Hsp40 Is Involved in Cilia Regeneration in Sea Urchin Embryos

Abstract

In a previous paper we demonstrated that, in Paracentrotus lividus embryos, deciliation represents a specific kind of stress that induces an increase in the levels of an acidic protein of about 40 kD (p40). Here we report that deciliation also induces an increase in Hsp40 chaperone levels and enhancement of its ectodermal localization. We suggest that Hsp40 might play a chaperoning role in cilia regeneration.

Both prokaryotic and eukaryotic organisms show rapid adaptive responses to physical and chemical changes in their environments. Among these physiological defense mechanisms, the stress response involves rapid reduction of bulk protein synthesis and induction of a set of cytoprotective proteins, conserved in evolution, known as heat shock proteins (Hsps).

Most Hsps are constitutively present in cells, also in the absence of any stress, and play a key role in protein folding and degradation. In addition, Hsps synthesis is enhanced by a wide variety of stressing conditions, such as temperature increase and/or exposure to heavy metals, amino acid analogues, or cytotoxic drugs, all of which lead to accumulation of aberrant non-native proteins. Although most of these stimuli primarily induce synthesis of members of the Hsp70 family of chaperones, each kind of stress affects the expression of specific combinations of stress proteins.

The Paracentrotus lividus sea urchin embryo provides a clear-cut example of differential stress response. After the blastula stage, both heat shock (Roccheri et al. 1981,1993) and EGTA treatment (Roccheri et al. 2000) induce in this embryo the synthesis of at least 16 Hsps. On the other hand, only seven Hsps are affected by culturing embryos in the presence of zincions (Roccheri et al. 1988,1993). All these different stresses induce an increase in Hsp70 expression.

Recently, we reported that deciliation of Paracentrotus lividus embryos can be considered a further kind of stress. In this case, however, the Hsp70 family did not show any change, although a significant modification in the concentration of an acidic stress protein of about 40 kDa (p40) was noticed (Casano et al. 1998,2003).

Two-dimensional electrophoresis has been previously used to analyze the repertoires of proteins induced by specific stresses. In the present study we used a modified two-dimensional electrophoresis (i.e., NEPHGE/SDS-PAGE), that also allows identification of more basic proteins, to better characterize proteins induced by embryo deciliation, Western and immuno-cytochemical analyses. We found that deciliation stress also causes an increase in Hsp40 levels, thus suggesting that this chaperone is involved in the assembly of newly growing cilia.

Materials and Methods

Culture of Embryos Under Physiological or Stressing Conditions

Adult sea urchins of the Mediterranean species Paracentrotus lividus were collected along the western Sicilian coast. Eggs were fertilized and embryos were grown at a concentration of 5000/ml in Millipore-filtered sea water containing antibiotics, in a beaker with a rotating propeller, until the desired stages were reached.

For animalizing treatment, zinc ions were added as 1.0 mM ZnSO₄, 30 min after fertilization. Embryos were then cultured for 22 hr until they reached an age corresponding to the gastrula stage of normal embryos (Lallier 1975).

Deciliation/Cilia Regeneration

Hypertonic treatment causes complete but reversible deciliation of sea urchin embryos (Auclair and Siegel 1966; Casano et al. 1998). Gastrula embryos were deciliated by adding 0.12 volumes of 4.45 M NaCl to the culture medium, thus increasing the saline concentration by 470 mM and reaching a final tonicity approximately twice that of normal sea water. After 2 min, 1.2 volumes of an artificial sea water, lacking NaCl but containing the other salts at a concentration 10% higher than in normal sea water, were added to restore normal sea water saline concentration. Embryos were then allowed to develop until the desired stage. In some experiments, embryos underwent repeated cycles (up to four) of deciliation, spaced by 1 hr grown in normal sea water, during which cilia started regrowing.

To label proteins, embryos were incubated with 50 μCi/ml of [³⁵S]-1-methionine (specific activity 1000 Ci/mM; Amersham Bioscience, Fair Lawn, NJ) for 30 min after restoration of normal saline conditions.

