# Source paper archive - Source title: Accessory B Chromosomes Alter Chromatin Accessibility and Gene Expression in Maize Pollen Nuclei - Author(s): Urbis, P., Hlouskova, L., Holusova, K., Dolezel, J., Bartos, J. - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.11.09.765325v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.11.09.765325v1.full.pdf - PDF SHA-256: c196ef4caafba3ce0f659c87e2160264b8e52a9d40fab0064e2603889a158294 - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) # 1 Accessory B Chromosomes Alter Chromatin Accessibility and Gene Expression in 2 Maize Pollen Nuclei 3 Petr Urbiš [1,2], Lucie Hloušková [1], Kateřina Holušová [1], Jaroslav Doležel [1], Jan Bartoš [1*] 4 5 1 Institute of Experimental Botany of the Czech Academy of Sciences, Centre of Plant Structural 6 and Functional Genomics, Šlechtitelů 31, Olomouc, 779 00, Czech Republic 7 2 Department of Cell Biology and Genetics, Palacký University, Olomouc, Czech Republic 8 9 - Corresponding author: Jan Bartoš; bartos@ueb.cas.cz 10 Petr Urbiš: urbis@ueb.cas.cz; ORCID: 0000-0002-2942-0581 11 Lucie Hloušková: hlouskoval@ueb.cas.cz; ORCID: 0000-0003-4800-8186 12 Kateřina Holušová: holusovak@ueb.cas.cz; ORCID: 0000-0002-4531-7170 13 Jaroslav Doležel: dolezel@ueb.cas.cz; ORCID: 0000-0002-6263-0492 14 Jan Bartoš: bartos@ueb.cas.cz; ORCID: 0000-0002-4154-8895 15 Date of submission: 9 November 2026 16 Number of tables: 1 17 Number of figures: 5 18 Word count: 5150 19 Supplementary data: Tables S1-S4, Figures S1-S5, Dataset S1-S10 20 21 **Abstract** 22 B chromosomes in maize ( _Zea mays_ L.) are supernumerary, nonessential chromosomes that 23 exhibit non-Mendelian inheritance through chromosome drive. During the second pollen 24 mitosis, they undergo nondisjunction, resulting in their preferential transmission via B 25 containing sperm cells. Although their transmission mechanisms have been extensively 26 characterized, the effects on chromatin organization and transcriptional regulation in 1 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 27 gametophytic cells remain poorly understood. Here, we combined flow cytometric sorting of 28 pollen nuclei with an assay for transposase-accessible chromatin sequencing (ATAC-seq) and 29 RNA sequencing (RNA-seq) to investigate the impact of B chromosomes on chromatin 30 accessibility and gene expression in mature maize pollen. Using pollen nuclei isolated from a 31 maize inbred line B73 carrying zero and two B chromosomes, we generated high-purity 32 datasets suitable for integrated chromatin and transcriptomic analyses. We identified 33 extensive chromatin accessibility and transcriptional activity at B-chromosome loci. 34 Importantly, the presence of B chromosomes was also associated with altered chromatin 35 accessibility and gene expression on the standard A chromosomes, revealing regulatory effects 36 extending beyond the B chromosome itself. Together, these findings contribute to a broader 37 understanding of effects of selfish genetic elements in plants. 38 39 **Teaser** 40 B chromosomes contain functional regulatory elements that drive active gene expression in 41 mature maize pollen. 42 43 **Introduction** 44 Pollen grains, the microscopic carriers of genetic material in plants, play a crucial role in the 45 reproductive process of angiosperms and gymnosperms and are vital for the continuation of 46 plant species. Consequently, isolating and studying the male gametophyte—a key factor in the 47 evolutionary success of flowering plants and a uniquely accessible cell type—is thus of great 48 interest (Honys and Twell, 2003; Chumak, 2015). To facilitate their study, flow cytometry (FCM) 49 has been found useful, despite the challenges posed by the chemically inert and physically 50 resilient pollen cell wall, which severely hinders the mechanical disruption required to release 51 intact nuclei (Chumak, 2015; Kron _et al.,_ 2021). Competition between male gametophytes for 52 the fertilization of a limited number of egg cells is common in sexual reproduction, and 53 pronounced differences in pollen tube growth rates reflect genetic differences among 54 individual microgametophytes (Mulcahy, 1979; Mulcahy _et al.,_ 1996; Honys and Twell, 2003; 55 Chumak, 2015). At the molecular level, these functional differences are increasingly linked to 56 epigenetic landscapes. 2 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 57 In addition to the standard karyotype, widespread populations of plant, animal and fungal 58 species contain nonessential, dispensable, supernumerary (i.e., “extra”) B chromosomes 59 (Longley 1927). These chromosomes are often preferentially inherited and deviate from the 60 Mendelian law of equal segregation via a process known as chromosome drive (Carlson _et al.,_ 61 2006; Wu _et al.,_ 2019). In maize, the persistence of the B chromosomes is governed by a two 62 component drive mechanism: nondisjunction at the second pollen mitosis (Roman, 1947; 63 Brennan _et al.,_ 2023) and preferential fertilization of the egg cell by the B-containing sperm 64 during double fertilization (Roman, 1948; B rennan _et al.,_ 2023). During the second pollen 65 mitosis, the sister chromatids of the B chromosome fail to separate at the centromeric region. 66 This nondisjunction occurs exclusively in the generative cell, yielding two sperm nuclei—one 67 carrying two B chromosomes and the other carrying none—while the vegetative nucleus 68 remains unaffected. 69 These B chromosomes have distinctive repeat-rich regions around the centromere (ZmBs 70 repeats, knob-like heterochromatin, etc.), so the nondisjunction is controlled by cis and trans 71 factors on the B chromosome (Birchler and Yang, 2021; Blavet _et al.,_ 2021). In pollen, B-specific 72 DNA is often localized to a particular tip or subregion of the sperm nucleus, giving that sperm 73 a unique spatial chromatin architecture that could influence how it interacts with the egg or 74 polar nuclei during double fertilization (Rusche _et al.,_ 1997; Blavet _et al.,_ 2021). In preferential 75 fertilization, the sperm harbouring two B chromosomes enters and fertilizes the egg cell at a 76 higher frequency compared to the sperm without Bs (Roman, 1948; Carlson, 1969; Carlson 77 and Phillips, 1986; B rennan _et al.,_ 2023), thus enhancing the transmission rate of B 78 chromosomes into the embryo and subsequent progeny (Roman, 1948; B rennan _et al.,_ 2023). 79 Mechanistically, preferential fertilization by B ‑ carrying sperm in maize seems to involve 80 specific sequences or structural features near the centromere ‑ proximal region of the B 81 chromosome, alongside egg ‑ cell-driven recognition of one sperm nucleus over the other, 82 although the exact molecular players remain largely unknown (González-Sánchez _et al.,_ 2003; 83 Brennan _et al.,_ 2023). Chromosome B-A translocation studies indicate that the determinant of 84 preferential fertilization maps to a region at or very near the B centromere, implying the 85 involvement of centromere-associated DNA, chromatin, or proteins (Beckett, 1991; Blavet _et_ 86 _al.,_ 2021). Furthermore, the B-centric heterochromatin contains a site required for 87 nondisjunction (Carlson and Chou, 1981; Carlson _et al.,_ 2006). This suggests that local repeats 3 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 88 or chromatin features may jointly promote both nondisjunction and biased fertilization, 89 possibly via altered centromere/kinetochore behaviour or nuclear positioning (Rusche _et al.,_ 90 1997; Blavet _et al.,_ 2021). New developments in analyzing sequence composition provide a 91 valuable opportunity to examine the underlying structure of B chromosomes. Under ordinary 92 conditions, B chromosomes do not bestow any advantages on the organisms that harbour 93 them but rather represent a specific type of selfish DNA (Jones and Rees, 1969; Carlson _et al.,_ 94 2006). The highest number of B chromosomes that a host species can bear is determined by 95 the chromosomes' origin and a balance between the B chromosome drive and the negative 96 impacts on fitness, particularly reduced fertility when Bs are present in higher numbers 97 (Bougourd and Jones, 1997; Wu _et al.,_ 2019). 98 In this study, we investigated the effect of B chromosomes on pollen nuclear chromatin 99 accessibility and the transcriptome. By comparing ATAC-seq and RNA-seq datasets from flow 100 sorted nuclei of mature maize B73 pollen grains with and without two B chromosomes 101 (+B/0B), we provide the first insights into the molecular behaviour of male gametophytes 102 carrying these accessory chromosomes. 103 104 **Results** 105 _RNA-seq_ 106 We performed RNA-seq on mature pollen of maize B73 (+B/0B) accessions in three biological 107 replicates. We generated on average ~35 million high-quality paired-end 150-bp reads 108 (Supplementary Table S1) and mapped them to the _Zea may_ s reference genome RefGen_v5 109 supplemented with the reference of the maize B chromosome (Hufford _et al.,_ 2021; Hloušková 110 _et al.,_ 2025). Out of 40,445 annotated genes, 8,679 were expressed (TPM >1) in B-negative 111 mature pollen (0B samples) and 9,462 in +B samples (Fig. 1A). Maize mature pollen grains 112 exhibit an overall twofold lower gene-expression level and fewer expressed genes with TPM 113 >1 than anthers and many other tissues (Shi _et al.,_ 2021; Han _et al.,_ 2022; Nelms and Walbot 114 2022; Zhang _et al.,_ 2022; Hloušková _et al.,_ 2025). However, they also contain a set of distinct 115 pollen-specific genes, characterized by a profound enrichment in mature pollen compared to 116 other types of tissue (Supplementary Dataset S1). We performed differential expression 117 analysis to investigate +B pollen, with particular attention paid to the expression of B 4 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 118 chromosome-encoded genes. Out of 689 predicted B chromosome-located genes, we 119 identified 103 transcribed genes defined as having transcripts per million (TPM) value greater 120 than 1 in mature pollen (Supplementary Dataset S2), 98 of these were shared with B 121 chromosome-encoded genes expressed in mature anthers with TPM >1 (Hloušková _et al.,_ 122 2025). 