Characterization of a gene family abundantly expressed in

12 Pages·2002·3.64 MB·English

The Plant Cell, Vol. 2, 263-274, March 1990 O 1990 American Society of Plant Physiologists
Characterization of a Gene Family Abundantly
Expressed in Oenothera organensis Pollen That Shows
Sequence Similarity to Polygalacturonase
Sherri M. Brown' and Martha L. Crouch
Biology Department, Indiana University, Bloomington, Indiana 47405
We have isolated and characterized cDNA clones of a gene family (P2)e xpressed in Oenothera organensis pollen.
This family contains approximately six to eight family members and is expressed at high levels only in pollen. The
predicted protein sequence from a near full-length cDNA clone shows that the protein products of these genes are
at least 38,000 daltons. We identified the protein encoded by one of the cDNAs in this family by using antibodies to
8-galactosidaselpollen cDNA fusion proteins. lmmunoblot analysis using these antibodies identifies a family of
proteins of approximately 40 kilodaltons that is present in mature pollen, indicating that these mRNAs are not stored
solely for translation after pollen germination. These proteins accumulate late in pollen development and are not
detectable in other parts of the plant. Although not present in unpollinated or self-pollinated styles, the 40-kilodalton
to 45-kilodalton antigens are detectable in extracts from cross-pollinated styles, suggesting that the proteins are
present in pollen tubes growing through the style during pollination. The proteins are also present in pollen tubes
growing in vitro. 60th nucleotide and amino acid sequences are similar to the published sequences for cDNAs
encoding the enzyme polygalacturonase, which suggests that the P2 gene family may functíon in depolymerizing
pectin during pollen development, germination, and tube growth. Cross-hybridizing RNAs and immunoreactive
proteins were detected in pollen from a wide variety of plant species, which indicates that the P2 family of
polygalacturonase-like genes are conserved and may be expressed in the pollen from many angiosperms.
INTRODUCTION
Although the male gametophyte of flowering plants has expressed by the pollen grain itself (Emerson, 1938). Also,
been reduced to the two or three cells of the pollen grain, mutations in a number of genes that have a detectable
these cells are highly differentiated and must perform the pollen phenotype, such as alcohol dehydrogenase {Freel-
complex functions of development, pollination, and fertili- ing, 1976), starch metabolism genes (Demerec, 1924),
zation. For example, late in its development, the pollen gametophytic factors (Jimenez and Nelson, 1965), and
grain accumulates storage materials and becomes able to certain restorers of male fertility (Buchert, 1961), exhibit a
survive some degree of desiccation. Desiccation may aid gametophytic pattern of inheritance, showing that their
in the grain's survival during its journey to the pistil. Once functions are expressed post-meiotically by the pollen
it arrives at the pistil, the pollen grain recognizes whether grain. The mature pollen grains in some species contain
it is on a compatible stigma and then the pollen tube grows stored mRNAs that encode proteins that function during
through the tissues of the pistil, often for long distances, germination and early tube growth; no new transcription
to enter the female gametophyte and release the sperm is required for these events (Mascarenhas, 1965, 1966).
cells for fusion with the egg and central cell. Approximately 20,000 different genes produce mRNAs
Although some functions are provided by sporophytic present in the mature pollen grain of Tradescantia palu-
gene products, the developing pollen grain expresses a dosa, as estimated by r,t analysis (Willing and Mascaren-
large number of genes and has considerable influence over has, 1984). These analyses also suggest that the majority
its own ontogeny. Gametophytic self-incompatibility, which (>64%) of these sequences in pollen are also present in
precludes self-pollination in a number of flowering plants vegetative tissues. This method of measuring overlap of
(including Oenothera organensis), is controlled by genes gene expression has some limitations; for example, related
genes cannot be distinguished. lsozyme studies also show
' To whom correspondence should be addressed. Current ad- extensive overlap between gametophyte and sporophyte
dress: Monsanto Company, 700 Chesterfield Village Parkway , St. gene expression (Tanksley, Zamir, and Rick, 1981; Sari
Louis, MO 63198. Gorla et al., 1986). Seventy-two percent of the enzymes
264 The Plant Cell
examined in corn and 60% in tomato were present in both dicted amino acid sequences of the cDNAs are similar in
sporophyte tissues and in pollen. Only a small PerCentage sequence to the enzyme polygalacturonase.
(5% to 6%) of the 30 to 34 sporophytic isozymes examined
were present only in pollen. Although isozyme studies can
examine the products of single genes, they are limited to IlESULTS
the relatively few gene products that can be analyzed in
this way.
lsolation of Pollen-Abundant Clones
Very little is known about the patterns of expression of
specific genes during pollen development. Early studies
Poly(A)+ RNA was isolated from pollen of O. organensis
showed that the 5s rRNA and tRNAs are synthesized
collected on the day of anthesis and used to construct a
early in microsporogenesis (Mascarenhas, 1975). Similar
cDNA library in the bacteriophage vector XgtlO as de-
results have been obtained for the 25s and 18s rRNAs;
scribed in Methods. To isolate cDNAs representing genes
synthesis peaks before pollen mitosis and no synthesis is
that are preferentially expressed in pollen, the cDNA library
detectable in the last 2 days of development (Steffenson,
was differentially screened. A portion of the library was
1966; Mascarenhas and Bell, 1970). However, several
plated, and replicate nitrocellulose filters were hybridized
papers describing total RNA synthesis note a large in-
with cDNA probes made from RNA from pollen, seedlings,
crease in RNA after the microspore mitosis and just before
and young flower buds. Severa1 clones that showed sig-
anthesis (Mascarenhas, 1975). These late RNAs are het-
nificantly higher signals with the pollen probe than with the
erogeneous in size and probably reflect mRNA synthesis
seedling and floral bud probes were chosen for further
late in pollen development. With the advent of the tech-
characterization. These clones fel1 into several categories
niques of molecular biology, it has been possible to ex-
based on their RNA accumulation patterns (S.M. Brown
amine the expression of genes during pollen development
and M.L. Crouch, manuscript in preparation). Pollen clone
in much finer detail. The isolation of cDNA clones corre-
2 (P2) was chosen for further analysis because it repre-
sponding to single- or low-copy genes expressed in mature
sents a class of mRNAs that is differentially expressed
pollen grains of T. paludosa, corn, and tomato has recently
and very abundant in mature pollen.
been described (Stinson et al., 1987; Hanson et al., 1989;
Ursin, Yamaguchi, and McCormick, 1989).
