Plant Cell, Vol. 11, 1307-1318, July 1999, Copyright © 1999, American Society of Plant Physiologists
ANI1: A Sex Pheromone –Induced Gene in Ceratopteris Gametophytes and Its Possible Role in Sex Determination
Chi Kuang Wen1,a,
Rachel Smitha, and
Jo Ann Banksa
a Department of Botany and Plant Pathology, Purdue University, West Lafayette, Indiana 47906
Correspondence to:
Jo Ann Banks, banks{at}btny.purdue.edu (E-mail), 765-494-5896 (fax)
 |
ABSTRACT |
|---|
Antheridiogen (ACE) is a pheromone that is required for the development of male gametophytes in the homosporous fern Ceratopteris richardii. Subtractive hybridization of cDNAs isolated from ACE-treated and non-ACE-treated gametophytes was used to isolate genes that are induced by the pheromone. The expression of one gene, ANI1 (for antheridiogen induced), was induced within 3 hr of ACE treatment, but its expression was transient. Patterns of ANI1 expression in wild-type and mutant gametophytes show that ANI1 expression inversely correlates with the predicted activity of one of the sex-determining genes, TRANSFORMER5 (TRA5). These data suggest that ANI1 transcription or transcript accumulation is directly or indirectly prevented by TRA5 in the absence of ACE and that ACE inactivates the TRA5 gene or its product, leading to the upregulation of ANI1. Cycloheximide (no ACE) induced the expression of ANI1, also indicating that ANI1 expression is subject to negative regulation in the absence of ACE. The sequence and inferred protein structure of ANI1 suggest that it is a novel, extracellular protein. The secreted portion of the ANI1 protein potentially forms a ß barrel with superficial similarities to lipocalins, which bind small hydrophobic molecules such as pheromones, steroids, and odorants. ANI1 may be an extracellular carrier of ACE that is required to initiate the male program of development as the sexual fate of the young gametophyte is determined.
|
|
INTRODUCTION |
|---|
The haploid gametophytes of many homosporous ferns, including Ceratopteris richardii, develop as either males or hermaphrodites (Figure 1A and Figure 1B). The sex of the gametophyte is determined by the pheromone antheridiogen, or ACE in Ceratopteris (for antheridiogen Ceratopteris) (Schedlbauer and Klekowski 1972 
; Banks et al. 1993 ). Ceratopteris spores develop as male gametophytes with numerous sperm-forming antheridia if they are continuously exposed to exogenous ACE from the time of spore germination. If ACE is removed from the medium surrounding an older male gametophyte, undifferentiated cells of the male prothallus will divide and differentiate a hermaphroditic prothallus, indicating that ACE is required to both initiate and maintain the male program of expression. In the absence of ACE, single spores develop as hermaphroditic gametophytes. These begin to produce and secrete ACE into the surrounding medium after they lose the ability to respond to the pheromone. The loss of sensitivity to ACE coincides with the development of a lateral, multicellular meristem, which is unique to the hermaphrodite (Figure 1A) (Banks et al. 1993 ). Exogenous antheridiogen thus performs several functions in the young, sexually undetermined gametophyte: it promotes the differentiation of sperm-forming antheridia and simultaneously suppresses the development of the meristem, the egg-forming archegonia, and the synthesis of antheridiogen (Banks 1997a ).
View larger version (125K):
[in this window]
[in a new window]
|
Figure 1.
The Ceratopteris Gametophytes.
(A) The hermaphrodite, which develops in the absence of ACE, forms a meristem (M), several egg-forming archegonia (AR), and sperm-forming antheridia (AN).
(B) The male, which develops only in the presence of ACE, has no meristem or archegonia. Almost all cells of the male prothallus differentiate as antheridia.
Both gametophytes are 14 days old. Bars = 250 µm.
|
|
Although the structure of ACE is unknown, several lines of evidence suggest that ACE is a gibberellin (GA). GA biosynthetic inhibitors block the male-inducing response, indicating that GA and ACE share a common biosynthetic pathway in Ceratopteris (Warne and Hickok 1989 ). Abscisic acid, a known antagonist of GA responses, also antagonizes the male-inducing ACE response (Hickok 1983 ; Warne and Hickok 1991 ), further implicating a GA-like molecule in sex determination in Ceratopteris. All known antheridiogens of other homosporous ferns have been identified as GAs and include GA109 (Wynne et al. 1998 ), GA73 methyl ester (Takeno et al. 1989 ; Yamauchi et al. 1996 ), antheridic acid (Yamauchi et al. 1991 ), 3-epi-GA63, and 3
-hydroxy-9,15-cycle-GA9 (Yamauchi et al. 1995 ). Finally, partial characterization of an antheridiogen of Ceratopteris indicates that it is a small hydrophobic molecule with a molecular mass of ~300 D (Koitabashi 1996 ), similar to other GAs.
Ceratopteris is a particularly useful system for studying antheridiogen and how it regulates sex determination in ferns because it is very easy to isolate and genetically characterize mutations that affect sex expression in this species. Although a genetic approach has been useful for identifying sex-determining genes and dissecting the sex determination pathway in Ceratopteris (Banks 1994 , Banks 1997b ; Eberle and Banks 1996 ), cloning these genes will be difficult due to the large genome size of Ceratopteris (~8 x 109 bp per haploid genome) and the lack of a transformation system. To understand how antheridiogen controls the sex of the gametophyte, we used a polymerase chain reaction–coupled subtractive hybridization technique (Wang and Brown 1991 ) to clone genes whose expression is induced by ACE during the brief period of time that the sex of the gametophyte is determined by the presence of ACE. One such gene, termed ANI1 (for antheridiogen-induced), has been cloned, and its expression in wild-type and mutant gametophytes has been assessed. The predicted structure of the protein encoded by ANI1 indicates that it may be an extracellular carrier of small hydrophobic molecules. Its possible role in sex determination is discussed.
