The B. napus cDNA and the A. thaliana gene encode proteins that are 73 ~o identical and are predicted to be 10.3 kDa and 11.6 .... 5-bromo-4-chloro-3-indolyl-/%D-glucuronide (X-. Gluc) and ... slides and stained with 1~o toluidine blue in 1 o.
Plant Molecular Biology 19: 611-622, 1992. © 1992 Kluwer Academic Publishers. Printed in Belgium.
611
The isolation and characterisation of the tapetum-specificArabidopsis thaliana A9 gene Wyatt Paul, Rachel Hodge, Sarah Smartt, John Draper and Rod Scott* Department of Botany, Leicester University, University Road, Leicester LEI 7RH, UK (*author for correspondence) Received 7 November 1991; accepted in revised form 19 March 1992
Key words: anther, Arabidopsis thaliana, Brassica napus, tapetum
Abstract
The Brassica napus cDNA clone A9 and the corresponding Arabidopsis thaliana gene have been sequenced. The B. napus cDNA and the A. thaliana gene encode proteins that are 73 ~o identical and are predicted to be 10.3 kDa and 11.6 kDa in size respectively. Fusions of an RNase gene and the reporter gene/%glucuronidase to the A. thaliana A9 promoter demonstrated that in tobacco the A9 promoter is active solely in tapetal cells. Promoter activity is first detectable in anthers prior to sporogenous cell meiosis and ceases during microspore premitotic interphase. The deduced A9 protein sequence has a pattern of cysteine residues that is present in a superfamily of seed plant proteins which contains seed storage proteins and several protease and e-amylase inhibitors.
Introduction
Male gametogenesis is a developmentally complex process requiring the formation and interaction of several cell types within the anther to successfully produce pollen [see 26, 37]. Two cell types of importance are the microsporocytes that divide meiotically to give haploid microspores and eventually pollen and the tapetal cells that form a single cell layer surrounding the central core of sporogenous cells. During meiosis, microspore formation and early maturation the tapetal cell layer is probably the most metabolically active
cell type in the anther. The tapetum is thought to provide nutrients and structural components for the developing microspores and it also has other roles, including the secretion of an enzyme that is required for the release of microspores from tetrads [28, 37]. The importance of the tapetal cells in anther development is underlined by the fact that natural male sterility is often linked to tapetal dysfunction [14]. Also male-sterile plants have been constructed by using a tapetum-specific promoter to express RNase in tapetal cells [27]. The complexity of anther development is reflected in the finding that many genes are
The nucleotide sequence data reported will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession numbers X61750, X61751 and X61752.
612 expressed exclusively in the anther. For example, it has been estimated by RNA solution hybridisation analysis that in tobacco up to 10000 transcripts are anther-specific [18]. Several groups have isolated cDNAs representing tapetumspecific transcripts [20, 29, 33, 36]. The functions of the products encoded by these tapetum-specific transcripts are unknown though TA29, isolated from tobacco, encodes a glycine-rich protein which was suggested to be a component of the microspore cell wall [20]. In Brassica napus a number of anther-specific cDNAs have been isolated from two cDNA anther libraries and analysed for temporal and spatial expression [33]. One of these cDNAs, A9, was suggested to represent a tapetum-specific transcript since it was expressed during developmental stages when the tapetum is metabolically active and in situ hybridisations showed that the A9 transcript appeared to be localised to the tapeturn [33]. In this work the A9 cDNA is further characterised and the homologous gene from Arabidopsis thaliana isolated. A9 is shown to be a highly expressed tapetum-specific gene whose predicted protein sequence suggests an ancestral relationship to a class of plant seed proteins that include seed storage proteins and inhibitors of proteases and co-amylase.
Materials and methods
Libraries and plasmids
The construction of the B. napus anther cDNA library is described previously [33]. The A. thaliana genomic library was constructed in the vector Lambda dash (Stratagene) by partially digesting DNA from the Landsberg erecta ecotype with Sau3A. pBluescript (Stratagene) or pTZ 18R/19R (Pharmacia) plasmids were used in subcloning and sequencing. Screening the A. thaliana genomic library
The library was plated out and plaques transferred to Hybond-N (Amersham) according to
the manufacturers' instructions. Hybridisation of the [32P]-labelled B. napus A9 probe to filters was performed in 6 x SSC (0.9 M NaC1, 0.09 M sodium citrate, pH 7), 5 x Denhardt's solution (0.1 ~o bovine serum albumin, 0.1 ~o Ficoll 400, 0.1 ~o polyvinylpyrrolidone), 0.5 ~o sodium dodecyl sulphate (SDS) and 100 #g/ml sheared herring sperm DNA, at 60 ° C overnight. Washes (at 60 °C) were twice in 3 x SSC, 0.1~o SDS for 15 min, followed by two washes for 15 min in 0.5x SSC, 0.1~o SDS.