We have chosen to examine pollen gene expression in P2 cDNA Represents a Medium-Sized Gene Family
O. organensis because it has a number of desirable char-
acteristics. It is easily grown in the greenhouse, where, To estimate the number of genes in the Oenofhera genome
under long day lighting conditions, it produces a steady that are related to the P2 cDNA, a DNA gel blot was
supply of flowers year-round. The flowers are extraordi- prepared from nuclear DNA that had been digested with
narily large, allowing easy dissection of individual floral several restriction endonucleases and probed with the P2
parts. The pollen grains are also large-more than 200 cDNA insert at moderate stringency. The size of the P2
pm (Dickinson and Lawson, 1975)-and approximately 7 gene family was approximated based on the number of
mg of pollen can be collected from a single flower. O. bands in each lane and the intensity of the bands compared
organensis has also been the classic species in genetic, with copy standards (described in Methods). Figure 1
biochemical,a nd immunologicals tudies of the role of pollen shows the results of the genomic DNA gel blot analysis.
in pollen-pistil interactions and self-incompatibility (re- We estimate that the P2 gene family contains approxi-
viewed in DeNettancourt, 1977). These studies have pro- mately six to eight genes.
vided valuable information about O. organensis pollen de- To determine whether the majority of the members of
velopment, germination, and growth. Although little infor- the P2 gene family are expressed in mature pollen, a
mation is available in the literature about its molecular portion of the cDNA library was rescreened to isolate
biology, O. organensis has proven to be a good source of additional cDNA clones. Approximately 1.2% of the se-
material for tissue culture and has been successfully trans- quences in the pollen cDNA library hybridized to the P2
formed in our laboratory using standard Agrobacferium cDNA insert probe. More than 50 additional clones were
methodology (C. Scelonge and M.L. Crouch, manuscript isolated; the 13 clones with the longest inserts were sub-
in preparation). cloned and mapped with restriction endonucleases. Figure
In this paper, we report tne identification of cDNAs that 2 shows a comparison of the 14 cDNAs. The cDNAs can
represent a gene family expressed primarily in pollen and be grouped into six different classes based on their restric-
present the results of analyses aimed at determining the tion maps. Four unique clones (P2, P22, P25, and P39)
possible roles that this gene family may play in pollen were isolated, each having a different combination of re-
development and/or germination. The temporal expression striction sites. The fifth class includes the five clones (P1 ,
patterns suggest that the products of these genes prob- P60, P96, P102, and P107) that have maps identical to
ably function late during pollen development and/or during the P1 cDNA. All are quite different from the unique clones.
germination and tube growth. The nucleotide and pre- Five other clones (P26, P34, P58, P67, and P85) form a
Polygalacturonase Expressionn i Pollen562
73 Nucleotide Sequences Show Similariotyt
'o Polygalacturonase
Q.
CO The nucleotide sequene cehet nfsot ire cDNA inserts were
obtained for two clones, P2 and P22, along with partial
sequence for several others, to obtain information about
Q.
possible functions for the P2 gene family. The two cDNAs
O
P E ' O show 87% sequence identity, 89% in the region in the
c ° ±
(Q O — open reading frame and 75% in the 3'-untranslated region
OQ UJ I
(data not shown). The open reading frame in the longer
P22 cDNA extends fre ofhimrts t nucleotido nteu cleotides
1o 018A t0s78 toA9p, Twcoh deaore n presumably ends
translatie oohnTto h. fweratr mes contain numerous stop
codons. Bece aufhisrtse t methionine codon tdoones
occur until nucleotide 322, this cDNA insert probably does
not contain the entire protein coding sequence. The P22
cDNA does encode more than 38,000 daltons of protein,
approan icv hieve ihnhosttg i zfoper oteins detected
immunologically (see below).
ehaTmino acid sequene chot efop en reading frameni
2 se2quPethne ce shows significae anmth isntimo oilartit y
BH
P22
P25
H PK P
P39 I ||
H P SG P
PI I I II
— P2 H P SG P
— P1
P6D I I II
H P SG P
P96 I 1 1
H P SG P
P102 I I II
H P SG P
P107 I I II I
P26
P34
B H K
. Oorg al feBnoGelon ts Ais MGeDnoFm ig.ic1u reS equences I I I
That Hybridize to the P2 cDNA. B H K
I I I
Five micrograms of nuclear DNA digested with the enzyme indi-
B H K
cated and the appropriate amount of copy standard were electro- I I I
phoresed, blotted, and probed with radiolabeled P2 cDNA insert,
as described in Methods. __ = 100bp
Figure 2. Restriction Endonuclease Maps of Pollen cDNAs Ho-
mologous to the P2 cDNA.
sixth class, each characterized by a restriction map iden-
The EcoRI insee hprtt foso llen cDNA clones were digested with
6 c2DPN eAh ttteih cos hadaitt t il fffoe rpeanmt f reohmt a variety of restriction endonucleases to generate the restriction
P1 class or the unique clones. These 14 cDNA clones maps shown. Abbreviations for restriction endonucleases are: B,
represent atl eas xisut nique mRNAs. BamHI; G, Bglll; H, Hindi; K, Kpnl; P, Pstl; S, Sstl.
266 The Plant Cell
PG-TOMATO VDKNGIKVINVLSFGAKGDGKT... e ohn2 slyte2 qddPuifnefe anaerce ehi nsdcetien oticta l
I I I I I I I I I I I I conservative change from the isoleucine in PG to the valine
Eifi. CTITNAQLFDITKYGAKGDGAT...
in P22. The predicted P22 amino acid sequence, like the
tomatoG P sequence, contains four glycosylation sites.
tTochonenss eee rsrn vipateeiod s sitieonh, te xrcoefp t
PG-TOMATO
0 9YDNIAFEQANNEACSSRTPVQFWPKNKNYLLKQITFSGPCRSS9t3SV1KIF Asn-lle-Thr (286 to 288 in the tomato sequence), which is
I I I I I II I II I I I III I I I I I I I I I I I I II present in the highly conserved 285 to 301 region.
