Nondisjunction in Favor of a Chromosome: The Mechanismof Rye B Chromosome Drive during Pollen MitosisW

Ali M. Banaei-Moghaddam,a Veit Schubert,a Katrin Kumke,a Oda Weib,a Sonja Klemme,a Kiyotaka Nagaki,b

Ji�rí Macas,c Mónica González-Sánchez,d Victoria Heredia,d Diana Gómez-Revilla,d Miriam González-García,d

Juan M. Vega,d Maria J. Puertas,d and Andreas Houbena,1

a Leibniz Institute of Plant Genetics and Crop Plant Research, 06466 Gatersleben, Germanyb Institute of Plant Science and Resources, Okayama University, Kurashiki 710-0046, JapancBiology Centre of the Academy of Sciences of the Czech Republic, Institute of Plant Molecular Biology, Ceske Budejovice 37005,Czech Republicd Departamento de Genética, Facultad de Biología, Universidad Complutense, 28040 Madrid, Spain

B chromosomes (Bs) are supernumerary components of the genome and do not confer any advantages on the organisms thatharbor them. The maintenance of Bs in natural populations is possible by their transmission at higher than Mendelianfrequencies. Although drive is the key for understanding B chromosomes, the mechanism is largely unknown. We providedirect insights into the cellular mechanism of B chromosome drive in the male gametophyte of rye (Secale cereale). We foundthat nondisjunction of Bs is accompanied by centromere activity and is likely caused by extended cohesion of the B sisterchromatids. The B centromere originated from an A centromere, which accumulated B-specific repeats and rearrangements.Because of unequal spindle formation at the first pollen mitosis, nondisjoined B chromatids preferentially become locatedtoward the generative pole. The failure to resolve pericentromeric cohesion is under the control of the B-specific nondisjunctioncontrol region. Hence, a combination of nondisjunction and unequal spindle formation at first pollen mitosis results in theaccumulation of Bs in the generative nucleus and therefore ensures their transmission at a higher than expected rate to the nextgeneration.

INTRODUCTION

Supernumerary or B chromosomes (Bs) are dispensable com-ponents of the genomes of numerous plant, fungi, and animalspecies. Because most Bs do not confer any advantages on theorganisms that harbor them, they are thought of as parasitic orselfish elements that persist in populations by making use of thecellular machinery required for the inheritance and maintenanceof A chromosomes (As) (reviewed in Jones and Houben, 2003).

The maintenance of Bs in natural populations is possible bytheir transmission at higher than Mendelian frequencies, and thisenables the maintenance of polymorphisms in populations(Kimura and Kayano, 1961). The variety of mechanisms, includingsegregation failure, by which B chromosomes gain advantage intransmission are known as accumulation or drive mechanisms.Depending on the species, B chromosome drive can be pre-meiotic, meiotic, or postmeiotic. In Poaceae, the drive is post-meiotic, occurring either in the first (rye [Secale cereale]), or in thesecond (maize [Zea mays]) mitosis of the pollen grain (reviewed inJones, 1991). Drive is the key for understanding B chromosomesand it occurs in many ways, but the molecular mechanisms re-main unclear (Jones, 1991; Burt and Trivers, 2006).

The behavior of rye Bs during pollen mitosis was first studiedby Hasegawa (1934), who described that the two chromatids ofthe B chromosome do not separate at anaphase of first pollengrain mitosis and in most cases are included in the generativenucleus. In the second pollen mitosis, the generative nucleusdivides to produce two sperm nuclei. The B chromatids, whichare often nondisjoined at first pollen anaphase, segregate nor-mally at second pollen mitosis. The irregular segregation of theBs during meiosis, forming univalents at metaphase I and lag-gards at anaphase I and II (Müntzing and Prakken, 1941), par-tially explains their non-Mendelian transmission, but the netincrease of B number in the progeny of 0B 3 +B crosses canonly be explained by the directed nondisjunction of the Bs to thegenerative nucleus during first pollen mitosis (Müntzing, 1946a).This mechanism has been observed in rye with up to six Bs(Kishikawa, 1965). A similar accumulation mechanism of Bs wasreported in other species of the tribe Triticeae (i.e., Aegilopsspeltoides; Mendelson and Zohary, 1972).In rye crosses 0B 3 2B (Bs transmitted via the male) or 2B 3

0B (Bs transmitted on the female side), plants with zero, two, orfour Bs are obtained, but plants with odd numbers of Bs are onlyrarely obtained in the progenies (Müntzing, 1945). Therefore,a similar drive process must be assumed to occur in male andfemale gametophytes. Indeed, Håkanson (1948) observed in theembryo sac during first postmeiotic division an anaphase cellwith lagging B chromosomes as previously observed at pollenanaphase.Nondisjunction works equally well when the rye B is introduced

as an addition chromosome into hexaploid wheat (Triticum

1 Address correspondence to houben@ipk-gatersleben.de.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Andreas Houben(houben@ipk-gatersleben.de).W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.112.105270

The Plant Cell, Vol. 24: 4124–4134, October 2012, www.plantcell.org ã 2012 American Society of Plant Biologists. All rights reserved.

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

aestivum; Lindström, 1965; Müntzing, 1970; Niwa et al., 1997;Endo et al., 2008), hypopentaploid Triticale (Kishikawa and Suzuki,1982), or Secale vavilovii (Puertas et al., 1985). Thus, the B con-trols the process of nondisjunction by itself (Matthews and Jones,1983; Romera et al., 1991).

Deficient Bs (defB) lacking the heterochromatic terminal re-gion of the long arm undergo normal disjunction at first pollenanaphase. Therefore, it seems that the accumulation mecha-nism of the B by nondisjunction requires a factor located at theend of its long arm (Müntzing, 1948; Håkanson, 1959; Endoet al., 2008). This factor can act in trans because if a standard B(Lima-De-Faria, 1962) or the terminal region of the long arm ofthe B (Endo et al., 2008) is present in the same cell containinga defB, nondisjunction occurs for both the standard and thedefB. Two B-specific repeat families E3900 (Blunden et al.,1993) and D1100 (Sandery et al., 1990) reside in the long armterminal region. Expression analysis revealed that D1100 andE3900 are highly transcriptionally active in anthers. In addition,the distal heterochromatin is marked with the euchromatin-specific histone modification mark H3K4me3 (Carchilan et al.,2007).

Here, we analyzed the fate of accumulating (standard Bs) andnonaccumulating Bs (defBs) during microgametogenesis. Thiswork provides direct insights into the cellular mechanism of ryeB chromosome drive. We identified a B-specific pericentromereorganization and suggest that the failure to separate sister chro-matids is caused by a B-specific pericentromere under the in-fluence of the nondisjunction control region. In combination withasymmetric geometry of the first pollen mitosis, the Bs ac-cumulate preferentially in the generative nucleus, which in turnenhances the transmission frequency of Bs to the next generation.If the nondisjunction control region of the B is missing, normalseparation of defB chromatids occurs and A and defB cen-tromeres intermingle in both nuclei of binucleate pollen. Hence,no accumulation of Bs occurs in generative nuclei.

