Analysis of the stability of the spermadhesin PSP-I ⁄ PSP-II heterodimer Effects of Zn 2+ and acidic pH Marı ´a Asuncio ´ n Campanero-Rhodes 1 , Margarita Mene ´ ndez 1 , Jose ´ Luis Sa ´ iz 1 , Libia Sanz 2 , Juan Jose ´ Calvete 2 and Dolores Solı ´ s 1 1 Instituto de Quı ´ mica Fı ´ sica ‘Rocasolano’, CSIC, Madrid, Spain 2 Instituto de Biomedicina de Valencia, CSIC, Valencia, Spain Proteins are designed to have a particular activity in a specific environment, and their fold and assembly are intimately related to this physiological function. Infor- mation on the organization of the protein structure, however, is usually acquired in simple buffer systems, far removed from the complex conditions encountered in intracellular and extracellular spaces and fluids. Besides the crucial influence of the local concentration of macromolecules, the presence of co-solutes may have a decisive effect on protein conformation and sta- bility [1]. Seminal plasma is a composite fluid, comprising secretions from the testes, epididymis and accessory sex glands. It is not merely a vehicle for the ejaculated sperm but it is also involved in numerous activities in the male and female reproductive tract, ensuring the viability and fertilizing capacity of spermatozoa. The seminal plasma contains abundant concentrations of different amino acids, peptides, lipids, fatty acids and various osmolytes, and it is an important source of cations [2]. In boar seminal plasma, for example, the concentration of Zn 2+ is surprisingly high Keywords heterodimer dissociation; PSP-I ⁄ PSP-II; spermadhesins; thermal stability; Zn 2+ Correspondence D. Solı ´ s, Instituto de Quı ´ mica Fı ´ sica Rocasolano, Serrano 119, 28006 Madrid, Spain Fax: +34 91 564 24 31 Tel: +34 91 561 94 00 E-mail: d.solis@iqfr.csic.es (Received 20 June 2005, revised 7 September 2005, accepted 14 September 2005) doi:10.1111/j.1742-4658.2005.04974.x Spermadhesins are a family of 12–16 kDa proteins with a single CUB domain. PSP-I and PSP-II, the most abundant boar spermadhesins, are present in seminal plasma as a noncovalent heterodimer. Dimerization markedly affects the binding ability of the subunits. Notably, heparin and mannose 6-phosphate binding abilities of PSP-II are abolished, indicating that the corresponding binding sites may be located at (or near) the dimer interface. Pursuing the hypothesis that cryptic binding sites in PSP-I ⁄ PSP-II may be exposed in specific physiological environments, we examined the influence of Zn 2+ and acidic pH on the heterodimer stability. According to near-UV CD spectra, the core native fold is preserved in the presence of physiological concentrations of Zn 2+ , a cation unusually abundant in boar seminal plasma. However, the thermostability of the heterodimer decreases significantly, as observed by CD and differential scanning calorimetry. The effect is Zn 2+ -specific and is reversed by EDTA. Destabilization is also observed at acidic pH. Gel filtration analysis using radioiodinated PSP-I ⁄ PSP-II reveals that dissociation of the heterodimer at low (nanomolar) protein concentrations is promoted by both Zn 2+ and acidic pH. Although the integrity of the heterodimer in seminal plasma seems to be guaranteed by its high concentration, dissociation may be facilitated in the female genital tract because of dilution of the protein in the intraluminal fluids of the cervix and the uterus, and the acidic fluid of the uterotubal junction. Such a mechanism may be relevant in the regulation of uterine immune reactions. Abbreviations DSC, differential scanning calorimetry. FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS 5663 (0.3–0.7 mm) [3,4], reaching the spermatozoa at ejac- ulation [5]. Seminal plasma also contains a large num- ber of different proteins that exert multiple effects on sperm function, including a diversity of enzymes, hormones, growth factors and transport proteins [6]. However, the precise role of most of the seminal plasma proteins in sperm physiology remains obscure. Spermadhesins are a family of 12–16 kDa proteins found in seminal plasma and ⁄ or attached to the sper- matozoal surface of a variety of mammalian species (e.g. boar, bull and horse) [7]. These proteins are composed