Changes in gravitational force cause changes in gene expression in the lens of developing zebrafish

Authors


Abstract

Gravity has been a constant physical factor during the evolution and development of life on Earth. We have been studying effects of simulated microgravity on gene expression in transgenic zebrafish embryos expressing gfp under the influence of gene-specific promoters. In this study, we assessed the effect of microgravity on the expression of the heat shock protein 70 (hsp70) gene in lens during development using transgenic zebrafish embryos expressing gfp under the control of hsp70 promoter/enhancer. Hsp70:gfp expression was up-regulated (45%) compared with controls during the developmental period that included the lens differentiation stage. This increase was lens specific, because the entire embryo showed only a 4% increase in gfp expression. Northern blot and in situ hybridization analysis indicated that the hsp70:gfp expression recapitulated endogenous hsp70 mRNA expression. Hypergravity exposure also increased hsp70 expression during the same period. In situ hybridization analysis for two lens-specific crystallin genes revealed that neither micro- nor hypergravity affected the expression level of βB1-crystallin, a non-hsp gene used as a marker for lens differentiation. However, hypergravity changed the expression level of αA-crystallin, a member of the small hsp gene family. Terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay analysis showed that altered-gravity (Δg) decreased apoptosis in lens during the same period and the decrease correlated with the up-regulation of hsp70 expression, suggesting that elimination of nuclei from differentiating lens fiber cells was suppressed probably through hsp70 up-regulation. These results support the idea that Δg influences hsp70 expression and differentiation in lens-specific and developmental period specific manners and that hsp family genes play a specific role in the response to Δg. Developmental Dynamics 235:2686–2694, 2006. © 2006 Wiley-Liss, Inc.

INTRODUCTION

Because Earth's gravitational field has remained constant throughout all of evolutionary history, it is possible that gravity plays a fundamental role in regulating/modulating gene expression and/or vertebrate development. To investigate the mechanisms by which gravity could influence these processes, many experiments have been performed both on the ground and in the Space Shuttle. However, unlike the studies under hypergravity conditions that can be accomplished by using centrifuges on Earth, the studies performed in a microgravity environment have been difficult to interpret. This difficulty arises because the experiments on the Space Shuttle or International Space Station have to be completed within a short period and cannot exclude the effects of space-specific physical stresses such as cosmic radiation, which can seriously damage DNA structure.

Our laboratory has developed an Earth-based technique for studying the effects of simulated microgravity on developing zebrafish embryos using a bioreactor that NASA designed for cells in culture (Moorman et al.,1999). Use of the bioreactor and a centrifuge allows us to investigate the effects of a broad range of gravitational forces on zebrafish development and gene expression. The zebrafish is the ideal animal for these experiments because, using gfp reporter gene technology, we can monitor gene expression and morphologic development in live embryos exposed to Δg for precise developmental time frames. Our previous experiments with transgenic embryos expressing the gfp gene under the control of the β-actin gene promoter have revealed that β-actin expression changed in tissue-specific and developmental period-specific manners (Gillette-Ferguson et al.,2003; Shimada et al.,2005), despite that β-actin has been referred to as a “housekeeping” gene.

To extend our study to the analysis of other genes, we selected the inducible heat shock protein 70 (hsp70) gene, because hsp70 is known as an indicator of various physical stresses. Hsp70 is induced by environmental stress and functions as a molecular chaperone to assist the folding of nascent polypeptide chains of intracellular proteins and to repair the altered or denatured proteins (Kiang and Tsokos,1998). Hsp70 is expressed specifically in lens from early embryogenesis (Blechinger et al.,2002a) and continues to be expressed even in the adult lens at the same level as in the embryonic lens (Dash et al.,1994). During development, hsp70 plays an essential role in lens formation (Evans et al.,2005). Later, when lens proteins such as crystallins play a role in the maintenance of lens cell transparency, Hsp70 is thought to protect the lens proteins from aggregation caused by environmental stresses such as heat and cold shock, ultraviolet-A (UV-A) irradiation, and oxidative stress (de Jong et al.,1986; Weinreb et al.,2001; Bagchi et al.,2002; Banh et al.,2003). Recent studies with cultured cells also demonstrated that microgravity exposure changed the expression level of hsp70 (Carlsson et al.,2003; Kumei et al.,2003). This finding suggested that microgravity could affect the expression of hsp70 in the lens during development.