RESULTS

A and B Centromeres Occupy Different Positions in theGenerative Nucleus of the Bicellular Pollen Grain

Bs significantly delay the progress of pollen mitosis. Anthersfrom 0B, 2B, and 4B plants revealed a mitotic index of 19, 32,and 52%, respectively (see Supplemental Figure 1 online). In 2Bplants, pollen prophases and metaphases showed a higher in-dex than in 0B, whereas anaphases and telophases showeda lower index. In 4B plants, the prophase index was much higherthan in 0B and 2B plants, and only a few cells were observed atmeta-, ana-, and telophase. This indicates that 2Bs delay thefirst pollen prophase and metaphase, and 4B delays the firstprophase so much that only a few pollen grains are able tofollow the mitotic process normally, resulting in aborted pollen.

To obtain direct insights into the mechanism of B accumulation,we analyzed the dynamics of chromosomes at microgameto-genesis, using tissue sections to maintain the natural arrangementof chromosomes. Pollen of +B rye (Figure 1) and of +B wheat(Figures 2A and 2B) undergoing the first mitotic division were

hybridized with the rye A and B centromere-specific probeBilby, and the B (peri)centromere-specific probe ScCl11-1, orthe B-specific probes E3900 to discriminate between As and Bs.In both species at anaphase, most Bs lagged behind the Asand exhibited nondisjunction. The resulting nuclei of the bicel-lular pollen are distinguished by contrasting chromatin config-urations: The generative nucleus is more condensed, whereasthe vegetative nucleus is characterized by diffuse chromatin(Houben et al., 2011). In 2B rye, the Bs were mostly observedin the generative nucleus (92%). In a few cases, the Bs showeddisjunction (1.5%) or formed micronuclei near the generativenucleus (6.5%). In 4B plants, B nondisjunction occurred be-cause more than two B signals can be observed in bicellularpollen. However, many pollen grains showed abnormalitiesand preferential distribution of the Bs often (30%) failed (seeSupplemental Figure 2 online).Notably, the distribution of centromeres differed between the

vegetative and generative nuclei in 0 and +B plants. In thevegetative nuclei, the A centromeres were loosely distributed,whereas in the generative nuclei, they were closely groupedtoward the pollen wall (Figure 1C). The A centromeres occupieda significantly smaller area (calculated based on a maximumintensity projection) in the generative nucleus (5.95 µm2) than in

Figure 1. Nondisjunction of Rye Bs.

(A) and (B) Pollen of rye +B at different stages of the first pollen mitosis insitu hybridized with the rye centromere-specific probe Bilby (in green)and the B centromere–specific probe ScCl11 (in red). At anaphase, asa result of nondisjunction, the Bs were unequally distributed (A) and/orlagged behind (B) the As and formed micronuclei. DAPI, 49,6-diamidino-2-phenylindole.(C) In the vegetative nucleus (V), the A centromeres are loosely distrib-uted, whereas in the generative nucleus (G), the A centromeres areclosely grouped and do not intermingle with B centromeres.Bs are arrowed. Bar = 10 µm.

B Chromosome Drive during Pollen Mitosis 4125

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

the vegetative nucleus (14.55 µm2; n = 12, t test P < 0.0003).Interestingly, centromeres of As and Bs did not intermingle andBs often occupied a different position toward the location of thecell plate.

To test whether the nondisjunction control region of the B isrequired for the contrasting distribution of As and Bs, we ana-lyzed the wheat-rye B addition line BS-2 (Endo et al., 2008),which is carrying a defB with the nondisjunction control regionmissing. Here, a random segregation of defBs was found at firstpollen mitosis, and the centromeres of As and Bs were in-terspersed (Figure 2C). Like in rye, the centromeres in the veg-etative nucleus were loosely distributed, whereas in the generativenucleus, the centromeres were associated. The same distributionof centromeres was found in pollen of 0B plants (Figure 2D). Hence,the nondisjunction control region of the B is required for the

accumulation and the interior positioning of Bs in the generativenucleus. However, this region does not control the differentialgrouping of A centromeres in the generative and vegetativenuclei.

Directed Nondisjunction of Rye Bs Is Accompanied byCentromere Activity and Unequal Cell Division

To test whether the absence of centromere activity causesnondisjunction of Bs, we assayed the centromeric histone H3variant CENH3 at microgametogenesis. CENH3 was selectedbecause in mammals (Howman et al., 2000) and Hordeum hy-brids (Sanei et al., 2011) its loss results in the failure of chro-mosome segregation. In parallel, the dynamics of microtubulefibers was visualized to test whether in rye and wheat the firstpollen mitosis represents an asymmetric cell division, as pre-viously reported for other plants (reviewed in Tanaka, 1997; Borget al., 2009).All centromeres of As and Bs were equally labeled by CENH3

(Figure 3; see Supplemental Figure 3 online). At first pollenprophase, all chromosomes together with the surrounding tu-bulin spindle apparatus were located toward the pollen cortex(see Supplemental Figure 3A online). Next, an asymmetric ana-phase spindle was formed. The peripheral generative spindlebundle is blunt and the pole is nearly in contact with the pollencortex. By contrast, the interior spindle bundle is long and pointed(Figure 3A; see Supplemental Figure 3B online). These are typicalfeatures of an asymmetrical cell division (Borg et al., 2009). At lateanaphase, in rye +B (Figure 3B) and wheat +B (Figure 3D), weidentified CENH3 signals of the same intensity as segregatedAs, corresponding with the position of lagging Bs (Figures 1Aand 2B). This suggests that centromere inactivity is not the causalreason for the lack of separation of B chromatids at first pollenmitosis. It is more likely that the cohesion between B chromatidsis higher than the microtubule traction force required for theseparation of chromatids.Microtubules reorganize into the bipolar phragmoplast array

between reforming nuclei at subsequent telophase; consequently,a curved cell plate is formed separating the generative and veg-etative cells (see Supplemental Figure 3C online). In both species,centromeres of As in the generative nuclei had a smaller dis-tance from each other than in the vegetative nuclei (Figures 3Cand 3E). A small number of CENH3 signals were observed towardthe interior of the generative nucleus localized within the sameregion where B centromeres were detected by fluorescence insitu hybridization (FISH) (Figures 1C and 2B). Likely, the observedasymmetrical microtubule spindle organization is causing a pref-erential accumulation of joined B chromatids in the generativenucleus. At the second pollen mitosis, the spindle organizationwas symmetric and both sperm nuclei were equally marked byCENH3 (see Supplemental Figure 3D online).

The Composition of Centromere Repeats Differs betweenAs and Bs

The centromeric sequences of rye A and B chromosomes werecompared because of the different segregation behavior of bothtypes of chromosomes. Therefore, different parts of Ty3/Gypsy-

Figure 2. Influence of the Nondisjunction Control Region on the Accu-mulation Process of Rye Bs.

Pollen of wheat ([A] and [B]) carrying a standard B of rye and of wheat(C) carrying a defB of rye, missing the control region of nondisjunction atfirst pollen anaphase (A) and at bicellular stage ([B] and [C]) in situ hy-bridized with the rye centromere-specific probe Bilby (in green), theB-specific probe E3900 (in red), and the centromere repeat CRW2 (inred). DAPI, 49,6-diamidino-2-phenylindole. Bs are arrowed. Bar = 10 µm.(A) and (B) At anaphase, due to nondisjunction, standard Bs lag behindthe As (A) and are placed into the generative nucleus (b).(C) If the nondisjunction control region is missing, normal segregation ofBs occurs irrespective of the differential grouping of A centromeres in thegenerative (G) and vegetative (V) nuclei.(D) Pollen at bicellular stage of 0B rye probed with Bilby.

