The Hsp70 chaperone is a major player in stress-induced transposable element activation

Abstract

Previous studies have shown that heat shock stress may activate transposable elements (TEs) in Drosophila and other organisms. Such an effect depends on the disruption of a chaperone complex that is normally involved in biogenesis of Piwi-interacting RNAs (piRNAs), the largest class of germline-enriched small noncoding RNAs implicated in the epigenetic silencing of TEs. However, a satisfying picture of how chaperones could be involved in repressing TEs in germ cells is still unknown. Here we show that, in Drosophila, heat shock stress increases the expression of TEs at a posttranscriptional level by affecting piRNA biogenesis through the action of the inducible chaperone Hsp70. We found that stress-induced TE activation is triggered by an interaction of Hsp70 with the Hsc70-Hsp90 complex and other factors all involved in piRNA biogenesis in both ovaries and testes. Such interaction induces a displacement of all such factors to the lysosomes, resulting in a functional collapse of piRNA biogenesis. This mechanism has clear evolutionary implications. In the presence of drastic environmental changes, Hsp70 plays a key dual role in increasing both the survival probability of individuals and the genetic variability in their germ cells. The consequent increase of genetic variation in a population potentiates evolutionary plasticity and evolvability.

Keywords: Hsp70; evolution; transposable elements.

Fig. 1.

Heat shock treatment impairs posttranscriptional silencing of TEs. (A) TE expression profiles in ovaries and testes from control and heat-stressed flies analyzed 2 d after the heat treatment. Changes in the levels of TE transcripts were determined by quantitative RT-PCR relative to Rp49 expression. Data are expressed as the mean ± SEM of 3 biological replicates and are normalized to the no heat shock controls (*P < 0.05; **P < 0.01). (B) MA plot for H3K4me3 and H3K9me3 marks on all of the TEs between the 2 conditions (heat-stressed versus control). Each point represents one TE, with the corresponding log₂ fold change and the average log₂ counts per million (cpm). Red dots highlight TEs with statistically significant changes (FDR < 0.05). Horizontal blue lines correspond to a fold change (FC) of 2.

Fig. 2.

Interaction of Ago3 with components of the Hsc70/Hsp90 chaperone machinery. (A) Ago3 immunoprecipitates from wild-type ovaries subjected to Western blotting analysis are probed with antibodies against the indicated cochaperones. Coimmunoprecipitation experiments clearly show that all of the proteins tested are direct interactors of Ago3; the asterisk indicates the IgG heavy chain. (B) Immunofluorescence analysis confirms the colocalization of Ago3 with the same cochaperones in the nuage. Images were obtained using a 40× magnification.

Fig. 3.

Components of the Hsc70/Hsp90 chaperone machinery are required for TE silencing. (A) Activation of various TEs by silencing cochaperone genes Droj2, Hsc70-4, Hop, and dFKBP59. Note that, for both Hop and dFKBP59, we analyzed 2 independent RNAi transgenic lines, while, for Hsc70-4, we tested the dominant negative variant DN-Hsc70-4. Germline expression was driven with nos-Gal4 (nosG4). To temporally control the expression of DN-Hsc70-4 and Droj2 RNAi knockdown, the tub-Gal80^ts system was used; the black line indicates the control value. TE expression levels from Hop and dFKBP RNAi samples were compared to nosG4 sample. For Droj2 RNAi and DN-HSC70-4, the same genotypes raised to the permissive temperature for GAL80^ts (18 °C) were used as controls. Results are presented as mean ± SEM (*P < 0.05; **P < 0.01). (B) Silencing of cochaperones in the male germline derepresses Stellate sequences, as visualized by the presence of crystals in primary spermatocytes. Images of squashed testes were acquired using a 40× magnification.

Fig. 4.

Heat shock affects piRNA loading onto Ago3. (A) qRT-PCR analysis of a panel of ovary-enriched piRNAs in ovary lysates from HHS and control (CTR) samples. Relative abundance of each piRNA was determined by the 2^-ΔΔCt method using 5S piRNA as an internal control. (B) Western blot assay showing Ago3-IP efficiency in both stressed and control samples. Ago3 signal intensity in each IP sample was performed to equalize the Ago3 protein in HHS and control immunoprecipitates. (C) Quantification by qRT-PCR of a specific set of piRNAs targeting different stress-induced TEs in Ago3 immunoprecipitates from HHS and control ovary lysates. Relative abundance of each piRNA was determined by the 2^-ΔΔCt method using, as internal control, Minisatellite#1 piRNA that is not modulated by HHS treatment. Data were expressed as the mean values ± SEM of 2 biological replicates (*P < 0.05).

Fig. 5.

Hsp70 interacts with Ago3 and the components of the chaperone machinery after heat shock. (A) One hour after heat shock, Hsp70 localizes in the nuage and (B) interacts with Ago3 and other chaperones, except dFKBP59. (C) Two days after heat shock, Hsp70 interacts with Hsp90 in forming cytoplasmic bodies outside the nuage (see arrows). Seven days after heat shock, Hsp83 is again localized in the nuage, while Hsp70 signals are absent. (D) Hsp830−Hsp70 cytoplasmic aggregates also include Ago3 and Hop and colocalize to lysosomes, as shown by their colocalization with the specific lysosomal marker LysoTracker (see arrows). (E) Solubility assay. Western blot of fractionated proteins obtained from adult ovaries after 1 d of recovery from HHS treatment. Image in A was obtained using a 63× magnification; images in C and D were acquired using a 40× magnification.