of 109–133 amino acids, show a 40–60% sequence identity, and contain a single CUB domain [8]. Members of the spermadhesin family have been shown to bind zona pellucida glycoproteins, serine proteinase inhibitors, phospholipids and ⁄ or sulfated glycosaminoglycans [9], suggesting that they may be involved in different steps of the complex fertilization process. In the boar, spermadhesins represent about 75% of the total protein content of seminal plasma, their concentration ranging from 0.6 to 7 mgÆmL )1 [10]. PSP-I and PSP-II, the most abundant boar spermadhesins, occur as a noncovalent heterodimer [11]. The secondary structure and stability of the PSP-I⁄ PSP-II heterodimer in solution has been investigated [12], and the crystal structure solved at 2.4 A ˚ resolution [13]. Both subunits consists of a compact ellipsoidal b-sandwich structure organized into two five-stranded (parallel and antiparallel) b-sheets. Accumulating evidence points to a role for PSP-I ⁄ PSP-II as an exogenous modulator of both sperm function and uterine immune activity, thus ensuring reproductive success. The PSP-I ⁄ PSP-II complex con- tributes to maintaining sperm with high viability, motility, and mitochondrial activity [14]. In addition, PSP-I and PSP-II are immunostimulatory for lympho- cyte activity in vitro [15]. Lymphocyte binding of PSP-I has been demonstrated [16]. Furthermore, the PSP-I ⁄ PSP-II heterodimer and its isolated subunits induce the recruitment of neutrophils into the peritoneal cavity of rats [17] and pigs [18]. The neutrophil migration- inducing activity of PSP-I ⁄ PSP-II, and possibly of the PSP-II subunit, is mediated by the stimulation of resi- dent macrophages, which release a neutrophil chemo- tactic substance [19]. In contrast, PSP-I appears to act directly on neutrophils [17]. The purpose of these immunostimulatory activities would be to prevent possible infections of the lower reproductive tract and to provide a foreign-cell-free uterine environment for the descending early embryos. The ligand-binding capabilities of the isolated sub- units have been investigated thoroughly. The PSP-II subunit exhibits mannose 6-phosphate and heparin binding abilities [20], whereas conflicting results on the heparin-binding ability of the PSP-I subunit have been reported [11,21,22]. These binding sites are nonetheless cryptic in the heterodimer, which is typically isolated from the nonheparin-binding fraction of boar seminal plasma [11], raising the question of their biological sig- nificance. In this context, it is noteworthy that the stimulatory activity of PSP-II on macrophages is selec- tively inhibited by mannose 6-phosphate [17]. Here we show that, in the presence of physiological concentrations of Zn 2+ , the stability of the hetero- dimer is significantly lowered, promoting at low pro- tein concentrations dissociation of the PSP-I and PSP-II subunits. Similar behaviour is induced by acidic pH. The results point to the possibility that the cryptic binding sites in the PSP-I ⁄ PSP-II heterodimer are exposed in the female genital tract environment. Results CD spectroscopy The far-UV CD spectrum of PSP-I ⁄ PSP-II exhibits a large positive band at % 202 nm and a negative region at 215 nm [12], as expected for the b-sandwich topol- ogy of the CUB domain [13]. In addition, the near-UV CD spectrum was dominated by the presence of a sharp positive band at 291 nm, in the tryptophan region (Fig. 1A). Furthermore, the spectrum showed a large negative region with minima around 287 and 268 nm. Thermal denaturation of PSP-I ⁄ PSP-II led to a decrease in the intensity of both the positive and negative bands (Fig. 1A) along with an increase in ellipticity below 250 nm. These changes reflect the loss of tertiary structure of the protein. Monitoring of the decrease with temperature of the ellipticity at 268 nm facilitated tracing of the denaturalization process. PSP- I ⁄ PSP-II thermal denaturation was irreversible [12], but the thermal denaturation profiles were practically scan-rate independent. Experimental curves were there- fore phenomenologically analyzed using a sigmoidal function (see Experimental procedures) from which a T 1 ⁄ 2 (temperature at which 50% of the protein is