|
|
RESULTS |
|---|
Timing of ACE Secretion and the Competence to Respond to ACE in Developing Gametophytes
Because this study was intended to identify early ACE-inducible genes, it was first necessary to define the period of time during which a population of developing gametophytes is competent to respond to exogenous ACE. This was determined by periodically transferring developing gametophytes grown in medium lacking ACE (fern medium [FM]) to medium containing exogenous ACE (conditioned fern medium [CFM]). If a gametophyte is competent to respond to ACE at the time of transfer, it will develop as a male gametophyte in the presence of exogenous ACE. If a gametophyte is not competent to respond to ACE at the time of transfer, it will develop as a hermaphrodite. Gametophytes that were transferred to CFM 80 hr after spore inoculation developed exclusively as males. Of those transferred to CFM 99 hr after spore inoculation, only 33% developed as males, whereas 16% of the spores transferred to CFM 104 hr after spore inoculation developed as males. These results indicate that if gametophytes have not perceived ACE after 104 hr, most of them will have lost the competence to respond to ACE and will therefore develop as hermaphrodites. Accordingly, adding ACE (CFM) to gametophytes <80 hr old should be sufficient to induce ACE-responsive genes in all members of a population.
When grown in the absence of exogenous ACE, some members of a population of gametophytes develop as males in response to endogenous ACE secreted by the faster growing, hermaphroditic members of the population. The presence of endogenous ACE in the culture medium makes it difficult to generate the population of gametophytes not exposed to ACE and thus not expressing ACE-induced genes required for this study. Therefore, it was necessary to establish when biologically effective amounts of ACE are first detectable in the medium supporting growing gametophytes. Medium lacking exogenous ACE that had supported the growth of gametophytes for 72, 82, 96, and 102 hr was added to wild-type spores. These spores developed exclusively as hermaphrodites, indicating that biologically effective amounts of ACE were absent in the medium. Medium that had supported the growth of gametophytes for 112 hr resulted in the development of 100% male gametophytes.
Based on these results, biologically detectable amounts of endogenous ACE are not present in the medium supporting the growth of gametophytes until between 102 and 112 hr after spore inoculation. Because the assay for ACE is biological, trace amounts of endogenous ACE may be present in media that have supported the growth of gametophytes for <102 hr, but the concentration is below that required to initiate male development. Thus, RNA isolated from spores germinated for up to 100 hr in FM should not contain messages from any ACE-inducible genes required for the initiation of male development.
Cloning of ACE-Inducible Genes by Subtractive Hybridization
Genes induced by exogenous ACE in gametophytes that were competent to respond to ACE were cloned using a polymerase chain reaction–coupled subtraction and hybridization procedure. This procedure removes sequences that are common between two populations of cDNAs generated from two populations of gametophytes. To generate the two populations of gametophytes, we initially cultured spores in medium lacking ACE (FM) for 72 hr. One population was then transferred to ACE-containing medium (CFM) and cultured for an additional 28 hr, and the other population was cultured for the same period of time but in the absence of exogenous ACE. cDNAs generated from CFM- or ACE-treated gametophytes are referred to as tracer cDNA; cDNAs generated from FM- or non-ACE-treated gametophytes are referred to as driver cDNA. After each round of hybridization/subtraction to remove driver cDNAs from the population of tracer cDNAs, the efficiency of subtraction was tested by probing subtracted tracer cDNA populations with Ceratopteris malate dehydrogenase (MDH) or elongation factor-1 (EF-1) probes that had been cloned previously (C.K. Wen and C. Juarez, unpublished results). Virtually no MDH cDNA remained after one round of hybridization and subtraction, whereas virtually all EF-1 cDNA was removed after two rounds of hybridization and subtraction (data not shown). After two rounds of hybridization and subtraction, the remaining tracer cDNAs were cloned. Six independent cDNA inserts were used to probe RNA gel blots; their corresponding messages were shown to accumulate in young gametophytes whether or not they had been treated with exogenous ACE (data not shown). The majority of the cloned cDNAs thus appeared to represent germination-related genes and not the desired ACE-induced genes.
To further eliminate germination-related genes from the population of cloned tracer cDNAs, we probed 20,000 Escherichia coli colonies, each containing a tracer cDNA insert, with labeled cDNAs originating from 78-hr-old gametophytes grown in FM (no ACE). The same colonies also were probed with a cDNA population that had been enriched for cDNAs present only in non-ACE-treated gametophytes by subtractive hybridization. After primary and secondary screenings using these two probes, 51 non-cross-hybridizing colonies were obtained. The inserts of each of the 51 clones were partially sequenced and then sorted into 15 unique groups based on their sequence. A representative cDNA insert of each group was then used to probe gel blots carrying RNA isolated from ACE-treated and non-ACE-treated gametophytes. Based on these results (data not shown), 12 of the 15 unique cDNA inserts appeared to represent ACE-inducible genes.
Any cDNA insert identified using this protocol is likely to represent only part of a full-length gene because the cDNAs were initially digested with AluI early in the hybridization and subtraction protocol. To determine how many genes were represented by the collection of enriched cDNA fragments, we used the inserts individually as probes to screen a cDNA library that was constructed from mRNA isolated from 100-hr-old gametophytes that had been grown in the presence of exogenous ACE for 28 hr before harvesting. Each cDNA insert hybridized with one of two unique recombinant plaques. Based on these results, the enriched cDNAs represent parts of two distinct genes or gene families, known as ANI1 and ANI2. The expression and characterization of ANI1 are described here.
