Genome-wide identification of the PIF gene family in broomcorn millet (Panicum miliaceum) highlights PmPIF5 as a candidate regulator of photoperiod responses Authors: Junjie Wang, Xiuyan Cui, Yiting Zhang, Chunyang Dai, Wei Zhou, Xiang Tian, Ling Chen, Zhijun Qiao, Haigang Wang Published: 2026 ABSTRACT AbstractBackgroundPhytochrome-interacting factors (PIFs) are central bHLH transcription factors that link light perception to developmental programs and stress response. Although PIFs have been characterized in several model and crop species, their family composition and potential roles in broomcorn millet remain unknown.ResultsBased on the AJ08 reference genome, publicly available transcriptome datasets, and resequencing variants from 516 accessions, we conducted an integrated analysis combining genome-wide identification, comparative genomics, expression profiling, RT–qPCR validation, and haplotype analysis. We identified 15 PmPIF genes that were unevenly distributed across 11 chromosomes and could be classified into four phylogenetic groups (Groups 1–4). All PmPIF proteins harbored a conserved bHLH domain and were predicted to localize in the nucleus; secondary structure prediction indicated that α-helices and random coils were predominant. Promoter analyses revealed abundant cis-acting elements related to light responsiveness, phytohormone signaling, and stress responses. Expression analyses under contrasting photoperiod treatments showed pronounced divergence among PmPIF members, with PmPIF5 consistently showing differential expression across multiple comparisons. KEGG enrichment implicated PmPIF5 in the circadian rhythm–plant pathway, and correlation analyses revealed significant associations between PmPIF5 and several light signaling–related genes, including PRR73, UVR8, GHD7, and COL13. Population-level haplotype analysis of PmPIF5 detected five haplotypes among the 516 accessions, with Hap1 being the predominant haplotype (74.42%).ConclusionsThis study provides a comprehensive resource for the PmPIF gene family in broomcorn millet and suggests that PmPIF5 is a promising candidate gene involved in photoperiod-related regulation. These findings lay the foundation for functional validation and support molecular breeding for photoperiod adaptation in broomcorn millet.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-026-08902-2.Keywords: Broomcorn millet, Circadian rhythm, Haplotype, Photoperiod, PIF(phytochrome-interacting factor), Transcriptome Background Broomcorn millet (Panicum miliaceum L.) is characterized by strong drought tolerance, adaptation to infertile soils, and a short growth cycle, and is therefore an important economic crop and a strategic reserve crop for disaster relief in the arid and semi-arid regions of northern China [1–3]. Following natural disasters such as flooding, drought, or hail, broomcorn millet can be replanted to efficiently exploit limited hydrothermal resources and achieve rapid yield formation, thereby contributing to regional food security [4–6]. Light is a pivotal environmental cue that fuels photosynthesis and broadly regulates photomorphogenesis, biological rhythms, and stress responses. Plants perceive light signals across different wavelengths through multiple photoreceptors, including phytochromes, cryptochromes, phototropins, and UV RESISTANCE LOCUS 8 (UVR8), among these, phytochromes are one of the earliest identified and most extensively characterized photoreceptor families [7, 8]. Phytochrome-interacting factors (PIFs) are a subgroup of basic helix–loop–helix (bHLH) transcription factors that constitute the core components of phytochrome signaling. PIF proteins typically harbor a conserved bHLH domain at the C-terminus, enabling them to bind G-box or E-box motifs in target promoters and modulate downstream gene expression [9]. Importantly, PIFs function beyond being simple downstream phytochrome effectors. Accumulating evidence indicates that PIFs act as multifaceted integrators [10] that participate in regulatory circuits involving the circadian clock (e.g., PRR genes) [11, 12], UV-B signaling (e.g., UVR8) [13], and photoperiod-sensitive regulators (e.g., GHD7) [14, 15]. In Arabidopsis thaliana, eight PIF proteins have been identified, all of which contain a DNA-binding bHLH domain [16]. This domain mediates homo- or heterodimerization with other bHLH proteins and enables DNA binding [17, 18]. In addition, the N-terminus of PIFs often includes a relatively conserved active phytochrome-binding (APB) motif, which is essential for interaction with the Pfr form of phytochrome B. PIFs have also been reported as key regulators