1 DSTQALTTAWKEACASASPSTILVPKG.NFAVGLITI.EGPCKSS1GLQ9L4Q
0 4G1SLEASSKISDYKDRRLWIAFDSVQNLWGGGGTINGNGQVWWPS9S8C1KIN
II I II I I I I I I I I I I I I I I I I P2 RNA Accumulation Patterns
05GTLKAPADPSKIKGLG.WINLNKIDLLTJFGGGVFDGQGKSAWVQND8C9HK
0 91KSLPCRDAPTALTFWNCKNLKVNNLKSKNAQQIHIKFESCTNVVASNL93M2I To determine hset patial accumulation patt2 ePtrr fano n-
II I I I I III I I I I I III III
99 NGPICKTLSMNLRLYAVTNSILRDVTTLDSKNFHVNVIGCKNLTFERFKI 148 scripts, an RNA gel blot was prepared from RNA isolated
from mature pollen, leaf, stigma, ovary,d na seedling tis-
0N4A2SAKSPNTDGVHVSNTQYIQISDT1IGTGDDCISIVSGSQNVQA9T8N2IT C
I I I I I I I I I I I I I I I I I I I I I I I I I I I I II I I sues and probed with the P2 cDNA insert. Figure 4 shows
149 SAAETSINTDGIHIGRSDGVNIINTEIKTGDDCISLGDGSKNINITN8I9T1 C
2 t1 r.a5Pn-e oksebn crlryhaip thdtt sae ttectn apboille len,
290 GPGHGISIGSLGSGNSEAYVSNVTVNEAKIIGAENGVRIKTW.QGGSGQA 338 not in the sporophytic tissues examined.
I I I I I I I I I I I I I I III I I I I I I I I I I I I I I I I
199 GPGHGISVGSLGRYKNEESWGIYVKNCTITGSONGVRIKTWPKSEPGEA 248 e timhf aoiTnccg2 muPmRuN lafAtoiso nd uring pollen
339 SNIKFLNVEMQDVKYPIIIDQNYCDRVEPCIQQFSAVQVKNWYENIKGT 388 development was also analyzed. In O. organensis, the
I I I I I I I III I I I I I I I I I I I I I I I I I I
249 SEMHFQDITMNSVGTPILIDQGYCPYNQCTAEVPSSVKLSKISFKN1KGT 298 length of the hypanthium is a very good indicator of a
variety of events in floral development, including the stages
9 83SATKVAIKFDCSTNFPCEGIIMENINLV.GESGKPSEATCKNVHFNNAE7H34
I I I II I I I I I I I I I I I I I I I I I I I III I of pollen development (S.M. Brown and M.L. Crouch,
992STTKEAVKLVCSKSFPCNGVELADIDLTYSGKGGPATSVCENIKPTIK84G3K
manuscript in preparation). The tetrad stage occurs when
438 VTPHCTSLEISEDEALLYNY* 457 the hypanthium is approximately 0.1 cm, the free micros-
I I
349 QIPAICSGSAAKAA- 362 pore stage extends from 0.15 cm, and the pollen are
binucleate in flowers with hypanthia greater than 1 cm. At
Figu. r3eC ompariso foPn redicted Amino Acid Sequences from
hypanthia lengths greater than 6 cm, the pollen begin to
O. organanesis Pollen cDNAs and a Tomato Fruit Polygalacturon-
reach functional maturity (can cause seed d sfenutal)l ,
e cDsNAa.
Thep ot panel showsa comparisone ht fo predicted amino acid
sequence of the P26 cDNA (which contains 66 nucleotides more
' 5 sequence than2 2P eht cDNA) withe ht predicted amino acid
sequa epn ofcolye galacturonase cDNA (Grierst oean l., 1986).
Each vertical line represenn tiasd entical amino aca c irdoo nserv-
ative change. The horizontal line highlights the first 4 amino acids
e mohtf ature polygalacturonase proten itnio mato fru elhoiTtw. er
panel shows a continuation of this comparison, now using the
predicted amino acid sequence of the P22 cDNA with the homol-
ogous region of the predicted amino acid sequence of the same
tomato fruit polygalacturonase cDNA (full-length sesqauewnce
not ob6 it2naPs ienhete rrotd)f . Each sequence ends witha s top
codon r (ob*Fr)e. e vnihtuytc ,leic acid se toqpnuree en-craes
sented. These sequences are available in Genbank or from the
— 1.5kb
authors.
acid sequencef o polygalacturonase from tomato fruit
(Grierson et al., 1986). Figure 3 shows the comparison of
the predicted amino acid sequences from the pollen and
fruit polygalacturonase cDNAs. Overall, 54% of the 382
residues e raidr 2ee2pPnrt diecnasrao e6l 2nPte ybd Figu2 rPme .4 RNAs Accumulate Onlyn i Pollen.
substituted with a similar amino acid in the tomato poly-
Five micrograA f mfotrNoos tmRa lvarious organs were electro-
galacturonase (PG) sequee nnhucTecl .eotide comparison
phoresed, blotted to a nylon membrane, and hybridized with
is similar; the cDNAs are identical over 58% of the length radiolabel2 ecPdD NA inses rdat, escribe ndiM etho edfhisTn. al
2 c2DPN Ae.h Tote hfsertarrei king regiof onids entical wash conditions were 1 x SSC, 60°C. The autoradiogram was
amino acidr soef; xamp ehrltee nig, ion from amino acids overexposed to emphasize the absence of the 1.5-kb transcript
era 71 f eG otsohemP qt6ua et1no ncie , 103 ot 582 e planht.int o thfero par ts
Polygalacturonase Expression in Pollen 267
maturity (can cause normal ratef oss eed ses tai) chieved A
by pollen in flowers with greater than 9-cm hypanthia. Final
hypathium length is an average of 15 cm.
