Background The calmodulin-regulated spectrin-associated proteins (CAMSAPs) belong to a conserved protein family, which includes members that bind the polymerizing mcrotubule (MT) minus ends and remain associated with the MT lattice formed by minus end polymerization. Only one of the three mammalian CAMSAPs, CAMSAP1, localizes to the mitotic spindle but its function is unclear. In Drosophila , there is only one CAMSAP, named Patronin. Previous work has shown that Patronin stabilizes the minus ends of non-mitotic MTs and is required for proper spindle elongation. However, the precise role of Patronin in mitotic spindle assembly is poorly understood. Results Here we have explored the role of Patronin in Drosophila mitosis using S2 tissue culture cells as a model system. We show that Patronin associates with different types of MT bundles within the Drosophila mitotic spindle, and that it is required for their stability. Imaging of living cells expressing Patronin-GFP showed that Patronin displays a dynamic behavior. In prometaphase cells, Patronin accumulates on short segments of MT bundles located near the chromosomes. These Patronin “seeds” extend towards the cell poles and stop growing just before reaching the poles. Our data also suggest that Patronin localization is largely independent of proteins acting at the MT minus ends such as Asp and Klp10A. Conclusion Our results suggest a working hypothesis about the mitotic role of Patronin. We propose that Patronin binds the minus ends within MT bundles, including those generated from the walls of preexisting MTs via the augmin-mediated pathway. This would help maintaining MT association within the mitotic bundles, thereby stabilizing the spindle structure. Our data also raise the intriguing possibility that the minus ends of bundled MTs can undergo a limited polymerization. Electronic supplementary material The online version of this article (10.1186/s12860-019-0189-0) contains supplementary material, which is available to authorized users.

Moonlighting proteins can perform one or more additional functions besides their primary role. It has been posited that a protein can acquire a moonlighting function through a gradual evolutionary process, which is favored when the primary and secondary functions are exerted in different cellular compartments. Transcription factors (TFs) and splicing factors (SFs) control processes that occur in interphase nuclei and are strongly reduced during cell division, and are therefore in a favorable situation to evolve moonlighting mitotic functions. However, recently published moonlighting protein databases, which comprise almost 400 proteins, do not include TFs and SFs with secondary mitotic functions. We searched the literature and found several TFs and SFs with bona fide moonlighting mitotic functions, namely they localize to specific mitotic structure(s), interact with proteins enriched in the same structure(s), and are required for proper morphology and functioning of the structure(s). In addition, we describe TFs and SFs that localize to mitotic structures but cannot be classified as moonlighting proteins due to insufficient data on their biochemical interactions and mitotic roles. Nevertheless, we hypothesize that most TFs and SFs with specific mitotic localizations have either minor or redundant moonlighting functions, or are evolving towards the acquisition of these functions.

The Drosophila Nonspecific Lethal (NSL) complex is a major transcriptional regulator of housekeeping genes. It contains at least seven subunits that are conserved in the human KANSL complex: Nsl1/Wah (KANSL1), Dgt1/Nsl2 (KANSL2), Rcd1/Nsl3 (KANSL3), Rcd5 (MCRS1), MBD-R2 (PHF20), Wds (WDR5) and Mof (MOF/KAT8). Previous studies have shown that Dgt1, Rcd1 and Rcd5 are implicated in centrosome maintenance. Here, we analyzed the mitotic phenotypes caused by RNAi-mediated depletion of Rcd1, Rcd5, MBD-R2 or Wds in greater detail. Depletion of any of these proteins in Drosophila S2 cells led to defects in chromosome segregation. Consistent with these findings, Rcd1, Rcd5 and MBD-R2 RNAi cells showed reduced levels of both Cid/CENP-A and the kinetochore component Ndc80. In addition, RNAi against any of the four genes negatively affected centriole duplication. In Wds-depleted cells, the mitotic phenotypes were similar but milder than those observed in Rcd1-, Rcd5- or MBD-R2-deficient cells. RT-qPCR experiments and interrogation of published datasets revealed that transcription of many genes encoding centromere/kinetochore proteins (e.g., cid, Mis12 and Nnf1b), or involved in centriole duplication (e.g., Sas-6, Sas-4 and asl) is substantially reduced in Rcd1, Rcd5 and MBD-R2 RNAi cells, and to a lesser extent in wds RNAi cells. During mitosis, both Rcd1-GFP and Rcd5-GFP accumulate at the centrosomes and the telophase midbody, MBD-R2-GFP is enriched only at the chromosomes, while Wds-GFP accumulates at the centrosomes, the kinetochores, the midbody, and on a specific chromosome region. Collectively, our results suggest that the mitotic phenotypes caused by Rcd1, Rcd5, MBD-R2 or Wds depletion are primarily due to reduced transcription of genes involved in kinetochore assembly and centriole duplication. The differences in the subcellular localizations of the NSL components may reflect direct mitotic functions that are difficult to detect at the phenotypic level, because they are masked by the transcription-dependent deficiency of kinetochore and centriolar proteins.

Centrosome-containing cells assemble their spindles exploiting three main classes of microtubules (MTs): MTs nucleated by the centrosomes, MTs generated near the chromosomes/kinetochores, and MTs nucleated within the spindle by the augmin-dependent pathway. Mammalian and Drosophila cells lacking the centrosomes generate MTs at kinetochores and eventually form functional bipolar spindles. However, the mechanisms underlying kinetochore-driven MT formation are poorly understood. One of the ways to elucidate these mechanisms is the analysis of spindle reassembly following MT depolymerization. Here, we used an RNA interference (RNAi)-based reverse genetics approach to dissect the process of kinetochore-driven MT regrowth (KDMTR) after colcemid-induced MT depolymerization. This MT depolymerization procedure allows a clear assessment of KDMTR, as colcemid disrupts centrosome-driven MT regrowth but not KDMTR. We examined KDMTR in normal Drosophila S2 cells and in S2 cells subjected to RNAi against conserved genes involved in mitotic spindle assembly: mast/orbit/chb (CLASP1), mei-38 (TPX2), mars (HURP), dgt6 (HAUS6), Eb1 (MAPRE1/EB1), Patronin (CAMSAP2), asp (ASPM), and Klp10A (KIF2A). RNAi-mediated depletion of Mast/Orbit, Mei-38, Mars, Dgt6, and Eb1 caused a significant delay in KDMTR, while loss of Patronin had a milder negative effect on this process. In contrast, Asp or Klp10A deficiency increased the rate of KDMTR. These results coupled with the analysis of GFP-tagged proteins (Mast/Orbit, Mei-38, Mars, Eb1, Patronin, and Asp) localization during KDMTR suggested a model for kinetochore-dependent spindle reassembly. We propose that kinetochores capture the plus ends of MTs nucleated in their vicinity and that these MTs elongate at kinetochores through the action of Mast/Orbit. The Asp protein binds the MT minus ends since the beginning of KDMTR, preventing excessive and disorganized MT regrowth. Mei-38, Mars, Dgt6, Eb1, and Patronin positively regulate polymerization, bundling, and stabilization of regrowing MTs until a bipolar spindle is reformed.