DNA and protein sequence analysis
DNA sequence was assembled and analysed using the GCG [4] and Staden computer packages. The program FASTA was used to screen the PIR-NBRF protein database (release 26), and TFASTA [31] used to screen the EMBL DNA database (release 25) for sequences homologous to the putative primary structure of A9. Alignment of sequences was performed with the aid of the program CLUSTAL using default parameters [12].
Construction of A9 promoter- G US fusion plasmids
The 329bp Hinc II-Rsa I fragment (positions 1105-1434 bp in Fig. 2) was cloned into Hinc IIcut pTZ18R forming pWP70A. DNA sequence analysis revealed the loss of a 'G' residue at the Rsa I, Hinc II junction which resulted in the recreation of the Rsa I site. The Hind III, Barn HI fragment of pWP70A containing the A9 promoter was cloned into Barn HI, Hind III-cut pBluescript forming pWP71. To reconstruct plasmids with larger A9 upstream regions the Eco RI, Hind III fragment of pWP71 was replaced with the 900 bp Hind III, Eco RI fragment of pWP64 (which contains the 1486 bp Acc I (position 498 bp in Fig. 2), Bgl II fragment cloned into Acc I, Barn HI-cut pTZ19R forming pWP72. Also the Eco RI, Hind III fragment of pWP71 was replaced with the 1397bp Hind III, Eco RI fragment of pWP55 (which contains a 3146 bp Xba I fragment from
613 G9.1 cloned into Xba I-cut pTZ19R) forming pWP73. The Hind III, Xba I fragments ofpWP71, pWP72 and pWP73 were cloned into Hind III, Xba I-cut pBI101.1 [ 15] forming pWP74, pWP75 and pWP76 respectively (Fig. 4a). Thus pWP74 contains a 329 bp A9 promoter fragment (positions 1108-1437 bp in Fig. 2), pWP75 a 934 bp A9 fragment (positions 501-1437 bp) and pWP76 a 1437 bp A9 fragment (positions 1-1437 bp) all fused to GUS.
Construction of A9 promoter-RNase fusion plasmids The oligonucleotide primers 5'-GGGTCTAGACCATGGGCACAGGTTATCAACACGTTTGACGGG-3' and 5'-GTAAAACGACGGCCAGTGCC-3' were used in a polymerase chain reaction (PCR) to generate a fragment encoding barstar and the mature barnase product (an RNase) from the plasmid pTG2 [13]. The first primer is homologous to nucleotides 95-221 bp of Fig. 1 in Hartley [11]. The second primer is homologous to a sequence immediately next to the Hind III site of pTZ18U (Pharmacia). The PCR product was then cloned between the 1437 bp A9 promoter fragment of pWP71 and a CaMV polyadenylation sequence in a vector derived from pJIT60 [10] forming pWP127 (Fig. 4b). A translational fusion of the A9 promoter to mature barnase (pWP128) was constructed by using an oligonucleotide primer to introduce an Nco I site around the initiating ATG of the A9 gene (Fig. 4b). Both chimaeric genes were transferred as Xho I fragments into Sal I-cut pBinl9 [2]. The promoter and coding sequence of barstar, a gene encoding a specific inhibitor of barnase, were included on these plasmids since mature barnase could not be cloned in its absence in Escherichia coli.
Transformation of tobacco Nicotiana tabacum was transformed using Agrobacterium pGV2260 as described by Draper et al.
[5].
Analysis of transgenic plants Plant material was histochemically stained with 5-bromo-4-chloro-3-indolyl-/%D-glucuronide (XGluc) and flurometrically assayed using 4methylumbelliferyl glucuronide (4-MUG) essentially as described by Draper et al. [5]. Protein concentrations were determined by the method of Bradford [ 3 ]. Semi-thick (0.18-0.4 #m) sections for light microscopy were prepared essentially as described by Grant et al. [9], mounted on glass slides and stained with 1~o toluidine blue in 1~o borax (disodium tetraborate). Results
Characterisation of the Brassica napus A9 cDNA The nucleotide sequence of the A9 cDNA and three other homologous clones (cDNAs A10, A13 1 1 q y l n k M E F L K S F T T I L CCTCTACAATATCTAAACAAAATGGAATTTCTCAAATCCTTTACAACTATTCTCTTTGTA
M
F
L
A
M
S
A
L
E
T
V
P
M
V
R
A
Q
Q
F
V 60
C
L
ATGTTTCTGGCCATGAGCGCTCTGGAGACCGTACCTATGGTTCGAGCTCAACAATGCCTA
D N L S N M Q V GACAATTTGAGCAATATGCAGGTGTGT
C
120
A P L V L P G A V N GCGCCGCTGGTTCTGCCTGGTGCAGTCAATCCA
P
A P N S N C C I A L Q A T N K D C l C GCCCCGAATTCA/hAT TGCTGCATTGCTCTCCAAGCAACTAACAAAGATTGTATATGTAAC ---C .................................