Recent studies with transgenic zebrafish embryos having the gfp gene under the control of the inducible-hsp70-4 promoter showed that cadmium exposure exhibited tissue-specific and dose-dependent up-regulation of gfp expression, and the endogenous hsp70 mRNA showed the same expression pattern (Blechinger et al.,2002b). This finding indicates that this system is advantageous for rapid and long-term profiling of embryos exposed to environmental stresses. Therefore, using the same transgenic embryos together with the NASA-designed bioreactor and a custom-built centrifuge should allow us to assess the effects of Δg on hsp70 gene expression during development.

We incubated zebrafish eggs in the bioreactor and on a centrifuge for defined durations beginning at specific developmental stages. The analysis of hsp70:gfp expression revealed that the effects were specific to the lens and that there was a developmental period that the lens was most susceptible to Δg. We also found that both micro- and hypergravity had similar effects on hsp70 expression and apoptosis in lens.

RESULTS

Tissue-Specific and Stage-Specific Susceptibility to Microgravity

The intensity of green fluorescent protein (GFP) fluorescence in the whole embryo did not show significant change in each group (Fig. 1B). On the other hand, the intensity in the lens showed a significant increase (45%; P = 0.05) during the period between 32 hours postfertilization (hpf) and 80 hpf (Fig. 1C), in contrast to 4% increase in whole embryo (Fig. 1B). This period overlaps the stage of early lens development. A change in GFP fluorescence intensity was not observed in other tissues (data not shown). To determine whether the changes in expression were due to general changes in developmental timing, we measured body length for embryos of each group. There was no significant difference in body length between the experimental groups and the controls (Fig. 1D), suggesting that microgravity does not affect the rate of zebrafish development. Hence, these results suggest that microgravity is a physical stressor of the developing lens, and there is a critical period during which the developing lens is susceptible to microgravity. Because simulated microgravity only influenced hsp70:gfp expression during the developmental period 32–80 hpf, this exposure time period and 32–56 hpf were used for all subsequent experiments.

Figure 1.

Effect of simulated-microgravity on green fluorescent protein (GFP) expression in lens. A: The bars indicate the developmental time, in hours postfertilization (hpf), that embryos were kept in the bioreactor or kept on the centrifuge. B,C: The GFP fluorescence intensity of whole body (B) and lens (C) in each group exposed to simulated microgravity at different stages for different time periods. The x-axis indicates each developmental time, and the y-axis indicates the intensity normalized to the mean of the 1g-data. D: Body length of each group. The x-axis indicates each developmental time, and the y-axis indicates the normalized body length. Errors bars = SEM. Numbers in parentheses indicate the numbers of embryos for the groups.

Simulated Microgravity Changes hsp70 mRNA Expression in the Developing Lens

Previous studies have demonstrated that, after exposure to various physical stresses the expression of the hsp70 gene is regulated not only at the transcription level but also at posttranscriptional levels, including mRNA stabilization and translation (e.g., Theodorakis and Morimoto,1987; Morimoto,1993). However, the effect of microgravity on these phenomena has not been characterized. Therefore, we thought it was important to demonstrate that hsp70:gfp expression recapitulated the expression of the native hsp70 mRNA. We isolated total RNA from control and microgravity-exposed embryos of each group and measured the expression level of the hsp70 mRNA by Northern blot analysis. When incubated in the bioreactor during the period between 32 hpf and 80 hpf, hsp70 expression was up-regulated compared with controls (70%; P = 0.002; Fig. 2). The increase correlated with the increase in the intensity of the GFP fluorescence in lens (Fig. 1C) rather than that of the whole embryo (Fig. 1B). This finding can be explained by the fact that the majority of hsp70 expression is detected in lens. These results indicate that this in vivo reporter system can recapitulate the native hsp70 expression under a simulated microgravity condition. During the time period between 32 hpf and 56 hpf, hsp70 mRNA expression decreased (−22%; P = 0.02; Fig. 2). The lower expression level might be due to a delay in lens development.

Figure 2.

Effect of simulated microgravity on hsp70 mRNA expression. After the exposure to simulated microgravity, 3 μg of total RNA samples from the embryos of each control group (C) and experimental group (mg) were loaded on each lane and Northern blot analysis for hsp70 mRNA was performed (upper panel). To confirm the integrity of RNA and the transfer efficiency, ribosomal RNAs on a nylon membrane stained with methylene blue are shown in the lower panel. The numbers on the bottom indicate the normalized densitometry of each group. This experiment was performed in triplicate.

Hypergravity Changes hsp70 Expression in the Developing Lens

Hypergravity exposure of Xenopus embryos causes the induction of both hsp70 and hsp60 expression, whereas microgravity does not (Rizzo et al.,2002). This finding led us to question whether hypergravity also affects hsp70 expression in the developing zebrafish lens during the same period that simulated-microgravity did.