in diverse species, with roles in chlorophyll biosynthesis [19], starch degradation [20], stomatal movement [21], organ development [22], abiotic stress responses [23–26], and hormone signaling pathways [27–29]. Moreover, PIF expression is frequently under circadian control and displays robust diurnal oscillations, reflecting the integration of light signaling with endogenous rhythmic programs [30–32]. While the PIF gene family has been systematically investigated in Arabidopsis, rice, and maize, a genome-wide understanding of PIFs in broomcorn millet, including their family composition, expression regulation, and potential functional mechanisms, remains limited. To address this gap, we integrated genomic, transcriptomic, and resequencing resources to perform a genome-wide identification of broomcorn millet PIF genes (PmPIFs) and systematically characterized their physicochemical properties, conserved motifs, phylogenetic relationships, and promoter cis-acting elements. We further examined PmPIF expression patterns under contrasting photoperiod regimes using RT–qPCR, and combined KEGG pathway enrichment and haplotype analyses to explore their putative pathway involvement and coordinated regulatory associations with photoperiod-related genes. In particular, we evaluated the potential breeding value of PmPIF5 in plants. This study provides a foundation for deciphering the molecular basis of photoperiodic responses and for developing molecular breeding strategies for broomcorn millet. Materials and methods Plant materials and photoperiod treatments The broomcorn millet cultivar AJ08 (pin a.i. shu 1 hao) was used in this study. The variety was obtained from the Medium-term Germplasm Resource Bank for National Characteristic Grain Crops (Taiyuan), and was officially approved as a new variety in Shanxi Province in 2025. Plants were grown in controlled-environment growth chambers in 2024 under three photoperiod regimes: long day (C16, 16 h light/8 h dark; lights on 05:00–21:00), short day (D8, 8 h light/16 h dark; lights on 09:00–17:00), and intermediate day (CK, 14 h light/10 h dark; lights on 07:00–21:00). The environmental conditions were maintained at 25 °C and 60% relative humidity. Plants were cultivated in uniform polyethylene pots (10 cm × 10 cm), with 10 fully developed seeds sown per pot, and photoperiod treatments were imposed immediately after sowing. At the two-leaf stage, five uniform individuals were retained per pot. Five biological replicates were established for each photoperiod treatment (15 pots). When leaf tissue was harvested at the five-leaf-and-one-heart stage, samples were collected at the same absolute zeitgeber time 4 h (ZT4) across all photoperiod treatments. Samples were excised rapidly, immediately frozen in liquid nitrogen, and stored at − 80 °C for subsequent RNA sequencing and quantitative analyses of PmPIF family members. Identification of PIF genes and chromosomal localization Arabidopsis PIF protein sequences were retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/) and used as queries for initial homology searches using BLAST implemented in TBtools (v2.371). To minimize the risk of missing candidate members, a hidden Markov model (HMM)–based search was subsequently performed. HMM profiles of PIFs were downloaded from the Pfam database (http://pfam.xfam.org/). TBtools software was used to search for PIF homologs in the broomcorn millet genome (P < 0.001). All obtained sequences were subsequently validated using SMART (http://smart.embl.de/) and NCBI-CDD search (https://www.ncbi.nlm.nih.gov/cdd/) with default parameters, to ensure that the bHLH and APB domains were present in the corresponding PIF protein sequences. All non‑redundant genes containing these domains were confirmed as members of the PIF gene family in broomcorn millet, and the complete set of family members was ultimately determined. Chromosome length information and physical positions of PmPIF genes were extracted using TBtools and visualized to generate a chromosomal distribution map. Physicochemical properties, gene structures, and conserved motif analyses Protein length, molecular characteristics (e.g., amino acid number and theoretical isoelectric point), and other physicochemical parameters of PmPIF proteins were predicted using the “Protein Parameter” function in TBtools. Conserved motifs were identified using the MEME suite, with the maximum number of motifs set to 15 and other parameters kept at default settings; results were exported in XML format. The genome annotation file (GFF3) was downloaded from the NCBI. Gene structures, motif compositions, and phylogenetic trees