Expression of ANI1 in Wild-Type and Mutant Gametophytes
RNA gel blots show that the ~1.3-kb ANI1 transcript is induced in young gametophytes between 30 min and 3 hr after treatment with exogenous ACE (Figure 2). A low basal level of ANI1 expression can be detected in 87- and 100-hr-old gametophytes that have not been exposed to exogenous ACE (Figure 2). In gametophytes that have been continuously exposed to exogenous ACE, the expression of ANI1 rapidly declines between 110 and 130 hr (Figure 2). Even after prolonged exposure of RNA gel blots, virtually no expression of ANI1 could be detected in a mixed population of mature wild-type male and hermaphroditic gametophytes (data not shown) or in sporophylls of the diploid sporophyte plant (Figure 2). The temporal pattern of ANI1 expression in wild-type gametophytes indicates that ANI1 is induced by ACE and that its expression is transient in the developing gametophyte. The transient expression of ANI1 correlates with the initiation, but not the maintenance, of the ACE-induced male program of development in wild-type gametophytes.
View larger version (45K):
[in this window]
[in a new window]
|
Figure 2.
ANI1 Is Upregulated by ACE in Young Gametophytes.
Total RNAs were prepared from variously treated wild-type gametophyte populations and sporophylls of the sporophyte plant (lane S). The period of time that gametophytes were grown in FM (lacking ACE) is indicated at top. Gametophytes were then transferred to CFM (containing ACE) for the period of time indicated below. Ethidium bromide–stained 28S rRNA bands are shown as loading controls.
|
|
The control of sexual identity in Ceratopteris gametophytes is ultimately controlled by two classes of key regulatory genes, the FEMINIZATION (FEM) and the TRANSFORMER (TRA) genes. Mutant fem gametophytes always develop as females, even in the presence of ACE, whereas mutant tra gametophytes always develop as males, indicating that the male program of expression is activated independently of ACE in the tra male. If ANI1 is involved in sex determination, its expression or accumulation is predicted to be affected by mutations that alter the sex of the gametophyte. To test the involvement of ANI1 in sex determination, we assessed the expression of ANI1 in fem1 and tra5 mutant gametophytes. Because both fem and tra mutant gametophytes cannot be self-fertilized (they make only one kind of gamete), it is impossible to produce a pure population of fem or tra gametophytes. Therefore, mixed populations of gametophytes of known genotypes were used to assess ANI1 expression in mutant backgrounds.
In a mixed population of gametophytes containing equal numbers of mutant fem1 TRA5 and fem1 tra5 double mutant gametophytes, ANI1 RNA accumulates to high levels when grown in the absence or presence of ACE, as shown in Figure 3. Therefore, a mutation in either the tra5 or fem1 gene leads to an accumulation of ANI1 in the absence of exogenous ACE. To determine which of these mutations results in ACE-independent expression of ANI1, we assessed the expression of ANI1 in mixed populations of either FEM1 and fem1 gametophytes or TRA5 and tra5 gametophytes by using RNA gel blot analysis. The population containing both FEM1 and fem1 gametophytes did not accumulate ANI1 when grown in the absence of ACE, whereas the population of TRA5 and tra5 gametophytes did accumulate ANI1 transcripts in the absence of exogenous ACE (Figure 3). These results indicate that a mutation of TRA5, but not FEM1, results in the accumulation of ANI1 in the absence of ACE. The expression of ANI1 in the FEM1 fem1 and TRA5 tra5 populations treated with ACE is uninformative because ANI1 is expressed in the wild-type members of each population in response to the added ACE.
View larger version (48K):
[in this window]
[in a new window]
|
Figure 3.
ANI1 Expression in Various Mutant Lines.
Total RNAs were prepared from gametophyte populations grown in FM lacking ACE for 90 hr (lanes - CFM) or grown for 72 hr in FM and then transferred to CFM and cultured for an additional 28 hr (lanes + CFM). Each population contained wild-type gametophytes or equal numbers of gametophytes of the given genotypes. Ethidium bromide–stained 28S rRNA bands are shown as loading controls.
|
|
Cycloheximide Treatment Results in ACE-Independent Expression of ANI1
Young gametophytes were pretreated with cycloheximide (CHX) before adding ACE to determine whether ACE-induced expression of ANI1 in wild-type gametophytes requires de novo protein synthesis. Preliminary studies, summarized in Table 1 and Table 2, demonstrated that 20 µM CHX was sufficient to inhibit de novo protein synthesis by 85% in young gametophytes after 30 min of CHX treatment. Gametophytes that were treated with 20 µM CHX for 30 min before adding ACE accumulated ANI1 at high levels, as determined by RNA gel blot hybridization (Figure 4). Thus, ACE-induced expression of ANI1 does not require the synthesis of new proteins. However, a 30-min treatment of gametophytes with 20 µM CHX alone (no ACE) also results in an accumulation of ANI1 RNA (Figure 4). This suggests that ANI1 can be induced independently of ACE by preventing the synthesis of a protein that acts as a repressor of ANI1, a destabilizer of the ANI1 transcript, or both.
View larger version (63K):
[in this window]
[in a new window]
|
Figure 4.
Effects of CHX on ANI1 Expression.
Total RNAs were prepared from wild-type gametophyte populations grown for 84 hr before adding CFM (lanes + CFM) and/or CHX (lanes + CHX). Gametophytes grown only in FM (no ACE) or without CHX are indicated by - CFM and - CHX, respectively.
In lane 1, gametophytes were cultured in CFM for 30 min before harvesting; in lane 2, gametophytes were cultured in CFM for 3 hr before harvesting; in lane 3, gametophytes were cultured in CFM for 8 hr before harvesting; in lane 4, gametophytes were cultured in 20 µM CHX (no ACE) for 3 hr before harvesting; and in lane 5, gametophytes were cultured in 20 µM CHX (no ACE) for 30 min, then transferred to CFM and cultured for 3 hr before harvesting. Ethidium bromide–stained 28S rRNA bands are shown as loading controls.
|
|
Structure of the ANI1 Gene and Inferred Protein
The nucleotide sequence of ANI1, based on the sequence of the longest cDNA clone obtained, is shown in Figure 5. The length of the ANI1 transcript (1.3 kb) determined by RNA gel blot hybridization is somewhat longer than is the longest cDNA clone (939 bp). Based on DNA gel blot analysis, shown in Figure 6, ANI1 is a member of a multigene family in Ceratopteris. The entire ANI1 gene hybridizes with two major and several minor genomic fragments when used to probe DNA gel blots containing genomic DNA digested with two different restriction enzymes and washed under stringent conditions. Hybridization patterns using a probe corresponding to the 5' end of the cDNA or the 3' untranslated region of the gene were similar in that both probes hybridized strongly to two DNA fragments and more weakly to other DNA fragments (Figure 6), indicating that ANI1 is a member of a gene family. Three independent ANI1 cDNA clones obtained from the same cDNA library were sequenced and shown to be >99% identical to one another (data not shown).
