pollen-stigma interaction in brassica. iii. hydration of ... · in germination media, hydration and...

Jf. Cell Sd. 76, 321-336 (1985) 321Printed in Great Britain © The Company of Biologists Limited 1985

POLLEN-STIGMA INTERACTION IN BRASSICA. III.

HYDRATION OF THE POLLEN GRAINS

M. I. ZUBERI* and H. G. DICKINSONDepartment of Botany, Plant Science Laboratories, The University of Reading,Whiteknights, Reading RG6 2AS, UJC

SUMMARY

A method is described by which the changes in shape that accompany hydration olBrassica pollengrains may be rapidly measured. Using this technique it has proved possible to chart the smallamount of hydration that takes place on anthesis, in addition to the response of pollen to a range ofrelative humidities in vitro and in vivo. Such measurements, together with pollen transfer exper-iments, indicate that under normal field conditions self-pollen undergoes a brief period of hydrationfollowed by some loss of water and that, in the course of this hydration, many pollen grains areinhibited from further growth. Raised levels of atmospheric water cause a variety of responses in self-pollen, ranging from tube growth through the pistil to the ovary, to tubes inhibited at the stigmaticsurface, accompanied by the formation of callose. Surprisingly, compatible cross-pollen is alsoaffected by high humidity, often developing extended tubes that are incapable of penetrating thestigmatic cuticle. The development of stigmatic callose is also stimulated by these tubes, as alsooccurs when other members of the Cruciferae are induced to germinate on Brassica stigmas by highlevels of atmospheric water. This formation of callose in response to challenge by all types of pollentube suggests models for the self-incompatibility response in Brassica that involve a direct linkagebetween S (incompatibility) gene products and the formation of callose may require some re-examination. Close study of the operation of the self-incompatibility system in a number of in-dividuals has revealed all aspects of the response to be heavily dependent on the particular S genespossessed by the plant.

INTRODUCTION

Many species of Brassica and other Cruciferae possess homomorphic sporophyticself-incompatibility (SI) systems by which self-pollen is inhibited on the 'dry' surfaceof the stigma (Kroh, 1964; Ockendon, 1972). It has long been appreciated that thehydration of grains that follows pollination is particularly significant, for a carefullymetered increase in hydrostatic pressure inside the grain is considered essential forgermination and tube growth (Kroh, 1966; Mascarenas & Bell, 1969; Heslop-Harrison & Shivanna, 1977; Heslop-Harrison, 1979). Although it has generally beenassumed that pollen grains coming in contact with compatible stigmas absorb waterthrough the papillar surface (Linskens & Kroh, 1970; Heslop-Harrison, 1979) andthat the passage of this water, driven first by matrix and then by osmotic potentialscauses rapid swelling of the dehydrated grain, no unequivocal experimental data existby way of confirmation. The probable reason for this lack of information is because

•Present address: Department of Botany, University of Rajshahi, Rajshahi, Bangladesh.

Key words: Brassica oleracea, hydration, pollen, self-incompatibility, stigma.

322 M. I. Zuberi and H. G. Dickinson

hydration of pollen is very difficult to investigate in vivo without disturbing thedifferent components of this intricately balanced system.

Following a self-pollination, pollen grains of self-incompatible Brassica either donot germinate on the self-stigma or they germinate but fail to penetrate the papillarwall. This penetration of the cuticle surface sometimes induces callose synthesis(Kanno & Hinata, 1967; Dickinson & Lewis, 1973; Roberts & Dickinson, 1983).Inhibition of incompatible pollen before germination points to the operation of the SIsystem at an early stage, with both 'recognition' and 'response' phases taking placewithin a few minutes of pollination. One of the simplest hypotheses proposed toexplain these phenomena involves the failure of pollen to hydrate on self-stigmasbecause S gene products in the outer layers of the cuticle of stigmatic papillae interactwith elements of pollen coat to provide a barrier to water-flow from the stigma to thepollen (Stead, Roberts & Dickinson, 1979; Dumas & Gaude, 1983; Dickinson &Roberts, 1983). That the availability of stigmatic water is central to the operation ofthe SI system is evident from the observation that the incompatibility barrier can beovercome in some cases by increasing the relative humidity and thus decreasing thevapour pressure deficit (VPD) acting on the pollen (Carter & McNeilly, 1975, 1976;Ockendon, 1978), and that water alone can activate pollen germination (Ferrari et al.1983). Nevertheless, strenuous attempts to demonstrate the existence of such a physi-cal barrier to water-flow from stigma to incompatible pollen have so far met withfailure (Roberts, Harrod & Dickinson, 1984).

A second alternative remains. It may be that the apparent lack of hydration ofincompatible pollen results from the incapacity of the pollen to use water madeavailable from the stigma surface. Certainly, Brassica pollen can absorb and losewater very rapidly (Stead et al. 1979), and it is thus of key importance to establish notonly whether or not water is available to self-pollen, but also the response of the grainsto this water.