4126 The Plant Cell

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

type centromeric retrotransposons like the rye-specific elementBilby (Francki, 2001) and the sequences originally identified inwheat centromeric sequences 6C6-3/-4 (Zhang et al., 2004),192-bp repeat (Ito et al., 2004), CCS1 (Aragón-Alcaide et al.,1996), and CRW2 (Liu et al., 2008), were in silico compared with454-sequence reads (Martis et al., 2012) from flow-sorted Asand Bs as well as from genomic DNA of 0B and +4B rye plants.The frequency pattern of all tested sequences was similar in thecompared sets of 454 reads of As and Bs (see SupplementalFigure 4 online). All A (peri)centromere-located sequences werealso identified at the B chromosome of rye. The in silico results

were confirmed by PCR using primers specific for the testedcentromeric sequences (see Supplemental Table 1 online) andgenomic DNA of 0B and +4B rye, 0B and +2B wheat, and DNAof sorted rye Bs (see Supplemental Figure 5 online). Ampliconsof expected size were obtained and no major sequence differ-ences (GenBank accession numbers JQ963501 to JQ963591)were identified between As and Bs.It has been reported for several species that the transcrip-

tional activity of repetitive DNA sequences located in the cen-tromeric heterochromatin is crucial for proper function of thecentromere (Ekwall, 2004; Morey and Avner, 2004; Bouzinba-Segard et al., 2006). To test the activity of A and B centromericrepeats, we analyzed the transcripts of root, leaf, and anthertissues of rye with and without Bs (see Supplemental Figure 6online). Each repeat, even different parts of the same repeat(e.g., ScCl11-1, ScCl11-2, and ScCl11-3 or Sc6C6-3 andSc6C6-4) displayed its unique expression pattern. Except forSc192 bp, ScCl11-1, and ScCl11-2, all centromeric transcriptswere detectable. The transcription level was highest in anthers.Bs did obviously not influence the expression of centromericrepeats.The centromere localization of all A and B centromere- and of

the B centromere-specific sequences was analyzed by in situhybridization and fluorescence wide-field microscopy in somaticmetaphases (Figures 4A and 4B; see Supplemental Figure 7online). Notably, Bs revealed more extended Bilby and ScCCS1signals along the centromeres than the As (Figures 4A and 4B).Using structural illumination microscopy (SIM) to achieve a res-olution below the limit of light microscopy, the differential dis-tribution of the rye-specific centromere sequence Bilby becameeven more obvious. All As displayed distinct ring-like hybrid-ization signals. By contrast, the centromere of the Bs revealedan extended and diffuse distribution of Bilby signals (Figure 4C).To confirm this finding, the volumes of A- and B-located Bilbysignals were comparatively quantified. Compared with As, thevolumes of B-localized Bilby signals occupied a 3.4-fold largervolume in interphase (n = 57 As and 12 Bs, P < 0.001) and a 2.8-fold larger volume in metaphase chromosomes (n = 157 As and12 Bs, P < 0.001).The reduced condensation level of pachytene chromosomes

enabled a quantitative length comparison of the higher orderorganization of centromeric repeats between As and Bs. In Bcentromeres, Bilby repeats are interrupted by insertion of theScCl11 B-specific centromeric repeats (Figure 4D). The middlepart of the centromeric region of the B was observed to benarrower than the rest, whereas this narrower region was notobserved in the As. To avoid possible differences in lengthmeasurements due to colors used for probe detection, bothprobes were pairwise detected with Cy3 and fluorescein. Themean length of the centromeric region was shorter in green Bilby(10.33 µm) than in red ScCl11 (13.83 µm), (t = 24.302, P value =0.0002; n = 14). However, the difference was not significantwhen the probes were detected with the opposite colors (meanred Bilby length = 13.13 µm, mean green ScCl11 length = 14.62µm; t = 21.465, P value = 0.15; n = 15). In 10 pachytene cells(five of each type of probe detection), the seven A centromerescould be individually observed. The mean length of the Bilbysignal on the As was significantly shorter than the mean length

Figure 3. B Centromeres Are Active during Pollen Mitosis.

Distribution of CENH3 (in red) and of a-tubulin (in green) at differentstages of the first pollen mitosis of rye +B ([A] to [C]) and wheat +B ([D]and [E]). DAPI, 49,6-diamidino-2-phenylindole. Apparent B positions arearrowed. Bar = 10 µm.(A) Formation of an asymmetric anaphase spindle. The peripheral gen-erative spindle (G) is blunt and the interior spindle (V) is long and pointed.(B) At late anaphase, CENH3 signals of lagging chromosomes arearrowed.(C) Binucleate pollen with a decondensed vegetative (V) and a con-densed generative nucleus (G). Note the differential grouping of CENH3signals.(D) Anaphase of wheat +B exhibiting CENH3 signals of lagging chro-mosomes (arrowed).(E) In binucleate pollen, the vegetative (V) and the generative nucleusshow ungrouped and clustered CENH3 signals, respectively. In mergeimages, 49,6-diamidino-2-phenylindole is shown in blue.

B Chromosome Drive during Pollen Mitosis 4127

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

of the Bilby signal on the B, independent of whether Bilby wasdetected in green (5.31 µm in the As versus 9.73 µm in the Bs ofthe same cells) or in red (7.79 µm in the As versus 13.46 µm inthe Bs of the same cells). Hence, the Bilby signal of the Bs issignificantly larger than the signal on the As irrespective of thecolor used for detection (t = 4.933; P value = 0.0001; n = 10).

We conclude that all A centromere sequences are also pres-ent in the centromeres of the Bs. However, the contrastingdistribution of Bilby signals suggests that the higher sequenceorganization of A and B centromeres differs. The pericentromere

of Bs is more extended and Bs have accumulated additionalsequences, like ScCl11 and mitochondrion-derived sequences.This suggests that the centromere of the B originated from anA centromere that underwent subsequent sequence accumu-lations and rearrangements. The B-specific extension of thepericentromere seems to be involved in the formation of aB-specific pericentric heterochromatin.

B Centromere–Specific Sequences Do Not Interact with theCentromere-Specific Histone Variant CENH3

After the identification of a specific composition of the B cen-tromere, we asked whether the B-located sequences ScCl11and mitochondrial DNA interact with CENH3-containing nucle-osomes and therefore contribute to an active B centromere.Chromatin immunoprecipitation (ChIP) was performed using

an anti-Os-CENH3 antibody and chromatin from +B rye. Nosignificant interaction was found between CENH3 and ScCl11 ormitochondrial sequences (Figure 5A). By contrast, a significantCENH3 interaction was found for Bilby. Other Ty3/Gypsy-typesequences, including Sc192 bp and ScCCS1, are targeted byCENH3 too, although at lower levels. This indicates that differentparts of the rye centromeric retrotransposon differently bind toCENH3, as previously shown in maize (Zhong et al., 2002)and rice (Oryza sativa; Nagaki et al., 2005). Hence, the rye Bcentromere–specific sequences identified so far are not com-ponents of the active B centromere. Consistently, after indirectimmunostaining of somatic chromosomes, no obvious differ-ences were found between the CENH3 signals of As and Bs(Figure 5B).

Figure 4. The Organization of Rye A and B (Peri)centromeric RegionsDiffers.

