The JIL-1 Kinase Affects Telomere Expression in the Different Telomere Domains of Drosophila

In Drosophila, the non-LTR retrotransposons HeT-A, TART and TAHRE build a head-to-tail array of repetitions that constitute the telomere domain by targeted transposition at the end of the chromosome whenever needed. As a consequence, Drosophila telomeres have the peculiarity to harbor the genes in charge of telomere elongation. Understanding telomere expression is important in Drosophila since telomere homeostasis depends in part on the expression of this genomic compartment. We have recently shown that the essential kinase JIL-1 is the first positive regulator of the telomere retrotransposons. JIL-1 mediates chromatin changes at the promoter of the HeT-A retrotransposon that are necessary to obtain wild type levels of expression of these telomere transposons. With the present study, we show how JIL-1 is also needed for the expression of a reporter gene embedded in the telomere domain. Our analysis, using different reporter lines from the telomere and subtelomere domains of different chromosomes, indicates that JIL-1 likely acts protecting the telomere domain from the spreading of repressive chromatin from the adjacent subtelomere domain. Moreover, the analysis of the 4R telomere suggests a slightly different chromatin structure at this telomere. In summary, our results strongly suggest that the action of JIL-1 depends on which telomere domain, which chromosome and which promoter is embedded in the telomere chromatin.


Introduction
Drosophila telomeres are the best-studied telomerase exception to date. The transposition of three special non-LTR retrotransposons; HeT-A, TART and TAHRE (HTT), buffers the receding chromosome ends when needed [1,2]. Therefore, in Drosophila the genes responsible for telomere elongation are embedded in the telomere chromatin and to understand telomere regulation in this organism it is necessary to understand gene expression from this genomic compartment.
Drosophila telomeres are organized in two different domains; the cap domain protecting the very end of the chromosome, and the telomere domain composed of the HTT array. Additionally, immediately adjacent to the HTT array exists the subtelomere domain, composed of the Telomere Associated Sequences (TAS) [3][4][5]. Although the subtelomere domain is not strictly part of the telomere, its closeness to the HTT array influences some of its functions.
Previous studies, based on the variegation of reporter genes, suggested that TAS sequences nucleate a compacted chromatin restrictive of gene expression at the telomeres, causing a telomere position effect (TPE) [6]. Later, molecular studies detected the presence of chromatin marks typical of repressive and compacted chromatin in TAS, such as H3K27me3 and Polycomb [3,4,6,7].
We aimed to better understand the regulation of gene expression in Drosophila telomeres. With this objective we assayed the influence of the JIL-1 kinase over Drosophila telomere and subtelomere domains. In Drosophila, JIL-1 is an essential gene responsible for the phosphorylation of the Ser10 of Histone H3 when cells are in interphase or when a stress response is needed [8][9][10]. Besides the telomeres, JIL-1 localizes at the interband region of polytene chromosomes and has been shown to be involved in maintaining chromosome structure and in protecting genomic regions from an excess of heterochromatinization [11][12][13][14][15]. JIL-1 has been found to localize among other genomic locations, at the HTT array and to be the first positive regulator of the expression of the telomeric retrotransposons [3,16]. In this study, we report that JIL-1 protects the telomere domain from the spreading of repressive chromatin from the flanking subtelomere domain and that JIL-1 action depends on which telomere domain, which chromosome and which promoter is embedded in the telomere chromatin.

Results
In order to assess the possible role of JIL-1 regulating gene expression at the HTT array, we chose several lines that contain a mini-white reporter gene at the telomere domain (HTT array), or at the subtelomere domain (TAS sequences) from different chromosomes [6,17,18]. The mini-white gene from the chosen reporter lines is expressed from the same basic promoter, which does not include the eye-specific enhancers (see references above). All flies in this study are in a white minus background (w-, white eyes). The amount of red color that is observed in the eyes of the reporter lines (EY08176, EY00453, EY09966, 39C-27, 39C-5, 39C-72 and 118E-5), depends on where in the chromosome the mini-white gene has been inserted. In order to evaluate the effect of the different mutations assayed in this study over the reporter lines, we have compared the descendants of the crosses with the JIL-1 and the Su(var)2-5 mutant flies (also in a w-background and with white eyes) with the descendants resulting from crossing the flies containing the reporter lines with a w 1118 strain (white eyes). This control cross is necessary in order to be able to evaluate the presence of the reporter gene in heterozygous state. We search for changes in eye color revealing a TPE effect.