The water content of pollen has long been appreciated to be important in itsviability, and Pfundt (1910) was the first to note that pollen germinability could beaffected by atmospheric humidity. Since then a series of authors have reported strongcorrelations between atmospheric water, hydration of the grain and pollen viability(Sartoris, 1942; Pfaler & Linskens, 1972; Barnabas & Rajki, 1976; Heslop-Harrison,1979). The swelling of the pollen grains that accompanies an increase in water con-tent, noted by Heslop-Harrison & Shivanna (1977) and Gilissen (1977), was studiedin Brassica oleracea by Stead et al. (1979), who demonstrated that hydration fromatmospheric water is accompanied by a change in shape of the grain from ellipsoid tospheroid, with little change in volume. In germination media, hydration and germina-tion of B. napus pollen has recently been clearly shown to be linked closely with anypre-exposure to water experienced by the grains (S. Corbet & J. Plumridge, personalcommunication). For example, those grains permitted to absorb water more slowlyduring pre-exposure are found to germinate better. Studies of the dynamics of pollenhydration have now been extended to work in vivo, and in the grasses it has provedpossible to draw up a 'balance sheet' of the water exchanges between pollen and stigma(Heslop-Harrison, 1979). However, the conditions faced by pollen on the 'dry' stigma

Hydration of Brassica pollen 323

of a crucifer with a SI system differ both physically and physiologically from those inany study so far made in vitro. In this paper we describe the changes in shape and thefate of pollen grains following intake of water in vitro and in vivo, and consider ourresults in terms of current theories that have been proposed to explain the operationof the SI system in Brassica.

MATERIALS AND METHODS

Plant materialPlants of B. oleracea possessing 5 genotypes 563,63 and 529,29 (kindly supplied by D. Ockendon,

N.V.R.S. Wellesbourne, Warwickshire), andB. campestris var. Toria with genotypes 53,3 and Si,,*(Zuberi, Zuberi & Lewis, 1981) were gTOwn in plastic pots in a heated, insect-proof glasshouse. Thetemperature varied from 15 °C to 28 °C and daylight was supplemented with artificial illumination.As the SI system was often found to be affected by variations in temperature, in atmospheric watervapour pressure deficit (VPD) and in the age of flowers and plants, stigmas used in the study wereselected from inflorescences and plants of comparable age and developmental stage (see Shivanna,Heslop-Harrison & Heslop-Harrison, 1978).

Methods of pollination and pollen removal

Pollinations were done on stigmas of freshly opened flowers, either on plants or on excised pistilsmaintained in small plastic pots, covered with Parafilm and containing modified White's medium(Roberts, Gaude, Harrod & Dickinson, 1983). The anthers and calices were carefully removed fromthe pistils before pollination. Pollinations were done by gently touching freshly dehisced longanthers to the stigma, thus applying a layer of pollen grains over the stigmatic papillae. At variousintervals after pollination, the pollen grains were removed from the stigma surface by touching thepollinated surface very carefully with the sticky surface of a piece of transparent adhesive tape(Sellotape), attached to a glass slide. For each stigma, three successive contacts were made in quicksuccession to adjacent points on the tape. The first contact point contained mostly the unhydratedor partially hydrated pollen grains, which may or may not have direct contact with the papillae; theother two contained the pollen grains that had been in direct contact with the stigma. If the pressureapplied during pollen transfer was not sufficiently gentle, pieces of stigmatic papillae becamedetached and adhered to the tape. The measurements of pollen grain diameter were taken im-mediately from the second and third contact points using a compound microscope equipped withan eye-piece graticule. To avoid overheating the pollen grains, the microscope was fitted with amirror rather than a tungsten light-source. When pollinations were done on plants, the microscopicexamination was carried out in the glasshouse. Many replicates of each 'treatment' were carried out,and from these results the mean long and short axes were calculated. Using the methods of Steadet al. (1979), the change in shape of hydrating pollen grains was obtained by calculating the ratiobetween the long axis and the short axis. In other experiments, pollen of B. napus, Aubrietiadeltoidea, Etysimitm sp. (Botanical Garden, University of Reading) and of Crambe sp. (kindlysupplied by D. Lewis) were used.

Treatments under different VPDsNormally pollinations were done either in the glasshouse or in the laboratory, where the VPD was

dependent on the prevailing weather conditions. During a sunny dry morning the VPD usuallyvaried between 11 and 14mbar, whereas higher humidity (VPD approx. 3-6-5 mbar) prevailedduring wet weather. Very low VPDs (ranging from 0 to 1-5 mbar) were created artificially insidecovered beakers with soaked tissue papers. Where necessary, stigmas were transferred to the beakersafter pollination. When pre-hydrated pollen grains were needed, flowers with dehisced anthers wereenclosed under low VPDs for 1-1-Sh before pollination. For pollen germination, a mediumsupplemented with 20 % polyethylene glycol was used (Roberts et al. 1983). Callose was detectedby ultraviolet fluorescence following rapid fixation (Dumas & Knox, 1983).