Anthers were collected from flowers at a variety of Developmental Stage
developmental stages, and the pollen was isolated from
(hypanthium length (cm))
e ahne ttr hhefeotsr t matey rfiibalt rl dasnteioadni menta-
tion through sucros sapw eArNeR. pared frome hit solated
pollen and an RNA gel blot was probed with the P2 cDNA
insert. FigureA 5 shows tha2 Ptt ranscripte sraf irs-edt
tectable at the 1 cm to 2 cm hypanthium stage. At this
stagf ode evelopmee nphott, llen have just undergoenhet
mitotic division that prod eughceentse d rvanetgiavee ta-
tive ce2 Pml leshRT. NA levels increase throughoeuhtt
maturation phase feo elplrohoawllteinn g ddniviasion
highest at anthesis.
Quantitative dot blot analyses were performed to deter-
mine the kinetics of accumulation of P2 transcripts. Total
RNA was immobilized on duplicate filters and hybridized
B
wit2 Ph echt DNA a roc lone containinga regioenht fo
ribosomal repeat froe rmhat dish nuclear genome (Delseny
et al., 1983). The radioactivity that hybridized to each dot
was quantitaty esbdc intillation counting. FiguB sr5eh ows
140
e htresultsf o these analyses. Each point indicateesht
amount of P2 mRNA relative to rRNA at each stage of
*~* 120"
Oenothera pollen developm eerhneTts. ults indicate that (£>
the level of P2 transcripts increases at least 200-fold during i 100-
o
their induction in pollen development.
RNA from mature pollen was used to determine the
absolute amount of P2 transcripts relative to total RNA
(data not shown). Total RNA was quantitated by UV ab- ^ 60
z
sorbancd ensa, erial dilutions were appliedn it riplicateot
QC 40-
nitrocellulose usinga commercial slot blotter. Serial dilu-
tions were also made using in vitro synthesized antisense
RNA2 sc PDmNaA d eein hsfretormt subcloned into n1 20-
Bluescript. The filters were probed with the P2 cDNA insert
e hht ydbnaridization sig snaqawul antitatey bsd canning
1-22-3 3-6 6-8 • 10cm
densitometry. These analyses indica2 ftPaem etihlhyatt
of mRNAs constitute 0.013% w/w of the total RNA in the Figur. 5eE xpres2 smPio RfonN As during Pollen Development.
mature pollen grain and that the P2 mRNA level does not
(A) Four micrograms of total RNA from various stages of pollen
change significantlyn i responseo tt emperature.
development (as indicated by the hypanthium length of the floral
bud) were electrophoresed, blotted to a nylon membrane, and
hybridized with radio2 cs ladDbPeaNeslcAen rd, Mibieetd hods.
Protein Identificd Aantciocanu mulation The final wash conditions were 0.1 x SSC, 65°C.
AeNxR p2Pre )Bs(sis oaqwnu at nobdt iAltoNaotRwtseT y. bd
An antibody agae inphsott lypeptide2 Pe nechotd yebd micrograms of denatured total RNA from various stages of pollen
s pcarDowdNuAo cidete edp nrhotitftey ine phrotd ufocts development were spotted onto nitrocellulose filters. Replicate
P2 gene family and to characterize their expression pat- filters were hybridized with radiolab2 ecPlDeNd A. Dilutions
ternsA . 750-bp fragment (frome hBt amHI '3s eihtt eot (1/1000) of the RNA samples were also spotted and hybridized
wa irtah diolabeled pr roroibbfoe A s(poNRmDEa1l 2eR)hNTA .
end of the clone) that encodes the carboxy-terminal 213
S52 dna S81 ee probeh ihncludteso tsb yontthhe stize u sed
amino ae hcotip dfoes n reading2 cPfDr eaNhmt Anei
rDNA from radish (Delseny, Cooke,d na Penon, 1983). Relative
was fused with the truncated /3-galactosidase gene in the
abus dneadtaewnrmy cse icbninetd illatioe ns phcoottus n.ftoing
three members of the pWR590 series of expression vec- The value rosef ach stage were determi enhrtee ybdla tive level
tors (Quo et al., 1982). Only the pWR590-2 construct of hybridization to the P2 and rRNA probes and corrected to the
producea dla rger fusion protein whea np ollen cDNA absolute valuf o0e .013% P2/rRNA determinr moefda ture pollen.
frag sminaeswne trted. This result confie romhptes n read-
g niframe predicted fromA ND s2P eht equence. Figure6
268 The Plant Cell
proteint on,w ith (3-galactosidase moietyr o with othe.Er
coli proteins.
ehpTreabsorbed antiserus amuws e odti dent2ifPy
a: ^ proten ineisx tracts from mature pollen gras insAhs .own
CC
3 o
g Q. O- in Figure 6, the antiserum reacts with a group of proteins
ranging in size from 40,000 D to 45,000 D. Several minor
bands (35,000 D to 40,000 D) were also detected when
more extract was loaded per lane. The relationships of the
— B-gal-P2 —
(90 KD) various polypeptides detected by the P2 antiserum were
B-gal — not examined.
(68 KD)
To de eathepe rptmtei hmfaid notiranean cgn cacue -
m2 Ppur leoahtttie ofionn s during pollen development,
P2 proteins i—
_|(40-45K D) extracts from immature pollen isolat vateadr ious stages
of development were examined by immunoblot analysis.
Figu7 rse ho nwiams munoblot probed e wphitrthe ab-
sorbed antiserum. The P2 antiserum did not detect the
cross-reacting polypeptides in the 1-cm- to 3-cm-stage
pollen, although P2 transcripts are detectable at this time
Coomassie Blue Stain Immunoblot (described abo2 pPv reeoh)Tt.e ie nfraisr st detectabltea
low leve eh3tl sn-i cmo -t7 -cm-stage polle dnnian, crease
Figure 6. Identification of P2 Family Polypeptides in Pollen.
in both the 7-cm- to 10-cm-stage and mature pollen.
