Loss of the Histone Pre-mRNA Processing Factor Stem-Loop Binding Protein in Drosophila Causes Genomic Instability and Impaired Cellular Proliferation

Background Metazoan replication-dependent histone mRNAs terminate in a conserved stem-loop structure rather than a polyA tail. Formation of this unique mRNA 3′ end requires Stem-loop Binding Protein (SLBP), which directly binds histone pre-mRNA and stimulates 3′ end processing. The 3′ end stem-loop is necessary for all aspects of histone mRNA metabolism, including replication coupling, but its importance to organism fitness and genome maintenance in vivo have not been characterized. Methodology/Principal Findings In Drosophila, disruption of the Slbp gene prevents normal histone pre-mRNA processing and causes histone pre-mRNAs to utilize the canonical 3′ end processing pathway, resulting in polyadenylated histone mRNAs that are no longer properly regulated. Here we show that Slbp mutants display genomic instability, including loss of heterozygosity (LOH), increased presence of chromosome breaks, tetraploidy, and changes in position effect variegation (PEV). During imaginal disc growth, Slbp mutant cells show defects in S phase and proliferate more slowly than control cells. Conclusions/Significance These data are consistent with a model in which changing the 3′ end of histone mRNA disrupts normal replication-coupled histone mRNA biosynthesis and alters chromatin assembly, resulting in genomic instability, inhibition of cell proliferation, and impaired development.


Introduction
Histones are a class of highly abundant nuclear proteins whose most basic function is to package and organize the genetic material. In addition to organizing DNA, histones play important roles in a number of other cellular processes critical for survival and development. These include DNA repair [1], chromosome segregation [2], regulation of transcription [3], and tissue differentiation [4]. In metazoans, there are two classes of histone proteins. The canonical, replication-dependent histones, H2a, H2b, H3, H4, and H1, are synthesized solely during S-phase, where they are utilized to package newly replicated DNA. The replication-independent histone variants are paralogs of the canonical histones which assemble into nucleosomes with specialized functions [5]. Unlike replication-dependent histones, histone variants can be synthesized and deposited into chromatin throughout the cell cycle [6].
Restriction of replication dependent histone biosynthesis to Sphase is conserved in all species of fungi, plants, and animals studied to date [7,8]. Yeast and Arabidopsis accomplish S-phase coupling through transcriptional regulation [7]. In metazoans, S-phase coupled histone production is controlled largely through post-transcriptional regulation of mRNA levels due to changes in pre-mRNA processing efficiency and mRNA half life [7]. In budding yeast, the production of histones in the correct stoichiometry is important for genome maintenance and successful cell cycle progression [2,9,10,11]. In human cells, the inhibition of histone gene expression leads to S-phase arrest [12,13]. These observations demonstrate that proper regulation of replicationdependent histone production is functionally important.
Metazoan replication-dependent histone mRNAs are not polyadenylated [14], and instead terminate with a conserved 39 stem-loop that is unique to histone mRNAs [15]. The regulatory properties conferred by the histone mRNA 39 end are likely to impact the rate of histone protein synthesis, histone stoichiometry, and the timing of histone synthesis during the cell cycle [7,8].
However, the precise connection between 39 end mediated regulation of histone mRNAs and histone protein production in vivo remain to be determined. Changes in the way that the 39 end affects mRNA processing, localization, or translation of histone mRNA could be reflected by changes in histone protein abundance or stoichiometry that alter properties of chromatin due to misregulation of chromatin assembly. In this study we test this hypothesis by analyzing mutations of the Slbp gene that disrupt normal histone mRNA 39 end formation and regulation.
Formation of the histone mRNA 39 end requires two unique sequence elements in the pre-mRNA. The first is the stem-loop, which remains part of the mature mRNA after pre-mRNA processing [16]. The stem-loop confers the specific coordinate regulation of replication-dependent histone mRNAs [17]. The second element is a purine rich histone downstream element (HDE), which is removed during the processing reaction [18]. Each of these sequences recruits factors that ultimately produce the single endonucleolytic cleavage between the stem-loop and HDE required for complete maturation of histone mRNAs [15].
