Figures
Abstract
Genes in multicellular organisms are expressed as part of a developmental program that is largely dependent on self-perpetuating higher-order chromatin states. The mechanism of establishing and maintaining these epigenetic events is well studied in Drosophila. The first known example of an epigenetic effect was that of (PEV) in Drosophila, which has been shown to be due to gene silencing via heterochromatin formation. We are investigating a process similar to Position Effect Variegation (PEV) using a mini-w transgene, called Pci, inserted in the upstream regulatory region of ci. The mini-white+ transgene in Pci is expressed throughout the adult eye; however, when other P or KP elements are present, a variegated eye phenotype results indicating random w+ silencing during development. This P element dependent silencing (PDS) can be modified by the haplo-suppressors/triplo-enhancers, Su(var)205 and Su(var)3–7, indicating that these heterochromatic modifiers also act dose dependently in PDS. Here we use a spontaneous derivative mutation of Pci called PciE1 (E1) that variegates like PDS in the absence of P elements, presumably due to an adjacent gypsy element insertion, to screen for second-site modifier mutations that enhance variable silencing of white+ in E1. We isolated 7 mutations in CG8878, an essential gene, that enhance the E1 variegated phenotype. CG8878, a previously uncharacterized gene, potentially encodes a serine/threonine kinase whose closest Drosophila paralogue, ballchen (nhk-1), phosphorylates histones. These mutant alleles enhance both PDS at E1 and Position Effect Variegation (PEV) at wm4, indicating a previously unknown common silencing mechanism between the two.
Citation: McCracken A, Locke J (2014) Mutations in CG8878, a Novel Putative Protein Kinase, Enhance P Element Dependent Silencing (PDS) and Position Effect Variegation (PEV) in Drosophila melanogaster. PLoS ONE 9(3): e71695. https://doi.org/10.1371/journal.pone.0071695
Editor: Barbara Jennings, University College London, United Kingdom
Received: April 17, 2013; Accepted: July 4, 2013; Published: March 10, 2014
Copyright: © 2014 McCracken, Locke. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by funding from a Discovery Grant from the Natural Sciences and Engineering Research Council (NESRC) of Canada (http://www.nserc-crsng.gc.ca). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
In Drosophila melanogaster, expression of the white+ gene (w+) is cell autonomous, and necessary for the import of pigment precursors for normal colour in the adult eye. In white− mutants, the absence of pigment results in a white-eyed phenotype that can be rescued with w+ containing P-element transgenes. However, in some insertion locations, expression of w+ is sensitive to the local chromatin environment, such as adjacent heterochromatin. For example, P{lacW}ciDplac (hereafter Pci), a transgene inserted proximally on chromosome 4 between Ribosomal protein S3A (RpS3A) and cubitus interruptus (ci), was originally isolated as an enhancer trap of ci [1]. Its w+ minigene is sensitive to changes in gene dosage for heterochromatin proteins HP1 and SU(VAR)3–7 [2].
The Pci transgene is also sensitive to the presence of P elements in the genome in a phenomenon called P element dependent silencing (PDS), which is phenotypically similar to heterochromatic position effect variegation (hPEV). In flies lacking P elements (M strains) the w+ transgene is expressed in a uniform manner (even red eye phenotype). However, in flies containing P elements (P strains) or KP s (derivative elements capable of mimicking some of the characteristics of P strains such as modifying P- repressor sensitive alleles, but not enabling P element transposition) variegated expression occurs resulting in a mosaic expression of white ommatidia on a red background in the eye [3]. PDS also occurs when other w+ transgenes are inserted near this location [2].
The random silencing of the w+ minigene in Pci during development indicates a phenomenon similar to heterochromatin spreading in hPEV. This is supported by Pci responding in a dose sensitive manner to Su(var) 205 and 3–7, like wm4 and other centromeric PEV lines [4], [5]. As with PEV, position is important as insertions of the same transgene in other locations do not display PDS. Also, chromosomal translocations of Pci away from its centromere-proximal location reduces PDS (Bushey & Locke 2004)2. The position dependence and variegated phenotype suggest PDS occurs via a similar mechanism as heterochromatic PEV.
During investigation of PDS at Pci two spontaneous mutants, P{lacW}ciDplacE1 and P{lacW}ciDplacE2 (hereafter E1 and E2), were recovered [2] that showed a variegated eye phenotype in the absence of other P elements and a complete white eye phenotype when combined with P elements, such as P{ry+ SalI}89D. PEV modifier loci, such as Su(var)205 and Su(var)3–7, suppressed variegation at E1 and E2. Molecular analysis of E1 and E2 revealed each had a novel gypsy element insertion approximately 1 kb distal from the Pci insert, but in opposite orientations 547 bp apart. Testing of E1 and E2 against mutants of su(Hw) and mod(mdg4) showed that variegation by E1 and E2 was not a result of the gypsy insulator function per se [2]; however, a possible interaction between the gypsy insulator and the wari element at the 3′ end of the w+ transgene can not be ruled out [6]. Both E1 and E2 trans-silence w+ expression of Pci on a paired homolog, but not when present on translocations. Therefore position or pairing contributes to trans-silencing similar to the dominant trans-inactivation of the wild-type homologue in bw+/bwD heterozygotes [2], [7], [8], [9].