View larger version (45K):
[in this window]
[in a new window]
|
Figure 5.
ANI1 cDNA and Deduced Amino Acid Sequence.
The DNA sequence of the longest ANI1 cDNA (GenBank accession number
AF113324) is shown, with the predicted amino acid sequence indicated underneath. Nucleotides in lowercase represent 5' and 3' untranslated sequences. The residues predicted to form a transmembrane region are shown in boldface. A potential cleavage site is indicated by a solid triangle, and a potential N-glycosylation site is indicated by a boldface asterisk adjacent to a shadowed N. The two 18–amino acid repeats are underlined, and the five highly conserved six–amino acid repeats are double underlined, with a sixth, less highly conserved repeat indicated by dashed underlines.
|
|
View larger version (39K):
[in this window]
[in a new window]
|
Figure 6.
DNA Gel Blot Analysis of ANI1.
Ceratopteris genomic was DNA digested with BamHI or HindIII.
(A) The 5' end of ANI1 (nucleotides 1 to 351) was used as a probe.
(B) The 3' untranslated region of ANI1 (nucleotides 723 to 939) was used as a probe. An ethidium bromide–stained 12-kb ladder is shown at right.
|
|
The deduced protein encoded by the largest open reading frame of the ANI1 cDNA (Figure 5) is 182 amino acids with a predicted pI of 9.4 and molecular mass of 22 kD. The nucleotide and predicted amino acid sequences of ANI1 share no significant similarity with other expressed sequence tags or proteins present in the databases, and thus ANI1 encodes a novel protein. The putative ANI1 protein contains two different repetitive motifs: two tandem copies of the 18–amino acid sequence PSTTVY(G/E)(K/E)EEKP(D/E)F(D/Y)(Y/R)YK and five tandem copies of the eight–amino acid consensus sequence VVIYKPKP. A sixth repeat (VVVYKEKE), which is less similar to the consensus sequence than it is to the others, precedes and is separated from the five tandem repeats. These repetitive sequences account for more than one-half of the total amino acid residues of ANI1.
The N-terminal residues (7 to 29) are hydrophobic and are predicted to form an inside–outside type II transmembrane domain, indicating that ANI1 is targeted to the outside of the plasma membrane. The sequence NASA, located at positions 25 to 28, corresponds to the consensus of N-glycosylation sites, indicating that ANI1 may be glycosylated at position 25. A potential signal sequence cleavage site (Nielsen et al. 1997 ) occurs at residues 28 to 29. If cleaved, ANI1 is likely to be an extracellular protein that is not anchored to the membrane. A hydrophilic region (positions 31 to 81) containing the two 18–amino acid repeats follows the transmembrane domain of the protein. Following the hydrophilic region, the protein is predicted to form a series of short, hydrophobic ß strands that alternate with hydrophilic loops/turns. The number of predicted ß strands varies, depending on the algorithm used, and ranges from five to eight. However, five of the six repeats are always predicted to form ß strands, regardless of the algorithm used. Making up these ß strands are the first four residues (usually VVIY) of the eight–amino acid repeat (Figure 5). One or usually two proline residues separate the ß strands and may cause a turn where they occur. Taken together, this region of the polypeptide could form a repeated anti-parallel structure with a +1 topology that folds into a ß barrel.
|
|
DISCUSSION |
|---|
The antheridiogens of homosporous ferns are small, hydrophobic, GA-like pheromones that are required to induce male and repress female development of the gametophyte at concentrations as low as 10-14 M (Yamane 1998 ). To understand how the antheridiogen of Ceratopteris functions at a molecular level, we isolated and characterized the gene ANI1, which is induced by ACE in young Ceratopteris gametophytes at a time when they are competent to respond to the sex-determining signal. The rapid induction of ANI1 by ACE and the ACE-independent expression of ANI1 in tra5 mutant gametophytes that constitutively express the male program of expression in the absence of ACE provide evidence that ANI1 functions in determining the sex of the developing gametophyte.
The ACE-independent expression of ANI1 in CHX-treated gametophytes indicates that ACE induces the expression of ANI1 by downregulating a repressor of ANI1 transcription or by downregulating a factor that destabilizes the ANI1 transcript. By preventing the synthesis of this factor or repressor, CHX treatment results in the accumulation of ANI1 in gametophytes grown in the absence of ACE. Accordingly, when ACE and CHX are not present in the medium, ANI1 transcription is repressed, or if transcribed, its transcripts are rapidly degraded in the developing gametophyte.
The proposed regulatory interactions between the major sex-determining genes in Ceratopteris that have been identified by mutation are shown in Figure 7A. By analyzing the expression of ANI1 in fem1 and tra5 mutant backgrounds, we have shown that the expression of ANI1 inversely correlates with the predicted activity of the TRA5 gene and shows no correlation with other sex-determining genes (Figure 7B). Based on this inverse correlation, it is likely that the repressor of ANI1 expression, or the factor that destabilizes its transcript, is either encoded by or regulated by the TRA5 gene. Therefore, we propose that ACE initiates male sex determination in the gametophyte by repressing or inactivating TRA5. Because TRA5 encodes the factor that directly or indirectly affects ANI1 expression, the repression or inactivation of TRA5 by ACE results in the accumulation of ANI1 transcripts.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 7.