Cilia Isolation

Gastrula embryos were deciliated once as described above and centrifuged at 700 × g for 3 min. To recover released cilia (first-generation cilia), the resulting supernatant was centrifuged at 10,000 × g for 10 min. Cilia were then washed once with sea water.

Sample Preparation and Electrophoretic Analysis

The whole or deciliated embryos and the isolated cilia were homogenized in O'Farrell (1975) urea lysis buffer. Protein concentration was measured by standard spectrophotometric procedures. Incorporation of labeled methionine into proteins was determined by TCA precipitation and counting in a Beckman scintillation counter (Beckman Coulter; Fullerton, CA). Standard aliquots (50 μg; verified also by Coomassie staining of a parallel gel) of homogenized samples were analyzed by electrophoresis on 10% polyacrylamide slab gels (SDS-PAGE) as described elsewhere (Roccheri et al. 1993).

Two-dimensional analysis included non-equilibrium pH gradient electrophoresis (NEPHGE) as first dimension, and SDS-PAGE as second dimension. NEPHGE was performed according to O'Farrell (1975) in the range of 3.5–10. Samples (150 μg each) were run at 400 V for 3.5 hr. pI of probable Hsp40 (pI 9.7–8.3; Hattori et al. 1993), was deduced from comparison with pI markers (calibration kits for pI determinations; Amersham Bioscience). Protein sizes were calculated from comparison with ¹⁴C-labeled molecular weight markers (Rainbow [¹⁴C]-methylated protein molecular weight markers; Amersham Bioscience). For fluorography, gels were soaked with 2,5-diphenyloxazole in dimethyl sulfoxide solution, dried, and finally exposed to Kodak X-Omat films.

Western Analyses

After mono-dimensional SDS-PAGE, proteins were electroblotted onto nitrocellulose filters (Hybond ECL; Amersham Biosciences) in a semi-dry apparatus (Novablot; Amersham Biosciences) at 0.8 mA/cm² for 2 hr. Filters were then incubated with blocking solution (3% bovine serum albumin, 5% horse serum, 0.02% sodium azide in PBS) for 3 hr and then overnight with human polyclonal anti-rabbit Hsp40 (StressGene Biotech; Glanford, Canada) at a dilution of 1:500. Filters were finally reacted with alkaline phosphatase-conjugated anti-rabbit IgG and stained with BCIP/NBT (Sigma; St Louis, MO).

Whole-mount Immunocytochemistry

Whole control and deciliated embryos were fixed for 12–16 hr with Bouin fixative, a mix of picric acid, 37% formaldehyde, and glacial acetic acid (15:5:1). Fixed embryos (on average 1000–2000/assay) were washed under gentle shaking in 100% ethanol and then in 100% methanol, containing 2.0 mM EDTA. After hydration, obtained by rinsing sequentially in 100%, 95%, and 70% ethanol, embryos were washed three times with PBST (PBS containing 0.1% Tween-20) and incubated overnight with one of the following antibodies in PBST containing 3% BSA: (a) monoclonal anti-Hsp60 (1:250; Sigma); (b) monoclonal anti-Hsp70 (1:200; Sigma); (c) polyclonal anti-Hsp40 (1:250; StressGene Biotech); (d) monoclonal anti-α-tubulin (1:1000; Sigma clone B-5–1–2) able to recognize sea urchin axonemal α-tubulin. As a control, some embryos were incubated without any primary antibody. Alkaline phosphatase-conjugated anti-mouse IgG secondary antibodies (Promega; Madison, WI) were used at a dilution of 1:1000. Staining was obtained by incubating embryos for various times (30 min-2 hr) with NBT/BCIP (Roche; Mannheim, Germany) in 100 mM NaCl, 100 mM Tris-HCl, pH 9.5, 50 mM MgCl₂ in relationship to the specific antibody. After rinsing twice with 50% ethanol to block staining reaction, embryos were finally collected in 80% glycerol, stratified on coverslips, observed under an Axioskop MC80 microscope (Zeiss; Oberkochen, Germany), and photographed with 100 ASA Kodak Gold film.