Table 1. Axis ratio (long: short) o/B. oleracea pollen grains in freshly dehisced anthersand those in anthers that had opened the previous day

5 genotype Freshly dehisced Open for 24 h

563,63 2-01 ±0-006 1-68 ±0-036Szs.zs 1-94 ±0-069 1-71 ±0-066

Measurements taken in the glasshouse, at an average VPD of 12mbar.

Table 2. Axis ratio (long: short) o/B. oleracea pollen (S&J,<SJ) exposed to a low VPD(approx. 0-1 mbar) for between 30 and 120 min, and the axis ratio of this pollen when

subsequently exposed to a normal VPD (approx. 12 mbar) for a period of 15min

Exposure timeat low VPD (min)

3060

120

111

Axis

Fresh

•47 ± 0-08•30 ±0-02•29 ±0 04

ratios

Old

l-56±01-39 ±01-29 ±0

•23•03•03

111

Axis ratios after 15 minfurther exposure to a

'normal' VPD

Fresh Old

•97 ±0-04 1-79 ±0-11•69 ±0-10 1-48 ±0-09•77 ±0-15 1-51 ± 011

Both freshly dehisced ('fresh') pollen and that from anthers opening the day previously ('old') wasmeasured.

See Table 1 for dimensions of freshly dehisced pollen.

RESULTS

The state of hydration of pollen contained in the anther

It is evident that pollen grains absorb water from the atmosphere whilst still in theanther, for there was a striking difference in the axis ratio of the pollen from freshlydehisced anthers compared with that of those that had opened on the previous morn-ing (Table 1). This difference in axis ratio was maintained even when pollen wasmeasured on days of very different humidity.

This observation is clearly of significance for experiments involving pollen fromplants grown in glasshouses or growth chambers where flowers retain pollen forseveral days.

Fig. 1. Brassica pollen showing different states of pollen hydration. A. B. oleracea pollenfrom freshly dehisced anthers (563,63)- ~ x 170. B. B. oleracea pollen from freshlydehisced anthers (563,63) exposed to a low VPD ( — 0 - 1 mbar) for 0-5 h. — X 170.c B. campestris pollen (53,3) removed from a self-stigma following 1 -75 h exposure in theglasshouse (VPD — 12mbar). — X 250. D. B. oleracea pollen from a compatible stigma(563 63 * 525 25) following l-75h at normal VPDs. Pollen tubes (arrows) are visible.- x 170.


Fig. 1


30

Time (min)

15 30 45

Time (min)

15 30 45Time (min)

Fig. 2

60

60


Hydration of pollen by atmospheric water

Pollen dusted on glass slides exposed to 'normal' VPD (approx. 14 mbar) on alaboratory bench for 5 h showed little change in axis ratio. However, when pollengrains were enclosed inside beakers with soaked tissue paper ('low' VPDs, approx.0-1 -5 mbar) pollen grains changed shape rapidly, indicating their ability to absorbwater from the atmosphere (Table 2; Fig. 1A, B). Pollen from freshly opened flowersbecame hydrated more quickly than those from older flowers, when placed underconditions of low VPD (Table 2). Pollen was also exposed to a 'normal' VPD for 15min following hydration, and the axis measurements were taken again. Thesehydrated grains were observed to lose water readily, as indicated by an increase in axisratio (Table 2). Moreover, pollen from freshly opened anthers became 're-hydrated'more readily than pollen from anthers that had dehisced the previous day.

When flowers with dehisced anthers were subjected to low VPDs, pollen in im-mediate contact with the atmosphere was hydrated readily and, within 1-5 h, becamefully hydrated. The subjacent grains, however, were hydrated more slowly. When leftto become hydrated in a saturated atmosphere for longer periods (4—5 h) in anthers,or for a shorter period (3 h) on a slide, pollen often lost its ability to germinate whentransferred to a compatible stigma. It was interesting to note that pollen fromB. oleracea plants possessing 5 genotype £25,25 was hydrated very rapidly, whereaspollen from Sz9,29 exhibited a low rate of hydration and retained its germinationability for longer periods. 563,63 pollen was intermediate in its behaviour (Fig. 2B).

Hydration of pollen following incompatible matings

Initial changes in pollen shape. Hydration of Brassica pollen on both compatibleand incompatible stigmas was studied in B. oleracea (genotype 563,63) andB.campestris (genotypes S^S* and 53,3) over a period of 3 months (April to June,1984). The changes in pollen shape on the stigma of some typical B. oleracea and B.campestris genotypes are given in Fig. 2A-C. Except under very dry conditions,dehydrated pollen from the anther always began hydrating immediately after comingin contact with the stigma papilla. Under very high VPDs the change in pollen shapewas comparatively slow, but under normal and low VPDs initial hydration was readilynoticeable from about 20-25 min after coming in contact with the stigma. Hydrationcould be monitored not only by the change in shape but also by the colour of the pollengrains. Hydrating pollen becomes more translucent while the dehydrated pollengrains remained comparatively opaque and dull brownish in colour (Fig. lc). Afterthis initial period of hydration, further development on an incompatible stigmadepended on the prevailing VPD.