(A) to (D) Mitotic metaphase chromosomes of rye + Bs (arrowed) labeledwith the A and B centromere-located sequences Bilby ([A] and [C]) aswell as with the B-specific repeats ScCl11 ([A] and [C]) and E3900 (B).FISH results were analyzed by fluorescence wide-field microscopy (WF)([A], [B], and [D]) and by SIM (C). In contrast with As, the centromere ofthe Bs (arrowed) revealed an extended and diffuse distribution of Bilbysignals. A centromeres displayed distinct ring-like hybridization signals.DAPI, 49,6-diamidino-2-phenylindole.(D) Pachytene chromosomes of rye + 2B (arrowed) with the A and Bcentromere-located sequences Bilby and with the B-specific repeatScCl11. The Bilby signal on the Bs is significantly larger that the signal onthe As.

Figure 5. Not All Sequences of the Rye B (Peri)centromere Interact withthe Centromere-Specific Histone Variant CENH3.

(A) ChIP analysis of (peri)centromeric sequences of rye using anti-Os-CENH3. The B-specific repeat E3900 and rDNA were used as negativecontrol. The columns and error bars represent the average percentage ofimmunoprecipitation (IP%) and SD from five independent experiments.**P < 0.01 and ***P < 0.001 in the Tukey’s honestly significant differencetest.(B) Immunostaining of a haploid mitotic metaphase cell of rye +2Bs(arrowed) with anti-Os-CENH3 (in red) as control. Bar = 10 µm.

4128 The Plant Cell

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

DISCUSSION

We provide direct insight into the cellular and molecular basisof the non-Mendelian accumulation mechanism of a selfishrye chromosome. Comparative analysis of the centromere re-peat composition revealed differences between A and B chro-mosomes of rye. Centromere sequences of As are also sharedby the B centromere. In addition, in the Bs, these sequencesare intermingled with and extended by ScCl11 repeats andmitochondrion-derived DNA. This indicates that the centro-mere of the B originated from an A chromosome and after-wards new sequences accumulated in the B centromere.However, ScCl11 and mitochondrial sequences are not partof the CENH3-containing chromatin but rather represent B-specificelements of the pericentromere. A comparable compositionwas also reported for the maize B centromere. Unlike maize Achromosomes, in which the centromeres are composed of aCentC satellite and centromere-specific retrotransposons, thecentromere of the maize B chromosome exhibits a differentsequence composition and organization. CentC and centromere-specific retrotransposons were disrupted by maize B-specificcentromeric sequences (Jin et al., 2005). The B-specific accu-mulation of repeats in the pericentromere is likely involved in theformation of pericentric heterochromatin, which is known to playa role in chromosome segregation (Yamagishi et al., 2008). Inparticular, it has been suggested that heterochromatin is requiredin sister chromatid cohesion. In fission yeast, repeats flanking thekinetochore are essential for cohesion (Bernard et al., 2001).

Unequal Cell Division: the Key for the Preferential Drive ofBs in the Generative Nucleus

An important clue comes from our finding that the microtubulespindle is asymmetrical during first pollen mitosis in rye andwheat. The asymmetry of this division plays a critical role in thedetermination and subsequent fate of the two unequal daughtercells, the vegetative and the generative (Twell, 2011). As a resultof the asymmetric cell division and of the subsequent formationof contrasting interphase structures, the centromeres of theAs are clustered in the more condensed generative nucleus,whereas in the decondensed vegetative nucleus, the centro-meres are distributed over a much larger area. Notably, the Aand B centromeres do not intermingle in the generative nuclei.Likely, the distinct interphase position of the Bs is a consequenceof B nondisjunction because defBs, lacking the nondisjunctioncontrol region, grouped together with the centromeres of the As.By contrast, the Bs exhibiting nondisjunction lagged behind theseparated A chromatids during anaphase and were preferentiallylocated toward the position of the former mitotic equatorial plate.

Considering the asymmetric geometry of the spindle, it isprobable that, as proposed by (Jones, 1991), inclusion of Bs inthe generative nucleus is caused by the fact that the equatorialplate is closer to the generative pole and lagging Bs are pas-sively included in the generative nuclei as the nuclear membraneis formed. Alternatively, the Bs may preferentially accumulate inthe generative nucleus due to a higher tension force on the Bcentromere toward the generative pole. It has been suggestedthat asymmetric spindle positioning in single-cell-stage of

Caenorhabditis elegans embryos is caused by an initial sto-chastic different net pulling force on the two microtubule asters.The increase of these unequal forces through an unknownmechanism will cause different positioning of the nucleus/centrosomal complex, which has similar initial positions relative toanterior-posterior cortical polarity cues (Grill et al., 2001; Sillerand Doe, 2009). Whether a comparable mechanisms exists inplants remains to be determined. However, it seems that asym-metrical spindles are also a key component of the premeiotic drivein, for example, the Asteraceae Crepis capillaris (Rutishauser andRothlisberger, 1966) and of the meiotic drive of Bs in, for example,the grasshopper Myrmeleotettix maculatus (Hewitt, 1976). Hence,the asymmetry of the mitotic spindle seems to be a major com-ponent of the B accumulation mechanisms.

Why Does Nondisjunction of Bs Mainly Occurin Gametophytes?

The discovery that some of the nondisjunction control region–specific repeats produce noncoding RNA predominantly inanthers (Carchilan et al., 2007) suggests an intriguing possibility,namely, that the nondisjunction of Bs occurs because thecontrol region provides RNA to maintain the cohesion in keyregions of B sister chromatids. One might imagine that thefailure in mitotic segregation reflects a failure to properly resolvethe pericentromeric heterochromatin during first pollen mitosis.The cell cycle type-specific segregation failure of Bs triggers thequestion: In which aspect the first pollen mitosis differs fromother mitotic events in other cell types? We argue that eithera haploid tissue type–specific expression of nondisjunctioncontrolling transcripts (Carchilan et al., 2007) and/or the for-mation of a contrasting chromatin composition during firstpollen mitosis (Houben et al., 2011) ensure cell type–specificaccumulation.Although no similarity between D1100/E3900 and B centro-

mere–located sequences has been found, it is notable that somesimilarity exists at the amino acid level between a part of theE3900 sequence (which encodes a partial reading frame for thegag protein of an long terminal repeat retrotransposon, mostclosely related to a highly conserved Ty3/gypsy family) andconserved centromeric sequences of CentC and osrch3 inmaize and rice, respectively (Langdon et al., 2000). However,sequence similarity between noncoding RNA and the targetregion seems not to be a functional requirement. For instance,dosage compensation in both flies and mammals requires non-coding RNAs, which spread in cis to coat the X chromosome. Theregions of the Xist RNA that are required for the localization on theinactivated X have no obvious sequence similarity (Wutz et al.,2002). Therefore, it seems possible that B-derived noncodingRNAs act as guide molecules to direct protein complexes tospecific genomic loci, such as the B pericentromere. Whetherthe B transcripts act directly or indirectly in the nondisjunctionprocess of Bs remains to be answered. However, recent identifi-cation of a number of rye B-located genic sequences (Martis et al.,2012) provides a basis to speculate about a role of protein-codinggenes in nondisjunction control as alternative possibility. Forinstance, a specific interaction of proteins and pericentromericsequences has been shown in murine fibroblastic cells infected

