Small heat shock proteins determine synapse number and neuronal activity during development

Loading metrics

Open Access

Research Article

Elena Santana, 

Teresa de los Reyes, 

Sergio Casas-Tintó


Published: May 21, 2020


AbstractEnvironmental changes cause stress, Reactive Oxygen Species and unfolded protein accumulation which hamper synaptic activity and trigger cell death. Heat shock proteins (HSPs) assist protein refolding to maintain proteostasis and cellular integrity. Mechanisms regulating the activity of HSPs include transcription factors and posttranslational modifications that ensure a rapid response. HSPs preserve synaptic function in the nervous system upon environmental insults or pathological factors and contribute to the coupling between environmental cues and neuron control of development. We have performed a biased screening in Drosophila melanogaster searching for synaptogenic modulators among HSPs during development. We explore the role of two small-HSPs (sHSPs), sHSP23 and sHSP26 in synaptogenesis and neuronal activity. Both sHSPs immunoprecipitate together and the equilibrium between both chaperones is required for neuronal development and activity. The molecular mechanism controlling HSP23 and HSP26 accumulation in neurons relies on a novel gene (CG1561), which we name Pinkman (pkm). We propose that sHSPs and Pkm are targets to modulate the impact of stress in neurons and to prevent synapse loss.

Citation: Santana E, de los Reyes T, Casas-Tintó S (2020) Small heat shock proteins determine synapse number and neuronal activity during development. PLoS ONE 15(5):

https://doi.org/10.1371/journal.pone.0233231Editor: Harm H. Kampinga, Universitair Medisch Centrum Groningen, NETHERLANDSReceived: February 18, 2020; Accepted: April 30, 2020; Published: May 21, 2020Copyright: © 2020 Santana et al. 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.Data Availability: All relevant data are within the paper and its Supporting Information files.Funding: We would like to declare that this research has been funded by grant BFU2015-65685P from the Spanish MICINN. 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.

IntroductionSynaptic dynamics remodel neuronal circuits under stress conditions [1]. The Heat Shock Protein family (HSPs) is involved in preserving cellular functions such as stress tolerance, protein folding and degradation, cytoskeleton integrity, cell cycle and cell death [2–7]. HSPs are molecular chaperones that represent an intracellular protein quality system to maintain cellular protein homeostasis, preventing aggregation and promoting protein de novo folding or refolding and degradation of misfolded proteins [8]. In addition, HSPs participate in developmental functions in a stress-independent manner [9, 10]. In Drosophila development small Heat Shock proteins (sHsps) have a specific temporal and spatial pattern of expression [10]. In particular, sHsp23 and sHsp26 show high expression levels in CNS during development, suggesting a role in neural development [10].
sHSPs include a large group of proteins represented in all kingdoms of life [11], with a conserved protein binding domain of approximately 80 amino-acid alpha crystallin [12]. These molecular chaperones were initially described as low molecular weight chaperones that associate early with misfolded proteins and facilitate refolding or degradation by other chaperones and co-factors [11] [13]. However, members of the sHSPs have diverse functions beyond the chaperon activity including cytoskeleton assembly [14], the suppression of reactive oxygen species, anti-inflammatory, autophagy, anti-apoptotic and developmental functions (reviewed in [2]). sHSPs represent the most extended subfamily of HSPs, albeit the less conserved [15]. sHSPs have a conserved primary structure divided in three elements required for their function: 1) a variable N-terminal long-sequence related to oligomerization, 2) the conserved α-crystallin domain required for dimmers formation that represents the main hallmark of sHsps family, and 3) a flexible short C-terminal sequence mediating oligomers stability [11, 16]. Posttranslational modifications in sHSPs shift the folding/degradation balance and, in consequence, alter dimer or oligomer formation and function [11, 17]. This chaperone control system modulates critical decisions for the folding or degradation proteins and a failure causes pathological conditions [17].
HSPs protect synaptic function in the nervous system from environmental insults or pathological factors [18–20] (reviewed in [21]), and are also associated to neurodegenerative diseases, aberrant protein-induced neurotoxicity and disease progression [13]. The sHSPs family is involved as a non-canonical role in Drosophila development and other biological processes such as synaptic transmission [22]. However, its implication in synaptic dynamics during development has not been described yet. Synapse number can be altered due to the influence of physiological parameters (aging, hormonal state, exercise) [23–26], pathological (neurodegenerative process) [27, 28] or induced conditions (mutants) [29] which alter cellular components and pathways [30]. The imbalance between the pro- and anti-synaptogenic pathways modulates the number of synapses [30]. The neuromuscular junction (NMJ) of Drosophila melanogaster is a stereotyped structure well established for the study of synapses [31]. Most of the molecules involved in synaptic transmission are conserved between Drosophila and vertebrates thus, this model system is well established for the study of synapses [32].
Here, we study the contribution of two sHsps, sHp23 and sHsp26 in the development of the CNS and synapse modulation. sHsp23 and sHsp26 are expressed in the CNS during the development [10, 33, 34] but their function remains unclear. In addition, we describe the function of CG1561, named Pinkman (Pkm), as a novel putative kinase that interacts with sHSP23 and sHSP26. Pkm regulates expression and protein stability and participates in the establishment of synapse number during development.