Fig. 2. Time course of changes in axis ratio (long : short) of Brassica pollen on self- andcross-stigmas under glasshouse conditions (VPD, -~ 12 mbar). Bars indicate standarderror, A. B. oleracea; S genotypes 563,63 selfed (O O) ; Szs,25 x $63,63 ( • • )•B. B. oleracea: S genotypes SM.W selfed (O O) ; Sw.w * 6̂3,63 ( d d ) .a B. campestris (var. Toria): S genotypes S4,4 selfed (O——O); S3,3 X S4,4 ( • • ) .


Changes in the shape of pollen at high VPDs. When the VPD was comparatively high(approx. ll-14mbar), pollen grains on an incompatible stigma became dehydratedand shrank visibly after a brief period of hydration. The time course of the typicalincompatible pollen-stigma interaction shown by a self-pollination of 563,63 underhigh VPD is given in Fig. 2A. It is, however, disappointing that the shrinkage of thepollen grains, readily detectable by eye, is not confirmed by these figures. The timerequired to complete the brief hydration and shrinkage varied dramatically withgenotypes. As seen in Fig. 2A-C, pollen grains of B. oleracea 563,63, 529,29 and B.campestris 54,4 all exhibited very different courses of hydration on self-stigmas undersimilar VPDs. Pollen of 563,63 genotype underwent a brief hydration and rapid shrin-kage (normally within an hour), while 529,29 pollen was hydrated very slowly andshrank only after a considerably longer period (2-5-3 h). Under these conditions nopollen germinated on incompatible stigmas and no callose was detectable when thesestigmas were examined with the u.v. microscope.

Changes in the shape of pollen at low VPDs. Under more humid conditions (VPDbetween 0 and 1*5 mbar), pollen grains were found to continue to become hydrated

Table 3. Pollen germination (%) o/Brassica S genotypes after various exposures onself- and cross-stigmas under different VPDs

Mating

B. oleracea563,63 self

525,25 self

5»,29 self563,63 X 525,25

563,63 x 529,29

525,25 X 563,63

•S25.25 X 529,29529,29 * 525,25

529,29 * 563,63

563,63 self

525,25 self529,29 self563,63 X 525,25

525,25 X 563,63

B. campestris53,3 self

53,3 X 54,4

53,3 self

53,3 X 54,4

VPD

Percentage of

l h

Normal f 0(approx. 11-14 mbar)

00

2500000

Low r 20(approx. 0-1 mbar)

Normal(approx. 11-14 mbar) <

Low(approx. 0—1 mbar) *

501

6050

0

20

r 10

I 50

2h

0

00

70000

300

50

70158075

0

45

30

60

pollen germinating

3h

0

50

801001

550

50

71208075

0

65

50

65

4h

0

100

80504

1180

5

65

80358080

0

75

60

65

Hydration ofBrassica pollen 329

on incompatible stigmas for a longer period and eventually to attain a fully hydratedmorphology. Moreover, these grains germinated, producing short or coiled tubes.These tubes produced a characteristic synthesis of callose when they came in contactwith the papilla surface. The time required by the self-pollen to become hydrated onincompatible stigmas and the proportion of pollen that actually germinated variedwith genotypes, as well as with any pre-exposure of pollen to low VPDs (Fig. 2; Table3). Thus, under conditions of low VPD 525,25 pollen became hydrated more rapidlyon self-stigmas and about 50% of the pollen germinated within 1 h, whereas underthe same conditions 563,63 pollen achieved only 20 % germination and 529,29 pollenless than 1 %.

The viability of shrunken pollen. To determine whether the pollen grains that shrinkafter initial hydration on incompatible stigmas have lost their ability to germinate anddevelop, several attempts were made to transfer pollen to compatible stigmas and ontogermination medium. In these experiments a small number of pollen grains (450-600per stigma) were applied with a single hair to ensure contact of all the pollen to stigmapapilla. While the number of pollen grains transferred was comparatively low, it wasevident that no 563,63 pollen grains managed to germinate on Szs,2s stigmas whentransferred after 1 h on the self-stigmas. Of the few pollen grains that were releasedfrom selfed stigmas into complete germination medium after 1 h, none produced

50 90 180Exposure to 'normal' VPD on self-stigma (min)

Fig. 3. Percentage of B. oleracea pollen germinating on self-stigmas when exposed to lowVPDs ( ~ 0 - l mbar) following various periods of incubation under normal VPDs (~ 12mbar).Preliminary tests snowed no further pollen development to take place after 3 h under the lowVPDs, so pollen was scored at or about this time. ( • • ) S63.63; (O O) S2s,2s; ( • •)529,29-


normal tubes. As an alternative approach, the selfed stigmas, after various exposuresto normal VPDs, were transferred to low VPDs in an attempt to induce incompatiblepollen grains to develop further. Results of such experiments (Fig. 3) indicated thatwith increasing length of exposure on the incompatible stigma, self-pollen lost itsability to germinate. The few grains still capable of germinating were probably not indirect contact with self-papillae and thus remained unaffected. When examined underthe microscope, the shrunken pollen grains were found to possess clumped andgranular cytoplasms.