JIL-1 is necessary for gene expression at the telomere domain (HTT)
By crossing flies containing mutant alleles of JIL-1 with lines with a mini-white gene inserted at the HTT array from different chromosomes, we screened for TPE effects in the eyes of the first generation ( Figure 1). In simple heterozygous combinations of JIL-1 alleles (Figure 1, 2 nd and 3 rd columns), we observed subtle enhancer position effect variegation in the telomere line EY08176 (1 st row) and no effect with the telomere line EY00453 (3 rd row). We did not observe any effect in the EY09966 line corresponding to the 4R telomere, likely because the reporter gene is already heavily repressed in this line (2 nd row). Nevertheless, when we combined both JIL-1 alleles in a trans-heterozygous mutant (JIL-1z60/JIL-1 h9 ) with the line EY08176, a clear enhancer effect of TPE was revealed (1 st row, 4 th column).
Following, we investigated if, in comparison with the expression in the wild type situation of the reporter line, a decrease in the expression of the HTT retroelements upstream to the mini-white insertion was observed in JIL-1 mutants. Quantitative real-time PCR was used to detect expression of the transcript containing both, the telomeric element and the mini-white reporter (Figure 2A and D). When compared with the expression detected in the reporter in a wild type background, JIL-1 heterozygous mutants show lower telomere transcription from the HTT element adjacent to the insertion in the 2 nd and in the 4 th chromosomes (EY08176, Figure 2A and D) (EY09966, Figure 2B and F). This reduction is in agreement with a reduction of the overall expression of HeT-A in the same reporter lines ( Figure 2B, 2D and [16]) suggesting a global role of JIL-1 over the HTT array.
In addition, we assayed the possible effect of flies containing mutant alleles of the Su(var)2-5 gene (HP1a), over the reporter lines inserted at the HTT array ( Figure 3). Interestingly, only a subtle change in mini-white expression was detected when we assayed the Su(var)2-5 05 mutation over the lines EY08176, EY09966 and EY00453 (Figure 3 and [18]).

JIL-1 acts as a suppressor of TPE over the TAS domain
JIL-1 z60 /+ and JIL-1 h9 /+ heterozygous alleles did not produce an appreciable effect over the mini-white inserted at the TAS domain of chromosomes 2R (39C-27) and 2L (39C-5) ( Figure  4A, 2 nd and 3 rd columns). Nevertheless, when the dose of JIL-1 was lowered in a trans-heterozygous combination, a significant increase in the mini-white expression was observed when compared with the control cross ( Figure 4A 1 st column), revealing a suppressor effect of JIL-1 over the TAS domain of these two telomeres ( Figure 4A, 4 th column). As expected, we did not observe any change in mini-white expression when the TAS lines (39C-27 and 39C-5) were combined with the HP1a allele Su(var)2-5 05 ( Figure 4A, 5 th column) [7].
In order to understand which changes in the TAS chromatin could explain the increase in mini-white transcription in the JIL-1 z60 /JIL-1 h9 allelic combination, we performed Chromatin Immunoprecipitation (ChIP) experiments. The chromatin environment at the TAS domain has been shown to be enriched in H3K27me3, which serves for targeting the Polycomb complex to the subtelomere domain [3,7]. We therefore developed the ChIP assay with an antibody against H3K27me3 ( Figure 4B). The ChIP data showed a significant decrease in the presence of the H3K27me3 mark inside the TAS domain in JIL-1 mutant alleles, suggesting a possible spreading of this mark into flanking domains. We have not been able to detect an increase in H3K27me3 at the HTT array indicative of the consequential spreading of this mark from the TAS towards the HTT array, ( Figure 4B and 5). Because the HTT array flanking the TAS domain at the 2R (39C-27) and 2L (39C-5) telomeres cannot be distinguished from the HTT arrays of other chromosomes, it is not surprising that the ChIP assay was not sensitive enough to reveal a slight increase in H3K27me3 in the vicinity of the HTT array.
The TPE assays with the JIL-1 mutant alleles have shown the same result for both TAS domains, TAS-L (2L, 3L telomeres, line 39C-5) and TAS-R (2R, 3R and XL telomeres, line 39C-27) [19], demonstrated by the increase in eye color of the resulting descendants ( Figure 4A). These results suggest that the protective boundary built by JIL-1 also exists in the other Drosophila telomeres.
As an additional control, we analyzed if the TAS reporter lines 39C-27 and 39C-5 behave as had been previously described for TAS sequences [6,7]. Therefore, we performed control crosses with mutant alleles of genes whose effect over the expression in the subtelomeric domain had been already reported (Table S1) and obtained results consistent with previous studies [6,7].