Fig. 6.

Hsp70 activation and colocalization with piRNA components to cytoplasmic bodies. Functional inactivation of chaperones induces Hsp70 expression at normal temperatures, and it colocalizes with Ago3 and the other chaperones to cytoplasmic bodies outside the nuage. (A) Ovaries from Hsp83 mutants (hsp83⁸⁴⁴⁵/hsp83^e4A) showing cytoplasmic bodies outside the nuage (see arrows for examples) in which Ago3 colocalizes with Hop and Hsp70 (see arrowheads for an example). (B) Hop-depleted ovaries [nosG4 > Hop^{RNAi (34002)}] showing Ago3 and Hsp83 in cytoplasmic bodies with Hsp70 (arrows for examples). (C) The quantitative reduction of chaperones Hsp83 or Hop, or other factors involved in piRNA biogenesis such as Spindle-E (hls^∆125/hls^spnE1), Ago3 (nosG4 > Ago3^RNAi), and Aubergine (aub^HN2/aub^QC42) induces activation of Hsp70 at normal temperature. (D) The expression of UAS-Hsp70 transgenic construct driven by nos-G4 at normal temperatures induces the formation of cytoplasmic bodies containing Ago3 and Hsp83 (arrows for examples). Images in A, B, and D were acquired using a 40× magnification.

Fig. 7.

Hsp70 is required for transposable element derepression after heat shock. TE transcript levels were analyzed in germline tissue of flies carrying a complete deletion of Hsp70 genes cluster [Df(3R)Hsp70A, Df(3R)Hsp70B] after 2 d of recovery from HHS. Data are represented as the mean ± SEM of 3 biological replicates quantified by qRT-PCR. Heat shock had no significant effect on TE transcripts.

Fig. 8.

Heat shock stress causes degeneration of ovaries in flies that lack Hsp70. (A) The ovaries from Hsp70-deficient flies appear normal without heat shock. (B) Twenty-four hours after HHS treatment, the ovaries contain degenerating egg chambers that are heavily stained by antibody against cleaved Caspase-3 (in green). (C) Western blot analysis shows that Hsp70 is not induced by heat shock in ovaries of females lacking the Hsf. Hsp70 can be induced by heat shock in hsf mutant females carrying a transgenic wild-type copy (hsf^+t8) of hsf gene. (D) No degeneration is observed without heat shock in ovaries from hsf transheterozygous mutants (hsf⁴/hsf¹), while (E) strong degeneration 24 h after heat shock is clearly visible as indicated by Caspase-3 staining. (F) In ovaries from hsf mutant females that also carry a transgenic wild-type copy of hsf, an increased amount of TE transcript is detectable after heat shock; results are presented as mean ± SEM (*P < 0.05; **P < 0.01). (G) Hsc70-4 (nosG4 > DN-Hsc70-4^71S) and Droj2 (nosG4 > Droj2^RNAi) mutant females show DAPI-stained ovaries with empty germaria arrested in stage 2, while Ago3 (nosG4 > Ago3^RNAi) mutant females show normal ovaries. (Scale bar, 100 μm.) (H) Bright-field images of freshly dissected whole testes from Hsc70-4, Droj2, and Ago3 mutant adult males show testes (arrows) that are abnormal compared to those of wild-type males and lack spermatocytes. (Scale bar, 200 μm.) Ovary images were acquired using a 10× objective. Whole testes were photographed using a stereomicroscope equipped with a NIKON D5000 camera (16× magnification).

Fig. 9.

Hsp70 transgene expression driven by nosG4 is able to induce TE and Stellate (Ste) activation in absence of stress. (A) qRT-PCR analysis shows a significant increase of TE transcripts in ovaries from transgenic females expressing 2 genes of the Hsp70 genes cluster (Hsp70Ab and Hsp70Bbb). Results presented as mean ± SEM (*P < 0.05; **P < 0.01). (B) Testes from control males stained by a specific anti-Stellate antibody. No signal is evident. (C and D) Testes from males expressing the (C) Hsp70Ab or (D) Hsp70Bbb transgenes show the presence of crystalline aggregates in spermatocytes after staining by anti-Ste antibody. The dashed line delimits spermatogonia. Images of squashed testes were acquired using a 40× magnification.

Fig. 10.

Model of cochaperone complex involved in piRNA biogenesis and TE silencing. (A) In normal conditions, the chaperone complex and Ago3 protein cooperate in mature RISC formation. Since, in normal conditions, Hsp70 gene cluster is completely repressed, the model is applicable also in case of flies carrying a complete deletion of such a cluster. (B) During heat shock, the chaperone−Ago3 complex is disrupted, and all of the factors are targeted to the lysosomes for degradation by the involvement of inducible Hsp70 chaperone. As a consequence, TEs are activated.