Hydration of pollen on compatible stigmas

Compatible pollen began hydration shortly after coming in contact with the stigmabut, unlike incompatible pollen, the process of hydration continued until germinationhad taken place (Fig. 2A-C). The course of change in pollen shape and the time takento germinate were found to depend both on the genotypes involved and the VPD towhich the pollen was pre-exposed (Fig. 2; Table 3). In the case of crosses 563,63 X £25,25(B. oleracea) and £3,3 X 54,4 (B campestris) the pollen was hydrated rapidly inconditions of normal VPD and germinated within 1 h. However, under identicalconditions in the crosses £25,25 X 563,63, and in those involving 529,29 pollen, the pollenwas hydrated slowly and germinated only after 4-6 h (Table 3). When pollinationswere done using pre-exposed pollen, or under conditions of lower VPD, changes inpollen shape were more pronounced and pollen germinated more rapidly than innormal VPD (Table 3). Under high VPDs the changes in pollen shape were lessevident (Fig. ID).

When compatibly pollinated stigmas were transferred to low VPDs (approx.0-1 mbar) pollen tube organization and penetration were adversely affected (Table4). In most cases pollen germinated under low VPDs but the tubes became

Table 4. Percentages of B. oleracea stigmas supporting germination and abnormalgrowth of pollen tubes under low VPDs (approx. 0—1 mbar)

Mating No germination

02005000

Percentage of stigmas

Tubes observed,Abnormal tubes, no some penetrations

penetration ofpapillae

10080

10060208038

(5-20/stigma),callose formed

000

35802062

563,63 self525,25 self5»,29 Self563,63 X 525,25

563,63 X 529,29

525,25 X 563,63

529,29 X 563,63

The number of stigmas scored ranged between 15 and 20 per mating excepting when the countwas particularly clear-cut, when between 5 and 8 were measured.


disorganized after emerging from the grains, failed to penetrate the papillar surface,and massive depositions of callose were produced in the papillae. In some crosses(Table 4), a few pollen tubes did manage to penetrate the papillar wall, but callose wasalso produced. Under low VPD the pollen tubes exhibited abnormal growth, manyof them being coiled, 'haustoria-like' and swollen. A similar adverse effect of highhumidity was also noted for .6. napus, which is a self-compatible species.

When the time required for the pollen to germinate under low VPDs was examinedit was found to be reduced by 20-30 min compared with that under a normal VPD.The process of pollen development and germination on compatible stigmas could bespeeded up further by pre-hydrating the pollen grains for 1-2 h in low VPDs. In thisway the time required for pollen to germinate on the stigma could be reduced to 25-35min whereas, under normal conditions, pollen might take 60—90 min to germinate.There were also differences in the rates of hydration of the three B. oleracea Sgenotypes studied; under low VPDs 529,29 pollen was hydrated more slowly than525,25 or 563,63- Moreover, when placed in low VPD on compatible stigmas, 5w,29pollen was less adversely affected, thus producing normal tubes more often thanothers (Table 4).

Hydration of pollen from other members of the Cruciferae

Pollen grains from other crucifer plants (B. napus, Aubrietia sp., Erysimum sp. andCrambe sp. were placed on B. oleracea stigmas and allowed to develop at normalVPDs. In all cases these pollens were hydrated and in many cases reached full hydra-tion by 1—1 -5 h. However, none of the grains germinated after 3—4h except for B.napus, which did so after 1-5 h at normal VPD. Stigmas of B. napus also induced B.oleracea pollen to become hydrated and germinate within 1 h. Under low VPDs,however, all the crucifer pollen tested germinated on B. oleracea stigmas, producingshort coiled tubes. These failed to penetrate the papillar wall and induced the forma-tion of callose.

DISCUSSION

The relationship between VPD and pollen hydration

The evidence presented here reinforces the view that the hydration of pollen playsa key role in the operation of the SI system. Even small changes in humidity appearto have far-reaching consequences for the ultimate fate of the pollen on the stigma.This complex interaction between the availability of water and the SI system is mostprobably the reason why this mechanism of cellular recognition has proved so resistantto investigation over the years, despite a considerable investment in time and effortby many groups.

Under relatively high VPDs (approx. 14 mbar) the normal SI reaction takes place,that is to say the incompatible pollen is recognized and inhibited on self-stigma beforegermination. In conditions of low VPDs (approx. 0-1-5 mbar), however, Brassicapollen readily obtains water from the atmosphere as well as from the stigmatic papilla,


and it seems reasonable to suppose that it is for this reason that incompatible grainsdo develop under low VDPs, thus bypassing a principal step in the SI response. Theirfuture, however, depends on the operation of another component of the response, thenature of which will be discussed later.