A previous screen for genetic modifiers of PDS [2] required tracking two chromosomes. In the work described here we have taken advantage of the E1 variegating PDS system, which is simpler, as all components are on the same chromosome. Furthermore, its intermediate variegated eye phenotype can visibly reveal second site genetic enhancers (white eye phenotype). We describe here the isolation and characterization of seven mutants, that enhance both PDS and hPEV at wm4. This gene, CG8878, was recovered along with mutations in trithorax and ash1 and appears to encode a novel type of kinase similar to the Vaccinia-Related-Kinase(VRK), Casein-Kinase (CK), and Tau-Tubulin-Kinase (TTK).
Materials and Methods
Drosophila stocks and mutations
Unless otherwise cited, D. melanogaster mutations were described previously [10]. The P{lacW}ciDplac (Pci) allele (Eaton & Kornberg 1990)1 is a P{lacZP\T.Ww+mC ampR ori = lacW} construct inserted ∼3 kbp upstream (distal) from the ci locus on chromosome 4. P{lacW}ci DplacE1(E1) has a gypsy element insertion ∼1 kb further upstream as previously described [2]. y1 w* P{lacW}3-76a is a lacW transgene inserted on the X-chromosome and was originally isolated by Y. N. Jan and provided by the Bloomington stock center, while In(1)wm4,wm4 was obtained from K.D. Tartof. Fly stocks were maintained at room temperature on standard yeast/cornmeal medium.
Mutagenesis used w−; dp−; e−; E1 males treated with 25 mM EMS as per [11] mated to y− w−; +/+ virgin females and screened for a dominant enhanced eye colour phenotype in the progeny. Putative mutants were mated to w−; dp−; e−; Pci flies to confirm transmission and segregation and to determine chromosomal location. Mutations were crossed inter se to establish recessive lethal complementation groups. Mutant CG8878 alleles were kept as balanced stocks with CyO.
Genetic Mapping
The dominant enhancer phenotype in an E1/Pci background was used for genetic recombination mapping because it gave a fuller red eye phenotype, which provided more room for enhancement and thus allowed a more reliable visual assessment of enhancement. Mutants were mapped relative to wgSp L Bc and Pin markers. Recombinants both left and right of the enhancer were collected and tested for retention of the enhancer phenotype by crossing males to w−; dp−; e−; E1 virgin females, and for retention of the recessive lethal phenotype by crossing to other members of the same complementation group. After establishing absolute linkage between the recessive lethal and dominant enhancer phenotypes, the position of the lethal locus was refined by complementation analysis against deficiencies in the region. At least 100 progeny were scored and if the heterozygous mutant/deficiency combination did not occur the combination was considered lethal.
DNA sequencing
A series of overlapping CG8878 gene segments were amplified by PCR and the product was sequenced. Point mutations were identified as double peaks on the chromatogram. All polymorphisms and mutations were confirmed by sequencing both strands.
Eye pigment assays
The amount of w+ gene activity was assayed by measuring the amount of brown eye pigment using a modification of the method of [12]. Heads from 5–9 day old adult flies were stored at −20° until extracted. For each genotype, three replicate samples of 10 heads were extracted in 200 µL of acidified ethyl alcohol (1% HCl in 30% ethanol) with shaking for 48 hours. Absorbance at 470 nm was then measured using a 96 well Costar flat bottom plate in a Bio-Tek PowerWave XS spectrophotometer. Photographs of representative adult flies eyes under mineral oil were taken using a Zeiss stereo microscope and a Nikon Coolpix 995 digital camera. For both the eye pigment assay and the adult eye photographs, the balancer chromosome CyO was used as the control.