dena- tured) of 62.2 °C can be estimated (Table 1). The far-UV and near-UV CD spectra of PSP-I ⁄ PSP- II were not affected by the presence of ZnCl 2 in the medium at concentrations up to 4 mm (data not shown). However, the stability of the heterodimer against thermal denaturation was significantly reduced, as evidenced by monitoring the variation with tem- perature of the ellipticity at 268 nm (Fig. 1B). At 0.5 mm ZnCl 2 , a concentration of Zn 2+ in the range of those reported for porcine seminal plasma, T 1 ⁄ 2 falls Stability of PSP-I ⁄ PSP-II heterodimer M. A. Campanero-Rhodes et al. 5664 FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS to 53.2 °C, and a further decrease was observed at higher Zn 2+ concentrations (Table 1). Differential scanning calorimetry (DSC) In a former study [12], the thermal stability of the PSP-I ⁄ PSP-II heterodimer was analysed by DSC, showing that the entire dimer constituted the cooper- ative unfolding unit. Thermal denaturation curves of PSP-I ⁄ PSP-II presented a single peak with a maximum at 60.5 °C and an apparent enthalpy change of 439 kJÆ(mol dimer) )1 [12]. We have since observed some differences among protein batches in the calori- metric enthalpy changes, with a mean ± SD DH cal of 405 ± 17 kJÆmol )1 (r, n ¼ 8). These variations are not related to the protein concentration or the scan rate used in the analysis. However, the T m values of the DSC transitions were quite reproducible from batch to batch (60.7 ± 0.3 °C), thus serving as a use- ful gauge of the heterodimer thermostability. DSC data confirmed that, in the presence of ZnCl 2 , the thermal stability of PSP-I ⁄ PSP-II was substantially reduced (Fig. 2A). As the Zn 2+ concentration was increased, a concomitant decrease in both the trans- ition temperature and the apparent enthalpy of dena- turation was observed (Table 1), and, at 4 mm ZnCl 2 , protein precipitation occurred above 65 °C. The desta- bilization induced by Zn 2+ was reversed by the addi- tion of EDTA to the sample (Fig. 2A). On the other hand, no significant decrease in the heterodimer stabil- ity was observed in the presence of 4 mm CaCl 2 (Table 1), emphasizing the specificity of the effect of Zn 2+ . Thermal destabilization of PSP-I ⁄ PSP-II was also noticed at acidic pH (Fig. 2B) in the absence of Zn 2+ cations. At pH 3.8 the apparent enthalpy of denatura- tion decreased % 75 kJÆmol )1 and the transition tem- perature was 8 °C lower (Table 1). Ultracentrifugation and chromatographic behaviour The sedimentation equilibrium data for PSP-I ⁄ PSP-II (0.25–0.5 mgÆmL )1 ) could be fitted to a single-ideal- component model with a weight-average molecular mass of 27 933 Da, confirming that PSP-I ⁄ PSP-II behaved in solution as a dimer. No influence of Zn 2+ at concentrations up to 4 mm on the average mole- cular mass of PSP-I ⁄ PSP-II was observed at this protein concentration range. On gel filtration chromatography, the elution time of PSP-I ⁄ PSP-II at concentrations of, or above, 0.01 mgÆmL )1 was 26 min, consistent with the time predicted for the dimer. However, analysis of the gel filtration behaviour using 125 I-labelled PSP-I ⁄ PSP-II revealed a broadening of the peak at lower protein concentrations (Fig. 3A), with the appearance of minor species at the elution volume of the isolated subunits. Fig. 1. Near-UV CD of PSP-I ⁄ PSP-II. Variation with temperature (A) and effect of Zn 2+ on the thermal denaturation (B) of the heterodimer. Spectra were obtained for 1 mgÆmL )1 PSP-I ⁄ PSP-II solutions in 20 mM Hepes, pH 7.0. (A) Representative spectra acquired at 25 °C(h), 50 °C (n), 56 °C(n), 62 °C(m), 70 °C(s)and77°C(d) °C. (B) Variation in ellipticity at 268 nm with temperature monitored in the absence (s)orin the presence of 0.5 (n)or4(h)m M Zn 2+ . The continuous lines correspond to the fit of the experimental data to a sigmoidal function. Table 1. Thermodynamic parameters of the thermal denaturation of PSP-I ⁄ PSP-II as determined by CD (T 1 ⁄ 2 ) and DSC (T m , DH cal ). ND, Not determined. pH Additive (m M) T 1 ⁄ 2 (°C) T m (°C) DH CAL (kJÆmol )1 ) 7 None 62.2 ± 0.5 60.7 ± 0.3 405 ± 17 ZnCl 2 (0.5) 53.2 ± 0.2 59.8 ± 0.1 260 ± 20 ZnCl 2 (0.5) +EDTA (1) ND 60.8 ± 0.1 440 ± 40 ZnCl 2 (4) 46.8 ± 0.2 51.8 ± 0.3 240 ± 10 CaCl 2 (5) ND 61.6 ± 0.1 460 ± 30 3.8 None ND 52.9 ± 0.6 330 ± 20 M. A. Campanero-Rhodes et al. Stability of PSP-I ⁄ PSP-II heterodimer FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS 5665 This behaviour was not related to the radioiodination of the protein because a 0.75 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II was eluted as a single sharp peak at 26 min when it was chromatographed in the presence of unlabelled protein (Fig. 3A). In contrast, the results suggested the existence of an association- dissociation equilibrium leading to dissociation of the heterodimer at protein concentrations in the low nano- molar range. The presence of 3 mm CaCl 2 did not modify the chromatographic behaviour of PSP-I⁄ PSP-II. In con- trast, the addition of 2 mm Zn 2+ intensified the deviation of the elution profile at low protein concentrations from that of the dimer. Thus, at PSP-I ⁄ PSP-II concentrations below 0.06 mgÆmL )1 , the radio- iodinated protein was eluted as a broadened peak, with a displacement of the maximum towards longer elution times and a decrease in the total area of the peak (Fig. 3B). At a given protein concentration, the changes in the profile became more intense when the sample was preincubated with Zn 2+ before the chromatography, as shown in Fig. 4A for a 6 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II analysed imme- diately after the addition of 2 mm ZnCl 2 or after an incubation period of either 2 h or 16 h. The composi- tion of the fractions eluted from the column was ana- lysed by RP-HPLC, using a protocol designed for the separation of the PSP-I and PSP-II subunits [11]. When a mixture of unlabelled and 125 I-labelled PSP- I ⁄ PSP-II was chromatographed under the above condi- tions, two radioactivity peaks were co-eluted with the unlabelled PSP-I and PSP-II subunits, together with a third radioactive peak, appearing at the void volume, which corresponded to free 125 I (Fig. 4B). A similar analysis of the material eluted from the gel filtration column revealed that the first fractions of the sample eluted immediately after the addition of Zn 2+ con- tained both PSP-I and PSP-II subunits, whereas the fractions eluted later were mainly composed of PSP-II, supporting the dissociation of the heterodimer (Fig. 4B). Preincubation of the 125 I-labelled PSP-I ⁄ PSP-II sample with Zn 2+ resulted in a gradual decrease in the amount of PSP-I eluted from the gel filtration column, so that, after incubation for 16 h, only the PSP-II subunit was detected by HPLC analy- sis. The 125 I-labelled PSP-I subunit became partially adsorbed to the vials used for preincubation, as revealed by radioactivity monitoring and SDS ⁄ PAGE followed by autoradiography of the material eluted Fig. 2. DSC profiles of the thermal denaturation of PSP-I ⁄ PSP-II. Effect of Zn 2+ (A) and pH (B). The excess heat capacity function (DC p ) of PSP-I ⁄ PSP-II was determined at a scanning rate of 20 °CÆh )1 in 20 mM Hepes, pH 7 (thick solid line in A and B) or (A) in the same buffer containing 0.5 m M Zn 2+ (thin solid line), 0.5 mM Zn 2+ plus 1 mM EDTA (dash line), 1 mM Zn 2+ (dash-dot line) or 4m M Zn 2+ (dot line) or (B) in 10 mM citric acid ⁄ sodium citrate, pH 3.8 (dot line). Fig. 3. Dependence on protein concentration of the gel filtration chromatographic behaviour of PSP-I ⁄ PSP-II. Effects of Zn 2+ (B) and acidic pH (C). A 0.75 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II alone (dot lines) or in the presence of 5.5 mgÆmL )1 unlabelled PSP- I ⁄ PSP-II (continuous lines) was chromatographed on a Superose 12 column equilibrated with 10 m M Tris ⁄ HCl (pH 7.8) ⁄ 0.15 M NaCl ⁄ 0.02% NaN 3 (Tris ⁄ NaCl), in the absence (A) or presence of 2m M ZnCl 2 (Tris ⁄ NaCl-Zn 2+ ) (B), or with 50 mM sodium acet- ate ⁄ acetic acid buffer (pH 4) ⁄ 0.15 M NaCl ⁄ 0.02% NaN 3 (C). In (B), the elution profile of a 0.06 mgÆmL )1 solution of 125 I-labelled PSP- I ⁄ PSP-II in Tris ⁄ NaCl containing 2 m M Zn 2+ is also shown (dashed line). Stability of PSP-I ⁄ PSP-II heterodimer M. A. Campanero-Rhodes et al. 5666 FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS from the vial with SDS⁄ PAGE sample buffer. The remaining 125 I-labelled PSP-I was nonspecifically