B Chromosome Drive during Pollen Mitosis 4129

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

by a virus strain (Mansuroglu et al., 2010). Here, the non-structural protein encoded by this virus interacts specifically withclusters of the pericentromeric g-satellite sequence of the hostgenome. This interaction induces sister chromatid cohesion andsegregation defects.As proposed by Masonbrink and Birchler (2010), nondis-

junction of maize Bs in the sporophyte and the gametophyte areprobably related, but the gametophyte may be a more tolerantenvironment for B nondisjunction. In maize, B nondisjunctionalso occurs during tapetal programmed cell death (Chiavarinoet al., 2000; González-Sánchez et al., 2004), and in the endo-sperm (Alfenito and Birchler, 1990) and in plants containingmultiple Bs, the nondisjunction rate in root meristems is in-creased (Masonbrink and Birchler, 2010). By contrast, in a numberof species, nondisjunction of Bs occurs regularly in somatic tissue(listed in Jones and Rees, 1982). However, as the drive mecha-nism of Bs is effective in gametophytes only, abnormal segrega-tion of Bs in somatic tissue will not affect the accumulation of Bs.It is tempting to speculate that the extended cohesion betweensister chromatids of Bs is also part of their unique behavior duringmeiosis if occurring as a univalent. Unlike A chromosomes, uni-valent Bs of many species (e.g., Poa alpina [Müntzing, 1946b],Anthoxanthum aristatum [Östergren, 1947], Godetia nutans[Hakansson, 1945], and rye [Manzanero et al., 2000]), split sisterchromatids much less frequently at anaphase I.

The Drive Mechanism of the B Chromosome of Rye: A Model

On the basis of the above-mentioned observations, we proposea model of how the accumulation mechanism works for the Bchromosome of rye (Figure 6). At all mitotic stages of micro-gametogenesis, the centromeres of As and Bs are active. How-ever, sister chromatid cohesion differs between As and Bs at firstpollen mitosis, and the B-specific pericentromeric repeats areinvolved in the formation of pericentric heterochromatin. The fail-ure to resolve the pericentromeric cohesion is under the control ofthe B-specific nondisjunction control element. Due to unequalspindle formation, joined B chromatids become preferentiallylocated toward the generative pole. As a consequence of amissing or delayed transport of Bs at anaphase, the B cen-tromeres locate opposite to the clustered A centromeres in theresulting generative nucleus. In the second pollen mitosis, thegenerative nucleus divides to produce two sperm nuclei, eachwith an unreduced number of Bs. Hence, a combination of non-disjunction and of unequal spindle formation at first pollen mitosisresults in the directed accumulation of Bs to the generativenucleus and therefore ensures their transmission at a higher thanMendelian rate to the next generation.As in rye, the accumulation mechanism in maize Bs requires

a factor located on the end of the long arm of the B that can actin trans (Roman, 1947; Carlson, 1978; Lamb et al., 2006), andthe B centromeric heterochromatin, irrespective of centromerefunction, is required for efficient nondisjunction (Carlson, 2006;

Figure 6. Model to Explain the Drive Mechanism of the Rye B Chro-mosome at Micro Gametogenesis.

(A) to (E) Pollen grain with 7As +1 standard B chromosome.(F) to (J) In the case that the nondisjunction control region of the B ismissing (defB), no accumulation of the defB occurs.(A) and (E) At first pollen mitosis, metaphase chromosomes are locatedtoward the pollen cortex and an asymmetric spindle is formed. The pe-ripheral generative (G) spindle bundle is blunt and the interior spindlebundle toward the vegetative pole (V) is longer and pointed.(B) In contrast with the separated A sister chromatids, the majority of Bsstay in between both spindle poles. The failure to resolve the pericentro-meric cohesion of the B is under the control of the B-specific non-disjunction control element. Likely, the B-specific centromere organizationis involved in the formation of a B-specific pericentric heterochromatin.Despite centromere activity, the cohesion between B chromatids in keyregions is probably stronger than the microtubule traction force required forthe separation of the chromatids.(C) The placement of Bs toward the generative nucleus is probably re-alized because the equatorial plate is closer to the generative pole andlagging Bs are included in the generative nucleus as the nuclear mem-brane is formed; alternatively, a higher spindle tension force toward thegenerative pole might preferentially pull the Bs to the generative nucleus.Unlike As, nondisjoined Bs accumulate preferentially at a position closeto the phragmoplast.(D) and (E) During the second pollen mitosis, the generative nucleusdivides to produce two sperm nuclei (S), each with an unreduced numberof Bs.(F) to (J) If the nondisjunction control region of the B is missing, normalseparation of defB chromatids occurs and A and defB centromeres in-termingle (H) in both nuclei of binucleate pollen.(J) Consequently, no accumulation of the defBs occurs. In addition, thenondisjunction control region of the B is required for the accumulation

and for the interior position of Bs in generative nuclei. However, thisregion does not control the differential grouping of centromeres of Achromosomes in generative and vegetative nuclei.

4130 The Plant Cell

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

Han et al., 2007). As the Bs of rye and maize originated in-dependently, comparable drive mechanisms in both speciesevolved in parallel. Although much of repetitive DNA evolutionis governed by neutral evolutionary processes (Charlesworthet al., 1994), we propose that some B-located repeats, like thoselocated in the Ab10 maize chromosome involved in neocentromeremeiotic drive (Mroczek et al., 2006), the satellites involved in seg-regation distortion of Drosophila melanogaster (Frank, 2000)or in centromere-associated drive in female meiosis (Fishmanand Saunders, 2008) are functionally involved in the regulationof chromosome segregation to ensure the maintenance of Bsin natural populations.

METHODS

Plant Material and Plant Cultivation

Plants with Bs from the self-fertile inbred line 7415 of rye (Secale cereale)(Jimenez et al., 1994) and from hexaploid wheat (Triticum aestivum) cvLindström with added standard rye Bs (Lindström, 1965) and hexaploidwheat cv Chinese Spring with added truncated rye Bs (line BS-2) (Endoet al., 2008) were grown at the same temperature, humidity, and lightconditions (16 h light, 22°C day/16°C night).

Genomic DNA and RNA Extraction, PCR, and RT-PCR

Genomic DNA was extracted from leaf tissue of rye and wheat with andwithout B chromosomes via the DNeasy plant mini kit (Qiagen) accordingto the manufacturer’s instructions. Rye Bs were flow-sorted and ream-plified as described by Kubaláková et al. (2003). Total RNA was isolatedfrom roots, leaves, and anthers (staged between meiosis and developmentof mature pollen bymicroscopy) using the Trizol method (Chomczynski andSacchi, 1987). The quality of RNA was checked on denaturing agarose gel,and possible DNA contamination was removed by treatment of total RNAwith DNase (Ambion TURBO DNase; Invitrogen). Absence of DNA con-tamination was confirmed by PCR using primers specific for Bilby (Francki,2001) or CRW2 repeats (Liu et al., 2008) (see Supplemental Table 1 online).cDNA was synthesized with 2.5 µg of DNase-treated total RNA using theRevertAid HMinus first-strand cDNA synthesis kit (Fermentas) according tothe manufacturer’s instructions. PCR and RT-PCR reaction was performedin a 25-mL reaction containing 100 ng of genomic DNA or cDNA, re-spectively,1 mM each primer (see Supplemental Table 1 online), buffer,deoxynucleotide triphosphates, and 1 unit of Taq polymerase (Qiagen).Thirty-five amplification cycles (45 s at 95°C, 1 min at correspondingannealing temperature [see Supplemental Table 1 online], and1min at 72°C)were run. Primers specific for the constitutively expressed GAPDH genewere used for RT-PCR control.