Results
Screen for second site enhancers of w+ variegation in E1
We screened ∼44,000 progeny from EMS treated fathers for dominant enhancement of w+ silencing in E1 and recovered 58 confirmed mutations. Inter se recessive lethal complementation analysis showed they fell into five simple and three more complex recessive lethal complementation groups, as well as many singles. Along with mutations in trithorax and ash1 (manuscript in preparation), there was a simple group, with 7 alleles (1a27a, 3a22a, 3a52a, 3a66a, 3a90a, 3a97a, 4a7a), which is described here. These seven were examined further and the dominant enhancer of E1/Pci phenotype for allele 4a7a genetically mapped to 2–65.4 (n = 490) by recombination relative to wgSp L Bc and Pin markers. Linkage between the dominant enhancer of E1/Pci and recessive lethal phenotypes was demonstrated as described in the Materials and Methods. Deficiency mapping of the lethal phenotype of this group refined its position to 48E2;48E4, within Df(2R)BSC199 (7,779,605: 8,059,989) but not Df(2R)BSC879 (7,779,605: 8,029,867), which includes the predicted CG8878 gene. This is the only candidate gene in this region predicted to influence gene expression in a heritable manner, expressed in the correct tissue at the correct time, and of a size likely to result in 7 independent mutations in a mutagenesis of this size. We note that both Hen1 and Prp8 might influence gene expression post-translationally, which should not lead to silencing in a clonal manner (variegation) as seen here.
DNA sequencing of the mutants
DNA sequencing spanning the entire predicted coding region of CG8878, in heterozygotes with the CyO balancer chromosome, showed that five alleles (1a27a, 3a22a, 3a52a, 3a66a, 3a97a) had a base pair change within CG8878 that altered the predicted amino acid coding sequence (Table 1, Figure 1). Three of the alleles (3a27a, 3a52a and 3a97a) had G/C to A/T transitions that resulted in premature stop codons; with 3a52a being at the amino terminal end of the first predicted STKc domain and therefore likely to be a null allele. Allele 3a66a had a single nucleotide deletion that caused a frame-shift leading to multiple premature stop codons while 1a27a had a G/C to A/T transition that predicts the loss of an intron donor splice site, a frame-shift and multiple premature stop codons. Two other alleles (3a90a and 4a7a) had identical nineteen base pair deletions in the 5′ upstream promoter region that included 4 base pairs of the proximal predicted E box and are thus presumptive transcriptional regulatory mutants.
Mutant designations are above the polypeptide backbone while the nature of the corresponding mutation is below. Regions of sequence similarity to ballchen (nhk-1) are shown as mauve bars below the CG8878 sequence. The five regions identified in the dot plots are shown above the polypeptide diagram.
Phenotypic characterization of the mutants
Visual pigment assessment for the dominant enhancement of white-eyed variegation in E1/+ heterozygotes indicated all mutant alleles were enhanced relative to the CyO control in both sexes, and frequently produced flies indistinguishable from w−. Representative photographs of mutant eyes are given in Figure 2a. The extent of enhancement was quantified by pigment assays. Both male and female mutants had less than half the pigment of non-mutant internal control flies (CyO balancer) from the same cross, with all being significant (95% confidence limit - Fig. 2b). All mutant alleles enhance silencing of E1/+.
a) Representative photographs of eyes from each class of progeny from heterozygous CG8878 mutants crossed back to the parental E1 stock used in this mutagenesis. Cy versus Cy+ flies were compared for each sex. Flies are posed facing right. (y− w−; dp− 3a52a/CyO, Cy dp− ♂ X w−; dp−; e−; E1). b) Pigment assays of heterozygous CG8878 mutants crossed back to the parental E1 stock used in this mutagenesis. Cy versus Cy+ flies were compared for each sex. (y− w−; dp− CG8878/CyO, Cy dp−♂ X w−; dp−; e−; E1
) Statistical significance between pairs (T-Test = 1-tailed, independent (unpaired, unequal variance)) is given above each mutant pair tested, * = p<0.05, ** = p<0.01, NS = not significant.
To see if the dominant enhancement (w+ silencing) was limited to the E1 allele, we quantitatively assayed the effect of the mutants on Pci/+ flies, which lack the gypsy element present in E1. All mutants reduced the amount of pigment compared to non-mutant internal control flies (CyO balancer) from the same cross (Figure 3b). However, only the putative null allele 3a52a showed significance in both sexes. Visually, variegation was less visible in this cross, with eyes displaying only a pattern of weak silencing starting at the posterior edge, with rapidly decaying anterior progression (Figure 3a).
a) Photographs of representative examples from each class of progeny from heterozygous CG8878 mutants crossed back to Pci, the parental stock from which E1 was derived. Cy versus Cy+ flies were compared for each sex. Flies are posed facing right. (y− w−; dp− 3a52a/CyO, Cy dp− ♂ X w−; dp−; e−; Pci). b) Pigment assays of heterozygous CG8878 mutants crossed back to the parental Pci stock E1 was derived from. Cy versus Cy+ flies were compared for each sex. (y− w−; dp− CG8878/CyO, Cy dp− ♂ X w−; dp−; e−; Pci
) Statistical significance between pairs (T-Test = 1-tailed, independent (unpaired, unequal variance)) is given above each mutant pair tested, * = p<0.05, ** = p<0.01, NS = not significant.