123 _Differentially expressed genes_ 124 Analysis of differentially expressed genes (using DESeq2) identified 88 differentially expressed 125 genes on the B chromosome and very low number (42) of upregulated and downregulated (3) 126 genes encoded by A chromosomes in the presence of two B chromosomes (Figs. 1B; 1C). 127 Expectedly, no downregulated B chromosome genes were detected by DESeq2 128 (Supplementary Dataset S3), as such category lacks biological meaning due to their complete 129 absence in 0B samples. In comparison, in maize anther with mature pollen, the DESeq2 130 analysis previously detected 244 upregulated genes expressed on B chromosomes alongside 131 808 upregulated and 348 downregulated genes encoded by A chromosomes under the same 132 condition (Hloušková _et al.,_ 2025). Our findings extend and refine the results of Hloušková _et_ 133 _al._ (2025; Supplementary Fig. S1), who reported that the expression of B-encoded genes is 134 most prevalent in reproductive organs. 135 Subsequently, we performed a gene ontology (GO) enrichment analysis to investigate the 136 functional roles of the 88 B chromosome- and 42 A chromosome-encoded upregulated genes. 137 These genes were compared against all annotated A- and B-localized genes as the respective 138 reference sets Where several significant terms were related, we reported the more specific 139 term (i.e., the one lower in the GO hierarchy) over its more general parent term to maximize 140 informativeness. The complete lists of tested GO terms, including non-significant ones, are 141 provided in Supplementary Dataset S4. 142 Among biological processes, A-chromosome genes were significantly enriched for “gene 143 expression” (GO:0010467), followed by “cellular response to stress” (GO:0033554) and 144 “macromolecule biosynthetic process” (GO:0009059). In the molecular function category, the 145 most enriched GO term was “ATP binding” (GO:0005524), followed by “ATP-dependent 146 activity” (GO:0140657). In contrast, no significantly enriched GO terms were identified within 147 the cellular component category (Fig. 2B). 5 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 148 Among biological process terms, genes on chromosome B were significantly enriched for 149 “chromatin organization” (GO:0006325), followed by “response to heat” (GO:0009408), 150 “trehalose metabolic process” (GO:0005991), “proteolysis” (GO:0006508), “disaccharide 151 biosynthetic process” (GO:0046351), “spindle assembly” (GO:0051225), “DNA duplex 152 unwinding” (GO:0032508), and “RNA-templated DNA biosynthetic process” (GO:0006278). In 153 the molecular function category, the most enriched GO term was “ATP hydrolysis activity” 154 (GO:0016887), followed by “microtubule motor activity” (GO:0003777), “DNA polymerase 155 activity” (GO:0034061), “fucosyltransferase activity” (GO: 0008417), “protein 156 methyltransferase activity” (GO:0008276), “O-acyltransferase activity” (GO: 0008374), 157 “histone binding” (GO:0042393), clathrin binding (GO: 0030276) and “microtubule binding” 158 (GO:0008017). In the cellular component category, “membrane” (GO:0016020) was the most 159 abundant GO term, followed by “microtubule associated complex” (GO:0005875), “lytic 160 vacuole” (GO: 0000323), clathrin-coated vesicle (GO: 0030136) and “P-body” (GO:0000932) 161 (Fig. 2A). 162 _ATAC-seq_ 163 We performed ATAC-seq on flow-sorted maize pollen nuclei (Supplementary Fig. S2) of the 164 inbred line B73 plants (+B/0B), using three biological replicates per condition. Across all six 165 libraries, read trimming retained nearly all reads (99.9 % on average; Supplementary Dataset 166 S4), and the removal of mitochondrial and plastid DNA also preserved the majority of reads 167 (96.15 % on average; Supplementary Dataset S5). Following downstream processing and 168 filtering against a blacklist of highly repetitive regions and unplaced contigs, the final number 169 of called peaks ranged from 71,184 to 80,052 per replicate; most libraries exhibited broadly 170 comparable peak numbers, with a slightly higher count observed in +B replicates 171 (Supplementary Dataset S5). 172 Overall, these data provided sufficient coverage to compare chromatin accessibility landscapes 173 between pollen with and without B chromosomes (Cherchame _et al._, 2025 ). The consensus 174 peak set consisted of 141,793 accessible chromatin regions (Supplementary Dataset S6). The 175 genomic distribution of peaks across chromosomes was broadly similar among replicates for 176 the standard A chromosomes (Supplementary Dataset S7), indicating that the core pollen 177 accessibility landscape was largely shared between 0B and +B samples. As expected, a clear 178 increase of peak number was observed for the B chromosome in +B replicates (Supplementary 6 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 179 Dataset S7). In 0B samples, on average 334 peaks were assigned to the B chromosome due to 180 mapping bias, whereas in +B samples chromosome B accounted for an average of 2,177 peaks 181 (Supplementary Dataset S7). This pattern is consistent with the presence of B-chromosomal 182 DNA only in the +B genotype and confirms that the B chromosomes maintain their own 183 accessible chromatin landscape in pollen nuclei. 184 Comparison of peak sets between conditions revealed that the majority of accessible regions 185 (72,375 peaks out of 141,793) were shared between 0B and +B pollen, which is consistent with 186 the common cellular identity of the samples. When restricting the analysis to peaks uniquely 187 associated with each condition, 31,803 peaks were specific to 0B and 37,615 peaks were 188 specific to +B (Supplementary Figs. S3A; S3B). This difference indicates that the presence of B 189 chromosomes is associated with an increased number of condition-specific accessible 190 chromatin regions. Genomic annotation revealed that in both conditions (+B and 0B) the 191 majority of peaks were located in distal intergenic regions (25,867-31,703), followed by 192 promoter-proximal (24,509-27,770), intronic (1,589-2,326), and exonic (1,904-2,918) 193 categories (Supplementary Table S2). Importantly, peaks were more abundant in +B samples 194 than in 0B samples peaks across most annotation classes, including promoter-associated 195 regions (4,472 for +B; 3,259 for 0B; Supplementary Table S3; Supplementary Table S4; 196 Supplementary Fig. S4; Supplementary Fig. S5). This suggests that the presence of B 197 chromosomes is associated not only with additional accessible intergenic regions, but also 198 with accessibility changes near genes and putative regulatory elements. To further 199 characterize the genomic context of accessible regions in the presence of the B chromosome, 200 we examined the distribution of peaks across regulatory features in the +B condition. As 201 expected, the majority of peaks (22,917) were located in distal intergenic regions; however, a 202 substantial number of peaks on the B chromosome were also found in promoter-proximal 203 regions (4,472), promoter regions (1-2 kb; 1,320), 5’ UTRs (247), 3’ UTRs (1,980), and gene 204 bodies (3,042 exon and 2,633 intron peaks) (Fig. 3A). 205 _Differential analysis of accessible chromatin regions (ACRs)_ 206 Differential accessibility analysis identified 2,905 regions as significantly more accessible (203 207 on A chromosomes and 2,702 on the B chromosome) and 12 regions as significantly less 208 accessible in the +B versus 0B comparison (Figs. 3B; 3C; 3D). Interpretation of chromosome 209 assignment in this context must consider the fact that the B chromosome is absent from 0B 7 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 210 plants. Therefore, peaks on the B chromosome can only contribute to the class of regions 211 gaining accessibility in the presence of the B chromosome, whereas a category of “less 212 accessible B-chromosome chromatin regions” lacks biological symmetry in this experimental 213 design. With this limitation in mind, the data show that the vast majority of more accessible 214 regions were mapped to the B chromosome (2,702 ACRs; Supplementary Dataset S8), while a 215 smaller number of accessible peaks were located on A chromosomes (203 ACRs; Figs. 3B - 3D). 216 Thus, the main accessibility gain associated with the presence of the B chromosome reflects 217 the addition of accessible B-chromosomal regions. Genome-wide peak density analysis 218 revealed localized regions of exceptionally high chromatin accessibility on chromosome 6 (Fig. 219 3D). Sequence annotation indicated that the region on chromosome 6 corresponds to 220 ribosomal DNA (rDNA) clusters, consistent with known nucleolar organizer regions (NORs) in 221 maize. Importantly, these regions did not exhibit strong differential accessibility between 0B 222 and +B samples, indicating that they reflect constitutively accessible genomic features (Liu _et_ 223 _al.,_ 2025) rather than B chromosome-associated regulatory changes. More interestingly, in 224 comparison with other A chromosomes, an increased number of differentially accessible 225 chromatin regions were observed on chromosome 5 (Fig. 3D). 226 Genome-wide visualization of peak density revealed distinct differences between A and B 227 chromosomes. When considering only differentially accessible regions (DARs), A 228 chromosomes showed relatively sparse and dispersed changes, whereas the B chromosome 229 displayed strong regional enrichment of peaks with increased accessibility in the +B condition. 