The U7snRNP, composed of the U7 snRNA and a heptameric ring of Sm and Lsm proteins, is targeted to the HDE via basepairing with the U7 snRNA [19]. Stem-loop binding protein (SLBP) directly binds the stem-loop of histone mRNAs and is necessary for correct processing of histone pre-mRNAs [20], and is absolutely required for processing Drosophila pre-mRNAs in vitro [21]. In addition, SLBP is required for the nuclear export and translation of histone mRNAs in mammalian cells [13,22]. To date, there is no evidence for functions of SLBP outside those directly involved in histone biosynthesis. Additionally, other factors common to the canonical cleavage and polyadenylation reaction are required for histone pre-mRNA processing [23,24,25].
The maturation of histone mRNAs is a well-conserved process and has been characterized at the molecular level in considerable detail [15,16]. Yet, despite the deep understanding of how the 39 ends of histone mRNAs mature, the significance of this unique 39 end in terms of replication-coupled histone protein production, genome integrity, and development have not been studied. This is due to difficulties in altering the 39 ends of histone mRNAs in metazoans. Knocking out factors involved in processing mammalian histone mRNAs leads to a severe reduction in transcript levels [13,26] making this approach unfruitful for specifically studying the significance of the 39 end. A second difficulty is that replication-dependent histone genes typically occur in large clusters, making strategies that alter or replace the 39 end in vivo problematic.
Investigation of this question in Drosophila provides a unique opportunity to overcome these difficulties. Unlike other organisms, the production of mature histone transcripts is not abrogated when normal processing of the pre-mRNA is prevented [27]. Null mutations in Slbp result in incorrectly processed histone mRNAs, due to read through and utilization of cryptic polyadenylation sites downstream of the normal cleavage site [27,28,29]. These transcripts contain both a stem-loop and a polyA tail and are not properly cell cycle regulated in some embryonic endocycling tissues, likely because they are not degraded at the end of S-phase [27,28,29].
Drosophila Slbp mutants support both DNA replication and chromatin assembly, and thus the aberrant histone mRNA in these mutants can be translated, though the rate and timing of translation is unknown. However, Slbp null mutants are developmentally delayed and die as pupae. [27,28]. Because the histone mRNA 39 end is thought to be involved in the coordinate regulation of replication-dependent histone production, and alterations in histone stoichiometry cause genomic instability [2], we considered the possibility that proper 39 end-mediated regulation of histone mRNA might be important for genomic stability. Here, we demonstrate that Slbp mutants, despite sufficient histones to support DNA replication, display several forms of genomic instability and cell cycle abnormalities that disrupt development.

Reduction of Slbp Causes Increased Frequency of Loss of Heterozygosity
To gain a generalized measure of genomic instability, we developed a loss of heterozygosity (LOH) assay. Our assay measured loss of function of a yellow + (y + ) locus, which is necessary for brown body pigment, from a 4 th chromosome translocation (the y + locus is normally located on the X chromosome). In flies heterozygous for this translocation, on an otherwise y genetic background, functional loss of the single y + locus on the translocation during development results in patches of yellow body color in adult flies, which are then scored as LOH events. To obtain a measure of LOH, we counted the frequency of yellow bristles found in the first twenty bristles of the anterior wing margin ( Figure 1A). Since viable adults were required for this assay, we utilized the Slbp 10 hypomorphic allele. Wild type (wt) and mutant genotypes were normalized relative to Slbp heterozygous siblings. The heterozygotes present in every experiment controlled We analyzed varying doses of the Slbp 10 hypomorphic allele. In females, we observed a trend toward increased LOH with decreased dosage of Slbp ( Figure 1B). In males, we found significant differences between wt and the Slbp 10 /+ heterozygous normalizing group, as well as between Slbp 10 homozygous mutants and heterozygous controls. The frequency of LOH in Slbp 10 /+ heterozygous males was ,1.9 fold that of wt siblings (p,0.012), while Slbp 10 homozygous mutants had ,2.3 fold the LOH of heterozygotes (p,0.011) ( Figure 1B). These results demonstrate that decreasing Slbp function increases LOH.