A Model of the Sex-Determining Pathway in Ceratopteris.
(A) The predicted regulatory interactions among the major classes of sex-determining genes identified by mutation are shown (described in detail by Banks 1997b ). Arrows indicate positive (activating) interactions; t-bars indicate negative (repressing) interactions. The her (hermaphroditic) mutants are hermaphroditic in the presence or absence of ACE; the tra (transformer) mutants are always male; the man1 (many antheridia1) mutant is male in the presence of ACE and hermaphroditic in the absence of ACE, but each hermaphrodite produces ~10 times more antheridia than does the wild-type hermaphrodite; and the fem1 (feminization1) mutant is similar to the her mutants except that the fem gametophyte produces no antheridia.
(B) The predicted activities of the sex-determining genes shown in (A), depending on the absence or presence of ACE and the genotype of the gametophyte. Genes in boldface indicate that the genes are active, whereas genes in italics indicate that the genes are not active. Also indicated is the expression of ANI1 in the same gametophytes (results of this study). ON indicates that ANI1 is expressed, and OFF indicates that it is not expressed. wt, wild type.
Because the model in (A) and predicted activities of the sex-determining genes in (B) are derived from genetic data only, the molecular mechanisms underlying these gene interactions or gene activities are unknown.
|
|
The induction of ANI1 by ACE in Ceratopteris is typical of other hormone-induced genes of flowering plants, including the auxin (Aux)/IAA family of genes, which are induced by indoleacetic acid (IAA) (reviewed in Abel and Theologis 1996 ), and the IBC6 and IBC7 genes, which are induced by cytokinin (Brandstatter and Kieber 1998 ). Their induction is relatively rapid (within 10 to 30 min after hormone treatment) and CHX insensitive. Each is also induced by CHX alone. Although the lag between ACE treatment and ANI1 induction is longer (between 0.5 and 3 hr of ACE treatment), ANI1 is induced by CHX, whereas its induction by ACE is CHX insensitive. The CHX insensitivity of these pheromone/hormone–induced genes suggests that all may be primary response genes in that their induction does not require the synthesis of new proteins. However, that all are also induced by CHX alone suggests that these genes are subject to negative regulation and that the hormone or pheromone acts to derepress their transcription or to prevent transcript degradation (Koshiba et al. 1995 ; Brandstatter and Kieber 1998 ). Because CHX alone induces ANI1 in Ceratopteris, it is not clear if the induction of ANI1 by ACE is insensitive to CHX or not.
Many of the auxin-, cytokinin-, and ethylene-induced genes in plants encode proteins that are components of their own signal transduction pathways. The Aux/IAA genes encode short-lived proteins that form homodimers or interact with other members of the Aux/IAA family of proteins to activate the transcription of auxin-responsive genes (Kim et al. 1997 ). One member, IAA24, is similar to ARF1, a transcription factor that binds to an early auxin-responsive element (Ulmasov et al. 1997 ). The recent demonstration that AXR3, a gene defined by mutations that result in increased auxin responses, is IAA17 provides direct evidence that the Aux/IAA genes are involved in auxin signaling (Rouse et al. 1998 ). The cytokinin-induced IBC6 and IBC7 genes encode proteins that are similar to the receiver domain of bacterial two-component regulators, indicating that these proteins are involved in cytokinin signaling. Although neither IBC6 nor IBC7 has been shown to bind cytokinin, further evidence that bacterial two-component regulator-like proteins are involved in cytokinin signaling comes from the recent cloning of CKI1 (Kakimoto 1996 ). The cki1 mutant of Arabidopsis is insensitive to cytokinin, and CKI1 encodes a protein homologous to histidine kinase and receiver domains of two-component regulators. Finally, three of the five members of the ethylene receptor gene family in Arabidopsis (ERS1, ETR2, and ERS2) are upregulated by ethylene (Hua et al. 1998 ). These genes also encode proteins homologous to bacterial two-component regulators. A fourth member of this family (ETR1) has been shown to bind ethylene in vitro (Schaller and Bleecker 1995 ).
Although the putative protein encoded by ANI1 shows no homology to other proteins, its predicted tertiary structure superficially resembles the lipocalin superfamily of proteins. Lipocalins are small (151 to 188 amino acids) extracellular proteins with eight antiparallel ß strands that fold into a ß barrel. Although lipocalin family members may share <20% overall sequence identity, they are unified by their similar tertiary structures (reviewed in Flower 1996 ). In animals, lipocalins bind small hydrophobic molecules, among them pheromones, retinoids, odorants, steroids, and prostaglandin. They function in olfaction, sterol, retinol, and pheromone transport, prostaglandin synthesis, cell regulation, immune modulation, and invertebrate coloration (Flower 1996 ). Most are synthesized by secretory tissues and transported through the plasma to target tissues. Some lipocalins, such as retinol binding protein (RBP), bind to cell surface receptors, at which either the retinol–RBP complex or retinol alone is transferred into the interior of the cell (Bavik et al. 1992 ; Smeland et al. 1995 ). Other lipocalins, such as the pheromone and odorant binding proteins (OBPs), are thought to concentrate specific odorants or pheromones at the site of the receptor. Although some OBPs have been shown to bind specific ligands (Lobel et al. 1998 ), proof that binding of ligand–OBP complex to specific receptors or evidence that OBPs concentrate odorants remains elusive.
Although the crystal structure of ANI1 is unknown at this time, its superficial resemblance to lipocalins suggests that ANI1 might also bind small hydrophobic molecules. Given that ACE is a small hydrophobic molecule (Koitabshi, 1996), we speculate that ANI1 is an extracellular protein that binds to ACE to facilitate the transport of antheridiogen from an aqueous medium to its receptor, to concentrate ACE at the site of the receptor, or to bind to the receptor itself. All of these possibilities place ANI1 early in the ACE signal transduction pathway that leads to the initiation of the male program of development in the gametophyte by ACE. Because ANI1 is only transiently expressed, it is unlikely that it functions in the maintenance of the male program of development, which requires continued exposure to ACE. Studies to determine the crystal structure of ANI1, its cellular localization, and its ability to bind ACE are currently in progress.