Materials and methods
Drosophila strains
Flies were maintained at 25°C in fly food in cycles of 12 hours of light and 12 hours of darkness. The following stocks were used: UAS.LacZ (gift from Dr. Wurz). Fly stocks from the Bloomington Stock Center: Gal4.D42 (BL-8816), UAS.Hsp23 (BL-30541), P{UAS-mLexA-VP16-NFAT}H2/TM3, Sb1 (BL- 66543), P{LexAop-CD8-GFP-2A-CD8-GFP}2; P{UAS-mLexA-VP16-NFAT}H2, P{lexAop-rCD2-GFP}3/TM6B, Tb1 (BL-66542). Fly Stocks from Vienna Stocks Center: Hsp26-GFP-V5-Flag (VDR318685), UAS.Hsp26RNAi and UAS.CG1561RNAi (VDR106503 KK (1), VDR32634 GD (2) and VDR32635 GD (3)). Fly Stocks from the FlyORF Zurich ORFeome Project: UAS.Hsp26 (F000796).

Drosophila dissection and immunostaining
Drosophila third instar larvae were dissected in phosphate-buffered saline (PBS) and fixed with PFA 4% in phosphate-buffered saline (PBS). Then, the samples were washed with PBST (PBS with 0.5% Triton X-100) and blocked with 5% bovine serum albumin (BSA) (Sigma) in PBST. We quantified the total number of active zones per NMJ of third instar larvae. We used the binary system Gal4/UAS (Brand & Perrimon, 1993) to drive all genetic manipulations to motor neurons (D42-Gal4). Actives zones were visualized using a mouse monoclonal antibody nc82 (1:20, DSHB, IA) which identifies the protein Bruchpilot, a presynaptic element. Neuronal membranes were visualized with rabbit anti-HRP (1:300, Jackson ImmunoResearch, West Grove, PA). Fluorescent secondary antibodies were Alexa 488 (goat anti-mouse, 1:500, Molecular Probes, Eugene, OR) and Alexa 568 (goat anti-rabbit, 1:500, Molecular Probes). Larvae were mounted in Vectashield medium (Vector Labs, Burlingame, CA). Synapse quantifications were obtained from the NMJ Drosophila model in muscle fiber 6/7 of the third abdominal segment only to regulate inter-individual data variability.
To localize sHSP23 or sHSP26, third instar larval brain or NMJ were dissected. We use an Hsp26-GFP-V5 fusion construct. sHSP23 was visualized using an anti-Hsp23 (Sigma-Aldrich S 0821) (1:500), and sHSP26 was visualized using anti-V5 (1:50) (Invitrogen 1718556) and anti-GFP mouse (1:50) (Invitrogen A11122). Drosophila brains were mounted in Vectashield with DAPI medium (Vector Labs, Burlingame, CA).

Image acquisition
Confocal Images were acquired at 1024×256 resolution as serial optical sections every 1 μm. We used a 63x objective with a Leica Confocal Microscope TCS SP5 II (Mannheim, Germany). We used IMARIS software (Bitplane, Belfast, UK) to determine the number of mature active zones with the ‘spot counter’ module.
We visualized Hsp23-Hsp26 co-localization and CaLex signal in ventral ganglia cells of third instar larva brains. We acquired brain images at 1024×1024 resolution as serial optical sections every 1μm at 20x objective. We acquired ventral ganglia cells images at 1024×1024 resolution, 63x objective with magnification of 2.5. We processed the images and analyzed them with LAS-AF (Leica Application Suite software).