Interestingly, we observed that the LOH phenotype was more severe in males. One possible explanation for this result might be the presence of the highly heterochromatic Y chromosome. This heterochromatic region of the genome sequesters histones, and has been shown to affect chromatin related phenotypes in other assays [30,31,32].

Slbp Mutants Exhibit Increased Frequency of Double Strand Breaks and Tetraploid Cells
Functional LOH can be caused by a number of processes including spontaneous mutation, mitotic crossovers events, chromosome loss, chromosome breakage, and heterochromatin spreading that reduces gene expression.
In order to determine which of these processes might contribute to the LOH we observed in Slbp mutants, we examined metaphase chromosome preparations of Slbp 15 null mutant larval brain neuroblasts. The karyotype of larval brain neuroblasts can be easily analyzed for chromosomal abnormalities, such as breaks and changes in ploidy.
First, we quantified the frequency of chromosome breaks. A normal Drosophila karyotype contains three pairs of autosomes and one pair of sex chromosomes ( Figure 2, category I). Nuclei were scored positive for breaks if there was at least one broken chromosome arm, relative to a normal karyotype, regardless of the extent of damage ( Figure 2, category II, IV). The baseline frequency of wt nuclei containing breaks was 2.46% in this assay ( Table 1, II + IV), which is slightly higher than, but comparable to, previous reports [33]. Cells containing more than a single break were extremely rare. In contrast, we observed a significant increase in chromosome breaks in the Slbp 15 (p,0.03) mutant, with 8.43% of nuclei containing breaks (Table 1, II + IV). In the mutant genotype, cells containing multiple breaks or other types of chromosomal abnormalities occurred at much greater frequency. The increase in chromosomal breaks in Slbp 15 mutants likely contributes to the increased LOH observed in these mutants. We next measured ploidy in Slbp 15 mutants. The frequency of cells containing one or more extra chromosomes was quantified for wt and Slbp 15 genotypes. We chose to determine polysomy rather than monosomy because it is impossible to distinguish true monosomy from loss of chromosomes due to preparation artifacts. Each normal mitotic nucleus possesses eight sister chromatid pairs ( Figure 2, category I). We found that Slbp 15 mutants were not statistically different from wt controls in percentage of polysomic cells (data not shown). Thus, it is unlikely that increased aneuploidy can account for the dramatic increase in LOH in Slbp mutants in our previous analysis. Consistent with this assertion, yellow bristles scored in the LOH assay did not have a Minute phenotype, which would result from loss of the entire translocated 4 th chromosome containing y + [34].
Unexpectedly, we detected the presence of tetraploid cells in Slbp 15 mutant brains ( Figure 2, category III). Tetraploidy can arise from several mitotic defects including failure of cytokinesis and prolonged arrest at the spindle checkpoint due to DNA damage, followed by subsequent cell cycle reentry [35]. Whereas tetraploidy was never detected in wild type cells, we observed an average of 7.29% tetraploid nuclei per brain in Slbp 15 mutants (p,0.03) ( Table 1, III + IV). Furthermore, a disproportionately large proportion of tetraploid cells also contained chromosomal breaks ( Table 1, category IV). This is consistent with tetraploidy in Slbp mutants being caused by prolonged arrest at the spindle checkpoint due to damage to the genome.