To assess this possibility, we placed zebrafish eggs on a centrifuge at 32 hpf and exposed them to 2g or 3g until either 56 hpf or 80 hpf. There was no significant difference in body length across groups (Fig. 3C), indicating hypergravity does not affect the rate of development. The intensity of GFP fluorescence for the whole embryo in each experimental group did not show any change compared with control (Fig. 3A). On the other hand, the intensity in lens was increased in both 2g (8.8%; P = 0.18) and in 3g (25%; P = 0.11) during the exposure period between 32 hpf and 80 hpf, No significant difference was observed during the exposure period between 32 hpf and 56 hpf (Fig. 3B). These results were similar to those for simulated microgravity (Fig. 1C), suggesting that hypergravity, especially 3g, affects hsp70 expression in developing lens in a manner similar to that of microgravity.

Figure 3.

Effect of hypergravity on green fluorescent protein (GFP) expression. A,B: The GFP fluorescence intensity of whole body (A) and lens (B) in each group exposed to hypergravity at different stages for different time periods. The x-axis indicates each developmental time, and the y-axis indicates the normalized intensity. C: Body length of each group. The x-axis indicates each developmental time, and the y-axis indicates the intensity normalized to the mean of the 1g-data. Errors bars = SEM. Numbers in parentheses indicate the numbers of embryos for the groups.

hsp70 mRNA Expression Correlates With hsp70:GFP Expression

Because hsp70 is known to play an essential role in lens formation (Evans et al.,2005), changes in both expression intensity and expression pattern are important. If Δg can cause changes in expression intensity, there might also be changes in expression pattern. To test this idea, we performed in situ hybridization with a probe for hsp70 transcripts after the exposure to both micro- and hypergravity during the period between 32 hpf and 56 hpf or 80 hpf (Fig. 4A). There was no difference in the expression pattern between control and experimental embryos in each group, suggesting that Δg does not cause defects in lens formation. To confirm that in situ hybridization gives results similar to both Northern blot analysis and gfp fluorescence analysis, we measured the intensity of the in situ hybridization signal in lenses of each group (Fig. 4B). The intensity was decreased in each group during the period between 32 hpf and 56 hpf (−14%; P = 0.03 in microgravity, −17%; P = 0.03 in 2g, and −47% P > 0.001 in 3g). However, intensity increased in microgravity (13.5; P = 0.002%) and in 3g (7.4; P = 0.02%) during the period between 32 hpf and 80 hpf. The hsp70:gfp results show significant correlation with the Northern blot data (Fig. 2), indicating that the change in the hsp70 mRNA expression in the whole embryo highly correlates with the expression level in lens. To clarify the mechanism of the reduction in hsp70 expression during the period between 32 hpf and 56 hpf, we determined the percentage of embryos expressing hsp70 mRNA compared with controls. The percentage of each group was lower than the controls (100%): 33% in microgravity, 63% in 2g, and 17% in 3g. These results indicate that Δg delays the onset of hsp70 expression in lens, and this delay is thought to be at least partly due to the decrease in the hsp70 expression. These results suggest that Δg changed the timing of lens development, that is, Δg delayed the timing first and then accelerated it, as the expression level of hsp70 changed.

Figure 4.

hsp70 mRNA expression in lens after Δg exposure. A: The regions of the eye in each embryo after in situ hybridization for hsp70 mRNA. Dorsal is up. le, lens. r, retina. B: The intensity of the signal in lens of each group (N = 12, 24 lenses) was measured and normalized to the mean of the 1g data. Errors bars = SEM.

Δg Does Not Affect the Timing of Lens Development

The βB1-crystallin gene is a lens differentiation marker gene that encodes a lens structural protein. βB1-crystallin expression is reported to accumulate as development progresses (Van Leen et al.,1987). The βB1-crystallin expression pattern and expression levels can be used to determine the developmental stage of the lens. Therefore, we performed in situ hybridization for βB1-crystallin mRNA in control embryos to determine the expression pattern and the expression level in the developing zebrafish lens. At 56 hpf, the lens showed strong expression in the posterior region (Fig. 5A, upper left). At 80 hpf, expression was observed in the posterior differentiating lens fiber region (Fig. 5A, lower left) similar to the expression pattern in other animals (McAvoy,1978; Van Leen et al.,1987), although even anterior proliferating lens epithelium displayed expression. Of interest, expression was also observed in the edge of the retina and in the future cornea (Fig. 5A, arrows). After exposure to Δg during the developmental periods between 32 hpf and 56 or 80 hpf, there was no difference in βB1-crystallin expression pattern (Fig. 5A) or expression intensity across groups (Fig. 5B). These results suggest that Δg did not affect the rate of lens development.

