm that Lat-B effectively disrupts the actin cytoskeleton in these cells. The results indicate that amyloplasts in hypocotyl endodermal cells settle more quickly compared with amyloplasts in root columella cells. F-actin disruption with Lat-B severely reduces amyloplast mobility within Arabidopsis endodermal statocytes, and these results suggest that amyloplast sedimentation within the hypocotyl endodermal cell is F-actin-dependent. Thus, a model for gravitropism in stem-like organs is proposed in which F-actin modulates the gravity response by actively participating in statolith repositioning within the endodermal statocytes.     

Amyloplasts as possible statoliths in gravitropic protonemata of the moss Ceratodon purpureus

The kinetics of gravitropism and of amyloplast sedimentation were studied in dark-grown protonemata of the moss Ceratodon purpureus (Hedw.) Brid. The protonemata grew straight up at a rate of 20–25 m·h^– in nutrient-supplemented agar. After they were oriented to the horizontal, upward curvature was first detected after 1–1.5 h and reached 84° by 24 h. The tip cells exhibited an amyloplast zonation, with a tip cluster of nonsedimenting amyloplasts, an amyloplast-free zone, and a zone with pronounced amyloplast sedimentation. This latter zone appears specialized more for lateral than for axial sedimentation since amyloplasts sediment to the lower wall in horizontal protonemata but do not fall to the basal wall in vertical protonemata. Amyloplast sedimentation started within 15 min of gravistimulation; this is within the 12–17-min presentation time. The data support the hypothesis that some amyloplasts function as statoliths in these cells.This work was supported by the National Aeronautics and Space Administration grant NAGW-780. We thank Professor E. Hartmann and J. Schwuchow for providing Ceratodon cultures, Dr. John Z. Kiss and Jeff Young for valuable discussions, and Professor Rainer Hertel (University of Freiburg, FRG) for bringing this material to our attention.     

Actomyosin mediates gravisensing and early transduction events in reoriented cut snapdragon spikes

We investigated the involvement of the actomyosin network in the early events of the gravitropic response of cut snapdragon (Antirrhinum majus L.) spikes. The effects of the actin-modulating drug, cytochalasin D (CD) and/or the myosin inhibitor, 2,3-butanedione-2-monoxime (BDM) on amyloplast displacement, lateral auxin transport and consequently on stem bending were examined. The inhibitory effect on cytoskeleton integrity was studied by using indirect immunofluorescence double-labeling of actin and myosin. Our results demonstrate that no organizational changes in actin filaments occurred in cortical and endodermal cells of the stem bending zone during reorientation. These results suggest that actin depolymerization is not required for amyloplast sedimentation. Unlike the chloroplasts in the cortex, the amyloplasts in the endodermis were surrounded by actin and myosin, indicating that amyloplasts may be attached to the actin filaments via the motor protein, myosin. This suggests the involvement of myosin as part of the actomyosin complex in amyloplast movement in vertical as well as in reoriented stems. This suggestion was supported by the findings showing that: (a) BDM or CD disrupted the normal organization of actin either by altering characteristic distribution patterns of myosin-like protein in the cortex (BDM), or by causing actin fragmentation (CD); (b) both compounds inhibited the gravity-induced amyloplast displacement in the endodermis. Additionally, these compounds also inhibited lateral auxin transport across the stem and stem gravitropic bending. Our study suggests that during stem reorientation amyloplasts possibly remain attached to the actin filaments, using myosin as a motor protein. Thus, gravisensing and early transduction events in the gravitropic response of snapdragon spikes, manifested by amyloplast displacement and lateral auxin transport, are mediated by the actomyosin complex.     

Microfilaments and microtubules control the morphology and movement of non-green plastids and stromules in Nicotiana tabacum

Plastid stromules are stroma-filled tubular extensions of the plastid envelope membrane. These structures, which have been observed in a number of species, allow transfer of proteins between interconnected plastids. The dramatic shape of stromules and their dynamic movement within the cell provide an opportunity to study the control of morphology and motion of plastids. Using inhibitors of actin and tubulin, we found that both microfilaments and microtubules affect the shape and motility of non-green plastids. Actin and tubulin control plastid and stromule structure by independent mechanisms, while plastid movement is promoted by microfilaments but inhibited by microtubules. The presence or absence of stromules does not affect the motility of plastids. Photobleaching experiments indicate that actin and tubulin are not necessary for the bulk of green fluorescent protein (GFP) movement between plastids via stromules.     