Slbp Mutants Exhibit Enhancement of Position-Effect Variegation
Silencing of a locus by surrounding heterochromatin could result in a functional LOH, even though the locus itself may be  For each genotype, the percentage of nuclei containing chromosomal breaks and tetraploidy was calculated for 6 individual brains and mean percentages determined. 23-132 mitotic nuclei per individual brain were analyzed. Four classes of abnormal karyotypes (see Figure 2): I) Normal karyotype with no breaks. II) Diploid nuclei containing breaks. II + IV) All nuclei containing breaks. III) Tetraploid nuclei without breaks. III + IV) Total tetraploid nuclei including those with breaks. IV) Tetraploid nuclei with breaks. doi:10.1371/journal.pone.0008168.t001 intact. Changes in histone gene dosage can alter the extent of heterochromatin [36,37]. Alternatively, mutations which promote relaxation of chromatin structure could also indirectly increase LOH by leaving the chromosome more susceptible to breaks. We reasoned that altered chromatin structure resulting from the misregulation of histone biosynthesis might account for some of the genomic instability we observed in Slbp mutants. Therefore, we hypothesized that changes in chromatin structure would correspond with the observed increases in genomic instability.
To analyze changes in chromatin structure, we measured position-effect variegation (PEV) in Slbp mutants. PEV is a phenomenon whereby gene expression is modulated by the structure of surrounding chromatin. Surrounding heterochromatin can spread into a normally transcriptionally active gene, silencing expression. In PEV, the expressed or silent state of gene expression is propagated clonally, producing a variegated phenotype. The extent of variegation can be modified when genes involved in the formation of chromatin structure are mutated. Mutations that inhibit heterochromatin formation suppress variegation, whereas mutations that increase the abundance of heterochromatin enhance variegation.
To measure PEV in Slbp mutants, we observed the extent of variegation of the w m4 inversion. The w m4 inversion is an Xchromosome inversion that relocates the w + gene from the tip of the X chromosome to a region proximal to the pericentric heterochromatin. Expression of the w + gene, which is necessary for pigment deposition in the eye, is modulated clonally by the state of this nearby chromatin. This enables both eye pigment variegation and total amount of pigment accumulation to be used as read-outs of chromatin structure.
We constructed a series of Slbp genotypes in conjunction with one copy of the w m4 inversion. By inspection of adult eyes, we observed that pigment expression decreased with decreasing dosage of Slbp ( Figure 3A). This indicates that Slbp mutation is an enhancer of PEV. To confirm this quantitatively, we utilized a spectrophotometer to measure the absorbance of eye pigment from samples of each genotype [38] ( Figure 3B). The absorbance of each sample was compared to that obtained from siblings from the same culture vial to control for parental and culture dependent effects.
We quantified pigment levels in Slbp mutants. Slbp 15 /Slbp 10 is the allelic combination with the smallest dose of SLBP compatible with adult viability. Pigment absorbance values in Slbp 15 /+ heterozygous sisters were ,2.3 fold greater than those of Slbp 15 / Slbp 10 females ( Figure 3C, left) (p,0.04). The Slbp 15 mutation also acts as a dominant enhancer of PEV since the wt absorbance was ,1.4 fold greater than that of Slbp 15 heterozygotes ( Figure 3C, middle) (p,0.0001). To ensure that these PEV phenotypes were due to the Slbp 15 mutation and not a component of the genetic background, we also examined the Slbp 12 null mutation, which was derived from a different P-element insertion [27]. We observed that Slbp 12 /+ females also exhibited PEV enhancement with wt pigment levels being ,1.4 fold greater than that of the heterozygotes ( Figure 3C, right) (p,0.0001). We were unable to analyze males quantitatively because Slbp mutation enhanced variegation so much that pigment could not be accurately measured.
These data demonstrate that Slbp mutation enhances PEV of the w m4 inversion, suggesting that Slbp mutation augments heterochromatin formation. Our data show that mutation of a factor required for proper processing and post-transcriptional regulation of histone mRNAs alters chromatin structure. Furthermore, we propose that changes in chromatin structure in Slbp mutants contribute to overall genomic instability.

