DOI: 10.1534/genetics.107.082446
Note
Cell Cycle Arrest of Stamen Initials in Maize Sex Determination
Jong Cheol Kim,*
,†He´le`ne Laparra,*
,1Alejandro Caldero´n-Urrea,*
,2John P. Mottinger,
‡Maria A. Moreno* and Stephen L. Dellaporta*
,3*Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520-8104,
†Environmental Biotechnology National Core Research Center, Gyeongsang National University, Jinju, 660-701,
Korea and‡Department of Cell and Molecular Biology, University of Rhode Island, Kingston, Rhode Island 02881
Manuscript received October 11, 2007 Accepted for publication October 17, 2007
ABSTRACT
The maize sex determination pathway results in the arrest of stamen in ear spikelets and the abortion of pistils in both the tassel spikelets and in the secondary florets of ear spikelets. Arrested stamen cells showed no signs of DNA fragmentation, an absence ofCYCLIN Bexpression, and an accumulation of the negative cell cycle regulatorWEE1RNA.
M
AIZE plants produce unisexual flowers, called florets in grasses, which are spatially separated, with staminate florets on the terminal inflorescence and pistillate florets on the axillary inflorescences (Figure 1). These florets are arranged in pairs (upper and lower) within a spikelet, the basic unit of grass inflorescence. Each floret is initially bisexual with three stamen initials and a central pistil. Unisexuality is achieved through the selective elimination of floral or-gans. All pistil primordia are eliminated in tassel spike-lets resulting in exclusively paired staminate florets. In most lines of maize, one or more axillary inflorescences (ears) contain spikelets with one pistillate floret and a sterile secondary floret through the elimination of all stamen primordia and the pistils in the secondary florets.Maize sex determination is under the control of cell death, cell protection, and phytohormone-mediated pro-cesses (reviewed by Irishand Nelson1989; Dellaporta
and Calderon-Urrea1995; Irish1996). Mutations that
affect sex include thetasselseed 1(ts1) andts2mutations, which cause the failure of primary and secondary tassel pistils to abort with the subsequent arrest of all tassel stamen primordia. The abortion of pistils in the second-ary ear florets is also blocked in these mutants resulting in
paired pistillate ear spikelets. The ts2 gene encodes a short-chain alcohol dehydrogenase/reductase (SDR) that is required for the abortion of pistil cells in both tassel spikelets and the secondary pistils of ear spikelets (Delonget al. 1993).TS2is also expressed in the
func-tional primary pistils of the ear spikelets even though these pistils do not undergotasselseed-mediated cell death. Primary ear pistils survive through the action of thesilkless 1(sk1) gene; mutations insk1block pistil survival causing these pistils to abort by a tasselseed-mediated cell death process (Veit et al. 1991; Irish and Nelson 1993;
Calderon-Urreaand Dellaporta1999).
We examined the process of the formation of pistil-late florets in both wild-type and in ts2mutant plants. The nuclear integrity in arresting stamen initials was examined by staining paraffin-embedded tissue sections with DAPI (Figure 2). At the organ level, stamen arrest was evidenced in wild-type ear florets by the lack of enlargement of stamen initials soon after the bisexual stage of floral development. In these growth-arrested stamens, cells continued to show nuclear integrity, as judged by DAPI staining (Figure 2A). As the primary pistils continued to mature, stamen initials showed no loss of nuclear integrity even at the stage wheretasselseed -mediated cell death had eliminated E2 pistils (Figure 2B). Unlike pistil abortion in tassel and secondary ear florets (Cheng et al. 1983; Calderon-Urrea and
Dellaporta 1999), nuclear degeneration is not an
indication of the stamen arrest process in ear florets. Mutations in ts2 cause pistillate florets to form in tassels instead of staminate florets. We examined the arrested stamens in the tassels of ts2 mutant plants. As with wild-type florets, the ts2 florets were initially 1Present address:Meristem Therapeutics, 8 rue des Fre`res Lumie`re,
63100 Clermont-Ferrand, France.
2Present address: Department of Biology, California State University, Fresno, CA 93740.
3Corresponding author: Yale University, Department of Molecular, Cellular and Developmental Biology, P.O. Box 208104, New Haven, CT 06520-8104. E-mail: [email protected]
bisexual (not shown) followed soon afterward by the arrest of stamen initials and the subsequent maturation of the pistil (Figure 2C). In arrested stamens, nuclear integrity was again evident with no indication of nuclear degeneration, and only at very late stages of inflores-cence development (.50 mm length) did any signs of a loss of nuclear integrity in arrested stamens become apparent (not shown). The process of stamen arrest in ts2tassel florets therefore appears to be similar to that
observed in wild-type ears, whereby an arrest of stamens occurs without a loss of nuclear integrity.