The left panel showa Cs oomassie Blue-sta% inSe8Dd S-poly- Interestingly, the various polypeptides do not accumulate
ac eriyhnltas fomol uliedbgel e proteins fro. m£c oli JM101 cells coordinately throughout pollen developm ehheigTnht .est
containine ghpt WR590-2 expression plasmid withoun iatn serrot
e7wh5 2i0inttP-hs bep rt (pWR590e -r2igh-PhT2t ) p.anel shows
an immunoblot of the insoluble proteins from the host E. coli cells
with on expression plasmids (JM101), containinge ht expression
plasmid withoutn a insert (pWR590-2), withe ht 750-bp2 P insert Hypanthium Length (cm)
(pWR590-2-P2e h)t dsn,a oluble proteins from mature. 0 orga-
nensis pollen (pollen)e h.T primary antibods yaw 1:500 preab-
sorbe 2adP ntiserume hst, econdas ra1wy: 2500 goat anti-rabbit
horseradish peroxidase, and the cross-reacting bands were visu- CO f-• CM
alized ewhsituthb strate 4-chloro-1-naps thadoel, scribneid
CO A
Methods.
48KD-
shows the proteins expressed in Escherichia coli contain-
ing the pWR590-2 expression plasmid with and without
the P2 insert. JM101 cells containing the pWR590-2
expression plasmid without insert produca e6 8,000-D
truncated /3-galactosidase protein. The cells containing the 36KD -
pWR590-2-P2 expression plasmid produca e9 0,000-D
e fuhstpie or sofnhiotz eTepi nr o.t seciion nsistent with
e hntucleotide 2 scPeDq eNu0heA0tn6 fco peel huts
amino acids contributed by the truncated (3-galactosidase.
26KD-
The 90-kD /3-galactosidase/P2 fusion protein was gel
purifd ienuads r eoidfm munizatie ohrnTe.s ulting poly-
clonal antiserum showed strong cross-reactivity agains-t0 Figure 7. Expression of P2 Family Polypeptides during Pollen
galactosi ddonatsa .hEcee or li prote oinrTesm .eohvte Development.
nonspecific antibode iehatsn, tises rapuwrme absorbed
Forty microgramf soc rude protein extract from various stagefso
with proteins isolated from pWR590-2-containing JM101 developing ps oie nlalhedh(ynit cp ayatnbetdh iume hletn fgoth
cells. The predominant protein from these cells is the floral bud) were electrophoresed through a 12% SDS-polyacryl-
68,000-D truncated /3-galactosidase protein. The immu- d enlaeca tlmreoigbdleo tted onto nitrocelle ulphorTsime .ary
non bFiloig tu6 rde emonstratese phthrteaa tbsorbed antibody was 1:1000 preabsorbed P2 antiserum; the secondary
antiserum reacts only with the 0-galactosidase/P2 fusion antibody and visualization were as described for Figure 6.
Polygalacturonase9 E6xp2ress nPiooin llen
P2 Homologe srEA xpressed ni Pollenn iO ther Plant
Species
Q
F0 sigh1uora web ls ot cA foronNtmai nRminag ture pollen
of several different taxa and probed with the P2 cDNA. A
UJ
(0
homologous transcript is present in all species examined.
UJ
ffi K
3 CO The sizee sht fo transcripte sht ni various speciesera
it !i I h- UJ sime ihlta1 ot.r 5-k. Obo rganens2 iPsm RtNonA tub,
I i
ui %
„, c. identical. There are also some higher molecular weight
speciesd . cpvAonritret anasn teta hina tte hhybtr idoizte
P2 probe.
Th2 eP antiserums awu sedo t probe immunoblots
jWWyrett
containing crude extracts of corn pollen proteins. The P2
antibodies deteo pwctrt otein bandn sci orn pollen thearat
PROTEINS I
approximately d 44n02a,,00 D D00(0d0t osanhtao wn).
Several minor bands were also detected, includinag
38,000-D band similar to the minor bands observed in O.
organeo nwtls adniarsg er polypeptides (approximately
53,000 D and 100,000 D) that have no detectable coun-
Figur 2PeP .8 olypeptidese rAP resent Onlyn i Pollend naP ollen
Tubes. terparts in O. organensis.
eht f fsoolouble p0 a r ^Fopgot aerri tn4e afrco hm lan e,
Oenothera plant were loaded onto a 12% SDS-polyacrylamide
DISCUSSION
gel, electrophoresed, and electroblotted onto nitrocellulose. Anti-
body incubations and visualization were as described in Figure 6.
We have demonstrated that a polygalacturonase-like gene
famils yih ighly expresse ndim aturing pollen gra.iOn fso
orgae ntheenmTspiso .ral patf toeetrxon pnres ssioin
molecular weight species appears first in the 3-cm- to 7-
unlike that observed previously for genes expressed pri-
cm-stage pollen, whereas the smaller polypeptides are not
abundant until the later stages.
The antiserum sawa lso used ot characterizee ht pres-
ence of P2 proteins in various parts of the plant and during
pollination. Figure 8 shows an immunoblot containing pro-
teins extracted froma varietf yo pae hrtpt fsola n.Ot.
organensis is a self-incompatible plant, and pollen tubes
are arrestede ht nis tigmasf o self-pollinated pistilosd dna
not enter the styles. P2 proteins were found only in pollen
d sntayles that contain pollen tua rb eseassuf oclt ross-
pollination, indicatin2 gpP r otehtheat eit prnarse sennit
pollen tubes as well as in pollen grains, but not in other
parts of the plant.
odeTtere mphr2 ianetnPes tei gnfen copne ois llen tubes
directly, pollen tubes grown in vitro were also examined.