230 Although this strong enrichment on B chromosome was captured via DESeq2, statistically a 231 high number of peaks were classified as significantly more open simply because they are 232 missing in 0B samples. Nevertheless, these results indicate that B chromosomes behave as an 233 accessible and dynamic chromatin feature. 234 ACRs, that were more accessible in samples with two B chromosomes, were associated with 235 738 nearby genes (168 genes on A chromosomes, after removal of duplicate peak-gene 236 assignments), and a substantial fraction of these genes were B-linked (570 genes, after removal 237 of duplicate peak-gene assignments; Supplementary Dataset S8). In contrast, peaks that were 238 less accessible were associated exclusively with A-chromosome genes (12 genes). Because B 239 chromosomes are absent from the 0B genotype, this asymmetry should not be interpreted 240 simply as evidence that B-linked loci are preferentially protected from repression. Rather, it 8 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 241 indicates that the presence of B chromosomes introduces an additional set of accessible loci, 242 most of which are positioned near annotated B-linked genes, while reductions in accessibility 243 are confined to loci present in both genotypes. However, the effect of the B chromosome on 244 the accessibility of chromatin in the vicinity of A-chromosome-encoded genes should not be 245 overlooked. 246 _Gene set enrichment analyis of genes near more open chromatin_ 247 We performed Gene Ontology (GO) enrichment analysis using the topGO package to 248 investigate the functional roles of genes located near more open chromatin regions (168 genes 249 on A chromosomes and 570 on the B chromosome). Where several significant terms were 250 related, we reported the more specific term (i.e., the one lower in the GO hierarchy) over its 251 more general parent term to maximize informativeness. The complete lists of tested GO terms, 252 including non-significant ones, are provided in Supplementary Dataset S4. Among biological 253 processes, genes on chromosome A were significantly enriched for “homeostatic process” 254 (GO:0042592) followed by “rRNA processing” (GO:0006364). In the molecular function 255 category, the most enriched GO term was “chromatin binding” (GO:0003682), followed by 256 “transcription coregulator activity” (GO:0003712), and “enzyme binding” (GO:0019899). In 257 contrast, no significantly enriched GO terms were detected within the cellular component 258 category (Fig. 4B). 259 For B-chromosome-located genes, the most prevalent biological process GO term was “RNA 260 templated DNA biosynthetic process” (GO:0006278), followed by “monoatomic cation 261 transmembrane transport” (GO:0098655), “DNA recombination” (GO:0006310), “proteolysis” 262 (GO: 0006508), “chromatin remodelling” (GO:0006338), “inorganic cation transmembrane 263 transport” (GO:0098662), “DNA duplex unwinding” (GO:0032508), “nucleotide-excision 264 repair” (GO:0006289) and “mitotic spindle organization” (GO:0007052). In the molecular 265 function category, the most enriched GO term was “ATP hydrolysis activity” (GO:0016887), 266 followed by “histone modifying activity” (GO:0140993), “DNA polymerase activity” 267 (GO:0034061), “damaged DNA binding” (GO:0003684), “single-stranded telomeric DNA 268 binding” (GO:0043047), “helicase activity” (GO:0004386), “peptidase activity” (GO:0008233) 269 and “fucosyltransferase activity” (GO:0008417). Finally. Within the cellular component 270 category, “membrane” (GO:0016020) was the most abundant GO term, followed by 271 “microtubule associated complex” (GO:0005875), “DNA replication factor A complex” 9 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 272 (GO:0005662), “spindle” (GO:0005819), “small-subunit processome” (GO:0032040) and 273 “microtubule” (GO:0005874) (Fig. 4A). 274 _Motif enrichment analysis_ 275 Transcription factors (TF) typically bind to short DNA sequences spanning 6-15 bp. We analysed 276 motifs within the differentially accessible chromatin regions (DARs) to predict TFs that 277 preferentially bind under the presence of B chromosomes. Motif enrichment analysis of the 278 strongly enhanced ATAC-seq peaks (2,905 DARs; adjusted P < 0.05 and log2 fold change ≥ 2) 279 identified several known motifs that were significantly enriched relative to the complete 280 consensus peak set of 141,793 accessible regions (Table 1; Supplementary Dataset S10). 281 Specifically, multiple independent TCP motifs showed a 1.3 to 2-fold enrichment 282 (At1g69690/TCP: 1.77-fold; At2g45680/TCP: 1.95-fold; TCP20: 1.45-fold; At1g72010 TCP: 1.38 283 fold; At5g08330/TCP: 1.35-fold; TCP7: 1.38-fold), providing a consistent regulatory signature 284 associate with the TCP-family, which is known to regulate reproductive development, cell cycle 285 dynamics, proliferation, and hormonal pathways (Viola and Gonzalez, 2023). A second 286 enriched group comprised MYB-related motifs (1.2- to 1.3-fold enrichment; involved in cell 287 wall biosynthesis, stress responses, and hormone signalling; Wu _et al._, 2024). Additionally, 288 significant enrichment of FHY3/FAR1- (1.27-fold), Trihelix- (1.1- to 1.3-fold), CAMTA- (1.2-fold), 289 C3H- (1.49-fold), GATA/ZML- (1.63- to 1.96-fold) and E2F-like (1.2- to 3.9-fold) motifs. 290 _Integration of ATAC-seq and RNA-seq_ 291 Numerous recent studies have demonstrated that the B chromosomes are not merely 292 structurally present, but also transcriptionally active. In anthers containing mature pollen, 293 hundreds of B-chromosome-encoded genes have been detected among expressed genes and 294 among significantly upregulated genes (Hloušková _et al.,_ 2025). In our mature pollen RNA-seq 295 dataset, the total number of expressed genes was lower, but B-linked genes still represented 296 a substantial fraction of the significantly upregulated genes. To identify loci showing 297 coordinated changes in chromatin accessibility and gene expression, we intersected the pollen 298 ATAC-seq and pollen RNA-seq datasets (Fig. 5B). Importantly, this overlap identified 87 genes 299 that exhibited both significantly increased chromatin accessibility and elevated transcript 300 abundance, the vast majority of which (83) corresponded to B-linked loci (Fig. 5B). By contrast, 301 no genes showed a reciprocal match between reduced accessibility and reduced expression. 10 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 302 Together, these results support the conclusion that the presence of B chromosomes in maize 303 pollen is associated with the establishment of accessible chromatin at B-linked loci and the 304 concomitant transcriptional upregulation of a subset of B-chromosome genes. 305 At the same time, the data suggest that the influence of B chromosomes is not restricted 306 entirely to the B genome itself. A portion of differentially accessible regions (DARs) and 307 differentially expressed genes were assigned to A chromosomes, however, only four A 308 chromosome genes were specifically highlighted as showing concordant changes in both 309 pollen ATAC-seq and pollen RNA-seq (Fig. 5B). These included loci associated with promoter 310 proximal as well as distal regulatory regions and encompassed genes with diverse predicted 311 functions: RNA processing ( _Zm00001eb048500_ ), helicase activity ( _Zm00001eb164280_ ), lipid 312 metabolism ( _Zm00001eb224090_ ), and DNA replication/repair ( _Zm00001eb330170_ ) 313 (Supplementary Dataset S9). 314 To evaluate the functional profile of all 87 coregulated genes, we performed a GO enrichment 315 analysis. Among biological processes, the most enriched was “response to heat” 316 (GO:0009408), followed by “trehalose metabolic process” (GO:0005991), “chromatin 317 organization” (GO:0006325), “disaccharide biosynthetic process” (GO:0046351), “spindle 318 assembly” (GO:0051225) and “RNA-templated DNA biosynthetic process” (GO:0006278) and 319 “microtubule-based movement” (GO:0007018). In the molecular function category, the most 320 enriched GO term was “ATP hydrolysis activity” (GO:0016887), followed by “microtubule 321 motor activity” (GO:0003777), “DNA polymerase activity” (GO:0034061), “fucosyltransferase 322 activity” (GO:0008417), “protein methyltransferase activity” (GO:0008276), “O 323 acyltransferase activity” (GO:0008374), “clathrin binding” (GO:0030276) and “microtubule 324 binding” (GO:0008017). Within the cellular component category, “membrane” (GO:0016020) 325 was the most abundant GO term, followed by “microtubule associated complex” 326 (GO:0005875), “lytic vacuole” (GO:0000323), “clathrin-coated vesicle” (GO:0030136), “P 327 body” (GO:0000932) and “proteasome complex” (GO:0000502) (Fig. 5B). Where several 328 significant terms were related, we reported the more specific term (i.e., the one lower in the 329 GO hierarchy) over its more general parent term to maximize informativeness. The complete 330 lists of tested GO terms, including non-significant ones, are provided in Supplementary Dataset 331 S4. 