To further characterize the process of stamen arrest, we performed TdT-mediated dUTP-biotin nick end labeling (TUNEL) assays (Gavrieliet al. 1992) to detect
signs of nuclear DNA fragmentation. In tassel and ear inflorescences up to 25 mm in length (bisexual stage), no DNA fragmentation was detected in the floral primordia in either pistil or stamen primordia (Figure 3, A and G). At later stages (inflorescences 30–50 mm length) as florets become unisexual, arrested stamens in ear florets showed no TUNEL signal (Figure 3, B and C) indicating a lack of detectable DNA fragmentation. As a positive control, we examined the process of elim-ination of pistil initials in wild-type tassel florets. As previously shown, pistil cells undergo a process of sub-epidermal tasselseed-mediated cell death detected by rapid loss of nuclei integrity (Calderon-Urreaand
Dellaporta 1999). In aborting pistils, subepidermal
cells were clearly positive in TUNEL assays (Figure 3, H and I) indicating extensive DNA fragmentation. TUNEL signals preceded the complete loss of nuclear integrity as judged by propidium iodide staining. Stamen arrest therefore was not a direct consequence of DNA frag-mentation or cell death, unlike the process of pistil cell death, which is accompanied by a rapid loss of nuclear integrity and DNA fragmentation.
Next, we examined the possibility that a block in stamen development was the result of cell cycle arrest. Thecyclin Bgene acts as a positive regulator in the G2/M phase of the cell cycle and its RNA is short-lived and present only in dividing cells in the G2/M transition (Fobertet al. 1994).In situRNA hybridization with the
maize B-type cyclin gene (Renaudinet al. 1994) is shown
in Figure 3. Thecyclin Bgene was highly expressed in a subset of cells in all floral organs in the early bisexual stage of inflorescence development, an indication of prevalent cell division in these tissues. As ear florets matured,CYCLIN Bexpression continued to be detected
Figure1.—Sex determination pathway in maize. Maize
flo-ral diagrams showing an immature spikelet (left) with two per-fect (bisexual) floral meristems, each with a central pistil initial (s), three stamen initials (d), a palea (–), lemma
( or ), and subtending glumes ( or ); lodicules are not shown. The transition to paired staminate spikelets in the tassel (top right) and to solitary pistillate spikelets in the ear (bottom right) is depicted. Secondary pistils also abort in ear florets (indicated by asterisk). Thetasselseed (ts) genes
1, 2, 3, and 5are required for pistil abortion in the tassel and in the secondary ear florets. Thedwarf(d)1,3,5, and
8andanther ear 1(an1) genes are required for stamen arrest in ear spikelets and ints2mutant tassel spikelets. Thesilkless 1
(sk1) gene protects primary ear pistils fromtasselseed-mediated cell death. Thepistillate(pi1, pi2) genes are required for E2 pistil abortion.
Figure2.—Nuclear integrity in arrested stamen
initials. Paraffin-embedded sections of wild-type ear spikelets were analyzed by DAPI staining. (A) Early unisexual stage of ear spikelet development showing developing upper (E1) pistils and ar-rested stamens with intact nuclei. Lower (E2) pistils have not yet undergone tasselseed-mediated cell death. (B) Later unisexual stage of ear spikelet de-velopment showing arrested upper (E1) stamens with intact nuclei and complete elimination of the lower pistil (red arrow). Arrested lower stamen initials also show intact nuclei. (C) Ints2mutant tassel spikelets, developing pistil and arrested up-per (T1) stamen initials show intact nuclei. Bars, 100mm. Inflorescences were fixed in 4% formalde-hyde, embedded in paraffin, and sections were prepared forin situhybridizations essentially as described ( Jackson1992) except that
Hemo-De (Fisher Scientific) was substituted for Histo-Clear (National Diagnostics, Atlanta). Nuclear integrity assays were performed by incubation of tissue sections in a 0.3 mmsolution of 49,6-diamidine-29-phenylindole dihydrochloride (DAPI; Roche Applied
throughout a subset of cells in developing pistils, but CYCLIN Bexpression was undetectable in any cells of the stamens (Figure 3, E and F), consistent with a cessation of cell division and arrested development of these stamens. Moreover, the lack ofCYCLIN B expres-sion occurred at the time that nuclei integrity remained high as judged by propidium iodide staining (Figure 3, B and C).