The subtelomere domain of the 4 th chromosome and the TAS domains have a different chromatin
The 4R telomeres of the reporter lines here analyzed (39C-72 and 118E-5) do not have a TAS domain flanking the HTT array. Instead, a small transition zone (≈ 5Kb) composed of decayed transposable elements together with pieces of the HTT elements lies at the subtelomere region of the 4R arm (Flybase GBrowse and [6,20]). Our experiments have revealed that the JIL-1 kinase and HP1a easily affect the 4 th chromosome subtelomeric lines 39C-72 and 118E-5 ( Figure 6). The effects that these two proteins exert at the 4R subtelomere are similar to the ones that they produce at the HTT array of the other chromosomes ( Figure 1 4 th column and Figure 2A, B, D, E and Figure 3). JIL-1 behaves as an enhancer of TPE while HP1a shows a strong suppressor effect in all cases where the reporter is inserted at the telomere domain. Additionally, we performed control crosses with insertions inside the arms of the 4 th chromosome, the lines 39C-12 and 39C-42 [21]. No change in the eye color of the JIL-1 mutant descendants was observed, while in Su(var)2-5 05 mutants the expected increase in red eye color with respect to the control cross was observed ( Figure  S1) [21]. Therefore, our results suggest that the subtelomere of the 4 th chromosome depends on JIL-1 for gene expression and on HP1a for gene silencing ( Figure 6).

JIL-1 is a positive regulator of gene expression at the telomere domain (HTT)
Both the experiments of telomere position effect on the miniwhite insertions (Figure 1) and the real-time PCR quantifications of the expression of the different telomeric elements (Figure 2), indicate that the lower amount of JIL-1 present in the mutant alleles JIL-1 z60 and JIL-1 h9 here assayed, results in a decreased expression of the HTT array. These experiments, in accordance with our recent published data [16], confirm that JIL-1 is necessary to obtain wild type levels of gene expression from the HTT array. EY08176 for the 2R telomere (1 st row), EY09966 for the 4R telomere (2 nd row), and EY00453 for the 3L telomere (3 rd row). Note that descendants from the JIL-1 heterozygous, JIL-1 z60 /+ or JIL-1 h9 /+, show a subtle enhancer effect in the telomere line EY08176 (1 st row) and no effect with the telomere line EY00453 (3 rd row). We did not observe any effect in the EY09966 line corresponding to the 4R telomere, likely because the reporter gene is already heavily repressed in this line (2 nd row). When the JIL-1 z60 /JIL-1 h9 transheterozygous allele is assayed with the EY08176 reporter line, an enhancer effect of TPE is observed (4 th column). Additionally, these experiments have also revealed that the effect of JIL-1 is stronger over the HeT-A promoter than over the mini-white promoter existent in the reporter lines used in this study. Similarly, when we assayed the effect of mutations of the Su(var)2-5 gene, known to greatly de-repress the expression of the HeT-A retrotransposon [16,[22][23][24], we only obtained a faint de-repression of the mini-white gene of the lines EY08176, EY09966 and EY00453 (Figure 3). In agreement with our observations, Frydrychova and colleagues found similar results using two additional Su(var)2-5 mutant alleles, Su(var)2-5 02 and Su(var)2-5 04 , over the same reporter lines [23].
In summary, JIL-1 and HP1a control gene expression from the telomere domain in Drosophila, being the HeT-A promoter especially sensitive to their effect.

JIL-1 protects the HTT array of the chromatin spreading from the TAS domain
Although JIL-1 has not been found to localize at the subtelomere domain, TAS, the observation that gene expression varies in this domain in a JIL-1 mutant background, suggests a role of this protein related with a putative boundary between the telomere and subtelomere domains. We have observed a pronounced increase in mini-white expression at the TAS domains from the 2L and 2R telomeres when placed in a JIL-1 trans-heterozygous (JIL-1 z60 /JIL-1 h9 ) mutant background ( Figure 4A, 4 th column). Accordingly, the ChIP data reveals a significant decrease of H3K27me3 upstream of the mini-white insertion when a JIL-1 trans-heterozygous background is present ( Figure 4B). Integrating these results, we propose a model in which the JIL-1 kinase acts as a  [16]). Moreover, our model also reflects the previously reported role of JIL-1 in the protection of the excessive spreading of heterochromatin to adjacent domains [13,14], and suggests that JIL-1 could exert a barrier function at the HTT-TAS boundary, based on the results obtained from this study.