Gravity perception in decapped roots of Zea mays

Time-lapse photography and light microscopy were used to determine whether or not sedimentation of the newly developed amyloplasts in the apex of Zea mays L. roots occurred at the time when geotropic responsiveness reappears following removal of the cap. All decapped roots exhibiting a geotropic response had some amyloplast sedimentation in the apical cortical cells. Exposing decapped roots to a centrifugal acceleration of 25 g for 4 h showed that amyloplasts of a similar size and development were not displaced within the cytoplasm when this treatment began 12 h after decapping, whereas displacement did occur when the treatment began 24 h after decapping. This finding indicates the occurrence of a change in the physical characteristics of the cytoplasm between 12 h and 24 h after removing of the cap, which allows amyloplast movement and thus restores gravity perception.     

Root cap angle and gravitropic response rate are uncoupled in the Arabidopsis pgm-1 mutant

The sedimentation of starch-filled plastids is thought to be the primary mechanism by which gravity is perceived in roots. Following gravity perception, auxin redistribution toward the lower flank of roots, initiated in the root cap, is believed to play a role in regulation of the gravity response. Amyloplast sedimentation and auxin flux, however, have never been directly linked. The overall aim of this study was to investigate the relationship among plastid sedimentation, gravitropism and auxin flux. Our data show that pgm-1 roots respond to gravity at one-third the rate of wild-type (WT) roots. Maintaining the root tip at a constant angle using image analysis coupled to a rotating stage resulted in a constant rate of response regardless of the angle of tip orientation in pgm-1 mutants, in contrast to the responses of WT and pin3-1 mutants, which showed increasing response rates as the tip was constrained at greater angles. To indirectly visualize auxin flux following reorientation, we generated a pgm-1 mutant line expressing the DR5::GFPm reporter gene. In WT roots a GFP gradient was observed with a maximum along the lower flank, whereas pgm-1 roots formed a GFP maximum in the central columella but lacked any observable gradient up to 6 h following reorientation. Our study suggests that the relationship between root cap angle and gravitropic response depends upon plastid sedimentation-based gravity sensing and supports the idea that there are multiple, overlapping sensory response networks involved in gravitropism.     

Distribution and activation of the Golgi apparatus in geotropism

We find a differential distribution of dictyosomes in the avascular tip cells of the oat (Avena sativa) coleoptile upon geostimulation. A differential activation (increased vesicle production) of the dictyosomes with respect to gravity also occurs, but only in the tip cells of the lower tissues. Similar differences in distribution and activation of dictyosomes occur also in cells subjacent to the avascular tip (1st and 5th millimeter from the apex) of both the upper and lower half-tissues. When only the outer epidermal cells below the apex are considered, the differential distribution and activation of dictyosomes occur only in the lower outer epidermis. The changes in distribution of dictyosomes begin at 6 minutes, or sooner, from the start of geostimulation, before the onset of geotropism. The kinetics of amyloplast sedimentation and Golgi movement do not appear to differ in the cells of the avascular tip. We suggest that the Golgi participates in, and possibly initiates, the differential elongation of cells of geotropically stimulated coleoptiles.     