Why plant hormones induce genes that encode proteins involved in their own signal transduction pathways is unknown. It has been suggested that the induction of some ethylene receptor genes by ethylene may be a mechanism for adapting to ethylene. If the different receptors have different affinities for ethylene, the induction of receptors with a low affinity for ethylene could allow the plant to become desensitized to ethylene such that higher and higher ethylene concentrations are needed to maintain a response (Hua and Meyerowitz 1998 ). In the case of cytokinin-induced genes that also display a basal level of expression in the absence of cytokinin, their upregulation by cytokinin may be necessary to amplify the signal and its response (Brandstatter and Kieber 1998 ). Given that ANI1 is expressed at low levels in non-ACE-treated gametophytes, it is possible that the induction of ANI1 is required to amplify ACE signaling or its response. Such amplification may be necessary to initiate male development in the gametophyte when antheridiogens are present at extremely low concentrations.
Given that there are at least two TRA loci (Banks 1997b ), two different active antheridiogens in CFM that are separable by thin-layer chromatography and HPLC (E. Strain, C. Chapple, and J.A. Banks, unpublished data), and multiple ANI1-like genes in Ceratopteris, the regulation of sex determination by ACE is likely to be complex. Continued studies to identify and integrate these components in the sex determination process in Ceratopteris should be useful in helping us to understand how pheromones regulate complex developmental processes in plants.
|
|
METHODS |
|---|
Strains and Growth Conditions
The origins of Hnn, the wild-type stain of Ceratopteris richardii, and the feminization1 (fem1) and transformer5 (tra5) sex-determining mutants as well as the conditions for gametophyte and sporophyte culture are described by Banks 1994 , Banks 1997b . Gametophytes were cultured at a density of 1 g L-1 liquid medium unless noted otherwise. Spores were cultured either in fern medium (FM) lacking Ceratopteris antheridiogen (ACE) or in conditioned fern medium (CFM) containing exogenous ACE, prepared according to Banks 1994 . CFM is a crude aqueous extract of medium that has supported the growth of wild-type gametophytes for 14 days.
Physiology of the ACE Response
The period of time that developing gametophytes are sensitive to exogenous ACE was determined by adding wild-type spores to FM in a single flask. Spores or developing gametophytes were periodically removed from the flask and placed individually into microtiter wells containing CFM. The sex of each gametophyte was scored 14 days after spore inoculation. The timing of ACE secretion was determined by adding wild-type spores to FM in a single flask. At varying periods of time, an aliquot of medium was removed, the spores or developing gametophytes removed by filtration, and the medium placed in individual microtiter wells. Fresh wild-type spores were then added individually to each well. The sex of each gametophyte was then scored 14 days after spore inoculation. In all experiments, 300 gametophytes were scored per time point or treatment.
DNA and RNA Isolation and Preparation of cDNA
Nucleic acids were isolated from gametophytes by mixing ground gametophytes in a buffer containing 4 M guanidine thiocyanide, 1 M ammonium thiocyanide, 0.3 M sodium acetate, pH 5.2, 1% sarcosine, 1% polyvinylpolypyrrlidone, and 0.5% ß-mercaptoethanol. After three chloroform–isoamylalcohol (24:1) extractions, nucleic acids were precipitated with 2.5 volumes of ethanol. RNA was further purified by hexadecyltrimethylammonium bromide (CTAB) precipitation (Murray and Thompson 1980 ), except that 2% CTAB and 0.84 M NaCl were used. DNA was isolated in the same way, except that 0.3 M sodium acetate in the initial extraction buffer was replaced with 0.1 M Tris-HCl, pH 7.5.
cDNAs were synthesized with an oligo(dT) primer, using the SuperScriptII system (Life Technologies, Gaithersburg, MD). cDNAs were either used in the gene expression screen or cloned into the 
Zip-Lox vector (Life Technologies).
Gene Expression Screen
Two populations of cDNAs, one derived from 80-hr-old ACE-treated gametophytes (the tracer cDNA) and the other from 80-hr-old non-ACE-treated gametophytes (the driver cDNA), were digested with AluI. The remaining subtraction/hybridization steps to remove driver cDNAs from tracer cDNA population were performed according to Wang and Brown 1991 with the following modifications. Two sets of adapters, rather than one, were ligated to the AluI-digested cDNA fragments. The sequences of the oligonucleotides used for linker preparation are as follows: J1, 5'-CTCTTGCTTGAATTCGGACTA-3'; J2, 5'-TAGTCCGAATTCAAGCAAGAGCACA-3'; CH1, 5'-ATCAGGCTTAAGTTCGTTCTC-3'; and CH2, 5'-GAGAACGAACTTAAGCCTGATCACA-3'. Biotin-labeled J1 or CH1 was used to generate driver cDNA, whereas nonlabeled J1 or CH1 was used to generate tracer cDNA. After three rounds of hybridization and subtraction, the remaining cDNA fragments were cloned into the pBluescript SK+ vector (Stratagene, La Jolla, CA).
Differential Screening by Colony Hybridization
Twenty thousand colonies containing single tracer cDNA inserts were transferred to nitrocellulose membranes and hybridized with 32P-labeled cDNA isolated from 80-hr-old gametophytes not treated with ACE. Replicate membranes were probed with labeled driver cDNA from the gene expression screen. The cloned inserts from colonies that did not hybridize to either probe were selected.