Not all the subtelomere domains have the same chromatin characteristics
We have not observed a different behavior in the control of gene expression when studying the HTT array of the 4 th and the 2 nd chromosomes. Interestingly, we did find a significant difference when looking at the subtelomere of the 4R telomere ( Figure 6). In the reporter lines 39C-72 and 118E-5 from the subtelomere domain of the 4 th chromosome, we have found that JIL-1 is important to allow gene expression and HP1a is necessary to repress it. This finding indicates that the chromatin of the subtelomere domain in the 4 th chromosome is not equivalent to the TAS chromatin in the other chromosomes.
In contrast, the results here described indicate that the chromatin at the 4R subtelomere of the reporter lines used in this study is less compacted and more permissive to gene expression than the TAS domains from the other chromosomes, and even than the HTT array. This scenario suggests that in the 4R telomere of these lines the compaction of the chromatin is in the opposite orientation than in the rest of the telomeres. It would be interesting to study if this reversed chromatin organization with respect to the other telomeres has a particular role in the general telomere function in Drosophila.

Conclusions
We report how the presence of the JIL-1 kinase is needed at the telomere in Drosophila in order to allow gene expression of the promoters embedded in this domain. The lack of JIL-1 causes an increase of silencing at the HTT array and a release of silencing at the subtelomeric domain likely by the spreading of heterochromatin towards adjacent domains. Finally, we discovered that the telomere and subtelomere domain of the 4R arm of some lines might have a different chromatin organization with respect to the other Drosophila telomeres.

Fly stocks and crosses
Fly stocks were maintained and crosses performed at 25°C on standard Drosophila corn meal medium. JIL-1 z60 /TM6 and JIL-1 h9 /TM6 stocks were provided by Kristin M. Johansen. Both JIL-1 mutations are hypomorph. The JIL-1 h9 allele contains a molecular lesion that results in a deletion inside the C-terminal of the protein and the JIL-1 z60 contains a mutation that eliminates the promoter and 5'UTR of the JIL-1 gene [9]. Su(var)2-5 05 contains a frame shift that results in a peptide containing only the first 10 amino acids of the original HP1 protein [22]. To assay for TPE modifications, males from homozygous reporter gene stocks, with a P-element inserted at a telomere or at the subtelomere domains, were crossed with virgin females from mutant stocks of the proteins of interest. The eye phenotype of the resulting F1 progeny with the mutation was compared with their siblings lacking the mutation. For each cross 20 flies were analyzed (males and females) and collected 2 days after eclosion and briefly frozen. Each cross was repeated three times. Images were obtained from mutant and control F1 males using Zeiss Discovery V.8 microscope with AxioCam MRE5 camera and Axiovision 4.7 software.

RNA Extraction and cDNA synthesis
Total RNA was isolated from ten whole third instar larvae and extracted using RNeasy Mini Kit (Qiagen) according to manufacturer's protocol. RNase Free DNase Set (Qiagen) was used to remove genomic DNA contaminations as follows: one on column during the extraction accordingly to manufacturer's protocol, and two in solution for 2 hours at 37°C. RNA was cleaned by precipitation and its quality was assessed using NanoDrop spectrophotometry.
One microgram of RNA was reverse transcribed into cDNA using Transcriptor First Strand cDNA Synthesis Kit (Roche) with Oligo d(T) primers, and the expression of the different transcripts analyzed by quantitative real-time PCR. For each fly strain, two independent RNA extractions were prepared and analyzed three independent times. Primers used for real time PCR: HeT-A_F (CCCCGCCAGAAGGACGGA) and HeT-A_R (TGTTGCAAGTGGCGCGCA) included in the 3'UTR,

Quantitative real-time PCR
Quantitative real-time PCR was performed to determine HeT-A expression and in ChIP experiments. The iQ5 Multicolor Real-Time PCR Detection System was used and the iQ TM SYBR ® Green Supermix (BioRad) was used to prepare the reactions. Relative levels of HeT-A expression were determined using the threshold cycle and normalized to actin levels (or RpL32 for ChIPs). Three independent experiments of two samples each strain were performed. Figure S1. JIL-1 control crosses. JIL-1 mutations have no effect on the expression of a reporter gene inserted in the arms of the 4 th chromosome (39C-12 and 39C-42). The Su(var)2-5 05 allele has the expected suppressor of variegation effect in these same lanes [21]. (TIF) Table S1. Subtelomeric domain control crosses. Previously reported mutant alleles from su(var)3-9, polycomb and trithorax were crossed with lines containing the reporter gene inserted in the subtelomeric domains of the 4 th (39C-72 and 118E-5) and 2 nd chromosomes respectively (39C-5 and 39C-27). (DOC)