Figure 5.

βB1-crystallin mRNA expression in lens after Δg exposure. A: The regions of the eye in each embryo after in situ hybridization for βB1-crystallin. Dorsal is up. a, anterior; p, posterior. B: The intensity of the signal in lens of each group (N = 12–16) was measured and normalized to the mean of the 1g data. Errors bars = SEM.

Δg Changes the Expression of Another hsp Family Gene

Another lens-specific crystallin, αA-crystallin, is a derivative of a small heat shock protein gene and can function as a molecular chaperone (Horwitz,1992). It can bind to other lens crystallins to prevent aggregation and light scattering in the aging lens (reviewed in Haslbeck,2002). The zebrafish αA-crystallin is thought to have a stronger chaperone-like activity than that of other animals (Dahlman et al.,2005). It is possible that Δg could affect αA-crystallin expression in a similar manner to that documented for hsp70 expression. To investigate this possibility, we performed in situ hybridization for αA-crystallin mRNA in embryos after exposure to both micro-and hypergravity during the developmental periods between 32 hpf and 56 or 80 hpf. Under each condition, no change in the expression pattern was observed compared with controls (Fig. 6A). We subsequently measured the intensity of the signal in the lens in each group (Fig. 6B). Although, simulated-microgravity did not affect the expression level, hypergravity at 3g caused a reduction in expression after each exposure duration −14%; P = 0.1 and −10%; P = 0.003). These results indicate that Δg also changes the expression of a small hsp gene. During the period between 32 hpf and 56 hpf, the percentage of embryos expressing αA-crystallin compared with controls was decreased to 31.1% at 3g, suggesting that Δg also delays the onset of αA-crystallin expression as well as hsp70 expression. That only hsp70 and αA-crystallin but not βB1-crystallin expression were affected by Δg suggests a specific role of hsp family genes in stress responses to Δg.

Figure 6.

αA-crystallin mRNA expression in lens after Δg exposure. A: The regions of the eye in each embryo after in situ hybridization for αA-crystallin mRNA. Dorsal is up. B: The intensity of the signal in lens of each group (N = 13–19) was measured and normalized to the mean of the 1g data. Errors bars = SEM.

Up-Regulation of hsp70 Expression Correlates With a Decrease in Apoptotic Nuclei

Hsp70 is known to prevent stress-induced apoptosis. In the lens, proteasome inhibition protects lens cells against apoptosis by up-regulating Hsp expression (Awasthi and Wagner,2005). As in other animals, the degradation of nuclei occurs in developing zebrafish lens to detach the lens from the surface ectoderm and eliminate nuclei from differentiating lens fiber cells (Cole and Ross,2001). This process raises the possibility that exposure to Δg causes a reduction in apoptotic nuclei in lens during the period where hsp70 expression is up-regulated. To test this idea, we performed terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay analysis with embryos exposed to micro- and hypergravity during the period between 32 hpf and 56 or 80 hpf and counted the number of TUNEL-positive nuclei in whole lens (Fig. 7). In all conditions, there was no difference in the morphology and the area of the apoptotic nuclei in lens between the control and the experimental embryos (data not shown). During the period between 32 hpf and 56 hpf, where the lens is diverging from surface ectoderm, the number of apoptotic nuclei increased in microgravity (21%) and decreased in 2g and 3g (−15% and −14%, respectively). These results suggest that Δg affects viability of lens cells during early lens development. Exposure to microgravity and 3g during the period between 32 hpf and 80 hpf, including the period where differentiating lens fiber cells are becoming anucleate, caused significant decreases in the number of apoptotic nuclei (−40% and −70%, respectively). Although, 2g exposure caused no change in the number of apoptotic nuclei. These changes correlated with the up-regulation of hsp70 expression during the same period, as shown in Figures 3B and 4B, suggesting that Δg suppresses the degradation of nuclei in developing lens through the up-regulation of hsp70 expression.

Figure 7.

Change in the number of apoptotic nuclei in lens after Δg exposure. The terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay was performed with embryos exposed to micro- and hypergravity during the period between 32 hours postfertilization (hpf) and 56 or 80 hpf. The number of the TUNEL-labeled nuclei was then counted and normalized to the mean of the 1g data. Each group: N = 10–16, 20–32 lenses. Errors bars = SEM.