Likewise, in the tassel inflorescence, CYCLIN B ex-pression was seen in a subset of cells throughout the floral meristem during early development (Figure 3J) and at later stages in the developing stamens (Figure 3K) but was absent from aborting pistils of wild-type tassels (Figure 3L). In mutantts2tassels,CYCLIN Bwas detected at all stages examined in a subset of cells of pistils (Figure 3, P–R), but not at later stages in stamens
Figure 3.—CYCLIN B1 expression and DNA
fragmentation in floral organs. TUNEL assays (green fluorescent signal) on paraffin-embedded section of wild-type ear (A–C), wild-type tassel (G–I), andts2 mutant tassel (M–O) florets and counterstained with propidium iodide (red fluo-rescent signal).In situhybridization withCYCLIN B1antisense probes to serial sections of wild-type ear (D–F), wild-type tassel ( J–L) andts2mutant tassel (P–R) florets. Note the absence of TUNEL signals (B and C) in wild-type stamen initials also showing no CYCLIN B1 expression (E and F). TUNEL signal is seen in wild-type pistils under-going tasselseed-mediated cell death (H and I). Bars, 100 mm. Total RNA from maize tissues was purified by the guanidine thiocyanate method (Chirgwin et al. 1979). First strand cDNA was
synthesized using SuperScript II (GIBCO BRL) and an oligod(T) primer according to the manu-facturer’s instructions. The cDNA clones were ob-tained by PCR amplification of first strand cDNA using CYCLIN B1 primers (P917 59-CCTGGACTCTGAGAA CAGCCTACCA-39 and P918 59-CCGACTCTGAGAA CAGCCTAGCAAA-39). PCR amplification was per-formed using the expand long template PCR system (Boehringer Mannheim) using the manu-facturer’s instructions under the following cy-cling conditions: 94° for 2 min, then 10 cycles at 94° for 10 sec, 65° for 30 sec, and 68° for 2 min, followed by an additional 20 cycles of PCR at 94°for 10 sec, 65°for 30 sec, 68°for 2 min. Am-plification products were cloned into the vector pCRII (Invitrogen), and sequenced to confirm their identity. For in situ riboprobes, plasmid DNA was linearized at a restriction site flanking the cDNA inserts by digestion with an appropri-ate enzyme that generappropri-ates a 59overhang. Linear-ized plasmid DNA was used as a template to synthesize digoxigenin-labeled RNA using T7 or T3 polymerase and 11-digoxigenin-dUTP accord-ing to manufacturer’s instructions (Roche Diag-nostics). The RNA probe was subject to mild alkaline hydrolysis for varying times in 0.1 n
NaOH to yield products in the 100–200 nucleotide range. The final concentration of riboprobe was adjusted to 10 ng/ml/kb with overnight hybridiza-tion at 55°. Tissue fixation, nuclear staining, and sectioning were as described above (Figure 2). Washing and detection of RNA hybridization signals was as previously described ( Jackson
1992; Calderon-Urreaand Dellaporta1999).
(Figure 3, Q and R), a result consistent with the arrest and cessation of cell division in these organs. Taken together, the data on nuclear integrity, TUNEL assays, andCYCLIN B expression suggest that the process of stamen arrest involves a cell cycle block rather than a cell death mechanism characteristic oftasselseed-mediated pistil elimination.
To further investigate the status of the cell cycle in the arrest of stamens, we examined the expression of several cell cycle regulators, including several other cyclins (Renaudin et al. 1994; Gutierrez et al. 2005),
retinoblastoma-related protein (Grafi et al. 1996),
cyclin-dependent kinase (CDK) (de la Paz Sanchez
et al. 2002; Sanchez Mde et al. 2005), the spindle
checkpoint protein MAD2 (Yu et al. 1999), CDK
in-hibitor (Coelhoet al. 2005), andWEE1(Xieet al. 1996;
Sun et al. 1999) genes (data not shown). Interestingly,
expression of a negative cell cycle regulator,WEE1, showed a striking accumulation of its RNA in arrested stamen cells but not in other floral organs, including dying pistils (Figure 4, A and C). The maizeWEE1mRNA was well expressed in the arrested stamens of both wild-type ear florets (Figure 4A) and in mutantts2tassel florets (Figure 4C). This period in whichWEE1RNA accumu-lation was observed coincided with the period in which CYCLIN Bwas absent in the cells of the arrested stamens (Figure 4, D and F). AlthoughCYCLIN B expression was absent (Figure 4C),WEE1RNA did not accumulate during pistil abortion in ear florets (Figure 4B). This result implicates a cell cycle block, possibly mediated by negative cell cycle regulatorWEE1, in the process of arresting stamen cells but apparently not in the process of pistil abortion.