The time-course of changes in the shape of pollen and its ability to germinate invitro have been shown to be affected by pre-exposure of pollen to high humidity (S.Corbet & J. Plumridge, personal communication). Very similar effects of pre-exposure to low VPDs on pollen in vivo are indicated by the present study. These datalend support to the hypothesis (Heslop-Harrison, 1979) that slow hydration by waterfrom either the atmosphere or the stigma enables the developing pollen grain toinitiate the process of cytoplasmic reorganization that is crucial for normal germina-tion and penetration of the stigma. The detailed ultrastructural evidence of thesechanges in cytoplasmic reorganization will be presented elsewhere (Elleman &Dickinson, unpublished data). Also underlined by this study is the importance ofselecting flowers with comparable histories of pre-exposure to VPDs in experimentsinvolving comparison of the rates of hydration and germination.

It is interesting that hydrating pollen grains for longer periods (2-3 h) in vitroresults in the diminution of their ability to germinate on compatible stigmas. Similareffects of hydration in vitro have also been reported by Roberts et al. (1983). Simplemicroscopic inspection reveals, however, that during these extended periods of ex-posure to a low VPD a considerable proportion of the pollen-coating flows onto theslide. Since contact between the pollen protoplast and the stigma surface is achievedand maintained through the coating, loss of this layer during pre-exposure might wellbe expected to result in low germination rates.

The variability in speed of hydration of different genotypes both in vitro and in vivois spectacular. For example, £25,25 pollen becomes hydrated more rapidly than529,29 and this difference is also reflected in the characteristics of the pollen—stigmainteraction involving these genotypes - following both compatible and incompatiblematings. While such observations support the concept of a direct relationship betweenhydration and the SI reaction, it should be remembered that the SI genotypes usedare far from isogenic in other respects.

Behaviour of pollen on incompatible stigmas

From the data presented here, it is now clear that, except under very high VPD (i.e.above approx. 14mbar), Brassica pollen undergoes a brief period of hydration afteralighting on an incompatible stigma. Under normal VPDs (approx. ll-14mbar)changes taking place during this hydration can render pollen inhibited. If we assumethat these events are under S gene control, the behaviour of incompatible pollen underconditions of lowered VPD becomes difficult to explain. For example, low VPDs havelong been known to induce self-pollen to overcome the SI system (Carter & McNeilly,1975, 1976; Ockendon, 1978) and, in saturated atmospheres, self-pollen will ger-minate to form long tubes, which often follow a tortuous path, only to be arrested atthe point of stigmatic penetration. This arrest is frequently accompanied by theformation of callose. Clearly, only two alternative hypotheses can explain these


events: either molecules produced as a result of 5 gene action are active at more thanone stage of pollen development, i.e. germination and tube penetration, or 5 geneproducts are involved only in controlling germination but, over the course ofevolution, their activity has become functionally linked with more general processesoperating in the plant. In the latter hypothesis, the arrest of the growth of the pollentube tip and the generation of callose would thus be caused by the interaction of 5 geneproducts during germination, but would not involve their participation at the pointof arrest.

Another important point emerging from the present study is that the 'speed' of theinhibition of self-pollen on the stigma varies dramatically with the genotypes involvedas well as with the VPD. So marked is this difference that under some VPDs thatinduced 563,63 or 525,25 pollen to be fully inhibited 529,29 pollen remained partiallyinhibited, in that many grains were still capable of germination when an appropriateenvironment was provided. Whether or not grains are irreversibly inhibited remainsopen to question. Recent preliminary studies in our laboratory suggest that, as mightbe expected, the time required for pollen to be 'released' from inhibition by exposureto low VDPs is proportional to the period of exposure on an incompatible stigma. Forexample, even after quite a short period on an incompatible stigma, some pollensrequired up to 3 days at a low VPD to begin germination. However, as also might beexpected, there does appear to be a period of exposure, strongly correlated with bothgenotype and environment, after which pollen can no longer be revived. This varia-tion in response most probably explains why Kroh (1964) and Roberts et al. (1983)failed to detect inhibition in their pollen-transfer experiments.

Although pollen grains from other genera of the Cruciferae were able to becomehydrated on Brassica stigmas, all failed to germinate at normal VPDs except for B.napus pollen, which developed readily. AsB. napus is an allotetraploid derived fromdiploid B. oleracea and B. campestris, it is likely that the genetic differences respons-ible for isolation at the generic level do not involve inhibition of pollen germination.It is, however, most interesting that in low VPDs the pollen of other Cruciferae cangerminate on B. oleracea stigmas but are later inhibited, accompanied by the forma-tion of callose. Under these circumstances there clearly cannot be linkage betweencallose formation and 5-gene-controlled events.