DISCUSSION

Our current research interest is how microgravity affects gene expression during development. We monitored hsp70 expression using transgenic zebrafish embryos expressing the gfp gene under the control of the hsp70 gene promoter/enhancer. We concentrated our efforts on the developing lens, because hsp70 and hsp70:gfp are expressed only in the developing lens and no other embryonic tissues until the embryo is exposed to some environmental stress. We used a NASA-designed bioreactor to simulate microgravity during specific developmental periods and found that hsp70:gfp expression was affected only in the developing lens and that the changes in expression were greatest during the developmental period between 32 hpf and 80 hpf, overlapping the stage of early lens development in the zebrafish (Fig. 1C). Interpretation of these findings depends on two assumptions; the expression of the transgene recapitulates the expression of the native hsp70 gene and the change in expression is not due to a change in developmental timing.

Comparability of Approaches

Northern blot and in situ hybridization analysis revealed that hsp70 mRNA expression was also up-regulated under the same condition. This finding suggests that the GFP-reporter system recapitulates the endogenous hsp70 gene expression under simulated-microgravity conditions. However, hsp70:gfp expression, Northern blot analysis, and in situ hybridization analysis all give different values for the magnitude of the up-regulation. We cannot exclude the possibility that stabilization of the hsp70 mRNA contributed to the difference in expression when hsp70:gfp expression is compared with either Northern blot or in situ hybridization data. The up-regulation of hsp70 mRNA expression during the developmental period between 32 hpf and 80 hpf might also be due to mRNA stabilization in addition to transcriptional activation. It has been suggested that hsp70 mRNA is stabilized after the exposure to some physical stresses (Theodorakis and Morimoto,1987; Kaarniranta et al.,1998; Alfieri et al.,2002), although its half-life is relatively short under nonstress conditions (Theodorakis and Morimoto,1987). However, mRNA stabilization does not explain the difference between the Northern blot and in situ hybridization data. A more likely explanation is that the different approaches have different sensitivities and different “noise” levels. This issue has also been raised when comparing microarray data, real-time polymerase chain reaction (PCR) data, and Northern blot data. The expression level of hsp70 mRNA, but not hsp70:gfp, was lower than control when exposed to Δg during the period between 32 hpf and 56 hpf (Figs. 2, 4B). This difference is probably due to the difference in the sensitivity of the signal detection. More important than the magnitude of the changes measured with each approach is that all of the approaches yield the same direction of expression change and the same comparative changes when expression levels are measured and compared for more than one gene.

Developmental Timing

One simple explanation of our results could be that Δg caused changes in developmental timing, possibly in a tissue-specific manner. We had discounted this possibility in our previous experiments because there were no changes in either overall morphological development or apparent change in morphological development of the specific organs being observed. However, we did not look at changes in timing of gene expression to document potential changes in timing of development. As in previous experiments, there was no effect of Δg on length of the embryo. Expression analysis for βB1-crystallin mRNA (Fig. 5B) shows that Δg did not affect the rate of lens development. Therefore, changes in either the overall developmental timing or timing of lens development do not explain the effects of Δg on the developing lens.

Microgravity vs. Hypergravity

Both simulated-microgravity and hypergravity exposure caused changes in hsp70 expression (Figs. 3B, 4). This finding suggests that both micro- and hypergravity have similar physical properties as stressors for embryos. Compared with the fact that Δg induces hsp70 expression only in lens, the same hsp70:gfp transgenic larvae exposed to cadmium showed a dose-dependent increase in the hsp70:gfp expression in skin epithelium, gill, liver, olfactory epithelium, and pronephric ducts (Blechinger et al.,2002b) and heat shock exposure induces ubiquitous hsp70:gfp expression in whole body (Halloran et al.,2000). These reports suggest that embryos recognize Δg as a different type of stressor from others. The effect of 2g on hsp70 expression was not significantly different compared with the effects at 3g, suggesting that there is a threshold for the effects of Δg on the developing lens.