Antibody generation

For biochemical assays, 5–10 adult fly heads were lysed in immunoprecipitation lysis buffer (NaCl 150 mM, 0,1% Tween-20 (Polyoxyethylene sorbitane monolaureate), TBS pH 7.5). We incubated Protein A/G agarose beads overnight at 4°C with 2 μl of the indicated antibody or control IgG (1:100), followed by incubation at 4°C for 1 h with supernatants. We washed the beads and resuspended in 1× SDS–PAGE loading buffer for western blot analysis in a 4%–12% gradient SDS-PAGE for the detection of sHSP23 and sHSP26. After electro-blotted onto nitrocellulose 0.45 μM (GE Healthcare) 100V for 1 hour, we blocked the membranes in TBS-Tween-20 buffer with 5% BSA. We incubated the membranes overnight at 4°C in constant agitation with anti-Hsp23 antibody (1:1000) (Sigma-Aldrich S0821), anti-Hsp26 (1:1000) (Abmart) We visualized the antibody-protein interaction by chemoluminescence using IRDye Secondary Antibodies anti-mouse (IRDye 800CW, LI-COR), anti-rabbit (IRDye 680 RD, LI-COR) and developed with Odyssey equipment. We used three RNAi tools to downregulate pkm expression to replicate this condition. pkm RNAi 2 was selected to do the rest of experiments due to the evidences we obtained in the blot assay.

Western blot
5–10 head samples were treated with lysis buffer (TBS1x, 150mMNaCl, IP 50x) and then homogenized and centrifuged 13500 rpm for 5 minutes. After selecting the supernatant we added NuPage 4x (Invitrogen by Thermo Fisher Scientific) and ß-mercaptoethanol 5%. Western blot analysis samples were run in a 4%–12% gradient SDS-PAGE for the detection of sHSP23 and sHSP26. After electro-blotted onto nitrocellulose 0.45 μM (GE Healthcare) 100V for 1 hour, we blocked the membranes in TBS-Tween-20 buffer with 5% BSA. We incubated the membranes overnight at 4°C in constant agitation with anti-Hsp23 antibody (1:1000) (Sigma-Aldrich S0821), anti-Hsp26 (1:1000) (Abmart) We visualized the antibody-protein interaction by chemo luminescence using IRDye Secondary Antibodies anti-mouse (IRDye 800CW, LI-COR), anti-rabbit (IRDye 680 RD, LI-COR) and developed with Odyssey equipment. Tubulin were used as a control.

Gene expression analysis with qPCR
10–15 head tissue samples were treated and homogenized with Trizol (Ambiend for Life techonologies). Chloroform was added and then centrifuged 13000 rpm at 4°C for 15 minutes. After discarding the supernatant, the RNA was treated with Isopropanol and then centrifuged 13000 rpm at 4°C for 10 minutes and washed with 75% Ethanol. RNA pellet was dissolved in DNAase RNAase free water. Then we performed a transcriptase reaction and a qPCR assay using Rp49 gene as a control. Primers for sHsp23, sHsp26, Pinkman and Cat were used: sHsp23 Fv (5′-3′) TGCCCTTCTATGAGCCCTAC, sHsp23 Rv (3′-5′) TCCTTTCCGATTTTCGACAC, sHsp26 Fv (5′-3′) TAGCCATCGGGAACCTTGTA, sHsp26 Rv (3′-5′) GTGGACGACTCCATCTTGGT, pkm Fv (5′-3′) TCGTGCTGGAGGATCTGTCTT, pkm Rv (3′-5′) CCCGGCCAATGATATAGCAT, Catalase Fv (5′-3′) TTCGATGTCACCAAGGTCTG, Catalase Rv (3′-5′) TGCTCCACCTCAGCAAAGTA, rp49 Fv (5′-3′) CCATACAGGCCCAAGATCGT, rp49 Rv (5′-3′) AACCGATGTTGGGCATCAGA.