were integrated and visualized using TBtools, and the final figures were refined using Adobe Illustrator. Phylogenetic analysis of PIF proteins from broomcorn millet, Arabidopsis, and rice PIF protein sequences from broomcorn millet, Arabidopsis, and rice were aligned using ClustalW in MEGA12. A phylogenetic tree was constructed using the neighbor-joining (NJ) method based on multiple sequence alignments. The resulting tree was visualized and annotated using iTOL (https://itol.embl.de/). Synteny and collinearity analyses Genome FASTA and GFF annotation files for broomcorn millet, rice, and Arabidopsis were obtained from the NCBI and Ensembl Plants (https://plants.ensembl.org/index.html). Species and inter-species collinearity analyses were performed using MCScanX, which was implemented in TBtools. Collinearity plots were generated using TBtools and polished using Adobe Illustrator. Promoter cis-acting element analysis For each PmPIF gene, a 2,000 bp sequence upstream of the transcription start site was extracted using TBtools (v2.351) and submitted to PlantCare (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) for the prediction of cis-acting element. Cis-elements related to light responsiveness, phytohormone signaling, and stress responses were retained and visualized using TBtools. RNA sequencing and transcriptome analysis RNA-seq was performed by Shanghai Paisennuo Biotechnology Co., Ltd. Total RNA was extracted from samples representing the two treatment conditions, with three biological replicates per condition. RNA concentration and purity were assessed using a NanoDrop spectrophotometer, and RNA integrity was evaluated using an Agilent 2100 Bioanalyzer. Poly(A)+ mRNA was enriched using the NEBNext Poly(A) mRNA Magnetic Isolation Module, followed by cDNA library construction. Libraries were quality-checked using an Agilent 2100 Bioanalyzer and sequenced using an Illumina platform to generate paired-end reads. Clean reads were aligned to the reference genome [33]. Transcript assembly was performed using Cufflinks, and assembled transcripts were compared with existing gene models. Sequences shorter than 200 bp were removed. Read counts mapped to each gene were obtained using HTSeq (v0.9.1) [34], and the expression levels were normalized as FPKM. Differentially expressed genes (DEGs) were identified using the thresholds of |log2(fold change)| > 1 and P < 0.05. GO enrichment was conducted using topGO based on the hypergeometric test, with P < 0.05 considered significant. KEGG pathway enrichment was performed using clusterProfiler (v3.4.4). Pathways with P < 0.05 were prioritized, and the top 20 enriched pathways shared across the treatments were summarized. Upregulated and downregulated genes within these pathways were examined to identify the most strongly enriched biological processes. Quantitative real-time PCR (RT–qPCR) validation RT–qPCR was conducted to quantify the expression of PmPIF family members and to validate the RNA-seq results. Total RNA was extracted using an RNA extraction kit (Tiangen Biotech, Beijing, China). RNA concentration was measured using an ultramicro UV spectrophotometer, integrity was checked on a 1% agarose gel, and cDNA was synthesized by reverse transcription. RT–qPCR reactions (20 µL) contained 10 µL 2×SuperReal PreMix Plus, 0.6 µL forward and reverse primers, 0.8 µL 50×ROX reference dye, 4 µL cDNA, and 4 µL RNase-free ddH2O. The cycling program was as follows: 95 °C for 15 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 32 s, with a final extension at 72 °C for 10 min. Actin (ID: 120667124) was used as the internal reference. Primers were synthesized by Sangon Biotech (Shanghai, China) (Supplementary Table 1). Each assay included three biological and three technical replicates. Relative expression levels were calculated using the 2^−ΔΔCt method. Statistical analyses and plotting were performed using GraphPad Prism (v10.1.2). Haplotype analysis Haplotype variation was analyzed using resequencing data from 516 broomcorn millet accessions. Haplotype calling and visualization were conducted as described by Zhou et al. [35]. using RStudio (https://www.rstudio.com/products/rstudio/download/) and Prism software. Heading date phenotypes were integrated to assess phenotypic differences among haplotypes, and statistical significance was evaluated using t-tests. The results are presented as boxplots. Statistical analysis One-way analysis of variance (ANOVA) was performed using SPSS (v22.0). Figures were generated using Origin 2024 (v10.1), GraphPad Prism (v10.4.0), and R (v4.50). Volcano and correlation plots were generated using R (v4.50), and KEGG-related visualizations were generated using the ggplot2 package. Results Physicochemical properties of PmPIF proteins The physicochemical features of the broomcorn millet PIF proteins are presented in Table 1. The predicted protein lengths ranged from 186 to 526 amino acids, with PmPIF1, PmPIF10, and PmPIF11 being the shortest and PmPIF8 being the longest. The molecular weights ranged from 21,287.17 to 56,253.49 Da, with PmPIF11 and PmPIF8 showing the minimum and maximum values, respectively. The theoretical isoelectric points (pI) ranged from 5.24 to 9.33, indicating that the PmPIF family includes both acidic and basic proteins, which may influence their charge states and interaction potential under different pH conditions. All PmPIF proteins exhibited instability indices > 40, suggesting that they are unstable in vitro. The aliphatic index ranged from 55.51 to 79.78, and the grand average of hydropathicity (GRAVY) values were < 1, indicating that PmPIF proteins are predominantly hydrophilic. Subcellular localization prediction consistently placed all PmPIF proteins in the nucleus. Table 1The physicochemical properties of the broomcorn millet PIF gene familyGene nameGene numberNumber of Amino AcidMolecular WeightTheoretical pIInstability IndexAliphatic IndexGrand Average of HydropathicitySubcellularlocalization PmPIF1 AJ81AG3303028931869.147.1678.5477.68-0.524Nucle PmPIF2 AJ81BG3637023025433.897.8479.2379.78-0.545Nucle PmPIF3 AJ82AG0232043646840.669.1173.6366.17-0.588Nucle PmPIF4 AJ82BG0335040041845.655.2474.1363.38-0.446Nucle PmPIF5 AJ83AG0818050152010.076.2865.0768.24-0.276Nucle PmPIF6 AJ83AG3690044647568.636.2476.1656.7-0.519Nucle PmPIF7 AJ83BG0985029730830.187.1870.161.45-0.318Nucle PmPIF8 AJ83BG4107052656253.498.970.6455.51-0.49Nucle PmPIF9 AJ84AG2145032335327.345.7178.2864.74-0.737Nucle PmPIF10 AJ87AG2233018621303.179.3393.4761.88-0.902Nucle PmPIF11 AJ87BG2363018621287.179.3393.7361.88-0.9Nucle PmPIF12 AJ89AG0586042044,8795.859.9960.69-0.591Nucle PmPIF13 AJ89AG1439050252116.775.9167.5465.48-0.291Nucle PmPIF14 AJ89BG0647042044818.985.9659.462.07-0.589Nucle PmPIF15 AJ89BG1603052053789.495.8671.261.73-0.366Nucle The physicochemical properties of the broomcorn millet PIF gene family Secondary and tertiary structure features of PmPIF proteins Secondary structure prediction using SOPMA showed that PmPIF proteins comprise α-helices, extended strands, β-turns, and random coils (Supplementary Table 2). Among these elements, α-helices were the predominant component, with PmPIF2 containing the highest proportion of α-helices and PmPIF3 containing the lowest. Tertiary structure modeling using SWISS-MODEL suggested that PmPIF10 and PmPIF11 share highly similar 3D conformations, and both display a relatively high α-helix content compared with other family members (Fig. 1). Tertiary structure sprediction of PIF protein in broomcorn millet Chromosomal distribution of PmPIF genes Chromosomal mapping revealed that the 15 PmPIF genes were unevenly distributed across 11 chromosomes (Fig. 2). Chromosomes 3 A, 3B, 9 A, and 9B each contained two PmPIF genes, whereas the remaining chromosomes harbored only one. Chromosome distribution of PIF gene family in broomcorn millet Phylogenetic relationships of PIF proteins across species To elucidate the evolutionary relationships of broomcorn millet PIFs, we constructed a phylogenetic tree using PIF proteins from Panicum miliaceum, Arabidopsis thaliana, Oryza sativa, Zea mays, Setaria italica, and Panicum virgatum (Fig. 3). The PIF family was clustered into four major groups. Group 1 contained five PmPIFs, three PvPIFs, three ZmPIFs, one OsPIF, and one AtPIF. Group 2 included four PmPIFs, one PvPIF, two ZmPIFs, two OsPIFs, two AtPIFs, and one SiPIF. Group 3 comprised six PmPIFs, four PvPIFs, one ZmPIF, two OsPIFs, one AtPIF, and two SiPIFs, representing the largest clade (16 members) in this study. Group 4 contained four AtPIFs, three OsPIFs, and one ZmPIF, but no broomcorn millet members, suggesting lineage-specific expansion and/or loss. Evolutionary tree of PIF proteins in broomcorn millet, Arabidopsis, Oryza sativa、 Zea Mays、Setaria italica、Panicum virgatum. Pm: Panicum miliaceum At: Arabidopsis thaliana;Zm: Zea mays, Os: Oryza sativ, Si: Setaria italic, Pv: Panicum virgatum Conserved motifs, domains, and protein structures of PmPIFs Motif analysis indicated that all 15 PmPIF proteins contained Motifs 1 and Motifs 2, whereas Motif 7 was present in 12 members (Fig. 4), suggesting that these motifs represent the core conserved features of the family. Domain annotation showed that most PmPIF proteins harbored a bHLH domain at the