We found that exposure to 3g during the periods between 32 hpf and 56 hpf or 32 hpf and 80 hpf decreased the expression of αA-crystallin, but not the non-hsp gene, βB1-crystallin in lens (Figs. 5, 6). This result suggests a specific role for hsp family genes in a stress-response to Δg. The developmental time period between 56 hpf and 80 hfp in zebrafish coincides with the stage of lens differentiation where the elongation of primary lens fiber cells occurs. Δg had different effects on hsp70:gfp expression during the period of 32–56 hpf than during the period 32–80 hpf. This finding suggests that Δg influences hsp70 expression differently in the differentiating lens. In this case, an increase of the molecular chaperone, the HSP70 protein, might play a role in preventing lens proteins such as crystallins from aggregation. This idea is supported by reports that heat shock and UV-A irradiation, known to cause aggregation of lens proteins (Blondin and Taylor,1987; Wong et al.,1978), caused an increase in the hsp70 expression in lens (Lang et al.,2000; Weinreb et al.,2001). The similar response of hsp70 and αA-crystallin to Δg is probably due to similar properties of their protein products in lens. The small hsp protein, αA-crystallin also acts as a molecular chaperone to protect lens proteins under thermal, UV, and oxidative stress conditions (Horwitz,1992; Wang and Spector,1995; Borkman et al.,1996). In addition, αA-crystallin stabilizes denaturing proteins in cooperation with other Hsps, including Hsp70 (Wang and Spector,2000). The common properties of the hsp family of genes have cooperatively protected fish's lenses from cataract formation that would be induced by a wide diversity of environmental stresses.

Potential Mechanisms

At least two potential mechanisms underlie the effects we have seen: Δg-related effects on the vestibular system, and/or Δg-related effects on heat shock transcription factors (HSFs). The highly conserved, vertebrate vestibular system (inner ear and vestibular nuclei in the brainstem) transduces the direction of gravity into meaningful neurological signals for proper orientation and balance. Goldfish with inner ear lesions show a lower up-regulation of HSP72 in muscle and skin when exposed to microgravity compared with goldfish with intact inner ears (Ohnishi et al.,1998). This finding suggests that an intact vestibular system is required for appropriate HSP72 regulation in response to Δg. This finding also suggests that the Δg up-regulated hsp70 expression in lens might, at least in part, be mediated through Δg related changes in the vestibular system. Our previous study indicated that there is a critical period where gravitational stimulus is require for functional vestibular development (Moorman et al.,2002). That period partly overlaps the period where hsp70 expression was up-regulated in the current study.

HSFs are direct regulators of HSP expression. Once activated by environmental stresses such as elevated temperatures, oxidants, heavy metals, and bacterial and viral infections, HSFs acquire the ability to bind to the heat shock element (HSE) located in the upstream enhancer region of hsp70 gene (Sorger,1991; Morimoto,1998). The hsp70 prompter/enhancer region in the hsp70:gfp reporter system we used contains several putative HSEs. HSF4 is required for cell growth and differentiation and its inactivation exhibits early postnatal cataract formation (Min et al.,2004; Fujimoto et al.,2004). Thus, HSFs also play important roles during lens development. Since hsp70:gfp expression in the lens changes after Δg, activation of HSFs and its binding to the HSEs undoubtedly played a role in this process. Therefore, HSFs probably play essential roles in the response of the developing lens to Δg.

Clinical Relevance

TUNEL assay analysis indicates that Δg caused a decrease in the number of apoptotic nuclei in lens cells (Fig. 7) in our experiments. The decrease in apoptosis coincided with the up-regulation of hsp70 expression, suggesting that Δg caused the reduction of apoptosis or the suppression of apoptotic pathway through the up-regulation of hsp70 expression. Differentiating lens fiber cells undergo denucleation by activating an apoptotic cascade (reviewed by Bassnett,2002). Hsp70 interacts with molecules involved in apoptotic pathways, such as caspases, apoptosis protease activating factor-1 (apaf1), apoptosis-inducing factor (AIF), and p53-inducible cell-survival factor (p53CSV; Beere et al.,2000; Garrido et al.,2001; Ravagnan et al.,2001; Komarova et al.,2004; Park and Nakamura,2005). In addition, proteasome inhibition causes up-regulation of HSP expression and suppression of both apoptosis and caspase activity in lens cells (Awasthi and Wagner,2005). Although we have not examined the effect of Δg on the expression of other genes related to lens development and differentiation, it is possible that the suppression of denucleation in differentiating lens under Δg conditions might be due to the change in expression of the genes in an apoptotic pathway.

Because differentiating lens fiber cells undergo denucleation by activating an apoptotic cascade, Δg might negatively affect postmitotic lens fiber differentiation. This idea is supported by reports that mice with a defect in DNase II–like acid DNase (DLAD), a member of the apoptotic cascade, develop cataracts and display a reduction in the response to light because of the defect in degrading DNA in the differentiating lens fiber cells (Nishimoto et al.,2003). If this is the case, a defect in nuclei elimination caused by Δg exposure could lead to lens opacity. Interestingly, astronauts have a higher incidence of cataracts than the age-matched general public (Cucinotta et al.,2001). Although the increase in the incidence of cataracts correlated with exposure duration for space radiation, the results would correlate equally well with exposure to microgravity. Our results suggest that exposure to microgravity might have predisposed astronauts to developing cataracts and might have exacerbated any effects of space radiation.