N
180
240
A L R A A T T F T T T C N L P S L D C G GCCCTTCGAGCAGCCACCACATTTACCACTACTTGC.AACCTCCCCTCTTTAGATTGTGGT ........................... T .................
300
I T I * ATAACCATATGAGTGGTTTCAGCAACGGTCAGTTCCGAGGATTTGGGGAGTTTGGTCTGC .........................................................
360
AAAAGACAACAAG .... A A T A A A A G T A T T A A A A T A A C G A G A A A G T G T G T G T G G--AATA ........... : . . . . . . . . . . .
.TTTTTTTT T .......
420
TTAATTTGGTCTGTTGTTCGGTTGGTTCAATA..CTTAAATATGACCCATCAATTAATAT ...................... AC
480
CGTTTTCATATTTATTATGGTAATATTTTT
510
Fig. 1. D N A sequence of the cDNA clones A9, A10, A13 and A30. A composite D N A sequence comprising of cDNAs A9, A10, and A13 is displayed, A9 extends from position 14 to 510bp, A10 from 10 to 489bp and A13 from 1 to 462bp. These sequences differ at position 413 bp where A9 lacks a T residue. The extent of cDNA A30 is shown by dashed lines underneath the A9/A10/A13 sequence; differences in the A30 sequence are indicated. The deduced primary sequence of the ORF encoded by these cDNAs is shown above the D N A sequence. The portion in lower case is probably not translated (see text). Putative polyadenylation signals are underlined.
614 and A30) is shown in Fig. 1. Three of these cDNAs, A9, A10 and A13, are identical over the region that they overlap (apart from at position 413 bp where A9 lacks a 'T' residue; since this is in a run of 9-10 'T' residues this difference could be due to a transcriptional error). The A9/A10/ A13 composite sequence is 504 bp in length and contains an open reading frame (ORF) extending from 1-309bp (Fig. 1). The A30 sequence is 305 bp in length and is 97 ~o homologous to the composite sequence. The deduced primary sequence of the ORF within A30 is identical to that encoded by A9/A10/A13 in the region they overlap despite two base changes between these sequences (Fig. 1). None of these cDNAs has a poly(A) tail though sequences that resemble a
polyadenylation signal (AATAAA) are present [ 17] (Fig. 1). This may explain, in part, why these cDNAs are smaller than the predicted size 550600 bp for the A9 transcript from an RNA gel blot (data not shown). Genomic D N A gel blots using B. napus A9 c D N A as a probe indicate that A. thaliana, B. oleracaea and B. campestris have a single A9 gene whereas B. napus has two (data not shown). The sizes of A9 hybridising bands in the Brassica species is consistent with B. oleracaea and B. campestris being the parents of B. napus. Thus A9 is probably encoded by two genes in B. napus one represented by cDNAs A9/A10/A13 and the other by c D N A A30, each derived from one of the parents of B. napus.