The Steady State Level of Replication-Dependent Histones Does Not Change in Slbp Mutants
Alterations in histone gene copy number and histone protein levels result in changes in chromatin structure and genome stability [2,10,37]. Given the known involvement of Slbp in histone mRNA biogenesis, as well as our observations that chromatin structure and genome stability are compromised in Slbp mutants, we determined whether the total amount of histone protein in Slbp mutants is different than wt. To do this, we measured histone H2b and H3 levels by western blot analysis of whole 3 rd instar larval extracts from homozygous Slbp 15 and Slbp 10 mutants, and compared the amount of histones to wild type counterparts in at least five separate experiments. We found no substantial or consistent change in level of histones H2b and H3 in Slbp 15 or Slbp 10 mutants ( Figure 4A). As an indication that we could detect a difference in the amount of core histone protein, we consistently observed a decrease in H2b and H3 in H2aV 810 mutant larvae. H2aV encodes a histone variant that when mutated causes misprocessing of histone mRNAs [25] and suppresses PEV [39]. We also measured steady state H2a and H3 abundance in histone extracts prepared from isolated nuclei, and again found no effect due to Slbp mutation (data not shown). These results indicate that the bulk amount of histone proteins present in Slbp 10 and Slbp 15 null mutants is similar to wild type. We cannot rule out either very subtle alterations in protein levels undetectable by our western blot assay or tissue specific differences in histone protein production. However, we conclude that the genomic instability and modification of PEV that we observed in Slbp mutants cannot be explained by large changes in the total amount of replicationdependent histone proteins

Distribution and Abundance of Euchromatic and Heterochromatic Markers Do Not Change in Slbp Mutants
Our results demonstrate that Slbp mutants exhibit genomic instability and changes in the extent of pericentric heterochromatin, but not detectable changes in the overall amount of histone protein. An alternative explanation for the observed genomic instability is that differences in timing of histone production during the cell cycle rather than the amount of histone synthesis leads to abnormal chromatin assembly. We previously showed that the misprocessed, polyadenylated histone mRNAs produced in Slbp mutants are not always properly cell cycle regulated and in some tissues accumulate in cells that are not synthesizing DNA [27,29]. Consequently, this misregulation could lead to production of replication-dependent histone proteins outside of S-phase, perhaps interfering with deposition of replication-independent histone variants. Two of these variants, H2aV and Cid, have roles in establishing various types of heterochromatin, while H3.3 is enriched in euchromatin [5]. This led us to hypothesize that the global euchromatin/heterochromatin balance would be shifted in histone pre-mRNA processing mutants.
To measure this, Western blots of whole 3 rd instar larval extracts were probed for euchromatic and heterochromatic markers. We chose dimethylation of H3-K4, a histone modification associated with transcriptionally active genes, as a marker of euchromatin [40,41]. H3-K9 dimethylation (H3K9-me2) was used as a marker for heterochromatin [42]. We observed no reproducible difference in the ratio of H3K4me2 to H3K9me2 between Slbp 15 and Slbp 10 mutant third instar larval protein extracts and wild type controls ( Figure 4B). Furthermore, we detected no substantial difference in any of our mutants when we considered either marker relative to a loading control ( Figure 4B).
It is possible that cell type specific changes in chromatin structure might be masked by assessing H3 methylation status in the whole animal. We therefore decided to ascertain the distribution of euchromatic and heterochromatic markers in a specific tissue, the larval salivary gland. This tissue was chosen because the large polytene chromosomes allow exceptional spatial resolution, and the wild type distribution of certain euchromatic and heterochromatic markers is well characterized. We again chose H3K4-me2 as our euchromatic marker. On polytene chromosomes, this modification is localized to euchromatic interbands which contain most active genes, and is mostly excluded from the chromocenter and other heterochromatic regions [43]. Heterochromatin protein 1 (HP1), a well-studied heterochromatin structural protein, was chosen as a marker of heterochromatin [44]. On polytene chromosomes, HP1 localizes to the chromocenter, telomeres, transposon arrays, and various heterochromatic bands ( Figure 5) [45].