Previous studies have shown that the elimination of pistils in tassel florets and secondary ear florets is a cell death process (Chenget al. 1983; Calderon-Urrea
and Dellaporta 1999). Pistil cell death is associated
with subepidermalTS2expression and nuclear degen-eration shortly after florets reach a bisexual stage of development (Delong et al. 1993; Calderon-Urrea
and Dellaporta1999). In contrast to this process, the
failure of stamens to mature in pistillate florets does not coincide with any early indications of cell death with growth of initials arrested shortly after the bisexual floret stage (Cheng et al. 1983; Delong et al. 1993;
Calderon-Urreaand Dellaporta1999). At this stage
of floral development there are no indications of a loss of nuclear integrity in arrested stamen cells as judged by DAPI (Figure 2, A and B) or propidium iodide (Figure 3, A–C) staining. The earliest cellular indica-tions of stamen arrest was the cessation of cell division, as determined by the absence of detectable expression of the G2/M regulatorCYCLIN Bin the arrested stamens of wild-type ear florets (Figure 3, E and F) and in arrested stamens in thets2mutant tassel florets (Figure 3, Q and R). A second indication of cell cycle arrest was the accumulation of the negative cell cycle regulator WEE1RNA only seen in arrested ear stamens (Figure 4, A and C) and arrestedts2tassel stamens (Figure 4C). In contrast, the pistil abortion process was not associated with WEE1 RNA accumulation (Figure 4B), although CYCLIN BRNA was also not detected in aborting pistils (Figure 3, K and L). In aborting pistils, however, both signs of nuclear degeneration (Figure 2C) and DNA fragmentation (Figure 3, H and I) were evident.
In eukaryotic cells, cyclin B regulates the transition to mitosis and its transcription is cell-cycle regulated, peaking at the G2/M transition and repressed in G1 (reviewed by De Veylderet al. 2003; Fungand Poon
2005).WEE1 is a known to be a negative regulator of mitosis, inhibiting cyclin B/Cdc2 activity and the pro-gression of cells from G2/M in the cell cycle (Nurse and Thuriaux1980; Russelland Nurse1986).WEE1
encodes a Thr/Tyr protein kinase that inhibits CDKs by phosphorylation (Mueller et al. 1995a,b; Murakami
and Vande Woude 1998). In light of our results,
including the lack of DNA fragmentation, the absence ofCYCLIN Bexpression, and the accumulation ofWEE1, our data point to a cell cycle arrest rather than a cell death mechanism that is responsible for the production
Figure 4.—WEE1 andCYCLIN B1 expression
in floral organs.In situhybridization of serial sec-tions usingWEE1 (A–C) and CYCLIN B1(D–F) antisense probes of wild-type ear (A and D), wild-type tassel (B and E), andts2mutant tassel (C and F) florets. Note accumulation ofWEE1
of pistillate florets in the maize sex determination pathway.
This work was supported by grants from the National Institutes of Health (NIH) (R01 GM-38148) to S.L.D., from the United States Department of Agriculture (9935301-8048) to J.P.M., and from the Korean Science and Engineering Foundation (KOSEF) to the Environmental Biotechnology National Core Research Center (R15-2003-012-01001-0). J.C.K. was supported from a postdoctoral fellowship from the KOSEF and from the NIH (R01 GM-38148).
LITERATURE CITED
Calderon-Urrea, A., and S. L. Dellaporta, 1999 Cell death and cell protection genes determine the fate of pistils in maize. De-velopment126:435–441.
Cheng, P. C., R. I. Grysonand D. B. Walden, 1983 Organ initiation and the development of unisexual flowers in the tassel and ear of
Zea mays. Am. J. Bot.70:450–462.
Chirgwin, J. M., A. E. Przbyla, R. J. Macdonaldand W. J. Rutter, 1979 Isolation of biologically active ribonucleic acid from sour-ces enriched in ribonuclease. Biochemistry18:5294.
Coelho, C. M., R. A. Dante, P. A. Sabelli, Y. Sun, B. P. Dilkeset al., 2005 Cyclin-dependent kinase inhibitors in maize endosperm and their potential role in endoreduplication. Plant Physiol.
138:2323–2336.
de laPazSanchez, M., A. Torres, M. B. Boniotti, C. Gutierrez and J. M. Vazquez-Ramo, 2002 PCNA protein associates to Cdk-A type protein kinases in germinating maize. Plant Mol. Biol.
50:167–175.
DeVeylder, L., J. Joubesand D. Inze, 2003 Plant cell cycle transi-tions. Curr. Opin. Plant Biol.6:536–543.