RNA and DNA Gel Blot Analyses
For RNA gel blots, 5 µg of total RNA prepared from gametophytes of various ages either treated or not treated with ACE was size fractionated on formaldehyde gels and blotted onto nylon membranes according to the manufacturer's instructions (Hybond N+; Amersham, Piscataway, NJ). For DNA gel blots, 10 µg of restriction-digested genomic DNA was size fractionated and blotted onto nylon membranes, according to the manufacturer's instructions (Hybond N+; Amersham). RNA and DNA gel blots were hybridized to probes labeled according to Mertz and Rashtchian 1994 by using hybridization conditions described by Church and Gilbert 1984 .
DNA Sequencing and Sequence Analysis
DNA sequencing was performed by the Iowa State University (Ames, IA) sequencing facilities, by the Purdue University sequencing facility, or by manual sequencing according to the manufacturer's instructions (Fidelity; Ventana Medical Systems, Tucson, AZ). Sequence similarities were sought using the BLAST system available through the National Center for Biotechnology Information (Bethesda, MD) (http://www.nim.nih.gov/cgi-bin/BLAST/). The TMpred program (Hofmann and Stoffel 1993 ) was used to predict membrane-spanning regions and their orientation. Predictions of protein secondary structure were made using the following programs available on the World Wide Web: Predator (Frishman and Argos 1996 ), Gibrat (Gibrat et al. 1987 ), Double Prediction (Deleage and Roux 1987 ), SSP (Solovyev and Salamov 1994 ), and GOR IV (Garnier et al. 1996 ).
Cycloheximide Treatment
The amount of cycloheximide (CHX) required to block in vivo protein synthesis was determined by culturing spores (400 mg per 400 mL of FM) for 72 hr and then adjusting the density of spores to 400 mg per 4 mL of FM. The spores were divided into four tubes, 100 µCi of 35S-L-methionine was added to each tube, and each tube was incubated for 20 min. CHX (0, 5, 10, or 20 µM) was then added, and the gametophytes were cultured for an additional 3 hr before extracting total protein (Hames 1990 ). To determine the minimum period of time that gametophytes must be exposed to CHX to block protein synthesis, similar experiments were performed, except that gametophytes were cultured in the presence of 20 µM CHX for varying periods of time before adding 100 µCi of 35S-L-methionine; cultures were then incubated for 3 hr before extracting total protein. Total protein was fractionated by SDS-PAGE and stained with Coomassie Brilliant Blue R 250 (Life Technologies) to ensure that each sample contained approximately the same amount of protein. The incorporation of 35S-L-methionine into proteins after various treatments was determined by autoradiography of gels and trichloroacetic acid precipitation according to Boefey 1990 .
|
|
FOOTNOTES |
|---|
1 Current address: Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD 20742. 
|
|
ACKNOWLEDGMENTS |
|---|
We thank Errol Strain and Brody DeYoung for their assistance in the greenhouse and Clint Chapple and Joe Ogas for critical reading of the manuscript. This work was supported by the National Science Foundation.
Received November 19, 1998; accepted April 6, 1999.
|
|
REFERENCES |
|---|
Abel, S., and Theologis, A. (1996) Early genes and auxin action. Plant Physiol. 111:9-17[CrossRef][ISI][Medline].
Banks, J. (1994) Sex-determining genes in the homosporous fern Ceratopteris. Development 120:1949-1958[Abstract].
Banks, J. (1997a) Sex determination in the fern Ceratopteris. Trends Plant Sci. 2:175-180[CrossRef][ISI].
Banks, J. (1997b) The TRANSFORMER genes of the fern Ceratopteris simultaneously promote meristem and archegonia development and repress antheridia development in the developing gametophyte. Genetics 147:1885-1897[Abstract].
Banks, J., Hickok, L., and Webb, M. (1993) The programming of sexual phenotype in Ceratopteris. Int. J. Plant Sci. 154:522-534[CrossRef].
Bavik, C., Busch, C., and Eriksson, U. (1992) Characterization of a plasma retinol-binding protein membrane-receptor expressed in the retinal-pigment epithelium. J. Biol. Chem. 267:23035-23042[Abstract/Free Full Text].
Boefey, S.A. (1990). 35S Labeling: In vivo labeling of plants. In Radioisotopes in Biology: A Practical Approach, R.I. Slater, ed (New York: IRL Press), pp. 127–136.
Brandstatter, I., and Kieber, J.J. (1998) Two genes with similarity to bacterial response regulators are rapidly and specifically induced by cytokinin in Arabidopsis. Plant Cell 10:1009-1019[Abstract/Free Full Text].
Church, G., and Gilbert, W. (1984) Genomic sequencing. Proc. Natl. Acad. Sci. USA 81:1991-1995[Abstract/Free Full Text].
Deleage, G., and Roux, B. (1987) An algorithm for protein secondary structure prediction based on class prediction. Protein Eng. 1:289-294[Abstract/Free Full Text].
Eberle, J., and Banks, J. (1996) Genetic interactions among sex-determining genes in the fern Ceratopteris richardii. Genetics 142:973-985[Abstract].
Flower, R.D. (1996) The lipocalin protein family: Structure and function. Biochem. J. 318:1-14.
Frishman, D., and Argos, P. (1996) Incorporation of non-local interactions in protein secondary structure prediction from the amino acid sequence. Protein Eng. 9:133-142[Abstract/Free Full Text].
Garnier, J., Gibrat, J., and Robson, B. (1996) GOR secondary structure prediction method version IV. Methods Enzymol. 266:540-553[ISI][Medline].
Gibrat, J.F., Garnier, J., and Robson, B. (1987) Further developments of protein secondary structure prediction using information-theory—New parameters and consideration of residue pairs J. Mol. Biol. 198:425-443.
Hames, B.D. (1990). One-dimensional polyacrylamide gel electrophoresis. In Gel Electrophoresis of Proteins: A Practical Approach, B.D. Hames and D. Rickwood, eds (New York: IRL Press), pp. 1–149.
Hickok, L. (1983) Abscisic acid blocks antheridiogen-induced antheridium formation in gametophytes of the fern Ceratopteris. Can. J. Bot. 61:888-892.