Polytene chromosome spreads of Slbp 15 mutant larvae were immunostained for H3K4-me2 and HP1. In wild type larvae, patterns of H3K4-me2 & HP1 distribution recapitulated previously published studies [43]. However, we observed no gross differences in the distribution or abundance of either marker in Slbp 15 mutant salivary gland chromosomes ( Figure 5). These data support the conclusion that the global distribution of euchromatin and heterochromatin in Slbp mutants is largely unaltered in larval endocycling tissues. However, this does not exclude the possibility that the distribution of euchromatin and heterochromatin might be altered in diploid tissues, such as brain and imaginal discs, as suggested by the PEV assay.

Slbp Mutants Exhibit Proliferation Defects and a Prolonged S-Phase
DNA replication and histone synthesis are coupled [12,46]. One advantage of this coupling might be to prevent DNA damage during S phase. Slowed assembly of chromatin behind the replication fork can result in collapsed forks [47], which can be processed into double strand breaks, causing genomic instability [47,48,49]. Mutations that prolong S-phase in Drosophila, such as in the Orc2, Orc5, PCNA, and Mcm4 genes, produce defects observable in mitotic chromosome spreads, including chromatid breaks [50,51]. If the special 39 end of histone mRNAs is needed for coupling rates of histone protein synthesis with DNA replication, then Slbp mutant cells may have chromatin assembly defects that cause prolonged S-phase, resulting in DNA damage and impaired cell proliferation.
To test whether Slbp mutants exhibit a proliferation defect, we performed a ''twin spot'' analysis in wing imaginal discs by inducing mitotic clones using FLP recombinase [52]. Induction of mitotic recombination in larvae heterozygous for the Slbp 15 mutation and a GFP expressing transgene produced a characteristic twin spot derived from the proliferation of wt and Slbp 15 mutant daughter cells ( Figure 6A). A range of clone sizes was produced, corresponding to the timing of each mitotic recombination event ( Figure 6B). Cell proliferation was assessed by comparing the number of Slbp 15 mutant to the total number of cells in each twin spot. The number of cells was determined using the two confocal images from a Z-stack that contained the largest two-dimensional area. To get an estimate of relative proliferation rate, we plotted both mutant and wt cell numbers vs. the total number of cells for each twin spot and fitted a trend line to each set of points ( Figures 6B). This analysis revealed a significant proliferation defect in Slbp 15 mutant imaginal disc cells (p,0.04) ( Figure 6B). Furthermore, this defect is more pronounced in larger clones, most likely because SLBP protein becomes increasingly depleted during each successive cell division. These results are consistent with the observation that Slbp 15 mutant larvae exhibit a developmental delay before the onset of lethality.
If DNA damage and genome instability are caused by impaired chromatin assembly during DNA replication, we would predict that progression of cells through S-phase would be slowed relative to other phases of the cell cycle. To assess whether the proliferation defect observed in Slbp mutant cells corresponds with impaired progression through S-phase, we utilized FACS to profile cell cycle phasing in wing imaginal disc cells of 3 rd instar wt and Slbp 15 mutant larvae ( Figure 6C). Using data from three separate experiments, we observed that cell cycle phasing in Slbp 15 wing disc cells was significantly different than wt (p,0.0001). In Slbp 15 mutant wing discs, 41.3% of cells were in G1, compared with 30.0% in wt discs. Similarly, 47.7% of Slbp 15 mutant cells were in S-phase, compared with only 36.2% in wt discs. Conversely, we observed relatively few cells with G2 DNA content in Slbp 15 mutant wing discs (11.3%), compared with wt (33.9%) ( Figure 6C). Because it is difficult to distinguish cells in G1 from very early S-phase by FACS, the increased G1 population in Slbp mutants may represent cells that are progressing very slowly through early S-phase. These data demonstrate that Slbp 15 mutant cells have a cell cycle defect, and are consistent with a model where impaired chromatin assembly capability impedes S-phase progression, producing DNA damage, genomic instability, and delayed mitosis.