Generation and Analysis of Rye Centromeric Sequences

The cereal centromeric sequences Bilby (Francki, 2001), 6C6-3/-4 (Zhanget al., 2004), 192-bp repeat (Ito et al., 2004), CCS1 (Aragón-Alcaide et al.,1996), and CRW2 (Liu et al., 2008) were compared with rye B-derivedsequence 454 reads (European Bioinformatics Institute–European Nu-cleotide Archive accession number ERP001061) by BLAST to identifysimilar B-located sequences for primer design (see Supplemental Table 1online). In addition, their frequencies on A and B chromosomes weredeterminedbyPROFREPalgorithm (Macas et al., 2007). PCRproducts from+B rye DNA were purified and used as FISH and/or dot blot hybridizationprobes. PCR products obtained from genomic rye 0B DNA and sorted rye Bchromosomes were cloned and sequenced. Sequences were deposited inthe GenBank database under accession numbers JQ963501 to JQ963591.

FISH and Indirect Immunostaining

Preparations of semithin pollen sections and of root tip meristems as wellas subsequent FISH or indirect immunostaining were performed ac-cording to Houben et al. (2011). FISH on pachytene chromosomes wasperformed as described (González-García et al., 2006). FISH probes werelabeled with ChromaTide Texas Red-12-dUTP, Alexa Fluor 488-5-dUTP(Invitrogen), or Cy5-dUTP (GE Healthcare Life Sciences) by nick trans-lation. The primary antibodies, mouse anti-a tubulin (Sigma-Aldrich;catalog number T 9026; diluted 1:100), and rabbit anti-Os-CENH3 (Nagakiet al., 2004) were used for indirect immunostaining. The anti-Os-CENH3antibody is able to recognize CENH3 of all tested grass species, includingrye and common wheat. Imaging was performed using an Olympus BX61microscope and an ORCA-ER charge-coupled device camera (Hama-matsu). Deconvolutionmicroscopy was employed to remove out-of-focusinformation. All images were collected in gray scale and pseudocoloredwith Adobe Photoshop 6. Maximum intensity projections were processedwith the program AnalySIS (Soft Imaging System).

Quantification of Fluorescence Signalsand Super-Resolution Microscopy

Quantification of FISH signals was performed with an epifluorescencemicroscope (Zeiss Axiophot) using a 3100/1.45 Zeiss a plan-fluar ob-jective and a three-chip Sony color camera (DXC-950P). The microscopewas integrated into a Digital Optical 3D Microscope system (Confovis) tomeasure the signals along x, y, and z axes as a basis to calculate the signalvolumes. ImageJ software (Collins, 2007) was used to determine thelength of centromeric regions of pachytene chromosomes. To measurethe areas occupied by CENH3 signals within the vegetative and gener-ative nuclei, maximum intensity projections from image stacks acquiredby the Olympus BX61 microscope and processed with SIS software(Olympus) were used. To achieve an optical resolution of ;100 nm, weapplied SIM using a 363/1.40 objective of an Elyra super-resolutionmicroscope system (Zeiss).

ChIP Assay

ChIP was performed as described by Nagaki et al. (2003) using young ryeleaves with Bs and the anti-Os-CENH3 antibody (Nagaki et al., 2004).ChIP without antibody served as a mock treatment. Briefly, after in-cubation with the antibody, the chromatin solution was separated intosupernatant (Sup) and pellet (Pel) fractions. DNA was blotted on HybondN+ membranes (GE) and used for dot-blot hybridization with rye (peri)centromere-localized sequences (Bilby, Sc6C6-4, Sc192 bp, ScCCS1,ScCRW2, ScCl11-2, and mitochondrial DNA [barley (Hordeum vulgare)BAC MmHB 0205G01]) and noncentromeric repeats (E3900 and 45SrDNA) (pTa71; Gerlach and Bedbrook, 1979). The probes were radio-labeled by [a-32P]dCTP using the DecaLabel DNA labeling kit (Fermentas)and hybridized to the ChIP DNA on the membranes. The signals werecaptured by a phosphor imager (Fujifilm FLA-5100) and quantified by TINA2.09 software. In each case, the percentage of immunoprecipitation[defined as Pel/(Pel + Sup)] of the mock experiments was subtracted fromthe percentage of immunoprecipitation of the antibody to CENH3 treat-ments (IP% = the percentage of immunoprecipitation [CENH3] 2 thepercentage of immunoprecipitation [Mock]). Each experiment was repeatedin five independent reactions. The data were analyzed statistically bypairwise comparison of the IP% of rye centromere-localized sequencesand the IP% of rDNA using the Tukey’s honestly significant difference test.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL datalibraries under accession numbers JQ963501 to JQ963591.

B Chromosome Drive during Pollen Mitosis 4131

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. Bs Significantly Delay the Progress of PollenMitosis.

Supplemental Figure 2. Pollen of Rye +4B Plants at Bicellular Stage.

Supplemental Figure 3. Distribution of CENH3 and of a-Tubulin atPollen Mitosis of Rye +B.

Supplemental Figure 4. Sequence Comparison of A and B Centro-mere-Located Repeats.

Supplemental Figure 5. PCR Amplification of the CentromericRepeats of Plants with and without Bs.

Supplemental Figure 6. Transcription Analysis of the (Peri)centro-meric Repeats of Rye with and without Bs.

Supplemental Figure 7. Distribution of Centromeric Sequences alongA and B Chromosomes of Rye.

Supplemental Table 1. Primer Sequences and PCR Conditions.

ACKNOWLEDGMENTS

We thank Takashi Endo for providing the wheat-rye B addition line BS-2

and Pavel Neumann, Ingo Schubert, and R. Neil Jones for fruitfuldiscussions. We also thank Karla Meier for excellent technical assis-tance. This work was supported by the Deutsche Forschungsgemein-schaft (HO 1779/10-1/14-1) and by Spanish Grants BFU 2006-10921,CCG10-UCM/GEN-5559, and AGL 2011-28542. M.G.-G. was supportedby a grant from the Ministry of Education of Spain.

AUTHOR CONTRIBUTIONS

A.H. designed the research. A.M.B.-M, V.S., K.K., O.W., S.K., K.N., J.M.,M.G.-S., V.H., D.G.-R., M.G.-G., J.M.V., and M.J.P. performed research.A.M.B.-M., M.J.P., and A.H. wrote the article.

Received September 17, 2012; revised September 17, 2012; acceptedSeptember 30, 2012; published October 26, 2012.

REFERENCES

Alfenito, M., and Birchler, J. (1990). Studies on B chromosomestability during development. Maydica 35: 359–366.

Aragón-Alcaide, L., Miller, T., Schwarzacher, T., Reader, S., andMoore, G. (1996). A cereal centromeric sequence. Chromosoma105: 261–268.

Bernard, P., Maure, J.F., Partridge, J.F., Genier, S., Javerzat, J.P.,and Allshire, R.C. (2001). Requirement of heterochromatin for co-hesion at centromeres. Science 294: 2539–2542.

Blunden, R., Wilkes, T.J., Forster, J.W., Jimenez, M.M., Sandery,M.J., Karp, A., and Jones, R.N. (1993). Identification of the E3900family, a second family of rye B chromosome specific repeatedsequences. Genome 36: 706–711.