C-terminus. Notably, PmPIF13 uniquely contained an additional APB domain at the N-terminus, together with a C-terminal bHLH domain, whereas the remaining members carried only the bHLH domain. Gene structure analysis revealed variations in exon–intron organization, with the number of exons ranging from 5 to 8. PmPIF1 and PmPIF3 had the highest number of exons and introns (8 exons and 7 introns), whereas other genes contained 5–7 exons and 3–6 introns. The marked differences in coding sequence length and predicted protein size among PmPIFs suggest potential functional diversification within this family. Conserved motifs in broomcorn millet PIF gene family Cis-acting elements in PmPIF promoters Cis-element analysis of the 2-kb promoter regions upstream of the PmPIF transcription start sites revealed a diverse repertoire of regulatory elements (Fig. 5). Light-responsive elements and multiple hormone- and stress-related elements were highly represented, including motifs associated with methyl jasmonate (MeJA), abscisic acid (ABA), and gibberellin (GA) responses. In addition, PmPIF3, PmPIF4, PmPIF5, PmPIF7, PmPIF12, and PmPIF14 contained putative MYB-binding sites, consistent with the role of MYB transcription factors in regulating light responses, secondary metabolism, stress responses, and hormone signaling. All PmPIF genes contain light-responsive elements that can regulate flowering time, floral organ morphology, and the overall flowering process in broomcorn millet. Collectively, these promoter features suggest that PmPIF genes may participate in light-regulated growth and development, and potentially contribute to hormone-mediated and stress-responsive processes. Cis-acting elements prediction in the PIF promoter of broomcorn millet Synteny and duplication patterns of the PmPIF family Gene duplication is a major force driving gene family expansion and functional diversification. In Panicum miliaceum, synteny analysis detected 16 collinear gene pairs among the 15 PmPIF genes (Fig. 6), suggesting that segmental duplication has substantially contributed to the evolution of this family. Inter-species synteny analysis further identified 24 collinear PIF gene pairs between Panicum miliaceum and Oryza sativa (Fig. 7), indicating considerable conservation of PIF loci across monocot grasses and supporting a close evolutionary relationship between Panicum miliaceum and Oryza sativa at the genomic level. Tra-species collinearity of the PIF gene family Collinearity between PIF gene families between Panicum miliaceum .L and Arabidopsis and Oryza sativa Transcriptome profiling under different photoperiod regimes identifies PmPIF5 as a consistently responsive gene RNA-seq analysis identified 5,645, 9,276, and 10,225 differentially expressed genes (DEGs) in the CK_vs_C16, CK_vs_D8, and D8_vs_C16 comparisons, respectively (Fig. 8). Within the PIF family, PmPIF5 and PmPIF6 were significantly downregulated in CK_vs_C16. In CK_vs_D8, PmPIF12 and PmPIF14 were downregulated, whereas PmPIF5, PmPIF7, and PmPIF8 were upregulated. In D8_vs_C16, PmPIF5, PmPIF6, PmPIF7, PmPIF8, and PmPIF9 were downregulated, whereas PmPIF12 and PmPIF14 were upregulated. The remaining genes (PmPIF1–4, PmPIF10–11, PmPIF13, and PmPIF15) did not show significant changes between comparisons. Notably, PmPIF5 (AJ83AG08180) was differentially expressed in all three comparisons, highlighting it as a strong candidate regulator involved in photoperiod responses in broomcorn millet (Fig. 8). Differential expression of PIF family members across photoperiod comparisons KEGG-based clustering implicates PmPIF5 in signal transduction modules To explore the pathways associated with PmPIF5, KEGG enrichment was performed using significantly enriched pathways (p < 0.05) across comparisons. The results indicated that PmPIF5 was primarily associated with signal transduction–related pathways (Supplementary Table 3), which included 118 DEGs. Hierarchical clustering of these DEGs revealed distinct expression patterns across photoperiod treatments and separated them into nine clusters (G-C1 to G-C9), each containing 4–19 genes (Fig. 9). For example, genes in G-C7 showed low expression in CK, strong upregulation under D8, and subsequent downregulation under C16. PmPIF5 belonged to the G-C7 clade, which contained 17 genes. In contrast, G-C2 displayed the opposite trend, with higher expression in CK, repression under D8, and restoration under C16 (Fig. 9). Clustering of DEGs in the signal transduction pathway, including PmPIF5 Co-expression analysis links PmPIF5 with photoperiod and light signaling–related genes To investigate the correlation between PmPIF5 (AJ83AG08180) and other genes, we selected 30 genes with the highest absolute expression correlation (|r|) with PmPIF5 (Fig. 10). The strongest negative correlation was observed with AJ89AG40440 (r = − 0.993), which encodes PRR73, a two-component response regulator implicated in photoperiod control. The strongest positive correlation was observed with AJ85BG17780 (r = 0.990), which encodes UVR8, a UV-B photoreceptor. These results suggest that PmPIF5 may act in concert with key light signaling components, including PRR73 (AJ89AG40440, AJ89BG44540), UVR8 (AJ85BG17780, AJ85AG16210), GHD7 (AJ89AG29410, AJ89BG32350), and COL13 (AJ89BG40420, AJ89AG36580), to coordinate photoperiod responses and thereby influence heading and flowering processes in broomcorn millet. Correlation analysis between PmPIF5 and photoperiod-related genes Haplotype variation of PmPIF5 and its association with heading date Resequencing of 516 broomcorn millet accessions identified seven SNPs within PmPIF5, defining five haplotypes (Hap1–Hap5) (Fig. 11A). Hap1 was the predominant haplotype, present in 74.42% of the accessions. Hap2 carried variants at two SNP sites, whereas the remaining haplotypes differed from Hap1 at only one SNP. Statistical comparisons were restricted to Hap1–Hap5. Observational data from three consecutive years (2019–2021) revealed significant differences in the average heading date among different haplotype cultivars, with Hap3 exhibiting significantly longer heading dates than other haplotypes across all years (Fig. 11B). Specifically, the heading dates for Hap1 to Hap5 were ~ 52 d, ~ 50 d, ~ 84 d, ~ 45 d, and ~ 48 d in 2019; ~57 d, ~ 49 d, ~ 84 d, ~ 48 d, and ~ 52 d in 2020; and ~ 46 d, ~ 39 d, ~ 74 d, ~ 39 d, and ~ 44 d in 2021, respectively. Although the heading date of the same haplotype fluctuated interannually, Hap3 consistently displayed the latest heading, suggesting that this haplotype is closely associated with delayed heading. Haplotype structure of PmPIF5 in 516 accessions and haplotype–phenotype association for heading date. Note: Different lowercase letters represents significant difference (P < 0.05) RT–qPCR validation of photoperiod-responsive expression patterns RT–qPCR profiling under different photoperiod regimes showed expression trends consistent with the RNA-seq results, supporting the reliability of transcriptome analysis. Under short-day conditions (8 h), the expression of PmPIF1–PmPIF9, PmPIF13, and PmPIF15 was significantly higher than that under the intermediate-day control (14 h), whereas that of PmPIF12 and PmPIF14 were significantly downregulated. Under long-day conditions (16 h), PmPIF4 and PmPIF15 were upregulated relative to the control, whereas PmPIF5, PmPIF6, and PmPIF12 were significantly downregulated. Compared with short-day conditions, PmPIF1–PmPIF3, PmPIF5–PmPIF9, and PmPIF13 were significantly downregulated under long-day conditions, whereas PmPIF4, PmPIF12, and PmPIF14 were upregulated. Together, these results indicate that PmPIF genes are broadly responsive to photoperiod changes and may contribute to photoperiod-mediated regulation in broomcorn millet (Fig. 12). Expression levels of the PIF gene in broomcorn millet under different light condition. Note: ****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant (P ≥ 0.05) Discussion Evolutionary expansion and structural conservation of the PmPIFs family PIFs are core members of the bHLH transcription factor superfamily, and individual light responses are often co-regulated by multiple PIFs acting in combination [36–38]. Here, we identified 15 PmPIF genes in broomcorn millet and systematically characterized their sequence features, evolutionary relationships, promoter architecture, and photoperiod-responsive expression patterns, with an additional emphasis on the population genetic architecture of PmPIF5. This study expands the PIF gene resources in crops and provides a framework for dissecting photoperiod sensitivity and genetic improvement in broomcorn millet. The number of PIF genes in broomcorn millet (15) exceeds that reported in Arabidopsis (8) [39], Brachypodium distachyon (7) [40], potato (7) [41], rice (6) [42], and maize (7) [43]. Their uneven chromosomal distribution and the presence of closely positioned, highly similar loci are consistent with genome-wide duplication and/or segmental duplication events during broomcorn millet evolution [40, 44]. Such duplications may increase the complexity of the regulatory network and facilitate adaptation to fluctuating light environments by retaining additional PIF paralogs. Phylogenetic analysis grouped PmPIFs into three major clades (Groups I–III), consistent with the classifications reported for pepper and rice [45, 46]. All PmPIF proteins contained the conserved bHLH domain (corresponding to Motif 1), a hallmark of PIF transcription factors [47], supporting the strong structural conservation of this family. Notably, only PmPIF13 carries an APB domain, suggesting a potentially specialized capacity for phyB interaction [48]. Exon–intron organization showed both conservation and divergence: several members (e.g., PmPIF3, PmPIF4, and others) appeared structurally conserved, whereas others (e.g., PmPIF1 and PmPIF2) may have experienced exon loss, implying possible functional divergence. The nuclear localization of all PmPIF proteins is consistent with previous studies in Arabidopsis, rice, and maize [49–51]. The predominance of α-helices may stabilize dimerization interfaces, whereas flexible coils allow conformational adaptability to environmental cues [52]. Promoter architecture supports integration of light, hormone, and stress cues Transcriptional regulation is strongly influenced by promoter architecture, particularly by cis-acting elements upstream of the coding regions [53]. The PmPIF promoters contained abundant light-responsive elements (e.g., G-box and GT1-motif) as well as multiple hormone-responsive motifs, including ABA-, GA-, MeJA-, and salicylic acid–related elements. This composite regulatory architecture supports the notion that PmPIFs may act as hubs that integrate light cues with endogenous hormone signaling and stress responses. In Arabidopsis, AtPIF1 modulates seed germination through the ABA and GA pathways [54, 55], whereas AtPIF4 and AtPIF5 integrate light and auxin signaling to regulate rhythmic growth [56]. In tomatoes, SlPIF4 enhances chilling tolerance by inducing JA- and ABA-related biosynthetic genes [57]. Together with the MYB-binding motifs observed in several PmPIF promoters, and the report that MYB30 can negatively regulate photomorphogenesis by promoting PIF4 accumulation [58], our data suggest that PmPIFs may participate in coordinated regulation involving hormone pathways and MYB transcription factors to fine-tune growth and stress responses in broomcorn millet [59]. This “light–hormone–stress” regulatory network may contribute to the ability of broomcorn millet to maintain reproductive success under drought and low-fertility conditions. In addition, the hypersensitivity of Arabidopsis pif1/3/4/5 mutants to high-light stress [60] raises the possibility that PmPIFs also contribute to photooxidative stress protection, consistent with observations that shade conditions can enhance drought tolerance [61]. PmPIF5 as a candidate node connecting photoperiod sensitivity and circadian regulation Broomcorn millet is a typical short-day crop, and photoperiod responses directly shape heading time and biomass accumulation. We observed pronounced photoperiod-dependent expression differences among PmPIF genes, supporting their involvement in diurnal and developmental regulation [62, 63]. Among them, PmPIF5 stood out as being consistently differentially expressed across all photoperiod comparisons. KEGG analysis further implicated PmPIF5-associated DEGs in the “circadian rhythm–plant” pathway. In model systems, PIF4 and PIF5 jointly regulate shade avoidance and morning gene expression dynamics [64]. The strong photoperiod responsiveness of PmPIF5 suggests that it may function as a candidate regulator that connects light signals with developmental timing, potentially via interactions with circadian oscillator components such as PRR73, which showed an exceptionally strong negative correlation with PmPIF5 in our data [65, 66]. In addition, the well-established antagonistic dynamics between phytochromes and PIFs (PIF accumulation in darkness versus light-triggered ubiquitination and degradation) provide a mechanistic framework for explaining the rapid transcriptional reprogramming observed under changing photoperiods [67, 68]. Population genetic variation of PmPIF5 and implications for breeding Haplotype analysis of 516 accessions revealed five haplotypes at PmPIF5, indicating detectable sequence polymorphisms at this locus. The dominance of Hap1 (74.42%) may reflect strong positive or purifying selection during domestication and improvement, whereby alleles favorable for agronomic performance and ecological adaptation are preferentially retained. The widespread prevalence of Hap1 across accessions suggests that it may confer advantages related to stable circadian regulation and broad adaptation. Functional validation of PmPIF5 alleles/haplotypes using genome editing approaches (e.g., CRISPR/Cas9), together with photoperiod/heading assays, will help to establish causal links between PmPIF5 variation and photoperiod-related traits, thereby facilitating precision breeding in broomcorn millet [69]. This study systematically identified the PmPIF gene family in broomcorn millet for the first time, revealing the expression profiles, haplotype variation of PmPIF5, and its associations with key photoperiod-related genes, thereby providing important candidate genes and a theoretical basis for understanding photoperiod regulation in broomcorn millet. However, several limitations remain due to the current stage of research. The RNA‑seq analysis was based on limited time points; more precise circadian regulation patterns would require dense time‑course sampling or free‑running experiments under constant light conditions. The observed expression correlations involving PmPIF5 gene/PmPIF5 protein require further validation through additional genetic interaction assays, such as yeast two-hybrid (Y2H), bimolecular fluorescence complementation (BiFC), electrophoretic mobility shift assay (EMSA), and chromatin immunoprecipitation sequencing (ChIP-seq), as well as complementary biochemical assays. The haplotype‑phenotype association is preliminary and should be confirmed in an independent and larger germplasm collection to verify the link between Hap3 and delayed heading. All inferences regarding PmPIF5 function are currently based on expression correlations and haplotype associations, and genetic causality has not been established. Future studies should combine dense time‑series sampling, molecular interaction techniques, and genetic transformation systems (e.g., CRISPR/Cas9) to thoroughly validate the function of this gene. Conclusions In this study, we identified 15 PmPIF genes in broomcorn millet at the genome-wide level. These genes are distributed across 11 chromosomes and encode proteins containing the conserved bHLH domain. Comparative analyses revealed substantial collinearity with rice PIF loci, supporting the conservation of PIF genes across grasses. Promoter analyses indicated that PmPIF genes are enriched for cis-elements associated with light responses, abiotic stress, and phytohormone signaling. Transcriptome profiling and RT–qPCR validation demonstrated that multiple PmPIF members respond to photoperiod changes, with PmPIF5 consistently differentially expressed across comparisons and linked to circadian rhythm–related pathways. Co-expression patterns further connected PmPIF5 with key photoperiod and light signaling genes (e.g., PRR73, UVR8, and GHD7), and haplotype analysis suggested that PmPIF5 variation may be associated with heading-date differences among accessions. Collectively, these findings provide a systematic framework for understanding PIF-mediated light signaling in broomcorn millet and offer candidate targets for breeding early maturing and broadly adapted cultivars with the desired phenotypes. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Authors’ contributions WJJ: writing – original draft, software, Investigation, analysis, conceptualization, editing. CXY: methodology, software, data curation. ZYT: Investigation, data curation. DCY: methodology, investigation. ZW: software, data curation. TX: Data curation, Conceptualization. CL: Resources, Data curation, Conceptualization. QZJ: Visualization, Resources. WHG: review & editing Supervision, Resources. Funding This work was supported by the Central Guiding Local Science and Technology Development Fund Project (YDZJSX20231A035). Data availability Sequence data that support the findings of this study have been deposited in the NCBI with the primary accession code PRJNA1457142.The datasets used and/or analysed during the current study available from the first author or corresponding author on reasonable request. [https://www.ncbi.nlm.nih.gov/bioproject/1457142]. Declarations Ethics approval and consent to participate The experiments in this study did not involve endangered or protected species. No specific permits were required for these locations/activities. We declare that all the materials and methods in this study complied with relevant institutional, national, and international guidelines and legislation. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Data Availability Statement Sequence data that support the findings of this study have been deposited in the NCBI with the primary accession code PRJNA1457142.The datasets used and/or analysed during the current study available from the first author or corresponding author on reasonable request. [https://www.ncbi.nlm.nih.gov/bioproject/1457142].