To analyze the data, we used GraphPad Prism 6 GraphPad Software, La Jolla, CA). Data are shown as mean ± SD. Statistical significance was calculated using D´Agostino & Pearson normality test and a Student’s two-tailed t-test with Welch‐correction. In case data were not normal, we performed a Student´s two-tailed t-test with Mann–Whitney-U correction. For multiple comparisons, we used One‐way ANOVA test with Bonferroni post‐test. *p value ≤ .05; ** p value ≤ .01; *** p value ≤ .001; **** p value .05 were not considered significant.

Technical considerations
Each experiment condition has its own control sample to reduce external variables.

Heat shock proteins modify synapses in CNS
To determine the effect of HSPs in synaptogenesis, we used the UAS/Gal4 Drosophila binary expression system [35] to modify Hsps expression in motor neurons using D42-Gal4 lines. We used UAS-RNAi lines knockdown sHsp20, sHsp22, sHsp23, sHsp26, sHsp27, sHsp40, Hsp67 Ba, sHsp27 Bc, Hsp70 Aa, Hsp70 Ba and Hsp90 (Fig 1A). To visualize the number of active zones in the NMJs we used anti-bruchpilot (brp) antibody. The quantification of the active zones revealed that the knockdown of sHsp20, sHsp22, sHsp26, sHsp27, sHsp40 and Hsp90 during development provoked a reduction in synapse number. In addition, we tested the effect in synapse number of sHsp23, sHsp26 and Hsp70 overexpression (Fig 1B). The results show that the upregulation of sHsp23, sHsp26 or Hsp70 decrease the number of active zones (Fig 1B).
Fig 1. Small heat shock proteins modulate synapses during Drosophila development.Synapses quantification screening with sHsps genetic tools under D42 driver expression. (A) Synapses modulation were detected by sHsp20 RNAi (sHsp20↓), sHsp22 RNAi (sHsp22↓), sHsp23 (sHsp23↓), sHsp26 RNAi (Hsp26↓), sHsp27 RNAi (Hsp27↓), Hsp40 RNAi (Hsp40↓), Hsp90 RNAi (Hsp90↓), (B) UAS.sHsp23 (Hsp23↑) UAS.sHsp26 (Hsp26↑) and UAS.Hsp70 (Hsp70↑) samples. One‐way ANOVA test with Dunn’s multiple comparisons post‐test. *p value ≤ .05; ** p value ≤ .01; *** p value ≤ .001. p value> .05 were not considered significant. Error bars show S.D. (C) Diagram of sHsp23 interactome and (D) diagram of sHsp26 interactome form http://flybi.hms.harvard.edu/results.php.

https://doi.org/10.1371/journal.pone.0233231.g001We focused on the role of two sHSPs, sHSP23 and sHSP26, due to their potential role as non-canonical-sHSPs in the CNS and their unexplored implication in synapses modulation. The upregulation of sHsp23 in presynaptic neurons causes a reduction in synapse number (Fig 1B). In addition, sHsp26 knockdown or upregulation induces a reduction in synapse number (Fig 1A and Fig 1B). Thus, the results suggest that sHSP23 is not required for synapse formation but in excess it is detrimental for the neuron and causes a reduction of synapse number during development. Besides, modification in any direction of sHsp26 expression affects to the correct establishment of synapse number during development, suggesting that sHSP26 fine control is required during development for synapse organization.
According to the interactome (flybase) both chaperones are predicted to physically interact with each other [36] (Fig 1C and Fig 1D). Furthermore, sHSP23 and sHSP26, both interact with: CG11534, CG43755 and Pkm (CG1561) proteins [36] (Fig 1C and Fig 1D).