retained on the FPLC column (results not shown). Overall, the results show Zn 2+ -enhanced dissociation of the PSP-I and PSP-II subunits at low heterodimer concentrations. No enhancing effect of Mg 2+ on the dissociation of 125 I-labelled PSP-I ⁄ PSP-II samples was observed at concentrations up to 30 mm. The heterodimer dissociation was also enhanced at acidic pH. Gel filtration of a 0.75 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II at pH 4 resulted in broaden- ing of the peak and the appearance of species at the elution volume of the isolated subunits (Fig. 3C). The addition of Zn 2+ at this pH did not induce additional changes in the chromatographic behaviour. Discussion The near-UV CD spectrum of PSP-I ⁄ PSP-II reflects the specific environment of chiral aromatic side chains in the tertiary structure of the folded protein, and the band intensities decrease in a sigmoidal way as ther- mal denaturation occurs. In particular, the spectrum is characterized by the presence of a sharp positive band in the tryptophan absorption region (Fig. 1A). Both PSP-I and PSP-II subunits contain a single tryptophan residue, which is accommodated within the hydrophobic core of the CUB domain. This core is conserved in the X-ray structures of proteins con- taining the CUB signature, including the mannan- binding lectin-associated protease-2 (MASP-2) [23], its alternative splicing product Map19 [24], and the C1s protease of the C1 complex of complement [25]. Thus, the Trp band can be regarded as a characteristic fingerprint of the native fold of PSP-I and PSP-II. The near-UV CD spectra of the isolated PSP-I and PSP-II subunits are also characterized by the presence of this band (data not shown), strongly suggesting that they preserve the overall fold of the CUB domain. In the presence of Zn 2+ concentrations resembling physiological total amounts in seminal plasma, the ter- tiary structure of native PSP-I ⁄ PSP-II is preserved. However, the thermal stability of the heterodimer is significantly lower than in the absence of this cation, as evidenced by a lower apparent enthalpy and trans- ition temperature of the thermal denaturation. This destabilization occurs with the dissociation of the het- erodimer at low protein concentrations. Nevertheless, the concentration of PSP-I⁄ PSP-II in seminal plasma is clearly high enough to guarantee the integrity of the dimer. In addition, it should not be overlooked that complexation by other Zn 2+ -binding molecules in sem- inal plasma definitely limits the level of free zinc avail- able. The neutral to alkaline pH of normal boar seminal plasma also prevents dissociation of the PSP- I ⁄ PSP-II heterodimer, and perhaps contributes to the reported protective action of this spermadhesin com- plex on sperm viability [14]. In fact, whereas free PSP- I has also been found in the heparin-binding fraction of boar seminal plasma [26], no free PSP-II has been detected, indicating that PSP-I is synthesized in excess over PSP-II, and that the PSP-II subunit is quantita- tively engaged in complex formation with PSP-I. Therefore, the heparin and mannose 6-phosphate bind- ing sites of PSP-II, which have been proposed to be located at the heterodimer interface [20], may not be exposed in the male genital tract. On the other hand, an acidic pH, such as that found in seminal vesicle dysfunction, may decrease the ther- mal stability of PSP-I⁄ PSP-II and favours its dissoci- ation at low protein concentrations. Previous DSC studies on the thermal denaturation of PSP-I⁄ PSP-II [12] showed that the whole dimer constituted the cooperative unfolding unit, suggesting that inter- subunit interactions may contribute critically to the thermal stability. The heterodimer interface is largely hydrophobic, consisting of a central, solvent-inacces- sible hydrophobic core flanked at both sides by a clus- ter of polar ⁄ charged residues and a solvent-exposed aromatic amino acid (Fig. 5) [13]. In addition to Fig. 4. Effect of incubation of PSP-I ⁄ PSP-II heterodimer with Zn 2+ at low protein concentration. Gel filtration behaviour (A) and analy- sis by RP-HPLC (B) of the composition of the fractions derived from the gel filtration column. (A) A 6 