ULTRASTRUCTURE OF PLASTID INHERITANCE: GREEN ALGAE TO ANGIOSPERMS

1. Plastid inheritance in most green algae and land plants is uniparental. In oogamous species, plastids are usually derived from the maternal parent; even when inheritance is biparental, maternal plastids usually predominate. Only a few species of conifer are known to have essentially paternal plastid inheritance. In spite of the overall strong maternal bias, there exists a spectrum of species in which plastid inheritance ranges from purely maternal to predominantly paternal. 2. Factors that influence the pattern of plastid inheritance operate both before (often long before) and after fertilization. For example, several different mechanisms for exclusion of plastids from particular cells, none of which is completely effective on its own, may operate sequentially during both gametogenesis and embryo-genesis. There appears to exist a general trend such that the more highly evolved the organism, the more numerous the mechanisms employed and the earlier they first come into operation. The pattern of plastid inheritance shown by a species represents the efficiency or lack of efficiency of these combined mechanisms. 3. In the newly-formed zygote of many unicellular algae, the plastids from both gametes are present and there is direct competition between them. Often the plastid from one mating type (usually the ‘invading’ male gamete, where this can be identified) quickly degenerates. Species such as Chlamydomonas are unusual in that the plastids from the two gametes fuse. In spite of this, inheritance of plastid DNA is normally uniparental. How this is accomplished remains unclear. In oogamous algae, the paternal plastids which enter the egg cell are frequently fewer in number and smaller in size than those contributed by the female gamete. The reduced contribution of paternal plastids can result from asymmetrical cell division or from differential timing of cell and plastid division during spermatogenesis. 4. In species ranging from unicellular algae to angiosperms, plastids may be partially or completely debarred from particular cells at critical stages during the reproductive cycle. An important factor in this form of plastid elimination is their postioning with respect to the nucleus prior to a cell division. When plastids closely encircle the nucleus, they are usually incorporated equally into the two daughter cells; when the plastids are concentrated at some distance from the nucleus, they are frequently excluded from one daughter cell. 5. Elimination of plastids from a gamete prior to plasmogamy prevents direct competition between the two types of plastid in the zygote or embryo. Perhaps the most effective method of excluding paternal plastids from the egg cell has been achieved by some lower land plants; the plastids migrate to the posterior part of the spermatozoid, and are discarded from there in a discrete vesicle before the egg is reached. 6. Plastid inheritance in conifers appears to be unique. In those species in which the derivation of plastids in the pro-embryo can be determined, it has been found that they come only from the male gamete. Maternal plastids are positively excluded from the pro-embryo and later degenerate. 7. In most angiosperm species plastid inheritance is maternal; in only a few species is it regularly biparental. The first step towards exclusion of paternal plastids often takes place in the uninucleate pollen grain where the plastids may be concentrated at the pole of the cell farthest from the site of the future generative cell. Any plastids that succeed in entering the generative cell may degenerate before the gametes are released from the pollen tube. Even if paternal plastids reach the egg, they are at a disadvantage because they are (a) entering an environment that is essentially alien, and (b) normally present in much smaller numbers than maternal plastids. Later, when the zygote divides, the few paternal plastids may fail to become incorporated in the small terminal cell which gives rise to the embryo proper. 8. There appears to be no consistent evolutionary progression in the use of more efficient mechanisms to influence plastid inheritance; most of the mechanisms associated with exclusion of paternal plastids in angiosperms, for example, can also be found in one or other species of green alga. The primary factors that influence plastid inheritance appear to be (I) direct competition in the zygote between plastids of the two parental types – the principal mechanism operating in isogamous algae, but also operating in some angiosperms; and (2) the divergent evolution of the two types of gamete - on the one hand a small male gamete with a minimum of cytoplasm which is capable of moving (spermatozoid) or being moved (pollen) efficiently, and, on the other hand, a large egg cell with numerous organelles, which is well able to act as ‘host’ for the future zygote. Many of the additional mechanisms that influence the pattern of plastid inheritance seem to be the more or less ‘accidental’ result of other evolutionary events.     

Isolation and characterization of amyloplast envelope membranes from Solanum tuberosum

Amyloplasts were separated from other subcellular organelles by sedimentation through a discontinuous sucrose density gradient. Purified amyloplast envelope membranes were similar in most characteristics to those from other types of plastids. A substantial proportion of the carotenoid content of these membranes was present in the esterified form. In contrast to published work on chloroplasts of photosynthetic tissue, our results showed that the acylase (acyl-CoA:sn-glycerol 3-phosphate acyltransferase) is firmly bound to the envelope membranes. Evidence was obtained to indicate that digalactosyldiacylglycerol, phosphatidylglycerol and sulpholipid, but not monogalactosyldiacylglycerol, are exclusively found in the cell as amyloplast lipids.     