11 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 332 These observations indicate that the presence of B chromosomes may also have secondary 333 effects on the regulatory state of selected A-chromosome loci. However, this effect appears 334 limited compared with the much stronger signal generated directly by B-linked accessible 335 chromatin and transcription. Overall, the results indicate that maize B chromosomes in pollen 336 are associated with a distinct accessible chromatin profile and with transcriptional activity, 337 particularly at B-linked loci. 338 The ATAC-seq and RNA-seq datasets capture different levels of the regulatory response (Fig. 339 5B; Supplementary Dataset S9). Specifically, differentially accessible regions show a strong 340 enrichment for TCP-family signatures (Table 1; Supplementary Dataset S10), whereas the most 341 strongly induced transcripts are dominated by stress- and proteostasis-related functions, RNA 342 metabolism, and several chromatin or transcription-associated factors (Supplementary 343 Dataset S9). Intriguingly, the only strongly induced, sequence-specific TF within the filtered A 344 chromosome DEG set is an AP2/ERF-family factor (Zm00001eb307550/ERF1B), whose motif 345 family is not significantly enriched among the opening chromatin regions (Supplementary 346 Dataset S9; Supplementary Dataset S10). 347 348 **Discussion** 349 We provide the first evidence that maize B chromosomes possess an accessible regulatory 350 landscape that is coupled to active transcription in mature pollen. While previous work has 351 extensively described the mechanism of B chromosome transmission through chromosome 352 drive, little was known about their chromatin organization and transcriptional activity in the 353 male gametophyte. By combining flow-sorted pollen nuclei with ATAC-seq and RNA-seq, we 354 demonstrate that B chromosomes harbour numerous accessible chromatin regions and 355 actively transcribed genes, indicating that they are functionally active components of the 356 pollen genome rather than genetically inert elements. Because the B chromosome is physically 357 absent from 0B plants, the strongest conclusion is not that chromatin accessibility is globally 358 redistributed between equivalent genomic compartments, but rather that the addition of B 359 chromosomes introduces a new set of accessible and transcriptionally active regions into the 360 pollen nucleus. Together, these findings indicate that such activity is predominantly restricted 12 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 361 to the B chromosome itself, as its influence on the standard A genome appears comparatively 362 modest. 363 While mature maize pollen grains show an overall lower gene expression level and a reduced 364 number of expressed genes (TPM > 1; Shi _et al.,_ 2021; Han _et al.,_ 2022; Nelms and Walbot, 365 2022; Zhang _et al.,_ 2022; Hloušková _et al.,_ 2025) relative to anther and other tissues, they 366 simultaneously maintain a subset of distinct, highly enriched pollen-specific genes (Honys and 367 Twell, 2003; Pina _et al.,_ 2005; Ma _et al.,_ 2006; Ling _et al.,_ 2015; Shi _et al.,_ 2021; Hong _et al.,_ 368 2022; Le Lievre _et al.,_ 2023; Supplementary Dataset S2). The less complex pollen transcriptome 369 observed here is consistent with previous studies and represents a conserved feature among 370 diverse species (Honys and Twell, 2003; Gómez _et al.,_ 2015; Han _et al.,_ 2022). Expression 371 profiling of tissues from B-positive plants revealed that 103 of the 689 predicted B 372 chromosome-encoded genes were transcribed in mature pollen (TPM > 1; Supplementary 373 Dataset S1), with 98 of these genes overlapping with those detected in the whole anther 374 transcriptome. In contrast to our results, Hloušková _et al._ (2025) previously reported that 352 375 B chromosome-located genes are actively transcribed across a wider selection of tissues 376 including 244 expressed B-genes in anther tissue containing mature pollen. Combined with 377 our detection of B-encoded genes in mature pollen, these data strongly suggest that B 378 chromosomes are preferentially expressed in reproductive organs, implying a potential 379 functional role for their genes in reproductive development, maturation, and transmission of 380 the B chromosome to the next generation. 381 The detection of substantial amounts of B-chromosome mRNA in mature pollen further 382 suggests that these transcripts are stored for subsequent use during pollen germination and 383 tube growth as previously described by Rutley and Twell (2015). The germination and guided 384 growth of pollen tubes are driven by a combination of stored transcripts and newly synthesized 385 mRNAs that facilitate signalling and maternal-tissue communication within the gynoecium 386 (Rutley and Twell, 2015). Consistent with the fundamental importance of pollen-pistil 387 communication, the pollen transcriptome was highly enriched for signalling components 388 known to facilitate female gametophyte-driven guidance of the pollen tube to the ovule 389 (Palanivelu and Tsukamoto, 2012). Conversely, processes related to energy metabolism, 390 transport, transcription, and translation remain under-represented (Rutley and Twell, 2015). 391 Compared to sporophytic tissues, the maize pollen transcriptome exhibits reduced total gene 13 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 392 diversity alongside a highly skewed expression profile characterized by exceptionally high 393 maximum transcript levels (Davidson _et al.,_ 2011). 394 Crit i cal to the contnuat i on of the B chromosome’s survival in the genet i i c lineage of maize is 395 its drive mechanism. W hile nondisjunction at the second pollen mitosis is controlled by cis and 396 trans factors on the B chromosome (Birchler and Yang 2021; Blavet _et al.,_ 2021) and occurs 397 when sister chromatids fail to separate at the centromeric region, the second component of 398 the drive is arguably more relevant to the biology of mature pollen. Current evidence suggests 399 this mechanism involves structural domains adjacent to the B centromere-proximal region 400 alongside egg cell-driven recognition of specific sperm nuclei, although the precise molecular 401 factors have yet to be identified (González-Sánchez _et al.,_ 2003; Brennan _et al.,_ 2023). 402 The B chromosome is distinguished by repeat-dense pericentromeric regions, notably ZmBs 403 repeats and knob-like heterochromatin (Birchler and Yang, 2021; Blavet _et al.,_ 2021). Within 404 the sperm nucleus, these B-specific sequences adopt a localized spatial organization at one 405 pole or subregion, imparting a distinct nuclear architecture that may influence gametic 406 recognition during double fertilization (Rusche _et al.,_ 1997; Blavet _et al.,_ 2021). When the egg 407 and central cell are both accessible to the two sperm, the egg appears to differentially interact 408 with the B-carrying sperm nucleus, likely via specific receptor-ligand, adhesion, or signalling 409 components on the egg plasma membrane or its surroundings. This mechanism may involve 410 egg-cell-expressed proteins or RNAs that recognize features associated with the B-centromere 411 region, such as centromeric proteins, unique chromatin marks, or nuclear envelope contacts, 412 leading to more efficient fusion or nuclear migration. Furthermore, since the B centromere 413 region is enriched in specific repeats related to knob heterochromatin, these regions replicate 414 late and can alter chromatin compaction or nuclear architecture, potentially changing how the 415 sperm nucleus interacts with the egg-cell nuclear envelope or cytoskeletal guidance. A 416 feasible, though not yet directly demonstrated, model is that egg-cell factors recognize nuclear 417 envelope or centromere/kinetochore proteins associated with the B region, allowing the B 418 carrying sperm nucleus to be preferentially anchored, imported, or fused within the egg, while 419 the other sperm nucleus associates with the central cell (González-Sánchez _et al.,_ 2003; Blavet 420 _et al.,_ 2021). 421 Alternatively, preferential fertilization might initiate earlier, within the pollen tube, if the B 422 carrying sperm exhibits altered interactions with the vegetative cell or tube cytoskeleton that 14 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 423 position it closer to the egg at discharge, or affect the sequence in which the two sperm are 424 released and fuse. B-linked factors could modulate the expression or localization of proteins in 425 the generative or sperm cells that influence motility-like behaviours inside the tube, such as 426 movement along actin filaments, thereby giving the B-rich sperm a kinetic advantage toward 427 the egg rather than the central cell (González-Sánchez _et al.,_ 2003; Blavet _et al.,_ 2021). Despite 428 cytological and genetic evidence pointing toward a model governed by nuclear architecture 429 and egg-cell recognition, the precise molecular machinery remains unknown. Specifically, the 430 genes, proteins, and direct interactions at the egg membrane, nuclear envelope, or pollen tube 431 that physically execute this preferential fertilization warrant further investigation. 432 Our results of differentially accessible chromatin extend observations of Hloušková _et al.,_ 2025 433 by demonstrating that B chromosome transcription is maintained in mature pollen - a highly 434 specialized cell type characterized by a relatively simple transcriptome, as described in _A._ 435 _thaliana_ (Honys and Twell, 2003; Pina _et al.,_ 2005), rice (Wei _et al.,_ 2010; Hong _et al.,_ 2022), 436 maize (Ma _et al.,_ 2006; Han _et al.,_ 2022) and cotton (Zhang _et al.,_ 2022). Typically, mRNA and 437 rRNA gene transcripts accumulate and transcription is subsequently repressed during pollen 438 maturation, with these transcripts stored for use in protein synthesis during pollen 439 germination (Mascarenhas, 1989; Schrauwen _et al.,_ 1990; Honys _et al.,_ 2000; Honys and Twell 440 2003). 