Pollen was germinated on solid culture medium, and a
"prins atmw" ady ebb riefly placinga damp piecef on itro-
cellulose one gthote rminating polle ephnTo. llen antigens
bound to the filter were assayed by incubation with the P2
antiserum, followey dbd etectioy nbg old-conjugated sec-
ondary antibod dsniielavse r enhancement. Figur9 ed em- Figure 9. P2 Antigens Are Present in Pollen Germinating in Vitro.
onstrates the presence of P2 antigens in pollen tubes
Antigens from pollen germinan tviniitg ro were "printed" onto
grow nnvi 2 piPt rreoohTt. eine prsar esent throughoeuhtt
nitrocellulose and detected by the P2 antiserum. Binding of the
ent tioar eepnp hpol eetollane dfgrno t hldt oucnbaael-s primary antibos vadwiys ualizey cbd olloida) lms geno c0lo1d(n d-
izea d sni pecific regiot onnp si t.oI ssibleo t determineyb ary antibodd sineialsv er enhancement. This photograph showas
this technique whether the antigens are present in the print of a single pollen grain and its tube. The P2 antigens appear
pollen tube cytoplasm, in the wall, or in both. in the grain and throughout the entire length of the tube.
270 The Plant Cell
E copy numbers (Stinson et al., 1987; Hanson et al., 1989;
3 Ursin et al., 1989). Six characteristic combinations of re-
striction sites were found in the cDNAs, suggesting that
e whave isolated clones representingt a leasx istd ifferent
CO
Q. o transcripts. The simplest interpretation is that the different
O '5J cDNA clones represent transcripts from different genes. If
0) C
Oe CO N *- CO this si true, me hotg fo sllea rton 2 Pef eahts nim ileyra
Oo E CQO. GCtmOO epxapttreersnsse fdo uinnd moantlyu roen cpeo lilne nth. eF 1o4u rc DcNloAness; t whoa do threesrt rciDctNioAn
classes were isolated five times each. The isolation of
diffen reeanf ccth lioon cnuD emsNbAe r sclass suggests
e tmhhatR tNAs produced fe rvohamtr ious gene family
members mighe pbt resent adt ifferent levelf soa bundance
in mature pollen.
The protein produc2 Ptg ehst feo ne family were identi-
fied by a polyclonal antiserum raised against a /i-galacto-
sidase/P2 fusion protein synthesized in E. coli. The P2
fam fiolpy olypeptids efiisr st detectabn flileo wers wi3th
cm to 7 cm hypanthium lengths, later than the period when
the mRNAs are first detectable. This lag in accumulation
l belot agnaly seAs NRmay se simefpnhlsoyi tirveti ftlye ct
versus immunoblot analyses. yAaltmer ngaatliv eehlyt,
indicate translationar ol post-translational regulatiosni tI.
interestie nvgah ritohtua st ta opcocnulymp euop-tdide s
late coordinately. The relationship of the various polypep-
tides d2 Peat neehtctist yeet bodrnc u lsmeieah Tr.
different polypepte hitd pebe ryaoms ductsf o different
members of the P2 gene family. Alternatively, the different
polypeptides coe uhtr leedb suf lot post-transcriptional
Figure 10. RNA Gel Blot of Homologous RNAs in Other Plant
changes or post-translational changes such as differential
Species.
protein processing, protein degradationr o, variationsni
One microgA rfaf rNotomom Rta lmature Oenothera pdonllean glycosylation. Transcripts fro2 mhP omologs were present
10 M9 of total RNA from mature Amaryllis vittata, maize (Zea
in pollen froa wm ide varf iosepty ec. i0eosr g.anesnsiis
mays), Spathophyllum Cleveland!/,d na Brassica napus pollen were
a dicot with bicellular pollen grains. The other species
electrophoresed through a 1 % formaldehyde gel and blotted to a
examined includea dd icot with tricellular po. llnBea(n pus),
nylon membrae nbhs eTahlo.w yt bridized wita hr adiolabele2Pd
a monocot with tricellular pollen am (odcnonoranc)o ,t
probe under permissive conditions (hybridization and final wash
7 -36e °fhiClTm). ss haeoxww4 pnwo resoeefkd s, with with bicellular pollen (A. vittata). The fact that the homolo-
m
intensifying screens, at —70°C. gous transcripts coe udblde tecten sid uch diverse taxa
suggests that polygalacturonase-like genee srca onserved
and may be expressed in the pollen from most angiosperm
species.
marily in pollen. Pollen-specific mRNAs from corn and 7. Hanson et al. (1989) noticed that the consensus AA-
paludosa bo aecgtcin umulae tep ohallfettend r mnitoasi s TAAAs aw locatedn a abnormally long distance (180
their levels peak in mature pollen grains (Stinson et al., bases) upse sthrf ieptote ao folmy adenye plhaottl ilnoenni -
1987). This temporal expression patts eaarwnl-sb oo specific cDNA Zmc13.e hT maize alcohol dehydrogenase
served for an unrelated Oenothera pollen cDNA (P3) that gene (ADH1), which is expressed in pollen as well as in
was isolated in the differential screen that identified the P2 other partse ht fo plant, alsoe ht sah AATAAAa long
cDNA. However, all pollen-specific genes are not coordi- distance from ehst itef oa ddition (Sacht sea l., 1986eh)T.
nately regulated; a third pollen cDNA (P6) isolated in this distance betw eAehAetTn e pAhoAtlyA ad dmneoantyi fl-
screen does not accumulate to detectable levels until just ation site in the P22 sequence is 126 bases. However,
before anthesis, several days after expression of the P2 other cDNAs in the P2 family, such as P1 and P2, do not
gene fams ifliiyr st observed (S.M. Bd rMonw.aLn. Crouch, have this feature; the distance between the consensus
manuscript in preparation). AAe f TsdophiAtonet Aly aAad d nm9eano1yti lf astioin
The P2 gene family was estimated to consist of six to 18 bases, respectively. The distance between the AATAAA
eight gen reenpsu cleus, unlike pollen-expressed genensi e ahddt idtimonnoa tsi n fmitie ost pl9an t± m 7R2N sAis
other species showne b otp resentn i singler o verywol nucleotides (Joshi, 1987).