To address the possibility that these mutations were acting directly on the w+ transgene, and not via any dependence on insertion position, we assayed pigment in y1 w* P{lacW}3-76a/+ flies with these mutants. The transgene in P{lacW}3-76a is the same as that in Pci but is located at 18A1 (60.7 cM) on the X chromosome and presents a full red eye. Of the three alleles tested, only 4a7a, a putative regulatory mutant, males showed a barely significant difference (p = .03) from non-mutant control flies (CyO balancer) of the same cross indicating that the dominant enhancement (silencing) is dependent upon chromosomal location and not on the w+ transgene Pci construct itself (Figure 4). The consistency of our results across all mutants tested indicates that this method of pigment determination is both accurate and precise.
a) Photographs of representative examples from each class of progeny from heterozygous CG8878 mutants crossed to P{lacW}3-76a, an insertion of the same transgene as Pci, but only at 18A1 on the X chromosome. Cy versus Cy+ flies were compared for each sex. Flies are posed facing right. (y− w−; dp− 3a52a/CyO Cy dp− ♂ X y1 w− P{lacW}3-76a ). b) Pigment assays of heterozygous CG8878 mutants crossed to P{lacW}3-76a. Cy versus Cy+ flies were compared for each sex. (y− w−; dp− CG8878/CyO Cy dp− ♂ X y1 w− P{lacW}3-76a
) Statistical deviation between pairs was insignificant (T-Test = 1-tailed, independent (unpaired, unequal variance)) except for 4a7a males which was significant at p<0.05 (* = p<0.05).
Next, we asked whether these mutants had an effect on classical hPEV by crossing y− w−; dp− CG8878*/CyO, Cy dp− males to virgin In(1)wm4; dp−; e− females. Variegation of wm4 was visibly enhanced by all three mutants tested (Figure 5 a) and quantitatively (95% and 99% confidence limits) enhanced in male and female flies respectively (Figure 5b).
a) Photographs of representative examples from each class of progeny from heterozygous CG8878 mutants crossed to wm4. Cy versus Cy+ flies were compared for each sex. Flies are posed facing right. (y− w−; dp− 3a52a/CyO, Cy dp− ♂ X wm4; dp−; e− ). b) Pigment assays of heterozygous CG8878 mutants crossed to wm4. Cy versus Cy+ flies were compared for each sex. (y− w−; dp− CG8878/CyO Cy dp− ♂ X wm4; dp−; e−
) Statistical significance between pairs (T-Test = 1-tailed, independent (unpaired, unequal variance)) is given above each mutant pair tested, * = p<0.05, ** = p<0.01, NS = not significant.
Amino acid sequence comparisons
Analysis of CG8878's predicted polypeptide sequence using SMART (University of Heidelberg) predicts two domains related to protein kinase separated by 194 amino acids (Figure 1). A comparison of CG8878's predicted amino acid sequence with eleven other Drosophila species reveals that homologs are present and highly conserved in all twelve species studied; this supports CG8878 being an essential gene (Figure 6, Figure S1). This cladogram parallels that already determined for these species [13].
Accession numbers given in Table S1 and the alignment in Figure S1. a) Score table showing degree of similarity between CG8878 homologues. b) Cladogram showing relative evolutionary distances between CG8878 homologues for twelve Drosophila species (http://www.ebi.ac.uk/Tools/msa/clustalw2/). Names are abbreviated using the capitalized first letter of the genus followed by the first three letters of the species.
The amino acid sequence of CG8878 shows the most similarity to D. melanogaster ballchen, and human orthologs, Vaccinia Related Kinases (VRK1, VRK2), which encode a nucleosomal histone 2a kinase (Figure 7, Figure S2). Amino acid sequence comparisons suggest both CG8878 and ballchen are derived from a common VRK like precursor. However, in CG8878, the VRK domain appears to have been split in two by an ∼194 amino acid insertion (Figures 1, 8). CG8878 shows 36% identity and 56% positive correlation to BALLCHEN over both parts of the kinase domain (http://blast.ncbi.nlm.nih.gov/Blast.cgi) indicating a functional conservation. Note, since our mutation 3a22a (R546Opal) is recessive lethal and enhances variegation at E1, Pci, and wm4 (Figures 2 b, 3 b, 5 b), it appears the second part of CG8878's split VRK-like kinase domain (Fig. 8) is essential for CG8878's function.
Abbreviations are as follows: Drosophila melanogaster (Dmel), Homo sapiens (Homo), Mus musculus (Mus) (http://www.ebi.ac.uk/Tools/msa/clustalw2/) Accession numbers given in Table S1.