441 Although little is known about the higher-order chromatin organization of maize B 442 chromosomes, recent work in other organisms suggests that supernumerary chromosomes 443 can influence nuclear architecture. A recent chromosome-level assembly of the cichlid fish 444 _Astatotilapia latifasciata_ (Jehangir _et al.,_ 2026) revealed that a 34 Mb region of B chromosome 445 acts as a major driver of nuclear architecture by reshaping chromatin organization. The 446 presence of the B chromosome in this species is associated with fewer topologically 447 associating domains (TADs), reduced chromatin loop formation, and altered genomic 448 compartmentalization, while B-chromosome protein-coding genes are predominantly 449 repressed relative to their A-chromosome paralogs. Furthermore, the B chromosome of the 450 cichlid fish influences long-range chromatin interactions and is linked to genomic 451 rearrangements within the host's standard A chromosomes (Jehangir _et al.,_ 2026). Similarly, in 452 wheat and rye, approximately 5-6% of A-derived transcripts were differentially expressed 453 when two Bs were present (Boudichevskaia _et al.,_ 2022). In maize, the presence of B 15 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 454 chromosomes has been shown to significantly alter the expression of over 3,000 genes located 455 on A chromosomes, as well as small RNAs (miRNAs; Shi _et al.,_ 2022; Hloušková _et al.,_ 2025). 456 Whether a gene is upregulated or downregulated often depends on the host species and tissue 457 type (Boudichevskaia _et al.,_ 2022). 458 Interestingly, several loci exhibited accessible chromatin in both genotypes but increased 459 transcript abundance only in the presence of B chromosomes. This observation suggests that 460 chromatin accessibility alone is insufficient to explain transcriptional activation and that 461 additional epigenetic or transcriptional regulatory mechanisms may contribute. Because B 462 chromosomes depend on preferential transmission for their long-term persistence, selection 463 may favour the retention of genes that enhance chromosome drive or ensure stable 464 inheritance, while genes unrelated to these processes may gradually degenerate. It follows 465 that candidate genes for the control of preferential fertilisation should be recruited from the 466 list of B-chromosome-encoded genes expressed in mature and/or active pollen. Alternatively, 467 some transcripts may simply result from the presence of accessible chromatin without 468 conferring an immediate functional advantage. Further functional characterization of the 469 expressed B-chromosome genes, particularly those associated with accessible promoter 470 regions (232 genes out of 540 B genes; Supplementary Dataset S7), will be necessary to 471 distinguish between these possibilities. 472 Beyond its biological findings, the present study demonstrates that flow cytometric 473 purification of mature pollen nuclei can be successfully combined with ATAC-seq and RNA-seq 474 to generate high-quality chromatin accessibility profiles. Mature pollen presents unique 475 technical challenges because of its highly condensed chromatin and resistant pollen wall, 476 making genome-wide epigenomic analyses difficult. Our workflow therefore provides a useful 477 framework for future studies investigating chromatin organization in the male gametophyte. 478 Collectively, our findings demonstrate that maize B chromosomes establish a distinct 479 chromatin accessibility landscape and contain transcriptionally active loci in mature pollen. 480 Furthermore, the enrichment of accessible chromatin near B-linked genes suggests that 481 regulatory activity is organized rather than randomly distributed across the B chromosome. 482 Ultimately, this structural and transcriptional characterization provides new insight into the 483 molecular organization of selfish chromosomes and establishes a foundation for future studies 16 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 484 aimed at identifying the molecular mechanisms underlying B-chromosome maintenance, 485 chromosome drive, and interactions with the host genome. 486 487 **Materials and Methods** 488 **Plant Material and Growing Conditions** 489 Mature pollen grains from maize ( _Zea mays_ L.) lines (B73 without the B chromosome - 0B; B73 490 possessing two B chromosomes - +B) were harvested from greenhouse-grown plants into 491 tubes. The original seed stock of reference inbred line B73 was provided by the Plant 492 Introduction Station (North Central Regional Plant Introduction Station, Iowa, USA) and the 493 original seed stock of +B line by prof. Birchler (University of Missouri, Missouri, USA). Seeds 494 were germinated at 25°C for 5 days, selected plants were grown in the greenhouse under 495 controlled conditions of 14h of light at 24°C and 10h of dark at 20°C. A direct PCR approach 496 was used to assess the presence of the B chromosome in individual seedlings. The +B seedlings 497 were further scored using droplet digital PCR (ddPCR) protocol to estimate the number of B 498 chromosomes (Svačina _et al.,_ 2023). Mature pollen (in three biological replicates for both - 0B 499 and +B line) was immediately used for nuclei isolation and subsequent fluorescence-activated 500 nuclei sorting by flow cytometry. 501 **Nuclei isolation and Flow cytometry** 502 Anthers with mature pollen grains were vortexed in LB01 buffer (Doležel _et al.,_ 1989). Isolation 503 was performed by bursting method according to Kron and Husband (2015). Suspension of 504 pollen and anthers was filtered through a 100 μm filter (Celltrics, Sysmex) to remove anthers 505 and through a 20 μm filter to capture pollen grains. The surface of the 20 μm filter with pollen 506 was then gently rubbed with glass rod for a few seconds to release nuclei from pollen grains 507 by washing down through the filter with LB01 buffer. Nuclei were stained with 4',6-diamidino 508 2-phenylindole (DAPI; 2 μg/ml) and loaded onto FACSAria II SORP flow cytometer and sorter 509 (Becton Dickinson, San Jose, USA) equipped with 390 nm laser for DAPI excitation. Sorting 510 gates were initially drawn on a dot plot of forward scattering (y-axis) and relative DAPI 511 fluorescence (x-axis) and subsequently on a dot plot of relative DAPI-W fluorescence (y-axis) 512 versus relative DAPI-A fluorescence (x-axis) to purify nuclei suspension. The individual fraction 513 of 50,000 of all pollen (vegetative and sperm) nuclei was flow-sorted for each sample into LB01 17 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 514 with protease inhibitor (cOmplete, Roche) for subsequent ATAC-seq analysis (Grandi _et al.,_ 515 2022). 516 **ATAC-seq library construction and sequencing** 517 Isolated nuclei were spun down at 500 x g for 5 minutes at 4 °C. During this time, tagmentation 518 master mix was prepared with the ATAC-Seq Kit according to the manufacturer’s manual 519 (Active Motif catalog No. 53150) to perform the tagmentation reaction that was held at 37 °C 520 for 30 min in a thermal shaker (Eppendorf® ThermoMixer® C) at 750 rpm. Tagmented DNA was 521 purified on purification column and PCR reaction for library amplification was performed 522 according to the manufacturer protocol. The libraries were size selected using SPRI beads. DNA 523 was quantified by Qubit HS DNA kit (Thermo Fisher, Q32854) and analyzed on Agilent 2100 524 BioAnalyzer for quality control. ATAC-seq libraries were sequenced on an Illumina NovaSeq 525 6000 platform (Illumina, San Diego, USA) which produced an average 200 million paired-end 526 reads per sample (2×150 bp paired-end reads). 527 **RNA Extraction and Sequencing** 528 Total RNA was isolated and sequenced for transcriptomic analysis according to Hloušková _et_ 529 _al._ (2025). RNA was isolated from pollen, in three biological replicates for reference B73 inbred 530 line without B chromosomes and three biological replicas for B73 inbred line containing two B 531 chromosomes, following the instruction provided in Monarch® Total RNA Miniprep Kit (New 532 England Biolabs, Ipswich, USA). The quality of total RNA was evaluated using Agilent 2100 533 Bioanalyzer and the indexed sequencing libraries construction was performed according to the 534 manufacturer’s protocol with NEBnext Ultra II RNA Library Prep Kit for Illumina (New England 535 Biolabs). Libraries were sequenced on an Illumina NovaSeq 6000 platform, using 2×150-bp 536 paired-end reads configuration, yielding 35 million paired-end reads per sample on average. 537 **ATAC-seq analysis** 538 The analysis of ATAC-seq data was performed in Miniconda (v24.4.0; 539 https://github.com/conda/conda), a cross-platform, open-source package and environment 540 management system, to ensure compatibility among the tools and data formats. First, FastQC 541 (v0.12.1; htps://www.bioinformat t [cs.babraham.ac.uk/projects/fastqc/](https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) i ) and MultiQC (v. 1.12; 542 Ewels _et al.,_ 2016) were used for the quality control of the raw sequencing data to ensure that 18 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 543 only high-quality reads are used in downstream analysis. The raw sequencing data were 544 further trimmed using trim galore (v0.6.2; https://github.com/FelixKrueger/TrimGalore), and 545 trimmed data were mapped to the _Zea may_ s reference genome RefGen_v5 (Hufford _et al.,_ 546 2021) supplemented with the reference of the maize B chromosome (Maize_Bchr_v2.0, 547 Hloušková _et al.,_ 2025), using a sensitive Bowtie2 v2.4.2 aligner (Langmead and Salzberg, 548 2012). All reads mapping to the mitochondrial and plastid DNA were removed from the .bam 549 files for subsequent analysis. Reads were sorted and indexed. Properly-paired end reads were 550 selected for next analysis using -f 3 parameter in samtools view (SAMtools v1.11; Li _et al.,_ 551 2009). PCR duplicates were removed by MarkDuplicates function in Picard (v2.9; 552 https://broadinstitute.github.io/picard/). Due to library size normalization, equivalently 27 553 millions of reads were then randomly selected by samtools view from each library (to achieve 554 size of the smallest available library - 27,298,261 reads; Supplementary Dataset S4) and 555 equivalently more reads (+ 6.2 %) were randomly selected for +B sample to compensate for 556 the B chromosome size. Before the subsequent analysis, .bam files were converted to .bedpe 557 files by bedtools bamtobed v2.26.0 (Quinlan, 2014). Significant broad peaks were then called 558 using MACS2 v2.2.6 (Zhang _et al.,_ 2008). To retrieve the black list of repetitive regions, 559 RepeatMasker (Smit, AFA, Hubley, R & Green, P. RepeatMasker Open-4.0. 