Hofmann, K., and Stoffel, W. (1993) TMbase—A database of membrane spanning protein segments. Biol. Chem. Hoppe Seyler 374:166.
Hua, J., and Meyerowitz, E. (1998) Ethylene responses are negatively regulated by a receptor gene family in Arabidopsis thaliana. Cell 94:261-271[CrossRef][ISI][Medline].
Hua, J., Sakai, H., Nourizadeh, S., Chen, Q.Q., Bleecker, B.A., Ecker, J.R., and Meyerowitz, E.M. (1998) EIN4 and ERS2 are members of the putative ethylene receptor gene family in Arabidopsis. Plant Cell 10:1321-1332[Abstract/Free Full Text].
Kakimoto, T. (1996) CKI1, a histidine kinase homolog implicated in cytokinin signal transduction. Science 274:982-985[Abstract/Free Full Text].
Kim, J., Harter, K., and Theologis, A. (1997) Protein–protein interactions among the Aux/IAA proteins. Proc. Natl. Acad. Sci. USA 94:11786-11791[Abstract/Free Full Text].
Koitabashi, R. (1996). Sex Determining Factor in Ceratopteris (Ceratopteris richardii). Masters Thesis (Tokyo, Japan: University of Tokyo).
Koshiba, T., Ballas, N., Wong, L., and Theologis, A. (1995) Transcriptional regulation of PS-IAA4/5 and PS-IAA6 early gene expression by indoleacetic acid and protein synthesis inhibitors in pea (Pisum sativum). J. Mol. Biol. 253:396-413[CrossRef][ISI][Medline].
Lobel, D., Marchese, S., Krieger, J., Pelosi, P., and Breer, H (1998) Subtypes of odorant-binding proteins—Heterologous expression and ligand binding. Eur. J. Biochem. 254:318-324[ISI][Medline].
Mertz, L.M., and Rashtchian, A. (1994) Nucleotide imbalance and polymerase chain reaction—Effects on DNA amplification and synthesis of high specific activity radiolabeled DNA probes. Anal. Biochem. 221:160-165[CrossRef][ISI][Medline].
Murray, M.G., and Thompson, W.F. (1980) Rapid isolation of high-molecular plant DNA. Nucleic Acids Res. 8:4321-4325[Abstract/Free Full Text].
Nielsen, H., Engelbrecht, J., Brunak, S., and von Heijne, G. (1997) Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Sci. 10:1-6[Free Full Text].
Rouse, D., Mackay, P., Stirnberg, P., Estelle, M., and Leyser, O. (1998) Changes in auxin response from mutations in an Aux/IAA gene. Science 279:1371-1373[Abstract/Free Full Text].
Schaller, G., and Bleecker, A. (1995) Ethylene binding sites generated in yeast expressing the Arabidopsis ETR1 gene. Science 270:1809-1811[Abstract/Free Full Text].
Schedlbauer, M.D., and Klekowski, E.J., Jr. (1972) Antheridiogen activity in the fern Ceratopetris thalictroides (L.) Brongn. Bot. J. Linn. Soc. 65:399-413.
Smeland, S., Bjerknes, T., Malaba, L., Eskild, W., Norum, K., and Blomhoff, R. (1995) Tissue distribution of the receptor for plasma retinol-binding protein. Biochem. J. 305:419-424.
Solovyev, V., and Salamov, A. (1994) Predicting -helix and ß-strand segments of globular proteins. Comput. Appl. Biosci. 10:661-669[Abstract/Free Full Text].
Takeno, K., Yamane, H., Yamauchi, T., Takahashi, N., Furber, M., and Mander, L. (1989) Biological activities of the methyl ester of gibberellin a73, a novel and principal antheridiogen in Lygodium japonicum. Plant Cell Physiol. 30:201-215[Abstract/Free Full Text].
Ulmasov, T., Hagen, G., and Guilfoyle, T. (1997) ARF1, a transcription factor that binds to auxin response elements. Science 276:1865-1868[Abstract/Free Full Text].
Wang, Z., and Brown, D.D. (1991) A gene-expression screen. Proc. Natl. Acad. Sci. USA 88:11505-11509[Abstract/Free Full Text].
Warne, T.R., and Hickok, L.G. (1989) Evidence for a gibberellin biosynthetic origin of Ceratopteris antheridiogen. Plant Physiol. 89:535-538[Abstract/Free Full Text].
Warne, T.R., and Hickok, L.G. (1991) Control of sexual development of Ceratopteris richardii: Antheridiogen and abscisic acid. Bot. Gaz. 152:148-153[CrossRef].
Wynne, G., Mander, L., Goto, N., Yamane, H., and Omori, T. (1998) Biosynthetic origin of the antheridiogen, gibberellin A(73) methyl ester, in ferns of the Lygodium genus. Tetrahedron Lett. 39:3877-3880[CrossRef].
Yamane, H. (1998) Fern antheridiogens. Int. Rev. Cytol. 184:1-32[ISI].
Yamauchi, T., Oyama, N., Yamane, H., Murofushi, N., Takahashi, N., Schraudolf, H., Furber, M., Mander, L., Patrick, G., and Twitchin, B. (1991) Biosynthesis of antheridic acid, the principal antheridiogen in Anemia phylliditis. Phytochemistry 30:3247-3250[CrossRef].
Yamauchi, T., Oyama, N., Yamane, H., Murofushi, N., Schraudolf, H., Owen, D., and Mander, L.N. (1995) 3-Epi-GA63, antheridiogen in Anemia phylliditis. Phytochemistry 38:1345-1348[CrossRef].
Yamauchi, T., Oyama, N., Yamane, H., Murofushi, N., Schraudolf, H., Pour, M., Furber, M., and Mander, L.N. (1996) Identification of antheridiogens in Lygodium circinnatum and Lygodium flexuosum. Plant Physiol. 111:741-745[Abstract].
ABSTRACT
INTRODUCTION