Summary

Taken together, the results support the idea that Δg influences hsp70 expression and differentiation in lens-specific and developmental period-specific manners and that members of the hsp family of genes play a specific role in the response to Δg. This study presents the first evidence of a link between Δg and effects on the lens and supports the role of ground-based space life science research in NASA's exploration agenda.

EXPERIMENTAL PROCEDURES

Animals

Zebrafish that express the gfp gene under the control of the promoter/enhancer of the zebrafish hsp70-4 gene (Halloran et al.,2000) were maintained at 28°C on a 14-hr-light and 10-hr-dark cycle. Eggs were collected within 3 hr after their production and fertilization and maintained at the same temperature until 8 hpf. Further incubation was carried out in 0.003% 1-phenyl-2-thiourea (PTU; Sigma, St. Louis, MO), an inhibitor of melanin pigment formation, at room temperature (20°C). We defined the stage of the embryos according to that of Kimmel (Kimmel et al.,1995).

Centrifuge

To achieve net 2g and 3g forces, a custom built slow-speed centrifuge was modified to accommodate horizontally oriented, water-filled tubes containing zebrafish eggs.

Bioreactor

A bioreactor (Synthecon, Houston, TX) that NASA designed to simulate microgravity for cells in culture (Jessup et al.,1993) was used to simulate many aspects of microgravity for zebrafish embryos (Moorman et al.,1999).

Exposure to Micro- and Hypergravity

Embryos at different stages were placed in the bioreactor or on the centrifuge at different developmental times for specific durations (Fig. 1A). To account for the effect of vibration, tissue culture dishes with the same number of control eggs were placed on the support frame of the bioreactor and on the center of the centrifuge (a position that does not rotate). We have already published extensive experiments demonstrating that there is no difference between control embryos placed on the support frame of the bioreactor compared with embryos incubated in the bioreactor rotating around a vertical axis (Moorman et al.,1999,2002; Gillette-Ferguson et al.,2003; Shimada et al.,2005). Each experiment was repeated four times using six embryos, each time, to obtain the images of GFP fluorescence.

GFP Fluorescence Imaging

All images were collected using a Leica DMRE microscope (Leica Microsystems Inc., Bannockburn, IL) equipped with a Ludl BioPrecision motorized stage (Ludl Electronic Products Ltd., Hawthorne, NY) and a Hamamatsu Orca-ER camera (Hamamatsu Photonics, Hamamatsu City, Japan). The microscope, stage, and camera were controlled using OpenLab software (Improvision, Lexington, MA) running on an Apple Dual-processor G4 computer. Before collecting fluorescence images, a brightfield image of each embryo/larvae was acquired using a ×5 objective. This image was used to measure rostral–caudal length as an indication of the age of the embryo/larva. After the brightfield image was acquired, a complete Z-series of fluorescence images was acquired. For this series, images were collected at 3-μm intervals using a ×10 objective. The entire stack of images was saved to disk. The camera gain, offset, and exposure time were kept constant for all homozygous embryos/larvae. Because heterozygous embryos/larvae had approximately half the fluorescence intensity as homozygous embryos/larvae, a second set of camera settings was used for all of the heterozygous embryos/larvae. Embryos older than 24 hpf were anesthetized using tricaine (0.04% 3-amino benzoic acidethylester: Sigma) during imaging to prevent movements of the embryo.

GFP Intensity Measurements

To measure the average intensity of fluorescence for the entire embryo/larva, an image was selected from the middle of the z-series stack and the average intensity for the entire image was calculated using the OpenLab software. An image was then selected where the outer edge of the lens was in focus (this yields a focal plane midway through the lens), a region of interest was drawn around the lens, and the software automatically calculated the average intensity within the region. For data analysis, individual measurements were normalized and the mean, standard deviation, and standard error were then calculated for each group. For statistical analysis, the means for the control and experimental groups were compared using a t-test.