2013-2015 560 <http://www.repeatmasker.org>) was used. For masking we used Maize TE Consortium 561 [(MTEC) curated library of repeats (htps://github.com/oushujun/MTECt](https://github.com/oushujun/MTEC) ) and custom-made 562 library of repeats of B chromosome sequence generated using DANTE and DANTE_LTR (Novák 563 _et al_ ., 2024). Peaks overlapping the blacklist of highly repetitive regions were filtered out using 564 bedtools intersect v.2.26.0 (Quinlan, 2014). Finally, annotation of accessible chromatin regions 565 and peak-gene assignment were done using ChIPseeker v1.48.0 (Yu _et al.,_ 2015) and 566 differential accessibility analysis was done using DESeq2 (v.1.12.3) package in R (RStudio 567 (v3.4.0), Love _et al.,_ 2014) with _P_ -value < 0.05 and Log 2 FoldChange > 2 or < -2 to identify 568 significant differentially accessible chromatin regions. For manual control, data were visualized 569 by Integrative Genomics Viewer (IGV; v2.17.4.; Robinson _et al.,_ 2011) 570 **Motif enrichment analysis** 571 Motif enrichment analysis of differentially accessible ATAC-seq regions was performed using 572 HOMER v.5.1 (Heinz _et al.,_ 2010). Regions with significantly increased accessibility (adjusted _P_ 573 < 0.05 and log2 fold change ≥ 2) were analyzed with findMotifsGenome.pl using the reference 19 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 574 genome of _Zea mays_ reference RefGen_v5 (Hufford _et al.,_ 2021) supplemented with the 575 reference of the maize B chromosome (Maize_Bchr_v2.0, Hloušková _et al.,_ 2025). Motif 576 searches were performed within 200-bp regions centered on each peak, testing motif lengths 577 of 6, 8, 10, and 12 bp. The complete set of 141,794 ATAC-seq peaks was used as the background 578 to control for sequence characteristics of accessible chromatin. Overlap of significant peaks 579 with the complete ATAC peak set was verified using bedtools intersect. 580 **RNA-seq analysis** 581 All RNA-seq reads were quality checked with FastQC (v.0.12.1; Andrews, 2010; 582 https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The fastp software (Chen _et_ 583 _al.,_ 2018) was used to filter out sequencing adaptors and trim low-quality sequences 584 (parameters: -q 30, -w 4, -l 50). Trimmed reads were mapped using STAR (Dobin _et al.,_ 2013) 585 to the reference genome of _Zea mays_ reference RefGen_v5 (Hufford _et al.,_ 2021) 586 supplemented with the reference of the maize B chromosome (Maize_Bchr_v2.0, Hloušková 587 _et al.,_ 2025). Mapped reads were counted using RSEM (Li and Dewey, 2011). Transcripts per 588 Million (TPM) values were calculated using Tximport (v.1.34.0) package in R (Soneson _et al.,_ 589 2015). Differential-expression analysis was performed using DESeq2 (v.1.12.3) package in R 590 (Love _et al.,_ 2014) with _P_ -value < 0.05 and Log 2 FoldChange > 2 or < -2 to identify significant 591 differentially expressed genes. 592 **Gene Set Enrichment Analysis** 593 Gene set enrichment analysis (GSEA) was performed according to the Hloušková _et al._ (2025). 594 Gene ontology (GO) terms provided for _Zea mays_ reference RefGen_v5 (Hufford _et al.,_ 2021) 595 were downloaded from MaizeGDB (https://www.maizegdb.org/). For the B chromosome 596 encoded genes, GO terms were obtained using Blast2GO pipeline (Hloušková _et al.,_ 2025). 597 TopGO (v.2.58.0) package in R (Alexa and Rahnenfuhrer, 2024) was used to run a gene set 598 enrichment analysis. GO terms were considered significantly enriched if the _P_ -value from 599 Fisher's exact test was below 0.05 and the term was represented by more than one gene. GSEA 600 was performed on subset of differentially upregulated genes located on A chromosomes and 601 B chromosome-located genes and genes located near significantly more open chromatin 602 regions on A chromosomes and the B chromosome and finally on all genes that overlapped 603 between RNA-seq and genes near accessible chromatin region of ATAC-seq. 20 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 604 605 **References** 606 **Beckett JB** . 1991. Cytogenetic, Genetic and Plant Breeding Applications of B–A 607 Translocations in Maize†. In: Gupta PK, Tsuchiya T, eds. Developments in Plant Genetics and 608 Breeding. Chromosome Engineering in Plants: Genetics, Breeding, Evolution, Part A. 609 Elsevier, 493–529. 610 **Birchler JA, Yang H** . _a_ . The supernumerary B chromosome of maize: drive and genomic 611 conflict. Open Biology **11**, 210197. 612 **Blavet N, Yang H, Su H,** _**et al.**_ 2021. Sequence of the supernumerary B chromosome of 613 maize provides insight into its drive mechanism and evolution. Proceedings of the National 614 Academy of Sciences **118**, e2104254118. 615 **Bojdová T, Hloušková L, Holušová K,** _**et al.**_ 2025. Sorghum embryos undergoing B 616 chromosome elimination express B-variants of mitotic-related genes. Genome Biology **27**, 8. 617 **Boudichevskaia A, Fiebig A, Kumke K, Himmelbach A, Houben A** . 2022. Rye B 618 chromosomes differently influence the expression of A chromosome-encoded genes 619 depending on the host species. Chromosome Research: An International Journal on the 620 Molecular, Supramolecular and Evolutionary Aspects of Chromosome Biology **30**, 335–349. 621 **Bougourd SM, Jones RN** . 1997. B Chromosomes: A Physiological Enigma. The New 622 Phytologist **137**, 43–54. 623 **Brennan VA, Yang H, Gao Z, Birchler JA** . 2023. Variation for maize B chromosome 624 preferential fertilization: a component of the B chromosome drive mechanism. The Nucleus 625 **66**, 305–310. 626 **Carlson WR** . 1969. Factors affecting preferential fertilization in maize. Genetics **62**, 543– 627 554. 628 **Carlson WR** . 2006. Unstable inheritance of maize B-type chromosomes that lack centric 629 heterochromatin. Genome **49**, 420–431. 630 **Carlson WR, Chou T-S** . 1981. B Chromosome Nondisjunction in Corn: Control by Factors 631 near the Centromere. Genetics **97**, 379–389. 632 **Carlson WR, Phillips RL** . 1986. The B chromosome of maize. Critical Reviews in Plant 633 Sciences **3**, 201–226. 634 **Chen S, Zhou Y, Chen Y, Gu J.** 2018. fastp: an ultra-fast all-in-one FASTQ preprocessor. 635 Bioinformatics 34, i884–i890. 636 **Cherchame E, Vromman M, Guégan J, Coutelier M.** 2025. ATAC-seq methods for 637 consensus peak generation to investigate differences in chromatin accessibility per sample 638 [group. Protocols.io https://www.protocols.io/view/atac-seq-methods-for-consensus-peak-](https://www.protocols.io/view/atac-seq-methods-for-consensus-peak-generation-to-36wgq326olk5/v1) 639 [generation-to-36wgq326olk5/v1](https://www.protocols.io/view/atac-seq-methods-for-consensus-peak-generation-to-36wgq326olk5/v1) 640 **Chumak N, Mosiolek M, Schoft VK** . 2015. Sample Preparation and Fractionation of 641 Arabidopsis thaliana Sperm and Vegetative Cell Nuclei by FACS. Bio-Protocol **5**, e1664. 642 **Davidson RM, Hansey CN, Gowda M,** _**et al.**_ 2011. Utility of RNA Sequencing for Analysis 643 of Maize Reproductive Transcriptomes. The Plant Genome **4** . 21 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 644 **Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M,** 645 **Gingeras TR** . 2013. STAR: ultrafast universal RNA-seq aligner. Bioinformatics **29**, 15–21. 646 **Doležel J, Binarová P, Lucretti S** . 1989. Analysis of Nuclear DNA content in plant cells by 647 Flow cytometry. Biologia Plantarum **31**, 113–120. 648 **Ewels P, Magnusson M, Lundin S, Käller M** . 2016. MultiQC: summarize analysis results 649 for multiple tools and samples in a single report. Bioinformatics **32**, 3047–3048. 650 **Gómez JF, Talle B, Wilson ZA** . 2015. Anther and pollen development: A conserved 651 developmental pathway. Journal of Integrative Plant Biology **57**, 876–891. 652 **González-Sánchez M, González-González E, Molina F, Chiavarino AM, Rosato M,** 653 **Puertas MJ** . 2003. One gene determines maize B chromosome accumulation by preferential 654 fertilisation; another gene(s) determines their meiotic loss. Heredity **90**, 122–129. 655 **Grandi FC, Modi H, Kampman L, Corces MR** . 2022. Chromatin accessibility profiling by 656 ATAC-seq. Nature protocols **17**, 1518–1552. 657 **Han Y, Hu M, Ma X, Yan G, Wang C, Jiang S, Lai J, Zhang M** . 2022. Exploring key 658 developmental phases and phase‐specific genes across the entirety of anther development in 659 maize. Journal of Integrative Plant Biology **64**, 1394–1410. 660 **Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng JX, Murre C, Singh** 661 **H, Glass CK** . 2010. Simple combinations of lineage-determining transcription factors prime 662 cis-regulatory elements required for macrophage and B cell identities. Molecular Cell **38**, 663 576–589. 664 **Hloušková L, Tulpová Z, Svačina R, Holušová K, Cápal P, Navrátilová P, Karafiátová** 665 **M, Bartoš J** . 2025. The maize B chromosome shapes the transcriptome throughout the entire 666 plant life cycle. Journal of Experimental Botany **76**, 6810–6825. 667 **Hong W-J, Lee SK, Kim S-H, Kim Y-J, Moon S, Kim E-J, Silva J, Jung K-H** . 2022. 