Discussion
In this report we identified four forms of genomic instability in Drosophila Slbp mutants: an increase in loss of heterozygosity, localized changes in heterochromatin structure as measured by modification of PEV, tetraploidy, and chromosomal breaks. Some of these measurable genomic defects may cause the lethality of Slbp null mutations. A large body of prior work indicates that SLBP participates in the processing of histone pre-mRNA resulting in formation of a unique mRNA 39 end that is not polyadenlylated [7,8]. In Drosophila, Slbp is essential for histone pre-mRNA processing, and all replication-dependent histone mRNAs are instead converted into polyadenylated mRNAs [27,29]. There is Figure 5. The global balance of euchromatin and heterochromatin remains unchanged in Slbp mutants. Polytene chromosome spreads of wt and Slbp 15 mutants were stained with antibodies for HP1 (red) and H3K4-me2 (green). DNA was stained with DAPI (blue). Leftmost panels are low magnification images of the entire genome. HP1 stains the chromocenter (white arrows) and telomeres, as well as other heterochromatic bands. Rightmost panels are high magnification images of a single homologous stretch of autosome for each genotype. Position of high magnification images are indicated on low magnification images with yellow arrows. doi:10.1371/journal.pone.0008168.g005 no evidence from genome wide studies that Slbp is capable of binding non-histone RNAs [53] and SLBP has never been implicated in the direct regulation of any genes other than the histone genes. Thus, the simplest interpretation of our data is that proper histone mRNA 39 end formation and its accompanying regulation are necessary for accurate replication and propagation of the genome. Changes in chromatin assembly and structure in Slbp mutants might generally increase the susceptibility of the genome to damage. Some parameters that we anticipated might be causal for these genomic instability phenotypes are relatively normal. We observed no changes in total histone protein levels, and no overt, global changes in chromatin structure as measured cytologically in polytene chromosomes using antibodies that recognize a histone modification enriched in euchromatin or heterochromatin. One possibility is that perturbations in the rate of histone protein expression during S phase or timing of expression relative to the cell cycle, rather than to changes in absolute amounts of histone protein, causes the genomic instability in Slbp mutants.
This possibility is consistent with our observations of impaired proliferation of Slbp mutant wing imaginal disc cells. Reduced histone protein synthesis may be compensated by slowing down the cell cycle, which would result in no observed change in histone protein abundance. In addition, reduced proliferation and slow organismal growth of Slbp mutants could be a direct consequence of impaired chromatin assembly in S-phase, which produces Sphase arrest and double-strand breaks in human cell lines [49]. In fact, work in mammalian cell culture demonstrates that the stemloop on histone mRNAs is necessary for coupling histone mRNA stability to DNA synthesis during S-phase by way of the DNAdamage responsive ATR pathway [54,55]. Any change in rate of histone protein synthesis producing a commensurate slow-down in cell proliferation might result in little or no change in assays which measure a snapshot of a dynamic state, such as global chromatin structure or total histone protein abundance.
Our data also suggest that slowed progression through S-phase contributes to the proliferation defect observed in Slbp mutants. This is similar to observations that RNAi knockdown of Slbp in mammalian cells results in delayed progression through S-phase [56]. Unlike Drosophila cells, mammalian cells depleted of SLBP produce very little polyadenylated histone mRNA, and instead fail to accumulate and export normal amounts of histone mRNA [13]. Thus, the common phenotype of S-phase delay may result from impaired production of histone protein, which may be caused by inefficient translation of polyA histone mRNA in the absence of SLBP, which is known to stimulate histone mRNA translation in vertebrates [57,58].