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II was chromatographed at 0.5 mLÆmin )1 on a Superose 12 column equilibrated with Tris ⁄ NaCl-Zn 2+ immediately after the addition of 2 m M ZnCl 2 (continuous line) or after incuba- tion for either 2 h (dash line) or 16 h (dot line) with the cation. Then 1-mL fractions were collected. The composition of selected frac- tions of 0 h (d, s) and 16 h (m, n) 125 I-labelled PSP-I ⁄ PSP-II-Zn 2 was subsequently analysed by RP-HPLC (B) on a C 18 column eluted with an acetonitrile gradient (indicated by the line), as described in Experimental procedures. Control 125 I-labelled PSP-I ⁄ PSP-II (h). M. A. Campanero-Rhodes et al. Stability of PSP-I ⁄ PSP-II heterodimer FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS 5667 hydrophobic contacts and van der Waals interactions, a salt bridge and a number of hydrogen bonds contrib- ute to stabilization of the heterodimeric association. Weakening of these polar interactions, substantiated by the increased tendency of PSP-I ⁄ PSP-II to dissoci- ate at low protein concentrations, because of protona- tion of the groups involved or as a result of Zn 2+ complexation undoubtedly plays a part in the decrease in heterodimer thermal stability. For example, proto- nation and ⁄ or the potential involvement of Asp2 in Zn 2+ coordination by PSP-I would prevent the forma- tion of two strong hydrogen bonds with residues Tyr108 and Ser110 from PSP-II [13]. The entry of semen into the female genital tract is associated with dilution of the PSP-I ⁄ PSP-II heterodi- mer, and the acidic environment of the cervical, uterine and intraluminal sperm reservoir fluids [18] may eventu- ally contribute to pH-induced destabilization of the qua- ternary structure of the spermadhesin complex. These changes, possibly in conjunction with other factors or conditions encountered in the female tract, may give rise to separation of the PSP-I ⁄ PSP-II subunits. As a conse- quence, the heparin and mannose 6-phosphate binding sites on PSP-II would be exposed. It is important to emphasize that the reported stimulatory activity of PSP- II on macrophages is selectively inhibited by mannose 6-phosphate [17], suggesting the involvement of this binding site in the proposed activity of PSP-II as a post-mating inflammation mediator. The neutrophil recruitment induced by PSP-I appears to use a different mechanism, acting directly on neutrophils [17]. Thus, the dissociation of the PSP-I ⁄ PSP-II heterodimer in the female genital tract may be of physiological significance. It may be of relevance for the regulation of the duration and magnitude of uterine immune reactions, particularly in the search of strategies to optimize fecundity in artifi- cial insemination. Experimental procedures Isolation and radioiodination of PSP-I ⁄ PSP-II The PSP-I⁄ PSP-II heterodimer was isolated from the non- heparin-binding fraction of boar seminal plasma by gel filtration chromatography as described [11]. The protein (300 lg) was labelled with 0.2 mCi 125 I using Iodogen (Pierce, Rockford, IL, USA), according to the manufac- turer’s recommendations. The radioiodinated protein was indistinguishable from the corresponding unlabelled one on SDS ⁄ PAGE and autoradiography. CD spectra PSP-I ⁄ PSP-II samples were dialyzed extensively against 20 mm Hepes buffer, pH 7, in the absence or presence of different concentrations of ZnCl 2 . CD spectra were recor- ded in a JASCO J-720 spectropolarimeter (Jasco Corp., Tokyo, Japan), fitted with a water bath thermostatted cell holder, or in a J-810 spectropolarimeter, equipped with a peltier temperature control system, using a band width of 0.2 nm and a response time of 2 s. Far-UV spectra were recorded in 0.02 and 0.1 cm pathlength quartz cells at a protein concentration of 1 and 0.2 mgÆmL )1 , respectively. Near-UV spectra were acquired at 1.0 mgÆmL )1 protein concentration in 1 cm pathlength cells. At least three differ- ent scans were acquired and averaged for each sample. For all CD spectra, the corresponding buffer baseline was sub- tracted. The observed ellipticities were converted into mean residue ellipticities using a mean molecular mass per residue of 127.4. This value was calculated by dividing the average molecular