TCTAGACATAACGGTGAGAGTTAATATTAAAATTTCAGGCGAGA_AAAATGATACTTGAAA AATATTATGATCGTTTTGGATATTCCTTACATCGAGTGAATGTTGGTTTGATTCATCTTC CAAGTGTTCTGCAAACGTATATTAAAGGTTTATTAACTGGTAAGAGATTAACCGGGTTTT GGTTCAGCATATACCATGATTGACTAACTGATCAAATAGTCTTTACTTATTATATAAAGA CGATACTATTGGTCATGCTACAAAATCAAGTCATACCATATCCTGAGAATGAATGTGGAG AATCGTTATAAGGCATAAGTGTGGGTATTGATCGTGGTACGAACAACCGCCTTGGCATCA ACATTAGCCACGATATCCAACATTTGAAGCATTGCCTATGGCGAGTGTTTGGTTGGTTTT GAAACTGATGATGATAACCAGAACGAGAAATGTCTTGTGAAGTATAATGTTCCGATGAAT TGGGATTATAATAATGTGTAGACATTGTAGGTTGGTTTTGATGATGATAAGTAATCATTG GAGAATTGTCTAACACATGCACTGGAGAATTATTGACTCTACCACGTTCTCTTTGATATT CCTCGATTTTCCTCGTGATTTCATCAGCCTCTCCGAAAAAGTAATTGTATCCACTAGAAC TTTGGGAATCTCCCATCTAATTTATGTATTAGAGAAGTTATAATATTTTGGGGAAATAGA TTTTCTCTACTGATTTTGTTGTGTGACATTATATTTTTATAAGTACATGTTTCTGTTTCG TTATATTGTTGTCGTGGTTGAGTCTTTATTAGAGCATGTAAATATGTTTATGAAATAAGC GAGAAAGGAATTAATTAAACGTATCGAGTGATAAATGCTTTAATGGATTCGAGATTTAGT ATTCTTAAATTTTTGTTTCATTATCATTGATTATAAAACTAAGTTATGTTGATCTCAAAT CCTTAATTATGTTCTCCTAAGAAGAGTACAAGTGGTGGGAACGAAAGATGAGTAAAATAC TAAAAATCTTTTCTCAAAAGTCAAATCGCATTAGTTAACAAAAACAAACCATGTGTTACC GTCAAATCAATGTGTTTAAAAGATGTTAACCACTAATCAAGCATTTACGTGTAACCGGAT CAACCGGATTTGGGTTTTGAATATGTTGTGGAGATGTATATAAATGATAAATTAATTGAA TATCTTAATTAATCTGTGAAAGAAACTACATCACACACTTTGTTATTTCCCCTAGCTTTT AGTTTTTTTATCATGCAAAACTTATGAAGTAACTAGATCAAGATCACAAAAAA_AAAGCAT CACTTCACTTCATGACCTAATTATTCTCGAAGCCCAkAACTATTTACATACACTTTTATT CTATAAATATAGATGATGGAATTCACCAATCCAAAAGTGAATAAAAAACACAAGTACAAA M V S L K S L A A I L V A CAATATAGTATCTAATTAGAATGGTATCTCTAAAGTCCCTTGCTGCTATTCTCGTTGCCA M F L A T G P T V L A Q Q C R D E L S N TGTTTCTTGCCACCGGACCTACGGTTCTAGCCCAGCAGTGCAGAGACGAACTGAGCAATG V Q V C A P L L L P G A V N P A A N S N TGCAGGTGTGCGCGCCGCTGCTTCTGCCCGGTGCGGTCAATCCTGCCGCGAACTCAAATT C C A A L Q A T N K D C L C N R L R A A GCTGCGCTGCCCTCCAAGCAACTAACAAAGATTGTCTATGTAACCGTCTTCGAGCAGCCA T T L T S L C N L P S F D C G K N I H R CCACACTTACCTCTCTTTGTAACCTCCCCTCTTTTGATTGTGGTAAGATGATCCATCGAT L K P F L L D F Y K L F H Q * TAAAACCTTTTTTACTAGATTTTTATAAATTATTCCATCAATAGTGTTTGTTTTATATTT GTTCTCATGATTTTTTGGACTTATGTTTTGTGAACTGTGCAGGCATAAGTGCCTAGTTGA ACAACATTCAGTTCCGAGGATTTGGGGAGTTTGGTCTGCAAACGACAAGACGAAT~
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080 1140 1200 1260 1320 1380 1440
1800 1860 1920
AAAATAATGAGAAATACACTATTTAGTGTTTT
1952
1500 1560 1620 1680 1740
Fig. 2. D N A sequence of the A. ~aliana A9 gene. The putative primary s~ucture of the A. ~ a ~ n a A9 gene is shown above the D N A sequence. Potential TATA boxes and possible p o l y a d e n ~ i o n sequences we underlined.