Plasmid Construction

Total RNA was isolated from 42 hpf zebrafish embryos using Trisol Reagent (Invitrogen, Carlsbad, CA) and then contaminating DNA was removed using DNase-free (Ambion, Austin, TX) according to the manufacture's instructions. The αA-crystallin and βB1-crystallin fragments were amplified by reverse transcription-PCR (RT-PCR) using upper primer 5′-CTGAATTCTCTATTAGCCTCCTACTTGC-3′ and 5′-GTGAATTCCTTCACCATGTCTCAGACC-3′ and lower primer 5′-GAGGATCCATACCACCACCTGGCTGTGG-3′ and 5′-GTGGATCCGAGGTGCTATGCTACAAGG-3′, respectively. After BamHI and EcoRI digestion, these fragments were inserted into the BamHI and EcoRI sites of pBluescript SKII+ plasmid (Stratagene, La Jolla, CA).

Northern Blot Analysis

To prepare the probe against hsp70 mRNA, an hsp70 fragment was amplified by RT-PCR using upper primer 5′-AGTCATCGCTGGACTGAACG-3′ and lower primer 5′-TGCTCTTGTTCAGTTCTCTGC-3′ from pBluescript SKII(+)-Hsp70 (provided by Dr. Patrick H. Krone). A DNA digoxigenin (DIG) probe was synthesized from the fragment with the same primers and DIG DNA labeling mix (Roche, Mannheim, Germany).

For Northern blot analysis, total RNA was isolated from each group (three to four repetitions of each experiment) using Trisol Reagent (Invitrogen, Carlsbad, CA) and then contaminating DNA was removed using DNase-free (Ambion) according to the manufacturer's instructions. Then 3 μg of the RNA was loaded on each lane and electrophoresis was performed through 1% formaldehyde agarose gels in MOPS buffer. The RNAs were then blotted onto Biodyne Transfer Nylon Membranes (PALL, Ann Arbor, MI) and UV-closslinked at 120,000 microjoules/cm2. Subsequently, the RNAs were hybridized with the hsp70 DNA DIG probe in 50% formamide/5xSSPE/0.1% N-lauroxl-sarcosine/2% Blocking Sol. (Roche)/0.1% sodium dodecyl sulfate/200μg/ml tRNA (Roche) overnight at 42°C. After washing, CSPD ready-to-use (Roche) was added and the blots were exposed to X-Omat film (Kodak, Rochester, NY) at 37°C. The integrity of the RNA and the transfer efficiency were confirmed by the staining with 5% methylene blue/0.5M NaOAc. Densitometric quantitation of the expression levels was carried out using AlphaDigiDoc RT software (Alpha Innotech, CA). For statistical analysis, the means of control and experimental mRNA levels were compared using a t-test.

Whole-Mount In Situ Hybridization

The hsp70, αA-crystallin, and βB1-crystallin RNA probes were synthesized from pBluescript SKII(+)-hsp70, pBluescript SKII(+)-αA-crystallin and pBluescript SKII(+)-βB1-crystallin plasmids, respectively, by Riboprobe in vitro Transcription Systems (Promega, Madison, WI). Whole-mount in situ hybridization was performed according to the protocol from Hashimoto (Hashimoto et al.,2004). In this process, to treat both control and experimental embryos in the same microtube, the groups were distinguished by removing the tip of the tails from the controls. For measurement of signal intensity, a brightfield image of each embryo was acquired and the intensity was analyzed using Openlab software. Net darkness of the signal was defined as a difference between the signal intensity in lens and the background intensity. For data analysis, individual measurements were normalized and then the mean, standard deviation, and standard error were calculated for each group. For statistical analysis, the means for the control and experimental groups were compared using a t-test.

TUNEL Assay

The TUNEL assay using the In Situ Cell Death Detection kit, TMR red (Roche Diagnostics, Penzberg, Germany) was performed as described by (Evans et al.,2005), with minor modifications. In the permeabilization process, 1 mg/ml collagenase (C-9891, Sigma) in PBS was used instead of proteinase K. The TUNEL-labeled nuclei were detected by using the Leica DMRE upright fluorescence microscope, and the number of labeled nuclei in the whole lens was counted in each group. For data analysis, individual measurements were normalized and then the mean, standard deviation, and standard error were calculated for each group. For statistical analysis, the means for the control and experimental groups were compared using a t-test.

Acknowledgements

We thank our senior technicians, Meghal K. Desai and Caroline M. Pignatelli, for technical assistance and fish care. We also thank Dr. John Kuwada for providing transgenic zebrafish having the hsp70-4:gfp gene and Denise Dehnbostel for editing the manuscript. N.S. performed all of the experiments, suggested the crystallin gene experiments, analyzed the data, and wrote the first draft of the manuscript. S.J.M. designed the initial experiments, rewrote the manuscript, and generated the financial support for the project.

Ancillary