mass obtained by MALDI MS (28 664 Da) by the number of amino-acid residues of the mature protein sequence (225 residues). Thermal denaturation experiments were carried out by increasing the temperature from 15 to 85 °C at a heating rate of 0.33 °CÆ min )1 , allowing the temperature to equili- brate for 5 min before recording the spectrum. Variations in ellipticity were monitored every 0.2 °C at 268 nm, and the complete spectrum was recorded every 5–15 °C, after an equilibration time of 1–5 min at the selected tempera- ture. No differences between the ellipticity values acquired at a given wavelength and those obtained from the spectra Fig. 5. Ribbon diagram of the PSP-I ⁄ PSP-II heterodimer showing the characteristics of the dimer interface. Residues of the hydro- phobic core are coloured in yellow, and hydrogen bonds formed at both sides by main-chain or side-chain atoms (coloured in CPK) of flanking polar residues are represented by dotted lines. The lateral chains of PSP-I Glu101 and PSP-II Arg43, which are involved in a salt bridge, are also shown. Residues are numbered according to the amino-acid sequence of the mature protein. In the lower part of the figure, PSP-I Asp2, a potential zinc ligand, forms two strong hydrogen bonds with residues Tyr108 and Ser110 from PSP-II. Stability of PSP-I ⁄ PSP-II heterodimer M. A. Campanero-Rhodes et al. 5668 FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS were observed. Thermal denaturation profiles were des- cribed in terms of the following sigmoidal function: HðTÞ¼H D ðTÞÀ½H D ðTÞÀH N ðTÞ=f1 À exp½AðT À T 1=2 Þ= RTT 1=2 g where T is the absolute temperature, T 1 ⁄ 2 is the half transition temperature, R is the gas constant, A is the temperature con- stant accounting for the ratio between the native and dena- tured states, and Q D (T)andQ N (T) are the ellipticity of the denatured and native states at temperature T. Q D and Q N were approximated as linear functions of temperature [Q i (T) ¼ Q i (T 0 )+m i (T ) T 0 ), where T 0 is the reference temperature and m i is temperature dependence of Q i for i ¼ N or D]. DSC For DSC, samples were dialyzed extensively against 20 mm Hepes buffer, pH 7, in the absence or presence of different concentrations of ZnCl 2 or CaCl 2 , unless otherwise stated. DSC measurements were performed using a Microcal MCS instrument (Microcal, Inc., Northampton, MA, USA) at a heating rate of 0.33 KÆmin )1 and under an extra constant pressure of 2 atm. The standard Microcal origin software was used for data acquisition and analysis. The excess heat capacity functions were obtained after subtraction of the buffer baseline. Reversibility of the transitions was checked by performing a second analysis after the first scan. Gel filtration chromatography Gel filtration was carried out on a Superose 12 HR 10 ⁄ 30 column (Pharmacia LKB Biotechnology, Uppsala, Sweden) equilibrated with 10 mm Tris ⁄ HCl (pH 7.8) ⁄ 0.15 m NaCl (Tris ⁄ NaCl), containing 0.02% (w ⁄ v) NaN 3 and, where sta- ted, ZnCl 2 or CaCl 2 at the indicated concentration. Alter- natively, the column was equilibrated with 50 mm sodium acetate ⁄ acetic acid buffer (pH 4) ⁄ 0.15 m NaCl ⁄ 0.02% (w ⁄ v) NaN 3 . The flow rate was 0.5 mLÆmin )1 , and the elution was monitored at 280 nm. Control proteins were chromatographed under similar conditions. For loading radioiodinated PSP-I ⁄ PSP-II on to the col- umn, the injection syringe was previously blocked for 3 h at 20 ° C with 10% (v ⁄ v) Tween 20 (Sigma, St Louis, MO, USA). Then 1-mL fractions were collected into vapex sam- ple tubes (PerkinElmer, Turku, Finland), similarly blocked with 0.5% (v ⁄ v) Tween 20 for 16 h at 20 °C, and their radioactivity was measured in an LKB MiniGamma counter (LKB Wallac, Turku, Finland). Composition of the frac- tions was monitored by HPLC analysis, as described below. RP-HPLC Fractions collected from the gel filtration chromatography of 125 I-labelled PSP-I ⁄ PSP-II were mixed with 250 lg unla- belled PSP-I⁄ PSP-II, and 500 lL of this mixture was ana- lysed by RP-HPLC on a 5-lm Hypersil ODS C 18 column (Sugelabor, Madrid, Spain), eluted at 1 mLÆmin )1 with an acetonitrile gradient in 0.1% (v ⁄ v) trifluoroacetic acid as follows: (a) 35% acetonitrile isocratically