To prepare thin sections, embryos were fixed in Carnoy's solution for 90 min and embedded in paraffin. Sections 5 μm thick were allowed to adhere to poly-l-lysine-pretreated slides. Deparaffinization was obtained by soaking the slides twice for 8 min with xylene. Hydration was obtained by successively rinsing the slides with 100% (twice for 3 min), 95% (twice for 3 min), and 70% ethanol (twice for 3 min). Slides were finally rinsed in deionized water. Hydrated sections (on average 3–4/experiment) were incubated with the same antibodies used for whole embryos as described above.

Results

It has recently been demonstrated that chaperones are common components of cilia and flagella, in which they are likely to be involved in axonemal protein dynamics (Stephens and Lemieux 1999). Moreover, in P. lividus embryos, deciliation is a kind of stress that does not involve Hsp70 increase (Casano et al 1998). Therefore, we wondered whether, in our system, chaperones other than Hsp70 might be involved in cilia formation. In particular, we focused on Hsp40 because its basic pI (Hattori et al. 1993) should have hampered its identification by the normal two-dimensional PAGE previously performed. To obtain a better resolution of basic proteins, we used non-equilibrium pH gel electrophoresis (NEPHGE) to analyze proteins synthesized after deciliation.

Sea urchin embryos at the early gastrula stage were deciliated and incubated with 50 μCi of [³⁵S]-methionine as described in “Materials and Methods.” Labeled proteins were then analyzed by NEPHGE. As shown in Figure 1, two proteins of about 40 kD increased after deciliation: (a) the already described acidic p40 (pI 6; Casano et al. 1998), indicated by a white arrow, and (b) a basic 40-kD protein (pI 8–9), indicated by a black arrow (Figure 1B) that could be Hsp40.

Basic Hsp40 (Hattori et al. 1993) is a member of the DnaJ family of proteins characterized by the presence of the highly conserved J domain. Hsp40s are involved in almost all aspects of protein synthesis, folding, and secretion and in degradation of misfolded proteins (Kelley 1998), and are known to function as co-chaperones of larger chaperones, such as Hsp70 and Hsp90, but also as independent chaperones (Cheetham and Caplan 1998). To study the previously unknown distribution of Hsp40 in P. lividus sea urchin embryos, we used anti-human Hsp40 antibodies to perform Western analyses and whole-mount in situ immunolocalization experiments.

Figure 1

Fluorograms of ³⁵S-labeled proteins analyzed by NEPHGE. (A) Control and (B) deciliated gastrula embryos. The two main 40-kD proteins induced by deciliation are indicated by arrows; white arrow points to the previously identified acidic p40 and black arrow points to a basic 40-kD protein (putative Hsp40).

As shown in Figure 2, Hsp40 is already present in control embryos (Lane 1), but increases after deciliation (Lane 2). Computerized quantitative analysis has in fact shown that deciliation causes about a twofold increase of Hsp40 in deciliated embryos. Moreover, it is highly concentrated in cilia isolated from control embryos (Figure 2, Lane 3). When we analyzed its distribution in the embryo by whole-mount immunolocalization, we confirmed the presence of Hsp40 in control embryos (Figures 3A and 3B; two different focal planes are shown) and found that Hsp40 is localized mainly in the ectodermal layer, as well as in cilia and the inner surface of the archenteron which, at this stage of development, presents short and stubby cilia (Hardin 1987). After one cycle of deciliation, immunostaining of ectoderm is enhanced (Figure 3C), and after four cycles staining is so intense that it is impossible to see the inside of the embryo even after only 10 min of staining reaction (Figure 3D). Taken together, these results indicate that deciliation induces Hsp40 synthesis and accumulation. To ascertain the specificity of induction, four-times deciliated embryos were also stained with either anti-Hsp70 (Figure 3E) or anti-Hsp60 (Figure 3F) antibodies, but we did not find any deciliation-dependent increase in the latter chaperones, thus confirming that Hsp70 (Casano et al. 1998) and Hsp60 (Rinaldi, personal communication) synthesis is not modified by deciliation.