Comparison symbols: * = identity, : = side groups with strongly similar properties, . = side groups with weakly similar properties. Amino acid color code: red = small hydrophobic, blue = acidic, magenta = basic, green = hydroxyl, sulfhydryl, amine, G. • Accession numbers given in Table S1.
Structure of CG8878
An amino acid sequence similarity is also found with human Casein Kinases and human Tau-Tubulin Kinases. This corresponds to a PcK kinase conserved domain (called PHA02882). However, a dot plot comparison of CG8878 with ballchen and hVRK1 (Figure 9) shows that the single, conserved PcK domain is split into two regions in CG8878. Thus the whole CG8878 sequence can be separated into 5 regions (1–96, 97–291, 292–516, 517–662, and 663–1004), with 2 and 4 corresponding to the PcK conserved domain.
The single PcK domain in ballchen and hVRK1 are split in CG8878. Horizontal, red dotted lines mark the five regions, with 2 and 4 corresponding to the conserved PcK functional domain. The vertical, green line separates ballchen and hVRL1 sequences. Setting for JDotter are: Window size: 50; Zoom factor: 1 base/pixel; Pixel factor: 48; Scoring matrix: BLOSUM62; GreyMap Tool: 0, 13; Maximum Plot size: 700 bases/pixel; Sliding window size for new plots: 50.
A ClustalW sequence alignment of D. melanogaster CG8878 with eleven other Drosophila orthologs all show a similar split PcK domain and 5 region organization (Figure S1). Comparison of sequence variation in each of the 5 regions shows that regions 2 and 4 are more conserved among these orthologs, suggesting evolutionary sequence conservation (Figure 10).
The twelve orthologs of the D. melanogaster CG8878 (see Figure S1) were aligned using ClustalW and the sequence similarities were counted as identical, conserved, similar, or no match for each of the five regions and is shown as a percent of the total in that region. Regions 2 and 4, which correspond to the PcK domain conserved with ballchen and hVRK1, are more conserved among these orthologs than the other regions.
Dot plot analysis, comparing D. melanogaster with D. virilis (Figure 11), also shows three regions of repeated sequence that are rich in glutamic and aspartic acid. Repeat 1 (position 300–353) has 42%, repeat 2 (421–449) has 46%, and repeat 3 (790–854) has 50% glutamic and aspartic acid content. Similarly positioned repeats can be found in three mosquito genes (A. aegypti, A. gambiae, and C. quinquefasciatus) supporting the contention that they are orthologs. Bombyx mori, a moth, has a similar gene (BGIBMGA000118-PA_www.silkdb.org) with a split PcK domain, but it lacks the three acid-rich regions. No other split PcK domain genes were identified; all other PcK containing genes had single, uninterrupted domains.
The three repeated, aspartic and glutamic acid rich regions are present in all Diptera but absent from B. mori, a moth. Setting for JDotter are as in Figure 9, but with Maximum Plot size: 1200 bases/pixel.
Additional dot plot analysis of CG8878 with hVRK1, hCK1, and hTTK1 (Figure 12) shows that hVRK1 and hCK1 have single PcK domains and lack a long C-terminal region. hTTK1 also has a single PcK domain, but contains an acid-rich repeat region, like CG8878. The D. melanogaster ortholog of hTTK1 is asator [14], which has a single PcK domain in the N-terminal quarter of the polypeptide, and lacks any acid-rich repeat regions in the C-terminal three-quarters end.
The three repeated, aspartic and glutamic acid rich regions are present in hTTK1 but not hVRK1 or hCK1. Setting for JDotter are as in Figure 11.
By examining the 3D hVRK1 structure as determined by Nuclear Magnetic Resonance [15] and the linear alignment of CG8878 with hVRK1 sequences, the splitting of the PcK domain in CG8878 would correspond to an insertion of Region 3 after part of the catalytic loop (just beyond the putative Ser/Thr kinase active site) but before the activation loop. This location is at the surface of the structure and corresponds to position 187–191 (YKNPD) in hVRK1 (Figure 13).
Regions of hVRK1 that are similar to D. melanogaster CG8878 are highlighted in yellow. Regions 2 (left) and 4 (centre) correspond to the conserved sequence regions. The location between these two regions, corresponding to the insertion of Region 3 (right) is on the surface of the polypeptide and highlighted in yellow.
NucPred [16] predicts, with a score of 1.0, a nuclear localization signal (NLS) in CG8878 at position 17–21 (RKRKR). This sequence is present in a similar position in all 12 Drosophila orthologs (Figure S1), suggesting it is conserved. Similar NLS sequences are present, but at different locations, in the three potential mosquito orthologs (score 0.95–1.0). Discrete NLS sequences appear absent from the B. mori gene (Score 0.87).