Dellaporta, S. L., and A. Calderon-Urrea, 1995 The sex determi-nation process in maize. Science266:1501–1505.
DeLong, A., A. Calderon-Urreaand S. L. Dellaporta, 1993 Sex determination geneTASSELSEED2 of maize encodes a short-chain alcohol dehydrogenase required for stage-specific floral organ abortion. Cell74:757–768.
Fobert, P. R., E. S. Coen, G. J. Murphy and J. H. Doonan, 1994 Patterns of cell division revealed by transcriptional regu-lation of genes during the cell cycle in plants. EMBO J. 13:
616–624.
Fung, T. K., and R. Y. Poon, 2005 A roller coaster ride with the mi-totic cyclins. Semin. Cell Dev. Biol.16:335–342.
Gavrieli, Y., Y. Shermanand S. A. Ben-Sasson, 1992 Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell. Biol.119:493–501.
Grafi, G., R. J. Burnett, T. Helentjaris, B. A. Larkins, J. A. DeCaprio
et al., 1996 A maize cDNA encoding a member of the retinoblas-toma protein family: involvement in endoreduplication. Proc. Natl. Acad. Sci. USA93:8962–8967.
Gutierrez, R., F. Quiroz-Figueroa and J. M. Vazquez-Ramos, 2005 Maize cyclin D2 expression, associated kinase activity and effect of phytohormones during germination. Plant Cell Physiol.46:166–173.
Irish, E., 1996 Regulation of sex determination in maize. BioEssays
18:363–369.
Irish, E. E., and T. Nelson, 1989 Sex determination in monoecious and dioecious plants. Plant Cell1:737–744.
Irish, E. E., and T. M. Nelson, 1993 Development of tasselseed2 inflorescences in maize. Am. J. Bot.80:292–299.
Jackson, D., 1992 In situhybridization in plants, pp. 163–174 in
Plant Molecular Pathology: A Practical Approach, edited by S. J. Gurr, M. J. McPhersonand D. J. Bowles. Oxford University Press, Oxford.
Mueller, P. R., T. R. Colemanand W. G. Dunphy, 1995a Cell cycle reg-ulation of a Xenopus Wee1-like kinase. Mol. Biol. Cell6:119–134. Mueller, P. R., T. R. Coleman, A. Kumagai and W. G. Dunphy, 1995b Myt1: a membrane-associated inhibitory kinase that phosphorylates Cdc2 on both threonine-14 and tyrosine-15. Sci-ence270:86–90.
Murakami, M. S., and G. F. VandeWoude, 1998 Analysis of the early embryonic cell cycles of Xenopus; regulation of cell cycle length by Xe-wee1 and Mos. Development125:237–248. Nurse, P., and P. Thuriaux, 1980 Regulatory genes controlling
mitosis in the fission yeast Schizosaccharomyces pombe. Genetics
96:627–637.
Renaudin, J. P., J. Colasanti, H. Rime, Z. Yuanand V. Sundaresan, 1994 Cloning of four cyclins from maize indicates that higher plants have three structurally distinct groups of mitotic cyclins. Proc. Natl. Acad. Sci. USA91:7375–7379.
Russell, P., and P. Nurse, 1986 cdc251functions as an inducer in the mitotic control of fission yeast. Cell45:145–153.
Sanchez Mde, L., S. H. Gurusinghe, K. J. Bradford and J. M. Vazquez-Ramos, 2005 Differential response of PCNA and Cdk-A proteins and associated kinase activities to benzyladenine and abscisic acid during maize seed germination. J. Exp. Bot.56:
515–523.
Sun, Y., B. P. Dilkes, C. Zhang, R. A. Dante, N. P. Carneiroet al., 1999 Characterization of maize (Zea mays L.) Wee1 and its activity in developing endosperm. Proc. Natl. Acad. Sci. USA
96:4180–4185.
Veit, B., B. Greene, B. Lowe, J. Mathern, N. Sinha et al., 1991 Genetic approaches to inflorescence and leaf develop-ment in maize. Dev. Suppl.1:105–111.
Xie, Q., A. P. Sanz-Burgos, G. J. Hannon and C. Gutierrez, 1996 Plant cells contain a novel member of the retinoblastoma family of growth regulatory proteins. EMBO J.15:4900–4908. Yu, H. G., M. G. Muszynskiand R. KellyDawe, 1999 The maize
homologue of the cell cycle checkpoint protein MAD2 reveals kinetochore substructure and contrasting mitotic and meiotic localization patterns. J. Cell. Biol.145:425–435.