A comparative analysis of results obtained by Western blotting and whole-mount immunolocalization suggested that deciliation could induce both increase and preferential localization of Hsp40. To clarify this point, we immunostained embryo sections. The results confirmed that Hsp40 was much more concentrated in fourfold deciliated (Figure 3H) than in control (Figure 3G) embryos and showed in addition that the protein accumulates at the apical surface of ectodermal cells.

To better clarify the possible involvement of Hsp40 in ciliogenesis, we performed whole-mount immunolocalization experiments on animalized embryos. Animalization of P. lividus embryos is induced by adding to sea water low concentrations of zinc ions from fertilization onward. Zinc-treated embryos have unusually long cilia resembling those in the apical tuft. We found that zinc-treated embryos accumulate Hsp40, especially in cilia and in their basal region (Figures 4A and 4B; two focal planes).

Figure 2

Western blot of SDS-PAGE of homogenized control embryos (Lane 1), deciliated embryos (Lane 2), and isolated cilia of control embryos (Lane 3) labeled with the anti-human Hsp40 antibody. Equal amounts of protein on different samples were loaded (50 μg).

Figure 3

Immunolocalization of Hsp40 in sea urchin embryos. (A,B) Whole-mount control embryos probed with anti-human Hsp40 antibodies and photographed at two different focal planes. (C) Whole-mount once-deciliated embryos probed with anti-human Hsp40 antibodies. (D–F) Whole-mount fourfold deciliated embryos probed with anti-human Hsp40 antibodies (D), with anti-human Hsp70 monoclonal antibodies (E), or with anti-human Hsp60 monoclonal antibodies (F). (G) Sections of control gastrula embryos. (H) Sections of fourfold deciliated gastrula embryos (from the same lots of embryos utilized for whole-mount experiments) probed with anti-human Hsp40 antibodies. Bar = 20 μm.

It is worth noting that cilia and their basal region are also immunostained by anti-axonemal α-tubulin antibodies (Figures 4C and 4D; two focal planes). Conversely, using the corresponding antibodies, we did not find any evidence of Hsp70 or Hsp60 accumulation (data not shown).

Because it has been reported that in P. lividus Hsp 70 is localized in the centrosomes (Agueli et al. 2001), we wondered whether these organelles also contain Hsp40. As shown in Figure 4E, which demonstrates the results of immunolocalization experiments performed on embryos at the first mitotic division, Hsp40 is in fact present in the centrosomes. Interestingly, in two-cell embryos Hsp 40 is present in the cortex as well. For comparison, immunostaining with anti-axonemal α-tubulin antibodies, which highlights the centrosome and the astral bipolar spindle, is shown in Figure 4F.

Figure 4

Immunolocalization of Hsp40 and α-tubulin in animalized sea urchin embryos and in two-blastomere embryos. Whole-mount zinc-treated embryos of an age corresponding to the gastrula stage of control embryos, probed with anti-human Hsp40 polyclonal antibodies (A,B) or with anti-sea urchin α-tubulin monoclonal antibodies (C,D) and photographed at two different focal planes. (E) Sections of two-blastomere embryos probed with anti-human Hsp40 polyclonal antibodies. (F) Sections of two-blastomere embryos probed with anti-sea urchin α-tubulin monoclonal antibodies. Bar = 20 μm.

Discussion

Sea urchin embryos can be uniformly deciliated by hypertonic shock. However, after a short lag, if deciliated embryos are allowed to develop in isotonic sea water, cilia are re-formed (Auclair and Siegel 1966). We previously demonstrated that in P. lividus embryos deciliation is a kind of stress. In addition to activation of the ciliary subroutine program (Stephens 1995), deciliation induces transient and significant reduction of bulk protein synthesis as a classical stress response. On the other hand, deciliated embryos show specific enhancement of the expression of an acidic (pI: 6) protein of about 40 kD (p40), whereas synthesis of Hsp70 is unchanged (Casano et al. 1998; Stephens and Lemieux 1999). We also demonstrated that this response is due to deciliation per se and not to the hypertonic treatment. If cultured in the presence of NaCl at a concentration just below that required for deciliation but for a longer period, embryos do not undergo deciliation and show a different stress response that does involve Hsp70 increase but not increase of p40 synthesis (Casano et al. 1998). Finally, we demonstrated that deciliation, like heat shock, is able to induce thermoresistance (Casano et al. 2003).