Low VPD and the development of compatible pollen

One of the most interesting observations on the effect of low VPDs concerns theiradverse effect on pollen tubes growing on compatible stigmas. These pollen tubesbecome disorganized and usually fail to grow into the compatible papillar wall. Gener-ally, cessation of tube growth is accompanied by a vigorous synthesis of callose in thesubjacent papilla. It seems reasonable to propose that this 'rejection' of the 'compat-ible' pollen tube must result from rapid hydration of the pollen grain, using water fromthe atmosphere as well as from the papillar wall. This in turn would generate a rapidaccumulation of hydrostatic pressure within the pollen, causing premature germina-tion and uncontrolled and rapid growth. Under these conditions, the normal synthesisand assembly of materials required for pollen tube growth might be sufficiently


disrupted to affect the precise compartmentation of the tube tip. Certainly tubeelongation, driven by hydrostatic pressure, would proceed far faster than normal, andpossibly more rapidly than the formation of the callose sheath along the tube. Theabsence of this wall could well result in the leakage of gametophytic molecules into thepapilla through the point of penetration and the stimulation of callose formation in anon-specific defence response, normally triggered by wounding or infection. Certain-ly, it is possible that normal pollen tubes do secrete defined gametophytic proteins incontrolled quantities for various purposes, but under these circumstances un-controlled leakage of cytoplasmic content would occur.

The SI system operating in Brassica

While our observation that incompatible pollen is normally hydrated on the stig-matic surface does not completely rule out the possibility that the development ofpollen is controlled only by the metering of water-flow to the grain, the fact thatdevelopment of much of this pollen is found to be inhibited probably does. It wouldthus seem more reasonable to assume that interaction between male and female 5 geneproducts results in the slow release of an inhibitor, either from the stigmatic surface(Roberts et al. 1983) or from the pollen. This inhibitor, possibly that isolated byHodgkin & Lyon (1984), must be secreted in carefully regulated quantities, for anydilution by the addition of atmospheric water apparently causes it to lose its effective-ness, and the pollen to germinate. While some of the pollen is certainly capable ofgrowth through the pistil to the ovary (Carter & NcNeilly, 1975, 1976; Ockendon,1978), much of it is not, and instead becomes arrested at the point of its penetrationof the stigmatic cuticle, accompanied by the formation of callose. The striking paral-lels in behaviour between this type of incompatible response and the development ofcompatible tubes under low VPDs, considered together with the formation of calloseon pollination by other members of the Cruciferae, indicates that this 'rejection' at thepapillar cuticle is completely or partially independent of 5 gene control, andrepresents a more general response to mechanical injury or challenge by 'foreign'proteins. It should, however, be emphasized that a considerable body of old andrecent data remains (Dickinson & Lewis, 1973; Heslop-Harrison, Heslop-Harrison& Knox, 1973; Kerhoas, Knox & Dumas, 1983) that suggests that fractions derivedfrom compatible pollen grains do not stimulate the formation of callose, whereasmaterials from incompatible and foreign pollen will. Clearly, until more is known ofthe molecules emitted by compatible and incompatible pollen tubes during their earlystages of growth under different conditions, any relationship between the formationof callose and expression of the 5 gene must remain hypothetical.

One of us (M.I.Z.) is grateful to the Association of Commonwealth Universities for the award ofa Fellowship. Part of this work is supported by the AF RC through a grant toH.G.D.Wealso thankProfessor Watkin Williams for allowing us to use his glasshouse, Gill Harrod for technical help andDr C. Elleman for helpful discussion.


REFERENCES

BARNABAS, B. & RAJKI, E. (1976). Storage of maize (Zea mays L.) pollen at —196°C in liquid N.Euphytica 25, 747-52.

CARTER, A. L. & MCNEILLV, T. (1975). Effects of increased humidity on pollen tube growth andseed set following self-pollination in Brussels sprouts (Brassica oleracea var. gemmifera).Euphytica 24, 805-813.

CARTER, A. L. & MCNEILLY, T. (1976). Increased atmospheric humidity post-pollination: Apossible aid to the production of inbred line seed from mature flowers in the Brussels sprout{Brassica oleracea var. gemmifera ) . Euphytica 25, 532-536.

DICKINSON, H. G. & LEWIS, D. (1973). Cytochemical and ultrastructural differences betweenintraspecific compatible and incompatible pollinations in Raphanus. Proc. R. Soc. Land. B, 183,21-38.

DICKINSON, H. G. & LEWIS, D. (1973). The formation of the tryphine coating the pollen grainsof Raphanus, and its properties relating to the self-incompatibility system. Proc. R. Soc. Land.B, 184, 149-165.

DICKINSON, H. G. & ROBERTS, I. N. (1983). Cell surface receptors in the pollen—stigma interactionof Brassica oleracea. In Receptors in Plants and Slime Moulds (ed. D. R. Garrod & D. M.Chadwick). New York: Marcel Dekker.

DUMAS, C. & GAUDE, T. (1983). Stigma-pollen recognition and pollen hydration. Phytomorphol-ogy 31, 191-201.