Borg, M., Brownfield, L., and Twell, D. (2009). Male gametophytedevelopment: A molecular perspective. J. Exp. Bot. 60: 1465–1478.

Bouzinba-Segard, H., Guais, A., and Francastel, C. (2006). Accumulationof small murine minor satellite transcripts leads to impaired centromericarchitecture and function. Proc. Natl. Acad. Sci. USA 103: 8709–8714.

Burt, A., and Trivers, R. (2006). Genes in Conflict: The Biology ofSelfish Genetic Elements. (Cambridge, MA; London: The BelknapPress of Harvard University Press).

Carchilan, M., Delgado, M., Ribeiro, T., Costa-Nunes, P., Caperta,A., Morais-Cecílio, L., Jones, R.N., Viegas, W., and Houben, A.(2007). Transcriptionally active heterochromatin in rye B chromo-somes. Plant Cell 19: 1738–1749.

Carlson, W.R. (1978). The B chromosome of corn. Annu. Rev. Genet.12: 5–23.

Carlson, W.R. (2006). Unstable inheritance of maize B-type chromo-somes that lack centric heterochromatin. Genome 49: 420–431.

Charlesworth, B., Sniegowski, P., and Stephan, W. (1994). Theevolutionary dynamics of repetitive DNA in eukaryotes. Nature 371:215–220.

Chiavarino, A.M., Rosato, M., Manzanero, S., Jiménez, G.,González-Sánchez, M., and Puertas, M.J. (2000). Chromosomenondisjunction and instabilities in tapetal cells are affected by Bchromosomes in maize. Genetics 155: 889–897.

Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloroform ex-traction. Anal. Biochem. 162: 156–159.

Collins, T.J. (2007). ImageJ for microscopy. Biotechniques 43(1,Suppl): 25–30.

Ekwall, K. (2004). The roles of histone modifications and small RNA incentromere function. Chromosome Res. 12: 535–542.

Endo, T.R., Nasuda, S., Jones, N., Dou, Q., Akahori, A., Wakimoto,M., Tanaka, H., Niwa, K., and Tsujimoto, H. (2008). Dissection ofrye B chromosomes, and nondisjunction properties of the dissectedsegments in a common wheat background. Genes Genet. Syst. 83:23–30.

Fishman, L., and Saunders, A. (2008). Centromere-associated fe-male meiotic drive entails male fitness costs in monkeyflowers.Science 322: 1559–1562.

Francki, M.G. (2001). Identification of Bilby, a diverged centromericTy1-copia retrotransposon family from cereal rye (Secale cerealeL.). Genome 44: 266–274.

Frank, S.A. (2000). Polymorphism of attack and defense. Trends Ecol.Evol. (Amst.) 15: 167–171.

Gerlach, W.L., and Bedbrook, J.R. (1979). Cloning and character-ization of ribosomal RNA genes from wheat and barley. NucleicAcids Res. 7: 1869–1885.

González-García, M., González-Sánchez, M., and Puertas, M.J.(2006). The high variability of subtelomeric heterochromatin andconnections between nonhomologous chromosomes, suggest fre-quent ectopic recombination in rye meiocytes. Cytogenet. GenomeRes. 115: 179–185.

González-Sánchez, M., Rosato, M., Chiavarino, M., and Puertas,M.J. (2004). Chromosome instabilities and programmed cell deathin tapetal cells of maize with B chromosomes and effects on pollenviability. Genetics 166: 999–1009.

Grill, S.W., Gönczy, P., Stelzer, E.H., and Hyman, A.A. (2001). Po-larity controls forces governing asymmetric spindle positioning inthe Caenorhabditis elegans embryo. Nature 409: 630–633.

Håkanson, A. (1948). Behaviour of accessory rye chromosomes in theembryo sac. Hereditas 34: 35–59.

Håkanson, A. (1959). Behaviour of different small accessry rye chromo-somes at pollen mitosis. Hereditas 45: 623–631.

Hakansson, A. (1945). Überzählige Chromosomen in einer Rasse vonGodetia nutans Hiorth. Bot. Not. 2: 1–19.

Han, F.P., Lamb, J.C., Yu, W.C., Gao, Z., and Birchler, J.A. (2007).Centromere function and nondisjunction are independent compo-nents of the maize B chromosome accumulation mechanism. PlantCell 19: 524–533.

4132 The Plant Cell

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

Hasegawa, N. (1934). A cytological study on 8-chromosome rye.Cytologia (Tokyo) 6: 68–77.

Hewitt, G.M. (1976). Meiotic drive for B-chromosomes in the primaryoocytes of Myrmeleotettix maculatus (Orthopera: Acrididae). Chro-mosoma 56: 381–391.

Houben, A., Kumke, K., Nagaki, K., and Hause, G. (2011). CENH3distribution and differential chromatin modifications during pollendevelopment in rye (Secale cereale L.). Chromosome Res. 19: 471–480.

Howman, E.V., Fowler, K.J., Newson, A.J., Redward, S.,MacDonald, A.C., Kalitsis, P., and Choo, K.H. (2000). Early dis-ruption of centromeric chromatin organization in centromere proteinA (Cenpa) null mice. Proc. Natl. Acad. Sci. USA 97: 1148–1153.

Ito, H., Nasuda, S., and Endo, T.R. (2004). A direct repeat sequenceassociated with the centromeric retrotransposons in wheat. Ge-nome 47: 747–756.

Jimenez, M.M., Romera, F., Puertas, M.J., and Jones, R.N. (1994).B-Chromosomes in inbred lines of rye (Secale cereale L). 1. Vigorand fertility. Genetica 92: 149–154.

Jin, W., Lamb, J.C., Vega, J.M., Dawe, R.K., Birchler, J.A., andJiang, J. (2005). Molecular and functional dissection of the maize Bchromosome centromere. Plant Cell 17: 1412–1423.

Jones, N., and Houben, A. (2003). B chromosomes in plants: Esca-pees from the A chromosome genome? Trends Plant Sci. 8: 417–423.

Jones, R.N. (1991). B-chromosome drive. Am. Nat. 137: 430–442.Jones, R.N., and Rees, H. (1982). B Chromosomes, 1st ed. (London;

New York: Academic Press).Kimura, M., and Kayano, H. (1961). The maintenance of super-

numerary chromosomes in wild populations of Lilium callosum bypreferential segregation. Genetics 46: 1699–1712.

Kishikawa, H. (1965). Cytogenetic studies of B-chromosomes in rye,Secale cereale L., in Japan. Agricultural Bulletin of Saga University 21: 181.

Kishikawa, H., and Suzuki, A. (1982). Cytological study on hypo-pentaploid Triticale with four B chromosomes of rye. Jpn. J. Genet.57: 17–24.

Kubaláková, M., Valárik, M., Barto, J., Vrána, J., Cíhalíková, J.,Molnár-Láng, M., and Dole�zel, J. (2003). Analysis and sorting ofrye (Secale cereale L.) chromosomes using flow cytometry. Genome46: 893–905.

Lamb, J.C., Han, F., Auger, D.L., and Birchler, J. (2006). A trans-acting factor required for non-disjunction of the B chromosome islocated distal to the TB-4Lb breakpoint on the B chromosome.Maize Genet. Coop. News Lett. 80: 51–54.

Langdon, T., Seago, C., Jones, R.N., Ougham, H., Thomas, H.,Forster, J.W., and Jenkins, G. (2000). De novo evolution of satelliteDNA on the rye B chromosome. Genetics 154: 869–884.