An Arabidopsis E3 ligase, SHOOT GRAVITROPISM9, modulates the interaction between statoliths and F-actin in gravity sensing

Higher plants use the sedimentation of amyloplasts in statocytes as statolith to sense the direction of gravity during gravitropism. In Arabidopsis thaliana inflorescence stem statocyte, amyloplasts are in complex movement; some show jumping-like saltatory movement and some tend to sediment toward the gravity direction. Here, we report that a RING-type E3 ligase SHOOT GRAVITROPISM9 (SGR9) localized to amyloplasts modulates amyloplast dynamics. In the sgr9 mutant, which exhibits reduced gravitropism, amyloplasts did not sediment but exhibited increased saltatory movement. Amyloplasts sometimes formed a cluster that is abnormally entangled with actin filaments (AFs) in sgr9. By contrast, in the fiz1 mutant, an ACT8 semidominant mutant that induces fragmentation of AFs, amyloplasts, lost saltatory movement and sedimented with nearly statically. Both treatment with Latrunculin B, an inhibitor of AF polymerization, and the fiz1 mutation rescued the gravitropic defect of sgr9. In addition, fiz1 decreased saltatory movement and induced amyloplast sedimentation even in sgr9. Our results suggest that amyloplasts are in equilibrium between sedimentation and saltatory movement in wild-type endodermal cells. Furthermore, this equilibrium is the result of the interaction between amyloplasts and AFs modulated by the SGR9. SGR9 may promote detachment of amyloplasts from AFs, allowing the amyloplasts to sediment in the AFs-dependent equilibrium of amyloplast dynamics.     

Cytological Basis of Plastid Inheritance in Phaseolus vulgaris-Investigations of Plastids,Mitochondria and Their DNA Nucleoids During Pollen Development

Electron microscopic and DNA fluorescence microscopic observations of the plastids, mitochondria and their DNA in the developing pollen of Phaseolus vulgaris L. have demonstrated that the male plastids were excluded during microspore mitosis. The formed generative cell was free of plastids because of regional localization of plastids in early developing microspore and the extremely unequal distribution during division. The fluorescence observations of DNA showed that cytoplasmic (plastid and mitochondria) nucleoids degenerated and disappeared during the development of microspore/pollen, and were never presented in the generative cell at different development stages. These results provided precise cytological evidence of maternal plastid inheritance in Phaseolus vulgaris, which was not in accord with the biparental plastid inheritance identified from early genetic analysis. Based on authors' previous observations in a variety of common bean that the organelle DNA of male gamete was completely degenerated, the early genetic finding of the biparental plastid inheritance was unlikely to be effected by genotypic difference. Thus those biparental plastid inheritance might be caused by occational male plastid transmission, and plastid uniparental maternal inheritance was the species character of Phaseolus vulgaris.     

Multiplication and Differentiation of Plastid Nucleoids during Development of Chloroplasts and Etioplasts from Proplastids in Triticum aestivum

The morphological changes of plastid nucleoids (pt nucleoids)in the shoot apex and along the axis of the leaf blade in Triticumaestivum L. cv. Asakaze were followed with fluorescence microscopyafter staining with 4'6-diamidino-2-phenylindole (DAPI) andquantified with supersensitive microspectrophotometry. Proplastidsin the shoot apex contained 1–10 spherical pt nucleoids.These pt nucleoids changed to a row of spherical and cup-shapedpt nucleoids in sausage-shaped plastids at the leaf base inboth dark and light conditions, in which active cell divisionwas observed. These structures have a higher copy number ofplastid DNA (pt DNA) (72–78 copies) compared to proplastidsin the shoot apex (32–45 copies) and, therefore, may reflectthat active pt DNA synthesis is in progression. In the dark,the cup-shaped pt nucleoids in the spherical etioplasts, whichoriginated from the sausage-shaped plastids, grew to form ring-shapedpt nucleoids. Each ring-shaped pt nucleoid is sub-divided intosmaller pt nucleoids. Under continuous illumination, similarmorphological changes of pt nucleoids occurred except for distributionof small pt nucleoids into young chloroplasts as well as inmature chloroplasts. However, pt nucleotids of leucoplasts inepidermal and vascular bundle sheath cells did not show conspicuouschanges along the axis of the leaf blade. The significance ofthese observations is discussed in relation to plastid differentiationand to the plastid division cycle. ⁴ Present address: Faculty of Science, University of Tokyo,Hongo, Bunkyo-ku, Tokyo, 113 (Received August 15, 1989; Accepted April 13, 1990)