We conclude the hybrid nature (kinase sequence most similar to ballchen/hVRK1 and presence of three acid rich repeats like that in hTTK1) and split PcK domain of CG8878 defines a novel kinase type to be added to the VRK, CK, and TTK groups.
Discussion
We induced, recovered, and characterized seven mutations that dominantly enhance the variable silencing (variegation) of E1, whose expression is similar to P element dependent silencing (PDS). The dominant enhancement genetically maps at or near the CG8878 locus and it could not be separated from the lethal phenotype by crossing over. The lethal phenotype deficiency maps to a very fine region that includes CG8878. Five alleles contain mutations resulting in stop codons; two at the amino terminal end of CG8878's amino proximal predicted STKc domain likely represent null alleles, one between CG8878's two predicted kinase domains, and two in the amino end of CG8878's carboxy proximal predicted kinase domain. Taken together, this shows that loss of the CG8878 gene function is responsible for the dominant enhanced silencing of w+ in E1 and a recessive lethal phenotype. Bioinformatic analysis of CG8878 indicates that it is likely a protein kinase, but the putative functional domain has been split in two. Furthermore, this split form appears limited to Dipterans.
CG8878 and Hen1
The CG8878 transcription unit is located entirely within the large (5.4 kb) second intron of another gene, Hen1 (formerly Pimet), in the antisense orientation. Hen1 has been shown to mediate 2′-O-methylation at the 3′ end of Piwi interacting RNAs in Drosophila [17], [18]. Piwi interacting RNAs are germ-line specific 24–30 nt RNAs that couple with PIWI proteins to silence invading transposable elements (reviewed by [19]). Given that Pci has P element terminal repeats and, at the 5′ end, a P element transposase lacZ fusion, we considered that Hen1, and not CG8878, might potentially be the enhancer identified in this screen, but several points argue against this: 1) all seven mutants had lesions in CG8878 coding or regulatory sequences; 2) all of these lesions are entirely inside Hen1's second intron, and predict no effect on Hen1 expression; 3) Hen1 is not an essential gene because PBac(WH)Hen1[f00810] is a null for Hen1 [18] but is not recessive lethal; 4) P{lacW}3-76a appears to be unaffected by our En(var)s despite being the same construct only at a different location; and 5) wm4, which is not P element derived, is significantly affected by our En(var)s. The most parsimonious explanation is that these mutations are due to lesions in CG8878, not Hen1, and that CG8878 is an essential gene and when mutated has a dominant En(var) phenotype.
Potential molecular function of CG8878
Although we have been unable to find split kinase domain CG8878 homologues outside of the order Diptera, CG8878 is highly conserved across Drosophila species (Figures 6, Figure S1). Nevertheless, the conservation of both VRK like kinase domains indicates that it is likely to encode a kinase of unknown, but essential, specificity. The closest Drosophila melanogaster paralog of CG8878 is ballchen (an NHK-1 homolog), with regions of maximum similarity coinciding with CG8878's putative kinase domains as shown in Figure 1b). ballchen has high affinity for chromatin and has been shown to phosphorylate Threonine 119 at the carboxy terminus of nucleosomal, but not free, H2A in Drosophila embryos. H2A T119 is phosphorylated during mitosis but not in S phase which coincides with NHK-1's chromatin association as shown by immunostaining and may be a component of the histone code related to cell cycle progression [20]. Ivanovska et al. [21] described a point mutation, Z3-0437, in the kinase domain of NHK-1 that led to female sterility due to defects in the formation of the karyosome. This led to metaphase I arrest as a result of failure of the synaptonemal complex to disassemble and to load condensin onto chromosomes in the mutant. Mitosis was also shown to be affected, as embryos laid by nhk-1−/− mutant females arrested with aberrant mitotic spindles and polar bodies. They also found a lack of Histone H4K5 and H3K14 acetylation in the karyosomes in nhk-1 mutant but not control oocytes, implying that Histone H2A threonine 119 phosphorylation is required for meiotic acetylation of these residues. Lancaster et al. [22] found that phosphorylation of barrier to autointegration factor protein (BAF) by NHK-1 was necessary for karyosome formation. Loss of NHK-1 or expression of nonphosphorylatable BAF resulted in ectopic chromosome-nuclear envelope association in oocytes leading the authors to propose that tethering of chromosomes to the nuclear envelope is disrupted by NHK-1 mediated BAF phosphorylation, allowing karyosome formation in oocytes.
CG8878's exact target and mode of action are yet to be determined, but sequence similarities suggest that Histone phosphorylation by CG8878 would readily explain its action as an En(var). For example, JIL1 phosphorylation of H3S10 blocks methylation of H3K9 allowing hyperacetylation of Histone 3 and promoting a transcriptionally active chromatin state [23]. CG8878's expression profile is consistent with it being a genome wide inhibitor of heterochromatin spread as it is expressed in all tissues, at all stages of development, with maxima at times of peak developmental change, such as early embryogenesis and prepupariation [24].