615 Isolation of the A9 gene from Arabidopsis thaliana
A 13 kb genomic clone (G9.1) containing the A. thaliana A9 gene was isolated by screening an A. thaliana genomic library with the B. napus A9 cDNA. A 3.1 kb Xba I fragment, containing the region hybridising to the A9 cDNA, was subcloned from G9.1 and partially sequenced as shown in Fig. 2. An ORF at position 14611781bp of Fig. 2 is 76~o identical to the A9 cDNA ORF at the nucleotide level. The putative products encoded by these ORFs, aligned in Fig. 3, are 73 ~o identical showing that this ORF encodes the A. thaliana A9 gene. Comparison of these sequences suggests that the initiating ATG of B. napus A9 is at position 22 bp in Fig. 1 and that the A9 gene contains no introns. The putative polypeptide encoded by B. napus cDNA is 96 amino acids in length with a calculated molecular mass of 10.3 kDa and that of the A. thaliana gene 107 amino acids with a mass of 11.6 kDa. Hydrophobicity plots of the putative A9 proteins indicate that both have a hydrophobic N-terminal sequence. These sequences are potential signal peptides since they conform to the criteria identified by von Heijne [39] in that they possess a positively charged N-terminal region, a central hydrophobic region and a more polar C-terminal region. Joshi [16] has aligned the promoters of sequenced plant genes and has suggested the consensus sequence TATATATA for plant promoter 'TATA' boxes. 70 bp upstream of the putative start of the A9 gene (positions 1382-1389bp) there is a sequence, TATAAATA, with similarity to this consensus sequence. Downstream of the
At
MVSLKSLAAILVAMFLA,
Bn
MEFLKSFTTILFVMFLAMSALETVPMVRAQQCLDNLSNMQVCAPLVLPGA
At
V N P A A N S N C C A A L Q A T N K D C L C N R L R A A T T L T SL C N L p S F D C G K M I H R L K 95
Bn
V N P A P N S N C C I A L Q A T N K D C Z C N A L R A A T T F T T T C N L pS L D C G i T i
At
PFLLDFYKLITHQ 107
i
Ill[
fin
iFJlF
i]
. i . .TGPTVLAQQCRDELSNVQVCAPLLLPGA
fiJl
l[l]iIlll
i r I II11
li lll[li
r
[ liB
[IIEI
45
rillJl
Jfll 50
III
I 96
Fig. 3. Alignment of the putative polypeptides encoded by the B, napus (Bn) A9 cDNA and the A. thaliana (At) A9 gene.
Residues shown in italics form putative signal peptides. The extent of these signal peptides was determined using the method ofvon Heijne [38].
A9 gene there are two potential polyadenylation signals (Fig. 2).
The A. thaliana A9 promoter retains both temporal and tissue-specific expression in transgenic tobacco
Three A9 promoter-transcriptional fusions to the reporter gene ¢?-glucuronidase (GUS), shown in Fig. 4a, were constructed in order to determine the extent of the promoter region upstream of the A9-coding region. These GUS fusion plasmids contained 1437 bp (pWP76), 934 bp (pWP75) or 329 bp (pWP74) of the region upstream of an Rsa I site which lies in the putative 5' untranslated region of the A9 gene (Fig. 4a). The three constructs were transformed into tobacco. Histochemical staining of transgenic plants and fluorometric assays revealed that G U S activity was localised to anthers. No GUS activity was detected in leaves, carpels, developing and mature seeds, pollen and whole seedlings at the two leaf stage. Plants transformed with the vector pBI101.1 showed no GUS activity. Sections of anthers of tobacco transformed with the A9-GUS fusion constructs confirmed that the A9 promoter was active in the tapetal cells of the anther (data not shown). Washed young microspores squeezed out of the anthers of A9-GUS transgenic plants (isolated at the stage when there is GUS activity in the tapetal cells) showed no GUS activity. Thus the A9 promoter is active only in tapetal cells. To investigate the temporal pattern of GUS accumulation in Ag-GUS transgenic plants more precisely, a series of buds of different sizes were isolated from each plant. The developmental stage of one anther of each bud was determined by microscopy and a second anther tested for GUS activity. Four plants transformed with pWP76, three with pWP75 and four with pWP74 were tested. Figure 5 displays the results from representative transgenic plants. GUS activity appears in the tapetum during early sporogenous cell meiosis, increases dramatically, reaches a plateau then falls precipitously and ceases in anthers with premitotic microspores. Therefore GUS expression in transgenic tobacco closely matches the
616
(a) A9 )do
R
A
I
R
R HH
I
I
I
I
1
RI R
Bg
II
I
',
,, -353
-960
I
2 kb
I
-24
G US-~
pWP74
-24
GUS " ~
pWP75
GUS----~
pWP76
,, ,, I
-146"i [/////////////////////////J
-24
i
/
(b) pTG2 Barnase --~
Barstar--~
[~'%*%~t~NN~NNNN \
Hd I
/
|
I
pWP127
J
A9
promoter -i~"
A9
p r o m o t e r --~
/'CaMV
No Nc
poly A
)