Lima-De-Faria, A. (1962). Genetic interaction in rye expressed atchromosome phenotype. Genetics 47: 1455–1462.

Lindström, J. (1965). Transfer to wheat of accessory chromosomesfrom rye. Hereditas 54: 149–155.

Liu, Z., Yue, W., Li, D., Wang, R.R., Kong, X., Lu, K., Wang, G.,Dong, Y., Jin, W., and Zhang, X. (2008). Structure and dynamics ofretrotransposons at wheat centromeres and pericentromeres.Chromosoma 117: 445–456.

Macas, J., Neumann, P., and Navratilova, A. (2007). Repetitive DNAin the pea (Pisum sativum L.) genome: comprehensive character-ization using 454 sequencing and comparison to soybean andMedicago truncatula. BMC Genomics 8: 427.

Mansuroglu, Z., Josse, T., Gilleron, J., Billecocq, A., Leger, P.,Bouloy, M., and Bonnefoy, E. (2010). Nonstructural NSs protein ofrift valley fever virus interacts with pericentromeric DNA sequencesof the host cell, inducing chromosome cohesion and segregationdefects. J. Virol. 84: 928–939.

Manzanero, S., Arana, P., Puertas, M.J., and Houben, A. (2000).The chromosomal distribution of phosphorylated histone H3 dif-fers between plants and animals at meiosis. Chromosoma 109:308–317.

Martis, M.M., et al. (2012). Selfish supernumerary chromosomereveals its origin as a mosaic of host genome and organellar se-quences. Proc. Natl. Acad. Sci. USA 109: 13343–13346.

Masonbrink, R.E., and Birchler, J.A. (2010). Sporophytic non-disjunction of the maize B chromosome at high copy numbers. J.Genet. Genomics 37: 79–84.

Matthews, R.B., and Jones, R.N. (1983). Dynamics of the B chro-mosome polymorphism in rye II. Estimates of parameters. Heredity50: 119–137.

Mendelson, D., and Zohary, D. (1972). Behavior and transmissionof supernumerary chromosomes in Aegilops speltoides. Heredity29: 329.

Morey, C., and Avner, P. (2004). Employment opportunities for non-coding RNAs. FEBS Lett. 567: 27–34.

Mroczek, R.J., Melo, J.R., Luce, A.C., Hiatt, E.N., and Dawe, R.K.(2006). The maize Ab10 meiotic drive system maps to supernumerarysequences in a large complex haplotype. Genetics 174: 145–154.

Müntzing, A. (1945). Cytological studies of extra fragment chromo-somes in rye; transmission and multiplication of standard fragmentsand iso-fragments. Hereditas 31: 457–477.

Müntzing, A. (1946a). Cytological studies of extra fragment chromo-somes in rye; the mechanism of non-disjunction at the pollen mi-tosis. Hereditas 32: 97–119.

Müntzing, A. (1946b). Different chromosome numbers in root tips and pollenmother cells in a sexual strain of Poa alpina. Hereditas 32: 127–129.

Müntzing, A. (1948). Cytological studies of extra fragment chromo-somes in rye. V. A new fragment type arisen by deletion. Hereditas34: 435–442.

Müntzing, A. (1970). Chromosomal variation in the Lindström strainof wheat carrying accessory chromosomes in rye. Hereditas 66:279–286.

Müntzing, A., and Prakken, R. (1941). Chromosomal aberrations inrye populations. Hereditas 27: 273–308.

Nagaki, K., Cheng, Z., Ouyang, S., Talbert, P.B., Kim, M., Jones, K.M., Henikoff, S., Buell, C.R., and Jiang, J. (2004). Sequencing ofa rice centromere uncovers active genes. Nat. Genet. 36: 138–145.

Nagaki, K., Neumann, P., Zhang, D., Ouyang, S., Buell, C.R.,Cheng, Z., and Jiang, J. (2005). Structure, divergence, and distri-bution of the CRR centromeric retrotransposon family in rice. Mol.Biol. Evol. 22: 845–855.

Nagaki, K., Talbert, P.B., Zhong, C.X., Dawe, R.K., Henikoff, S.,and Jiang, J. (2003). Chromatin immunoprecipitation reveals thatthe 180-bp satellite repeat is the key functional DNA element ofArabidopsis thaliana centromeres. Genetics 163: 1221–1225.

Niwa, K., Horiuchi, G., and Hirai, Y. (1997). Production and charac-terization of common wheat with B chromosomes of rye from Korea.Hereditas 126: 139–146.

Östergren, G. (1947). Heterochromatic B-chromosomes in Anthox-anthum. Hereditas 33: 261–296.

Puertas, M.J., Romera, F., and Delapena, A. (1985). Comparisonof B-chromosome effects on Secale cereale and Secale vavilovii.Heredity 55: 229–234.

Roman, H. (1947). Mitotic nondisjunction in the case of interchangesinvolving the B-type chromosome in maize. Genetics 32: 391–409.

Romera, F., Jimenez, M.M., and Puertas, M.J. (1991). Factorscontrolling the dynamics of the B-chromosome polymorphism inKorean rye. Heredity 67: 189–195.

Rutishauser, A., and Rothlisberger, E. (1966). Boosting mechanism of B-chromosomes in Crepis capillaris. Chromosomes Today 1: 28–30.

B Chromosome Drive during Pollen Mitosis 4133

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021

Sandery, M.J., Forster, J.W., Blunden, R., and Jones, R.N. (1990). Iden-tification of a family of repeated sequences on the rye B-chromosome.Genome 33: 908–913.

Sanei, M., Pickering, R., Kumke, K., Nasuda, S., and Houben, A.(2011). Loss of centromeric histone H3 (CENH3) from centromeresprecedes uniparental chromosome elimination in interspecific bar-ley hybrids. Proc. Natl. Acad. Sci. USA 108: E498–E505.

Siller, K.H., and Doe, C.Q. (2009). Spindle orientation during asym-metric cell division. Nat. Cell Biol. 11: 365–374.

Tanaka, I. (1997). Differentiation of generative and vegetative cells inangiosperm pollen. Sex. Plant Reprod. 10: 1–7.

Twell, D. (2011). Male gametogenesis and germline specification inflowering plants. Sex. Plant Reprod. 24: 49–60.

Wutz, A., Rasmussen, T.P., and Jaenisch, R. (2002). Chromosomalsilencing and localization are mediated by different domains of XistRNA. Nat. Genet. 30: 167–174.

Yamagishi, Y., Sakuno, T., Shimura, M., and Watanabe, Y. (2008).Heterochromatin links to centromeric protection by recruiting shu-goshin. Nature 455: 251–255.

Zhang, P., Li, W.L., Fellers, J., Friebe, B., and Gill, B.S. (2004). BAC-FISH in wheat identifies chromosome landmarks consisting of dif-ferent types of transposable elements. Chromosoma 112: 288–299.

Zhong, C.X., Marshall, J.B., Topp, C., Mroczek, R., Kato, A.,Nagaki, K., Birchler, J.A., Jiang, J., and Dawe, R.K. (2002).Centromeric retroelements and satellites interact with maize kineto-chore protein CENH3. Plant Cell 14: 2825–2836.

4134 The Plant Cell

Dow

nloaded from https://academ

ic.oup.com/plcell/article/24/10/4124/6101558 by guest on 14 O

ctober 2021