Our mutants suggest the predicted kinase domains are essential for function. The enhancer (of E1 and wm4) phenotypes and recessive lethal phenotypes of 3a66a, which results in a premature stop codon between CG8878's two predicted kinase domains, and 3a22a, and 3a97a, which result in a premature stop codon in the amino end of CG8878's carboxy proximal predicted kinase domain, all argue that this latter predicted kinase domain is essential for CG8878 function.
The putative Kinase coding region of CG8878 is most similar to hVRK1, but is split into two segments (Regions 2 and 4). The conserved NLS sequence supports nuclear localization and thus a possible role in chromatin modification. The conserved presence of the aspartic and glutamic acid rich repeats suggest possible interaction sites. These are lacking in hCK1, a cytosolic protein, only present once in hTTK1, and absent in the D. melanogaster asator (TTK1 ortholog). Together, this suggests that CG8878 encodes a protein Kinase that modifies chromatin structure.
CG8878 acts at the ci locus
Pci was isolated as an enhancer trap at the ci locus since the enhancer-trap reporter accurately mimicked that of ci RNA with both being expressed specifically in anterior compartment cells of the imaginal discs [1]. The w+ transgene in Pci (and the E1 gypsy element) are inserted in the ci distal regulatory region. Pci is a recessive allele of ci because it exhibits ci wing phenotype when heterozygous with ci57g (a deletion upstream of Pci in the regulatory region) and ci1 (a gypsy insert upstream of Pci). All our mutant CG8878 alleles enhance variegation (reduce w+ expression in the transgene) in E1 and E1/Pci (unpublished observation), but have little effect on P{lacW}3-76a, the same construct at a different location. Thus the silencing is location dependent and is thus not likely due to a direct interaction with the white promoter, but with the ci regulatory region itself. Since Pci reporter expression is approximately halved when 3a52a is present, and does not depend on the presence of E1, we infer that CG8878 normally acts at the ci regulatory region to impede the spread of heterochromatin into this region, likely in a dose sensitive manner.
Supporting Information
Figure S1.
Pairwise alignment of CG8878 and 12 Drosophila homologues. Species names are abbreviated using the capitalized first letter of the genus followed by the first three letters of the species. Comparison symbols: * = identity, : = side groups with strongly similar properties, . = side groups with weakly similar properties. Amino acid color code: red = small hydrophobic, blue = acidic, magenta = basic, green = hydroxyl, sulfhydryl, amine, G (http://www.ebi.ac.uk/Tools/msa/clustalw2/. Note: for D. persimilis a nucleotide was removed (five A's to four A's – a presumed sequencing error) to facilitate amino acid alignment. Accession numbers given in Table S1.
https://doi.org/10.1371/journal.pone.0071695.s001
(TIF)
Figure S2.
Pairwise alignment of BALLCHEN and VRK1 from mouse and humans. Symbols are the same as Figure S1. Comparison symbols: * = identity, : = side groups with strongly similar properties, . = side groups with weakly similar properties. Amino acid color code: red = small hydrophobic, blue = acidic, magenta = basic, green = hydroxyl, sulfhydryl, amine, G (http://www.ebi.ac.uk/Tools/msa/clustalw2/.) Accession numbers given in Table S1.
https://doi.org/10.1371/journal.pone.0071695.s002
(TIF)
Table S1.
Polypeptide accession numbers used. Note: the persimilis sequence was modified by the removal of one nucleotide from the DNA sequence to alter the amino acid reading frame to facilitate alignment.
https://doi.org/10.1371/journal.pone.0071695.s003
(PDF)
Author Contributions
Conceived and designed the experiments: AM JL. Performed the experiments: AM JL. Analyzed the data: AM JL. Contributed reagents/materials/analysis tools: AM JL. Wrote the paper: AM JL.
References
- 1. Eaton S, Kornberg TB (1990) Repression of ci-D in posterior compartments of Drosophila by engrailed. Genes Dev 4: 1068–1077.
- 2. Bushey D, Locke J (2004) Mutations in Su(var)205 and Su(var)3–7 Suppress P-Element-Dependent Silencing in Drosophila melanogaster. Genetics 168: 1395–1411.
- 3. Sameny A, Locke J (2011) The P-element-induced silencing effect of KP transposons is dose dependent in Drosophila melanogaster. Genome 54: 752–62.
- 4. Sun FL, Cuaycong MH, Craig CA, Wallrath LL, Locke J, et al. (2000) The fourth chromosome of Drosophila melanogaster: interspersed euchromatic and heterochromatic domains. Proc Natl Acad Sci USA 97: 5340–5345.