668 Comparative transcriptome analysis of pollen and anther wall reveals novel insights into the 669 regulatory mechanisms underlying anther wall development and its dehiscence in rice. Plant 670 Cell Reports **41**, 1229–1242. 671 **Honys D, Combe JP, Twell D, Čapková V** . 2000. The translationally repressed pollen672 specific ntp303 mRNA is stored in non-polysomal mRNPs during pollen maturation. Sexual 673 Plant Reproduction **13**, 135–144. 674 **Honys D, Twell D** . 2003. Comparative Analysis of the Arabidopsis Pollen Transcriptome. 675 Plant Physiology **132**, 640–652. 676 **Hufford MB, Seetharam AS, Woodhouse MR,** _**et al.**_ 2021. De novo assembly, annotation, 677 and comparative analysis of 26 diverse maize genomes. Science **373**, 655–662. 678 **Jehangir M, Ahmad SF, Oliveira JIN, Cardoso AL, Wolf IR, Valente GT, Srikulnath K,** 679 **Martins C** . 2026. Genome assembly of Astatotilapia latifasciata uncovers B chromosome– 680 linked chromatin reorganization. Heredity, 1–13. 681 **Jones RN, Rees H** . 1969. An anomalous variation due to B Chromosomes in rye. Heredity 682 **24**, 265–271. 683 **Kron P, Husband BC** . 2015. Distinguishing 2N gamete nuclei from doublets in pollen using 684 flow cytometry and pulse analysis. Cytometry. Part A: The Journal of the International 685 Society for Analytical Cytology **87**, 943–957. 22 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 686 **Kron P, Loureiro J, Castro S, Čertner M** . 2021. Flow cytometric analysis of pollen and 687 spores: An overview of applications and methodology. Cytometry. Part A: The Journal of the 688 International Society for Analytical Cytology **99**, 348–358. 689 **Langmead B, Salzberg SL** . 2012. Fast gapped-read alignment with Bowtie 2. Nature 690 Methods **9**, 357–359. 691 **Le Lievre L, Chakkatu SP, Varghese S, Day RC, Pilkington SM, Brownfield L** . 2023. 692 RNA-seq analysis of synchronized developing pollen isolated from a single anther. Frontiers 693 in Plant Science **14** . 694 **Li B, Dewey CN** . 2011. RSEM: accurate transcript quantification from RNA-Seq data with or 695 without a reference genome. BMC bioinformatics **12**, 323. 696 **Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G,** 697 **Durbin R, 1000 Genome Project Data Processing Subgroup** . 2009. The Sequence 698 Alignment/Map format and SAMtools. Bioinformatics **25**, 2078–2079. 699 **Ling S, Chen C, Wang Y, Sun X, Lu Z, Ouyang Y, Yao J** . 2015. The mature anther700 preferentially expressed genes are associated with pollen fertility, pollen germination and 701 anther dehiscence in rice. BMC Genomics **16**, 101. 702 **Liu Y, Gao X, Liu H,** _**et al.**_ 2025. Constraint of accessible chromatins maps regulatory loci 703 involved in maize speciation and domestication. Nature Communications **16**, 2477. 704 **Longley E. A** . 1927. Supernumerary chromosomes in Zea mays. Journal of Agricultural 705 Research **35**, 769–784. 706 **Love MI, Huber W, Anders S** . 2014. Moderated estimation of fold change and dispersion for 707 RNA-seq data with DESeq2. Genome Biology **15**, 550. 708 **Ma J-J, Liu S-W, Han F-X, Li W, Li Y, Niu S-H** . 2020. Comparative transcriptome analyses 709 reveal two distinct transcriptional modules associated with pollen shedding time in pine. 710 BMC Genomics **21**, 504. 711 **Ma J, Morrow DJ, Fernandes J, Walbot V** . 2006. Comparative profiling of the sense and 712 antisense transcriptome of maize lines. Genome Biology **7**, R22. 713 **Mascarenhas JP** . 1989. The Male Gametophyte of Flowering Plants. The Plant Cell **1**, 657– 714 664. 715 **Mulcahy DL** . 1979. The rise of the angiosperms: a genecological factor. Science **206**, 20–23. 716 **Mulcahy DL, Sari-Gorla M, Mulcahy GB** . 1996. Pollen selection — past, present and 717 future. Sexual Plant Reproduction **9**, 353–356. 718 **Nelms B, Walbot V** . 2022. Gametophyte genome activation occurs at pollen mitosis I in 719 maize. Science **375**, 424–429. 720 **Novák P, Hoštáková N, Neumann P, Macas J** . 2024. DANTE and DANTE_LTR: lineage721 centric annotation pipelines for long terminal repeat retrotransposons in plant genomes. NAR 722 genomics and bioinformatics **6**, 113. 723 **Palanivelu R, Tsukamoto T** . 2012. Pathfinding in angiosperm reproduction: pollen tube 724 guidance by pistils ensures successful double fertilization. Wiley Interdisciplinary Reviews. 725 Developmental Biology **1**, 96–113. 23 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 726 **Pina C, Pinto F, Feijó JA, Becker JD** . 2005. Gene Family Analysis of the Arabidopsis 727 Pollen Transcriptome Reveals Biological Implications for Cell Growth, Division Control, and 728 Gene Expression Regulation. Plant Physiology **138**, 744–756. 729 **Quinlan AR** . 2014. BEDTools: The Swiss-Army Tool for Genome Feature Analysis. Current 730 Protocols in Bioinformatics **47**, 11.12.1-34. 731 **Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, Mesirov** 732 **JP** . 2011. Integrative Genomics Viewer. Nature biotechnology **29**, 24–26. 733 **Roman H** . 1947. Mitotic Nondisjunction in the Case of Interchanges Involving the B-Type 734 Chromosome in Maize. Genetics **32**, 391–409. 735 **Roman H** . 1948. Directed Fertilization in Maize. Proceedings of the National Academy of 736 Sciences of the United States of America **34**, 36–42. 737 **Rusche ML, Mogensen HL, Shi L, Keim P, Rougier M, Chaboud A, Dumas C** . 1997. B 738 chromosome behavior in maize pollen as determined by a molecular probe. Genetics **147**, 739 1915–1921. 740 **Rutley N, Twell D** . 2015. A decade of pollen transcriptomics. Plant Reproduction **28**, 73–89. 741 **Schrauwen JA, de Groot PF, van Herpen MM, van der Lee T, Reynen WH, Weterings** 742 **KA, Wullems GJ** . 1990. Stage-related expression of mRNAs during pollen development in 743 lily and tobacco. Planta **182**, 298–304. 744 **Shi Y, Li Y, Guo Y, Borrego EJ, Wei Z, Ren H, Ma Z, Yan Y** . 2021. A Rapid Pipeline for 745 Pollen- and Anther-Specific Gene Discovery Based on Transcriptome Profiling Analysis of 746 Maize Tissues. International Journal of Molecular Sciences **22**, 6877. 747 **Shi X, Yang H, Chen C, Hou J, Ji T, Cheng J, Birchler JA** . 2022. Effect of aneuploidy of a 748 non-essential chromosome on gene expression in maize. The Plant Journal: For Cell and 749 Molecular Biology **110**, 193–211. 750 **Soneson C, Love MI, Robinson MD** . 2015. Differential analyses for RNA-seq: transcript751 level estimates improve gene-level inferences. F1000Research **4**, 1521. 752 **Svačina R, Hloušková L, Karafiátová M, Bartoš J** . 2023. Scoring the number of B 753 chromosomes in Zea mays L. using droplet digital PCR assay. Plant Methods **19**, 43. 754 **Viola IL, Gonzalez DH** . 2023. TCP Transcription Factors in Plant Reproductive 755 Development: Juggling Multiple Roles. Biomolecules **13**, 750. 756 **Wei LQ, Xu WY, Deng ZY, Su Z, Xue Y, Wang T** . 2010. Genome-scale analysis and 757 comparison of gene expression profiles in developing and germinated pollen in Oryza sativa. 758 BMC genomics **11**, 338. 759 **Wu X, Xia M, Su P, Zhang Y, Tu L, Zhao H, Gao W, Huang L, Hu Y** . 2024. MYB 760 transcription factors in plants: A comprehensive review of their discovery, structure, 761 classification, functional diversity and regulatory mechanism. International Journal of 762 Biological Macromolecules **282**, 136652. 763 **Yu G, Wang L-G, He Q-Y** . 2015. ChIPseeker: an R/Bioconductor package for ChIP peak 764 annotation, comparison and visualization. Bioinformatics **31**, 2382–2383. 765 **Zhang Y, Liu T, Meyer CA,** _**et al.**_ 2008. Model-based analysis of ChIP-Seq (MACS). 766 Genome Biology **9**, R137. 24 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 767 **Zhang R, Zhou L, Li Y,** _**et al.**_ 2022. Rapid Identification of Pollen- and Anther-Specific 768 Genes in Response to High-Temperature Stress Based on Transcriptome Profiling Analysis in 769 Cotton. International Journal of Molecular Sciences **23** . 770 771 **Supplementary data** 772 Supplementary Table S1. Summary statistics of Illumina NovaSeq transcriptome sequencing 773 Supplementary Table S2. Distribution of accessible chromatin regions among replicates 774 Supplementary Table S3. Distribution of accessible chromatin regions uniquely in 0B/+B 775 conditions 776 Supplementary Table S4. Distribution of accessible chromatin regions including common 777 regions for 0B/+B condition 778 Supplementary Figure S1. Chromosomal distribution of differentially expressed genes in maize 779 anthers with pollen 780 Supplementary Figure S2. Flow cytometric analysis of pollen nuclei from maize inbred line B73 781 with and without B chromosomes 782 Supplementary Figure S3. Distribution of accessible chromatin regions (ACR) after DESeq2 783 among replicates. 784 Supplementary Figure S4. Genomic annotation of accessible chromatin regions (ACR) uniquely 785 in 0B (A) and +B (B) condition on chromosomes. 786 Supplementary Figure S5. Genomic annotation of accessible chromatin regions (ACR) including 787 common regions for 0B (A) and +B (B) condition on chromosomes. 