sHSP23 and sHSP26 colocalize in neurons and interact physically
To determine the expression and subcellular localization of sHSP23 and sHSP26 proteins in larval brain we used a green fluorescent reporter tagged form of sHSP26 (HSP26-GFP-V5) and we generated a monoclonal specific antibody against sHSP26 (S1 Fig). We dissected third instar larvae brain and visualized both sHSPs. The data show that sHSP23 and sHSP26 localize in the cytoplasm of CNS cells, in particular in the optic lobes and the central nerve cord (Fig 2A and 2B`). The co-localization of both proteins occurs in neuroblasts and also in ganglion mother cells and differentiated neurons, compatible with a general role in nervous system development.
Fig 2. sHSP23 and sHSP26 colocalize in CNS.(A-F) Confocal microscopy images of 3rd instar Drosophila larval brain and NMJs. (A) sHSP23 is labeled with anti-GFP antibody driven by D42-Gal4 to visualize its expression in brain regions (magenta). Scale bar size 100 um. (B) sHSP26 is stained with anti-sHSP23 (green). (A`-B`) Magnification images of larval brain. Arrows indicate neuroblast, arrowheads indicate ganglion mother cells and asterisk indicate neurons where sHSP23 and sHSP26 colocalize in the cytoplasm. Scale bar size 100 um. (C-F) sHSP26 is labeled with anti-GFP antibody driven by D42-Gal4 to visualize its expression in NMJ (red), sHSP23 is stained with anti-sHSP23 (green) and neuronal membrane is detected with anti-HRP staining (magenta). Scale bar size 50 um (C`-F`) Magnification images of synaptic boutons in NMJ. (G) Co-Immunoprecipitation assay membrane revealed with sHSP26 (green, arrow) and sHSP23 (red) antibodies in control samples. Fly heads were lysed in immunoprecipitation lysis buffer and incubated with protein A/G agarose beads previously treated with sHSP23 or sHSP26 antibody and IgG antibody as a control. The samples were prepared for western blot analysis. The antibody-protein interaction is visualized by chemoluminescence. Molecular weights are indicated in all the membrane images. * Unknown/unspecific band (H) sHSP23 and sHSP26 interaction diagram.

https://doi.org/10.1371/journal.pone.0233231.g002To further analyze the presence and co-localization of sHSP23 and sHSP26 we analyzed larvae NMJs (Fig 2C–2F`). The confocal images show an accumulation and colocalization of sHSP23 and sHSP26 throughout the NMJ but particularly intense in the synaptic buttons (Fig 2C–2F`). This observation is compatible with a role in synaptic activity as most of the active zones are in the synaptic buttons.

sHSP23 and sHSP26 interact physically
In general, sHSPs proteins exhibit regions susceptible of posttranslational modifications (PTMs) which favor their oligomerization and alter the affinity of interaction by co-chaperones [17, 37]. Since, this mechanism maintains the activity of sHSPs it has been proposed that it regulates their function [17].
The results show that both proteins are localized in the same sub cellular compartments. To determine if both chaperones interact physically, we performed a co-immunoprecipitation assay. We used head protein extracts that were incubated with specific antibodies to specifically immobilize each sHSP in Protein A/G agarose beads. Samples were pre-cleared with untagged beads to avoid unspecified binding. Agarose beads were incubated with HSP23 or HSP26 antibody overnight. Head extract proteins and antibody-bound beads were incubated 1 hour. We revealed the western blot membranes with sHSPs antibodies and the results show that sHSP23 (Fig 2G lane 2) immunoprecipitation also precipitates HSP26, and vice versa (Fig 2G lane 3). Both specific bands are corroborated in the input positive control (Fig 2G lane 1) and the lack of signal in the negative control (Fig 2G lane 4) 22c10 antibody). These results confirm the physical interaction between sHSP23 and sHSP26 (Fig 2H), and it is consistent with their co-localization in the motor neuron buttons.

Pkm interacts with sHSP23 and sHSP26 and modulate synapse number
CG11534, CG43755 and Pkm proteins have been postulated that interact with both sHSP23 and sHSP26 (Flybase). They have unknown functions but predicted to have protein kinase like activity (Flybase, http://flybi.hms.harvard.edu/results.php). HSPs posttranslational modifications modulate their function [17] and therefore, we quantified the number of active zones in the NMJ upon knockdown of each candidate gene. The knockdown of pkm in motor neurons increases synapses number while we could not find any significant change for CG11534 and CG43755 knockdown (Fig 3A). In consequence, we focused our study in pkm as a candidate gene to interact with sHsp23 and sHsp26 in nervous system development.
Fig 3. Pkm does not affect to sHSP23-sHSP26 interaction.(A) Quantification of synapse active zones in the NMJ is shown for the knockdown of all candidate genes genotypes: CG43755 RNAi (CG43755↓), CG11534 RNAi (CG11534↓) and pkm RNAi (pkm↓). One‐way ANOVA test with Bonferroni post‐test* P
Read More

Show More

Related Articles

Leave a Reply

Your email address will not be published.

Back to top button