- 5. Wallrath LL, Elgin SCR (1995) Position effect variegation in Drosophila is associated with an altered chromatin structure. Genes & Development 9: 1263–1277.
- 6. Chetverina D, Savitskaya E, Maksimenko O, Melnikova L, Zaytseva O, et al. (2008) Red flag on the white reporter: a versatile insulator abuts the white gene in Drosophila and is omnipresent in mini-white constructs. Nucleic Acids Research 36: 929–937.
- 7. Dernburg AF, Broman KW, Fung JC, Marshall WF, Philips J, et al. (1996) Perturbation of Nuclear Architecture by Long-Distance Chromosome Interactions. Cell 85: 745–759.
- 8. Csink AK, Henikoff S (1996) Genetic modification of heterochromatic association and nuclear organization in Drosophila. Nature 381: 529–531.
- 9. Belyaeva ES, Koryakov DE, Pokholkova GV, Demakova OV, Zhimulev IF (1997) Cytological study of the brown dominant position effect. Chromosoma 106: 124–132.
- 10.
Lindsley DL, Zimm GG (1992) The Genome of Drosophila melanogaster. Academic Press, New York.
- 11.
Ashburner M (1989). Drosophila, A laboratory manual Cold Spring Harbor Press, Cold Spring Harbor, New York
- 12. Ephrussi B, Herold JL (1944) Studies of eye pigments of Drosophila. I. Methods of extraction and quantitative estimation of the pigment components. Genetics 29: 148–175.
- 13. Hahn MW, Han MV, Han S-G (2007) Gene Family Evolution across 12 Drosophila Genomes. PLoS Genet 3(11): e197.
- 14. Qi H, Yao C, Cai W, Girton J, Johansen KM, et al. (2009) Asator, a Tau-Tubulin Kinase Homolog in Drosophila Localizes to the Mitotic Spindle. Developmental Dynamics 238: 3248–3256.
- 15. Shin J, Chakraborty G, Bharatham N, Kang CB, Tochio N, et al. (2011) NMR Solution Structure of Human Vaccinia-related Kinase 1 (VRK1) Reveals the C-terminal Tail Essential for Its Structural Stability and Autocatalytic Activity. J Biol Chem 286: 22131–22138.
- 16. Brameier M, Andrea Krings A, MacCallum RM (2007) NucPred—Predicting nuclear localization of proteins. Bioinformatics 23: 1159–1160.
- 17. Saito K, Sakaguchi Y, Suzuki T, Siomi H, Siomi MC (2007) Pimet, the Drosophila homolog of HEN1, mediates 2′-O-methylation of Piwi- interacting RNAs at their 3′ ends. Genes Dev 21: 1603–1608.
- 18. Horwich MD, Li C, Matranga C, Vagin V, Farley G, et al. (2007) The Drosophila RNA Methyltransferase, DmHen1, Modifies Germline piRNAs and Single-Stranded siRNAs in RISC. Current Biology 17: 1265–1272.
- 19. Saito K, Nishida KM, Mori T, Kawamura Y, Miyoshi K, et al. (2006) Specific association of Piwi with rasiRNAs derived from retrotransposon and heterochromatic regions in the Drosophila genome. Genes & Development 20: 2214–2222.
- 20. Aihara H, Nakagawa T, Yasui K, Ohta T, Hirose S, et al. (2004) Nucleosomal histone kinase-1 phosphorylates H2A Thr 119 during mitosis in the early Drosophila embryo. Genes Dev 18: 877–888.
- 21. Ivanovska I, Khandan T, Ito T, Orr-Weaver TL (2005) A histone code in meiosis: the histone kinase, NHK-1, is required for proper chromosomal architecture in Drosophila oocytes. Genes Dev 19: 2571–2582.
- 22. Lancaster OM, Cullen CF, Ohkura H (2007) NHK-1 phosphorylates BAF to allow karyosome formation in the Drosophila oocyte nucleus. J Cell Biol 179: 817–824.
- 23. Zhang W, Deng H, Bao X, Lerach S, Girton J, et al. (2006) The JIL-1 histone H3S10 kinase regulates dimethyl H3K9 modifications and heterochromatic spreading in Drosophila. Development 133: 229–235.
- 24. Gelbart WM, Emmert DB (2011.8.23) Calculation of RPKM to generate quantitative expression data: read-length values for modENCODE developmental timecourse RNA-Seq data. FlyBase analysis
- 25. Brodie R, Roper RL, Upton C (2004) JDotter: A Java Interface to Multiple Dot Plots Generated by Dotter. Bioinformatics 20: 279–281.
- 26. Wang Y, Geer LY, Chappey C, Kans JA, Bryant SH (2000) Cn3D: sequence and structure views for Entrez. Trends Biochem Sci 25: 300–302.