Eukaryotic cilia and flagella are built up and maintained through a process called intraflagellar transport (Marshall and Rosenbaum 1999), which involves rapid particle movement along the axonemal microtubules. Transport and assembly of tubulin and other proteins at the distal tips of cilia/flagella require the chaperoning function of different Hsps. Chlamydomonas flagella contain an Hsp70-related, axoneme-associated protein (Bloch and Johnson 1995). Homologues of both Hsp70 and Hsp90 are present in Tetrahymena cilia and are also complexed with tubulin in the cytoplasm (William and Nelsen 1997). Sea urchin embryonic cilia and sperm flagella contain a 78-kD Hsp70 cognate (Stephens 1997; Stephens and Lemieux 1999) that shows an axonemic distribution similar to that observed in Chlamydomonas flagella. Hsp70 cognates were also detected in ctenophore comb plate, molluscan gill, and rabbit tracheal cilia (Stephens and Lemieux 1999). The wide distribution of chaperones in cilia and flagella strongly suggests their involvement in axonemal protein dynamics and/or their catalytic role (as in the case of TCP-1α) in monomer incorporation and/or turnover (Stephens 2000,2001).

Interestingly, the amount of Hsc78 remains constant before, during, and after deciliation (Casano et al. 1998; Stephens and Lemieux 1999).

In this study we show that Hsp40 is constitutively expressed in sea urchin embryos and that its synthesis is further enhanced by deciliation. In addition, we report its preferential localization in the ectodermal apical layer and in the basal region of the cilia, thus suggesting its specific involvement in cilia regeneration. These data are consistent with an involvement of the Golgi apparatus in cilia regeneration (Stephens 2001) and suggest the existence of post-translational modifications of Hsp40 proteins that could anchor them to specific microdomains of the plasma membrane. Although not yet demonstrated, post-translational modifications could be in agreement with literature data. Several members of the DnaJ/Hsp40 family have potential prenylation sites at their C-terminal ends and DnaJ homologues can be farnesylated in both lower and higher eukaryotes (Caplan et al. 1992; Zhou et al. 2000; Bozidis et al. 2002). Moreover, recent evidence indicates that the human homologue of Hsp40 associates with endoplasmic reticulum membranes and is luminally oriented (Yu et al. 2000), and that the mouse homologues mDj7 and mDj9 have transmembrane domains (Ohtsuka and Hata 2000).

Interestingly, we also found Hsp 40 in the centrosomal region, in two-cell embryos. The presence of chaperones in the centrosome has been already documented. In particular, members of the Hsp70 family have been described as cell cycle-specific components of the centrosome in HeLa cells (Rattner 1991), in the dinoflagellate C. cohnii (Perret et al. 1995), and in the ciliated protozoan Euplotes (Fleury et al. 1998). Moreover, Hsp70 has been detected in centrosomes isolated from P. lividus embryos at the first mitotic metaphase (Agueli et al. 2001), suggesting involvement of Hsp70 in chaperoning centrosome and mitotic spindle assembly. Given the well-known function of Hsp40 as a co-chaperone of Hsp70, our results are in good agreement with those of Agueli et al. (2001), although the specific role of Hsp40 must be further clarified.

It is also worth noting that, before hatching, each sea urchin embryonic blastomere carries a single cilium and that, after telophase, a mitotic centriole pair (diplosome) moves towards the apical membrane and, on reaching it, one of the centrioles serves as basal body (Stephens 1994).

In conclusion, our data demonstrated for the first time both the presence of Hsp 40 in P. lividus embryos and its preferential localization in the ectodermal apical region and in centrosomes.

In addition, we found that Hsp40 concentration increases after deciliation, as well as (along with axonemic α-tubulin) in ciliary basal regions of animalized embryos. Although we cannot yet define the kind of structure to which the protein localizes (centrioles, basal bodies, or others), the data reported suggest a direct involvement of Hsp40 both in the regrown and in the normal turnover of embryonic cilia.