In conclusion, our data suggest that proper histone mRNA 39 end formation is necessary for maintaining genomic stability and normal cell cycle progression. We propose a model in which inefficient chromatin assembly during S-phase in histone mRNA processing mutants causes DNA damage, genomic instability, and problems with cell proliferation, leading to impaired development.

Loss of Heterozygosity (LOH) Analysis
Flies of the following genotypes were used for LOH analysis: y;Slbp 10 ; T(1;4)y + , y;Slbp 10 /TM3, Sb; T(1;4)y + , y;+/TM3, Sb; T(1;4)y + . Wings were dissected and the first 20 bristles of the anterior wing margin were imaged on a standard light microscope. The number of yellow bristles was counted from each image. Statistical analysis was conducted using InStat (Graph Pad), comparing numbers of yellow bristles per wing among classes. Two-tailed p-values from the non-parametric Mann-Whitney test are reported.

Metaphase Spreads
Metaphase spreads were conducted as described [63]. Briefly, brains from wandering 3 rd instar larvae were dissected in saline. Swelling was induced with 0.5% sodium citrate. Neuroblasts were fixed for 10 seconds in a solution of 46% acetic acid, 46% methanol, and 8% ddH 2 O. Brains were then incubated in 45% acetic acid and squashed under a siliconized coverslip, dehydrated in 95% EtOH, rehydrated in 2X SSC, and stained with 0.1mg/ mL DAPI. In experiments assessing chromosomal abnormalities, brains were incubated for 90 minutes in 0.1 mM colchicine to induce mitotic arrest prior to sodium citrate treatment. Individual nuclei were imaged and the karyotype was ascertained. For each brain, the percentage of nuclei in each category was calculated. Microsoft Excel was used to perform unpaired t-tests between groups.

PEV Assay
Flies of the following genotypes were utilized for analysis: In(1)w m4 /w;Slbp 10 /Slbp 15 , In(1)w m4 /w; Slbp 15 /TM3, Sb, In(1)w m4 /w; +/TM3, Sb. Pigmentation was quantified as described [38]. Briefly, newly eclosed flies were collected and aged for four days. Flies were decapitated and pigment from 30 heads per sample was extracted in 1 ml of 30% Acidified Ethyl Alcohol (AEA) over the course of 3 days. Absorbance readings were obtained at 480 nm on an Eppendorf spectrophotometer. P-values from two-tailed paired t-tests were ascertained using InStat (Graph Pad). For each experiment, 10 sample pairs were compared except when comparing Slbp 15 /Slbp 10 and Slbp 15 /+ flies. These transheterozygotes did not eclose at a Mendelian frequency, thus only 3 sample pairs were analyzed.

Clonal Analysis
Clones were generated using the Flp recombinase system described previously [52]. First and second instar larvae were heat shocked for 35 minutes. Wing discs were fixed 3-4 days after heat shock in 4% paraformaldehyde for 20 minutes and stained for Armadillo (Developmental Studies Hybridoma Bank) (1:50), as well as DAPI. Clones were imaged on a Zeiss 510 confocal microscope, and cell counts were taken from the two slices which represented the largest total area. P-values for twin spot analysis were obtained using a two tailed paired t-test.

FACS Analysis
FACS analysis on wing discs was performed as described previously [65,66]. Briefly, wandering 3rd instar wing imaginal discs were dissected in PBS during a stage matched one hour developmental window. Discs were dissociated with 10X trypsin-EDTA (Sigma) and DNA was labeled with 1X Hoescht 33342 (Acros Organics) in 1X PBS for 3 hours rocking. Flow cytometry was performed using a LSR II (BD), and the data was analyzed with FloJo version 7.2.5 software (FloJo). Percentages of G1, S, and G2 were calculated using the ModFit LTTM software (Verity Software House). P-values were obtained using a x 2 independence test.