788 Supplementary Dataset S1. TPM values of genes in 12 maize plant tissues in +B/0B conditions 789 (genes expressed predominantly in pollen) 790 Supplementary Dataset S2. TPM values assigned for chromosome-encoded genes (the green 791 shading indicates differences in expression) 792 Supplementary Dataset S3. Results of differential gene expression analysis in the pollen tissue 793 in 0B/+B condition 25 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 794 Supplementary Dataset S4. Summary statistics of gene ontology enrichment analysis 795 Supplementary Dataset S5. Summary statistics of reads and peaks in ATAC-seq workflow 796 Supplementary Dataset S6. ATAC-seq consensus peak set 797 Supplementary Dataset S7. Distribution of accessible chromatin regions on chromosomes 798 Supplementary Dataset S8. Results of differential chromatin accessibility analysis in the pollen 799 nuclei in 0B/+B condition 800 Supplementary Dataset S9. Gene description of differential gene expression analysis (RNA 801 seq), genes near differentially accessible chromatin analysis (ATAC-seq) and genes overlapping 802 both methods 803 Supplementary Dataset S10. Homer known Motif enrichment results 804 All supplementary files are available at: 805 [htps://zenodo.org/records/24077189?previewt](https://zenodo.org/records/24077189?preview=1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc5NDIxNTU1MywiZXhwIjoxNzk4Njc1MTk5fQ.eyJpZCI6IjZmM2RhMmZiLWJlYWItNGFlYy04NmFmLTM3YjQwOTBiMzlmNSIsImRhdGEiOnt9LCJyYW5kb20iOiI3OTkwNzNhYmY2YjU5MDJjNmE1YzljMzkzOTc4OWQ2OSJ9.ZapwDv8MTpmCU_9vYlCWhnYRxjHhpesXM2D_VP-eSvVG6-4JCqQclY0peC51dve2em-v8FA14Z5EyT5jX3wHPw) =1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MT 806 [c5NDIxNTU1MywiZXhwIjoxNzk4Njc1MTk5fQ.eyJpZCI6IjZmM2RhMmZiLWJlYWItNGFlYy04Nm](https://zenodo.org/records/24077189?preview=1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc5NDIxNTU1MywiZXhwIjoxNzk4Njc1MTk5fQ.eyJpZCI6IjZmM2RhMmZiLWJlYWItNGFlYy04NmFmLTM3YjQwOTBiMzlmNSIsImRhdGEiOnt9LCJyYW5kb20iOiI3OTkwNzNhYmY2YjU5MDJjNmE1YzljMzkzOTc4OWQ2OSJ9.ZapwDv8MTpmCU_9vYlCWhnYRxjHhpesXM2D_VP-eSvVG6-4JCqQclY0peC51dve2em-v8FA14Z5EyT5jX3wHPw) 807 [FmLTM3YjQwOTBiMzlmNSIsImRhdGEiOnt9LCJyYW5kb20iOiI3OTkwNzNhYmY2YjU5MDJjNmE](https://zenodo.org/records/24077189?preview=1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc5NDIxNTU1MywiZXhwIjoxNzk4Njc1MTk5fQ.eyJpZCI6IjZmM2RhMmZiLWJlYWItNGFlYy04NmFmLTM3YjQwOTBiMzlmNSIsImRhdGEiOnt9LCJyYW5kb20iOiI3OTkwNzNhYmY2YjU5MDJjNmE1YzljMzkzOTc4OWQ2OSJ9.ZapwDv8MTpmCU_9vYlCWhnYRxjHhpesXM2D_VP-eSvVG6-4JCqQclY0peC51dve2em-v8FA14Z5EyT5jX3wHPw) 808 [1YzljMzkzOTc4OWQ2OSJ9.ZapwDv8MTpmCU_9vYlCWhnYRxjHhpesXM2D_VP-eSvVG6-](https://zenodo.org/records/24077189?preview=1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc5NDIxNTU1MywiZXhwIjoxNzk4Njc1MTk5fQ.eyJpZCI6IjZmM2RhMmZiLWJlYWItNGFlYy04NmFmLTM3YjQwOTBiMzlmNSIsImRhdGEiOnt9LCJyYW5kb20iOiI3OTkwNzNhYmY2YjU5MDJjNmE1YzljMzkzOTc4OWQ2OSJ9.ZapwDv8MTpmCU_9vYlCWhnYRxjHhpesXM2D_VP-eSvVG6-4JCqQclY0peC51dve2em-v8FA14Z5EyT5jX3wHPw) 809 4JCqQclY0peC51dve2em-v8FA14Z5EyT5jX3wHPw 810 811 **Acknowledgements** 812 We thank Jitka Weisserová, Helena Tvardíková and Magdaléna Stejskalová for technical 813 assistance. Computational resources were provided by the e-INFRA CZ project (ID:90254), 814 supported by the Ministry of Education, Youth and Sports of the Czech Republic. 815 816 **Funding** 817 This work was supported by the Czech Science Foundation (project no. 23-04887S) and by the 818 Operation Program Research, Development, and Education project: Plants as a tool for 819 sustainable global development (No.CZ.02.1.01/0.0/0.0/16_019/0000827). 820 26 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 821 **Author contributions** 822 PU, JB – investigation; PU, LH, KH, JB – formal analysis; PU – writing – original draft; PU, LH, KH, 823 JD, JB – review and editing; JB – conceptualization, project administration and funding 824 acquisition. 825 826 **Competing interests** 827 The authors declare they have no competing interests. 828 The authors declare that the research was conducted in the absence of any commercial or 829 financial relationships that could be construed as a potential conflict of interest. 830 831 **Data availability** 832 Raw sequence reads are available at the European Nucleotide Archive 833 (https://www.ebi.ac.uk/ena/browser/home) under accession number PRJEB129204 (ATAC 834 seq and RNA-seq). 27 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 835 **Figures** 836 837 **Figure 1. Gene expression and chromosomal distribution of expressed and differentially** 838 **expressed genes in maize pollen. (A)** Number of A-chromosome genes with TPM > 1 detected 839 in different maize tissues. Tissues are: FN, first stem node; HUSK, husk; IMMC, unpollinated 840 immature cob; MT, meiotic tassel; IT, immature tassel; PL, first leaves at the one-leaf stage 841 (V1); PR, primary root; SILK, silk; ST, leaf tip of the youngest actively growing leaf at the five842 leaf stage (V5); WS, whole germinated seedling; ANT, anthers with mature pollen; and Pollen, 843 mature pollen. Data for all tissues with exception of Pollen were originally published by 844 Hloušková _et al_ . (2025). **(B)** Volcano plot shows differential gene expression between 2B and 845 0B pollen. The x-axis represents log2 fold change and the y-axis represents -log10 (adjusted _P_ 28 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 846 value). Dashed lines indicate the thresholds used to define differentially expressed genes. A 847 total of 42 A-chromosome genes were upregulated and three genes were downregulated, 848 while 10,274 genes were not significantly differentially expressed. **(C)** Distribution of expressed 849 (TPM > 1) genes along chromosomes 1-10 and the B chromosome. Gene density is shown 850 along each chromosome, and the numbers indicate the number of 851 upregulated/downregulated differentially expressed genes identified on the respective 852 chromosome. 29 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 853 854 **Figure 2. Gene set enrichment analysis of upregulated genes.** **(A)** Gene ontology (GO) 855 enrichment analysis of genes upregulated on the B chromosome. **(B)** GO enrichment analysis 856 of genes upregulated on the A chromosomes. Bar plots show significantly enriched GO terms 857 ( _P_ < 0.05) for molecular function, cellular component, and biological process categories. Bar 858 length represents the number of upregulated genes associated with each GO term, while bar 859 colour indicates the significance level ( _P_ value). 30 860 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 31 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 861 **Figure 3. Distribution and genomic annotation of accessible chromatin regions and** 862 **differentially accessible regions in maize pollen.** **(A)** Distribution of accessible chromatin 863 regions across the 0B/+B pollen nuclei and their genomic annotation. **(B)** Differential analysis 864 of accessible chromatin regions on the A chromosomes. Volcano plot shows differential 865 accessibility between +B and 0B pollen, with differentially accessible regions identified 866 according to the applied significance and fold-change thresholds. The x-axis represents log2 867 fold change and the y-axis represents -log10 (adjusted _P_ value). Dashed lines indicate the 868 thresholds used to define differentially expressed genes. A total of 203 regions were more 869 accessible and 12 regions were less accessible, while 137,260 regions were not significantly 870 differentially accessible. **(C)** Genomic annotation of differentially more accessible chromatin 871 regions (DARs) on the A chromosomes, showing their distribution among genomic features. 872 **(D)** Distribution of DARs along chromosomes 1-10 and the B chromosome. Peak density is 873 shown along each chromosome, and the numbers indicate the number of more/less 874 differentially accessible regions identified on the respective chromosome. 32 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 875 876 **Figure 4. Gene set enrichment analysis of genes near more open chromatin regions.** **(A)** Gene 877 ontology (GO) enrichment analysis of genes near more open chromatin regions on the B 878 chromosome. **(B)** GO enrichment analysis of genes near more open chromatin regions on the 879 A chromosomes. Bar plots show significantly enriched GO terms ( _P_ < 0.05) for molecular 880 function, cellular component, and biological process categories. Bar length represents the 881 number of upregulated genes associated with each GO term, while bar colour indicates the 882 significance level ( _P_ value). 33 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 883 884 **Figure 5. Gene set enrichment analysis of overlapping upregulated RNA-seq genes and genes** 885 **near more open chromatin regions.** **(A)** Gene ontology (GO) enrichment analysis of of 886 overlapping upregulated RNA-seq genes and genes near more open chromatin regions. Bar 887 plots show significantly enriched GO terms ( _P_ < 0.05) for molecular function, cellular 888 component, and biological process categories. Bar length represents the number of 889 upregulated genes associated with each GO term, while bar colour indicates the significance 890 level ( _P_ value). **(B)** UpSet plot showing intersection of the genes located near the more open 891 chromatin regions in the presence of the B chromosome (ATACseq) and genes upregulated in 892 the presence of the B chromosome (RNAseq). The localization of genes on A chromosomes 893 and B chromosome is indicated by corresponding suffix. 34 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.09.765325; this version posted November 11, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.09.765325) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 894 **Tables** 895 **Table 1: HOMER known motif enrichment analysis of ATAC-seq regions showing increased** 896 **chromatin accessibility in the presence of B chromosomes.** 897 The 13 highest-ranked known motifs identified by HOMER are shown for regions with 898 significantly increased accessibility. Motif occurrence in the target regions (n = 2,907) was 899 compared with the background set of accessible chromatin regions (n = 132,729). Motifs are 900 ranked according to enrichment _P_ -value. Target and background frequencies indicate the 901 proportion of sequences containing each motif, and enrichment was calculated as the ratio of 902 target to background motif frequencies. 35