DUMAS, C. & KNOX, R. B. (1983). Callose and pollen-pistil incompatibility. Theor. appl. Genet.67, 1-10.

FERRARI, T. E., COMSTOCK, P., MORE, T. A., BEST, B., LEE, S. S. & WALLACE, D. H. (1983).Pollen-stigma interactions and intercellular recognition in Brassica: Pathways for water uptake.In Pollen: Biology and Implications for Plant Breeding (ed. D. L. Mulcahy & E. Ottaviano), pp.243-249. New York, Amsterdam, Oxford: Elsevier Biomedical.

GILISSEN, L. J. W. (1977). The influence of relative humidity on the swelling of pollen grains invitro. Planta 137, 2299-2301.

HESLOP-HARRISON, J. (1979). Aspects of the structure, cytochemistry and germination of thepollen of rust (Secale cereale. L). Ann. Bot. 44, Suppl. 1, 1-47.

HESLOP-HARRISON, J., HESLOP-HARRISON, Y. & KNOX, R. B. (1973). The callose rejectionreaction: a new bioassay for incompatibility in Cruciferae and Compositae. Incompatibility News-letter 3, 75-76.

HESLOP-HARRISON, Y. & SHIVANNA, K. R. (1977). The receptive surface of the angiospermstigma. Ann. Bot. 41, 1233-1258.

HODGKIN, T. &G. D. LYON (1984). Pollen germination inhibitors in extracts of Brassica oleracea(L.) stigmas. New Phytol. 96, 293-98.

KANNO, T. & K. HINATA (1969). An electron microscopic study of the barrier against pollen tubegrowth in self-incompatible Cruciferae. PI. Cell Physiol., Tokyo 10, 213-216.

KERHOAS, S., KNOX, R. B. & DUMAS, C. (1983). Specificity of the callose response in stigmas ofBrassica. Ann. Bot. 52, 597-602.

KROH, M. (1964). An electron microscopic study of the behaviour of Cruciferae pollen afterpollination. In Pollen Physiology and Fertilisation (ed. H. F. Linskens), pp. 221-224. Amster-dam: North Holland.

LINSKENS, H. G. & KROH, M. (1970). Regulation of pollen tube growth. Curr. Top. devl Biol. 5,89-113.

MASCARENHAS, J. P. & BELL, E. (1969). Protein synthesis during germination of pollen. Studieson polyribosome formation. Biochim. biophys. Acta 179, 199-203.

OCKENDON, D. J. (1972). Pollen tube growth and the site of incompatibility reaction in Brassicaoleracea. New Phytol. 71, 519-522.

OCKENDON, D. J. (1978). Effect of hexane and humidity on self-incompatibility in Brassicaoleracea. Theor. appl. Genet. 52, 113-119.

PFALER, P. F. & LINSKENS, H. F. (1972). In vitro germination and pollen tube growth of maize(Zea mays L.) pollen. VII. Storage temperature and pollen source effects. Theor. appl. Genet.42, 136-140.

336 M.I. Zuberi and H. G. Dickinson

PFUNDT, M. (1910). Der Einflus der Luftfeuchfligkeit auf der Lebensdauer des Blutenstaubes. Jb.wiss. Bot. 47, 1-40.

ROBERTS, I. N. & DICKINSON, H. G. (1983). Intraspecific incompatibility on the stigma ofBrassica. Phytomorphology 31, 165-174.

ROBERTS, I. N., GAUDE, T. C , HARROD, G. & DICKINSON, H. G. (1983). Pollen-stigma inter-actions in Brassica oleracea: a new pollen germination medium and its use in elucidating themechanism of self-incompatibility Theor. appl. Genet. 65, 231-238.

ROBERTS, I. N., HARROD, G. & DICKINSON, H. G. (1984). I. Infrastructure and physiology of thestigmatic papillar cells, J. Cell Sci. 66, 241-253.

SARTORIS, G. B. (1942). Longevity of sugar cane and corn pollen - a method for long distanceshipment of sugar cane pollen by airplane. Am. J. Bot. 29, 395—403.

SHIVANNA, K. R., HESLOP-HARRISON, Y. & HESLOP-HARRISON, J. (1978). The pollen-stigmainteraction: bud pollination in the Cruciferae. Acta bot. neeri. 27, 107-119.

STEAD, A. D., ROBERTS, I. N. & DICKINSON, H. G. (1979). Pollen-pistil interactions mBrassicaoleracea: Events prior to pollen germination. Planta 146, 211-216.

ZUBERI, M. I., ZUBERI, S. & LEWIS, D. (1981). The genetics of incompatibility in Brassica. I.Inheritance of self-compatibility in Brassica campestris L. var. Toria. Heredity 46, 175-190.

(Received 17 November 1984 - Accepted, in revised form, 13 February 1985)

pollen-stigma interaction in brassica. iii. hydration of ... · in germination media, hydration and...

Documents