PolygalacturonaseE xpression in Pollen 271
Complementary DNAs representing the P2 gene family P2 proteins in pollen tubes is consistent with these hy-
show extensive sequence similarity to the enzyme poly- potheses. We are currently assaying the P2 proteins to
galacturonase (PG: poly(l,4-a-~-galacturonide)glycano- determine whether they have polygalacturonase activity
hydrolase, EC 3.2.1.15), which has been studied during and whether they can degrade the pistil cell walls.
fruit ripening in tomato. Polygalacturonase cDNA clones
have been isolated by several groups, and the sequences
of two PG cDNA clones have been determined (Grierson METHODS
et al., 1986; Sheehy et al., 1987). The similarities of the
predicted amino acid sequences from the tomato fruit and
the pollen cDNAs appear throughout the lengths, but are Plant Material
especially striking in a few regions.
The products of the P2 gene family have several other Oenotbera organensis Munz (Emerson)p lants (kindly provided by
Dr. Adolph Hecht, Washington State University) were grown in a
similarities to the PG gene products. The mRNA corre-
greenhouse using supplemental lighting (mercury vapor or incan-
sponding to the P22 cDNA is 1531 nucleotides long (ex-
descant lamps at 1O 0 to 200 wEm-*sec-', 4 AM to 8 AM and 4 PM
cluding the polyA tail) as determined by DNA sequence
to 8 PM) to extend the natural daylength to 16 hr. Temperatures
and primer extension analyses (data not shown). This
during the tissue collection periods usually ranged from 65OF to
length is similar to the 1.6-kb size of the tomato PG mRNA. 82°F. All tissue was collected from a population of plants vege-
The PG mRNA is very abundant when at its maximum tatively propagated from a single individual.
level in ripe tomato fruit, where it constitutes 1.2% of the Mature pollen was collected from O. organensis flowers in the
mRNA (Bennett and DellaPenna, 1987). P2 mRNAs are at evening, just as the flowers opened, using a suction apparatus
a similarly high level (0.013% of total RNA) at their peak in consisting of a Pasteur pipet attached to a water aspirator. Pollen
mature O. organensis pollen. The mature tomato PG pro- from the dehisced anthers was trapped within the pipet by 4
layers of cheesecloth. lsolated pollen was immediately frozen in
tein is predicted to be 42,000 D and appears to be ap-
liquid NPa nd stored at -7OOC.
proximately 46,000 D in size on SDS-PAGE, a size similar
lmmature pollen at the appropriate stages as determined by
to that for the P2 family of pollen proteins, which appear
hypanthium length (S.M. Brown and M.L. Crouch, manuscript in
to be 40,000 D to 45,000 D on SDS-PAGE.
preparation) was isolated by gently disrupting anthers in 15%
One major difference between the P2 family and PG sucrose using a loose-fitting sintered glass homogenizer at 0°C.
gene in tomato is their genome organization. The tomato The majority of the anther tissue was removed by filtration through
genome appears to have a single copy of the PG gene a single layer of cheesecloth. The immature pollen was pelleted
(DellaPenna, Alexander, and Bennett, 1986), whereas the away from remaining debris by centrifugation (<lOOOg, 4OC) and
O. organensis P2 gene family was estimated at approxi- washed several times with 15% sucrose. The final pellet was
mately six to eight copies by DNA gel blot analysis. It is frozen at :7O0C until use.
possible that multiple PG genes also exist in the tomato Pollen from Spatbophyllum clevelandii, Amaryllis vittata, and
corn (Zea mays) was collected from greenhouse-grown plants
genome, but were not detected on DNA gel blots with the
and stored at -7OOC.
fruit cDNA probe because they have diverged significantly
from the gene that is expressed in fruit.
The role of PG in tomato fruit is thought to be digestion
RNA lsolation
of the middle lamellar region of cell walls, which may
contribute to softening of the fruit during the ripening
Total RNA from all tissues was isolated by phenol extraction as
process. Polygalacturonase activity has been described
reported by Finkelstein et al. (1985). RNA from Brassica napus
recently in the pollen of a number of grass species, includ- (Topaz) mature pollen was a generous gift of R. Nolan. Poly(A)+
ing maize (Pressey and Reger, 1989). It is likely that the RNA was isolated from total RNA using oligo(dT)-cellulose( Sigma)
P2 family of proteins has polygalacturonase activity, given according to Maniatis, Fritsch, and Sambrook (19 82).
their extensive similarities to the tomato PG protein. If so,
the proteins may function by depolymerizing pectin in the
cell walls of the pistil during pollination to allow penetration Construction of cDNA Library
by the pollen tubes and/or to provide wall precursors for
the rapidly growing tube. Another possible function of An O. organensis mature pollen cDNA library was constructed by
pollen PG is to act on its own wall to facilitate cell elonga- modifying published procedures (Huynh, Young, and Davis, 1985;
Gasser et al., 1989). First-strand cDNA was synthesized from
tion. In plant/pathogen interactions, polygalacturonase ac-
poly(A)+ RNA by incubating 20 ng/wL RNA; 20 ng/pL oligo(dT)l2-
tivity has been shown to release pectic wall fragments
18 (Pharmacia LKB Biotechnology Inc., Piscataway, NJ); 50 mM
that, when added exogenously, are capable of stimulating
Tris-HCI, pH 8.3, 40 mM KCI, 8 mM MgCI,, 40 mM DTT, 0.5 mM
physiological changes such as ethylene biosynthesis, pro-
each of four deoxynucleotide triphosphates, and 0.6 units/wL
teinase inhibitor l, and phytoalexin activation (Giovannoni reverse transcriptase (Seikagaku) for 90 min at 42OC. Homopol-
et al., 1989). Release of oligosaccharides may also be ymer dG tails were added to the purified products of the first-
important in pollen/pistil interactions. The presence of the strand synthesis in a reaction mixture consisting of 90 mM Na-
272 The Plant Cell
cacodylate, pH 7.0, 0.9 mM dGTP, 1.8 mM COCI,, and 50 units/ were prepared by random oligonucleotide priming (Feinberg and
25 LI terminal transferase (Boehringer-Mannheim, Indianapolis, Vogelstein, 1983; Hodgson and Fisk, 1987) or nick translation
IN) at 37°C for 2 hr. The reaction was terminated by the addition (Maniatis et al., 1982). Final wash conditions were 0.1 x SSPE,
of an equal volume of 10 mM Tris-HCI, pH 7.2, 4 mM EDTA, 0.1'% SDS at 65°C.
followed by treatment with RNase A (0.25 19/50 pL) to destroy Copy number standards were prepared by calculating the
the mRNA template. The oligo(dC)-primed second-strand synthe- amount of cDNA insert that would correspond to a single copy
sis reaction contained 0.6 pg of oligo(dC)12-18 (Pharmacia), 35 gene in the Oenothera genome [2C = 3.04 pg, as determined by
mM MgCI,, 0.3 mM each of four deoxynucleotide triphosphates, flow cytometry of nuclei from seedlings (D. Galbraith, personal
and 40 units/60 pL Klenow fragment of E. coli DNA polymerase I communication)]. The appropriate amount of each subclone was
(New England Biolabs, Beverly, MA) and was incubated for 2 hr digested with EcoRl to release the cDNA insert from its plasmid
at 37°C. lhe duplex cDNA was methylated without further puri- vector, added to sheared salmon sperm DNA, then electropho-
fication using EcoRl methylase (New England Biolabs), and the resed, blotted, and probed along with the Oenothera genomic
methylated cDNA was repaired with T4 DNA polymerase (Be- DNA.
thesda Research Laboratories,G aithersburg, MD) under standard
conditions (Maniatis et al., 1982) to produce blunt ends. EcoRl
linkers were ligated to the double-stranded, methylated cDNA RNA Analysis
using T4 DNA ligase (New England Biolabs) under standard
conditions (Bethesda Research Laboratories).T he products were Total RNA was initially quantitated spectrophotometrically. If the
digested with an excess of EcoRl and the cDNA, then size spectrophotometric AZ60/A280ra tio was less than 1.7 (as it fre-
selected and purified by electrophoresis in low-melting-pointa ga- quently was with Oenothera tissues other than pollen), the RNA
rose (Seaplaque, FMC). cDNA fragments between 0.5 kb and 12 was also quantified using formaldehyde gel electrophoresis (Man-
kb were excised from the gel and purified by extraction and Elutip iatis et al., 1982) ora dot assay modified for use with RNA (Sharp,
chromatography (Schleicher & Schuell, Keene, NH). 1985). The dot assay was also used to confirm the spectropho-
Approximately equimolar ratios of the size-fractionatedc DNA tometric quantitation of poly(A)+ RNA. Concentrations of total
and Xgtl O arms (Stratagene, La Jolla, CA) were ligated as above RNA samples for dot-blot or RNA gel blot analysis were confirmed
and packaged in vitro using Gigapack Gold (Stratagene) according using an rDNA clone (pRE12, Delseny et al., 1983) as a hybridi-
to the manufacturer's instructions. zation probe for dot-blot analysis.
RNA gel electrophoresis was performed in the presence of
formaldehyde essentially as described by Maniatis et al. (1 982),
lsolation of Pollen-Abundant Clones
and the RNA was transferred to nylon membranes (GeneScreen,
Du Pont-New England Nuclear, Boston, MA) according to the
3'P-labeled cDNA probes from poly(A)+ RNA were prepared using manufacturer'si nstructions. RNA dot-blot analysis was performed
a random priming technique as described by Gasser et al. (1989) as described by Finkelstein et al. (1985). Prehybridization and
and used to perform a differential screen of the pollen library. In hybridizationc onditions were similar to those described for gen-
the primary screen, the unamplified library was probed (Maniatis omic DNA gel blots except that 50% formamide was included and
et al., 1982) with radiolabeled cDNA prepared from RNA of either the temperature was 42°C. RNA dot blots were quantitated by
mature pollen or 1O -day-old seedlings. cDNA clones that hybrid- excision of the individual dots and liquid scintillation counting.
ized more strongly to the pollen probe than to the seedling probe
were subjected to a secondary screen with cDNA probes from
either pollen or immature floral bud RNA. lhe immature buds
Nucleotide Sequencing
were collected at least 2 weeks preanthesis and, therefore, in-
cluded premeiotic anthers. P2 was chosen for further analysis
DNA sequencing was carried out by the dideoxynucleotide
because of its strong hybridization to the pollen probe and lack
method using a kit containing modified 17 DNA polymerase
of hybridizationt o the seedling or immature floral bud probes. lhe
(Sequenase, United States Biochemicals, Cleveland, OH) accord-
1.2-kb EcoRl insert was subcloned into a plasmid vector (Blue-
ing to the manufacturer's instructions. Plasmids for double-
script, Stratagene) for further analysis.
stranded templates were prepared as described by Maniatis et al.
(19 82). Preparation of single-stranded templates was performed
as recommended by Stratagene. The sequence of both strands
Genomic DNA Gel Blot Analysis
was obtained for the majority of the sequences presented. In
those regions where both strands were not sequenced, single-
Nuclear DNA was isolated as described by Scofield and Crouch
strand sequence was obtained from at least two different overlap-
(19 87). Oenothera genomic DNA was digested with restriction
ping subclones. Computer analysis of sequence information was
endonucleases, separated by electrophoresis, and transferred to
performed using the University of Wisconsin Genetics Computer
nitrocellulose (Maniatis et al., 1982). Prehybridization was per-
Group programs (Devereaux, Haeberlil, and Smithies, 1984).
formed in 5 x SSPE (1 x SSPE = 0.15 M NaCI, 10 mM NaH,PO,,
1 mM EDTA), 1 X PE [50 mM Tris-HCI, pH 7.5, 0.1% sodium
pyrophosphate, 5 mM EDTA, 1% SDS, 0.2% PVP (40,000 D),
Construction of Expression Plasmid
0.2% Ficoll (40,000 D), 0.2% BSA], 10% dextran sulfate, 50 pg/
mL DNA (salmon sperm), and 50 pg/mL tRNA yeast at 65°C for
4 hr to 12 hr. Hybridization was performed at 65°C for 12 hr to The P2 expression plasmids were constructed by inserting a
16 hr in the fresh prehybridizations olution. Hybridization probes portion of the P2 cDNA into each of the three members of the