for 5 min; (b) 35–40% (v ⁄ v) for 5 min; (c) 40–50% for 80 min; (d) 50–70% (v ⁄ v) acetonitrile for 10 min. The column was re-equilibrated with 35% (v ⁄ v) acetonitrile for 20 min before application of a new sample. The elution was moni- tored at 280 nm, and 3 mL fractions were collected. The elution position of the radioiodinated PSP-I and PSP-II subunits was checked by analysing control 125 I-labelled PSP-I ⁄ PSP-II under the same conditions. Analytical ultracentrifugation Sedimentation equilibrium experiments were performed by centrifugation of 80-lL samples of concentration 0.5 mgÆmL )1 , at 30 000 g and 20 °C, in an Optima XL-A analytical ultracentrifuge (Beckman Coulter Instruments, Inc., Richmond, CA, USA) equipped with UV-Vis optics and An50Ti analytical rotor. Data were collected using 12 mm pathlength double-sector six-channel centre pieces with quartz windows. Under these conditions, equilibrium was reached before 12 h of centrifugation. Baseline offsets were determined from radial scans of the samples run for 6 h at 160 000 g. Weight-average molecular masses, M w , were calculated with the xlaeq program, using the signal conservation algorithm [27]. Acknowledgements We thank DGICYT (BQU2000-1501-C02-02, BQU2003- 03550-C03-03, BIO2003-01952 and BFU2004-1432) for financial support. We also thank Professor Heriberto Rodrı ´ guez-Martı ´ nez (Faculty of Veterinary Medicine, Clinical Centre Ultuna, Uppsala, Sweden) for critical reading of the manuscript and helpful discussions. References 1 Minton AP (2005) Influence of macromolecular crowd- ing upon the stability and state of association of pro- teins: predictions and observations. J Pharm Sci 94, 1668–1675. 2 Mann T & Lutwak-Mann C (1981) Biochemistry of semi- nal plasma and male accessory fluids: application to andro- logical problems. In Male Reproductive Function and Semen. Springer-Verlag, Berlin, Heidelberg, New York. 3 Boursnell JC, Baronos S, Briggs PA & Butler EJ (1972) The concentrations of zinc in boar seminal plasma and vesicular secretion in relation to those of nitrogenous substances, citrate, galactose and fructose. J Reprod Fertil 29, 215–227. M. A. Campanero-Rhodes et al. Stability of PSP-I ⁄ PSP-II heterodimer FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS 5669 4 Arver S & Eliasson R (1980) Zinc and magnesium in bull and boar spermatozoa. J Reprod Fertil 60, 481–484. 5 Rodrı ´ guez-Martı ´ nez H, Ekwall H, Kvist U, Malmgren L & Plo ¨ en L (1987) X-ray microanalysis of boar sper- matozoa: changes in the elemental composition at ejacu- lation. Proceedings of the 40th Annu. Meet. SCANDEM, Bergen 1 , p. 34. 6 Shivaji S, Scheit K-H & Bhargava PM (1990) Proteins of Seminal Plasma. Wiley & Sons, New York. 7 Calvete JJ, Sanz L, Dosta ` lova ` Z&To ¨ pfer-Petersen E (1995) Spermadhesins: sperm-coating proteins involved in capacitation and zona pellucida binding. Fertilita ¨ t 11, 35–40. 8 Romero A, Roma ˜ o MJ, Varela PF, Ko ¨ lln I, Dias JM, Carvalho AL, Sanz L, To ¨ pfer-Petersen E & Calvete JJ (1997) The crystal structures of two spermadhesins reveal the CUB domain fold. Nat Struct Biol 4, 783–788. 9To ¨ pfer-Petersen E, Romero A, Varela PF, Ekhlasi- Hundrieser M, Dosta ` lova ` Z, Sanz L & Calvete JJ (1998) Spermadhesins: a new protein family. Facts, hypotheses and perspectives. Andrologia 30, 217–224. 10 Dosta ` lova ` Z, Calvete JJ, Sanz L & To ¨ pfer-Petersen E (1994) Quantitation of boar spermadhesins in accessory sex gland fluids and on the surface of epididymal, ejacu- lated and capacitated spermatozoa. Biochim Biophys Acta 1200, 48–54. 11 Calvete JJ, Mann K, Scha ¨ fer W, Raida M, Sanz L & To ¨ pfer-Petersen E (1995) Boar spermadhesin PSP-II: location of posttranslational modifications, heterodimer formation with PSP-I glycoforms and effect of dimeriza- tion on the ligand-binding capabilities of the subunits. FEBS Lett 365, 179–182. 12 Mene ´ ndez M, Gasset M, Laynez J, Lo ´ pez-Zumel C, Usobiaga P, To ¨ pfer-Petersen E & Calvete JJ (1995) Analysis of the structural organization and thermal stability of two spermadhesins. Calorimetric, circular dichroic and Fourier-transform infrared spectroscopic studies. Eur J Biochem 234, 887–896. 13 Varela PF, Romero A, Sanz L, Roma ˜ o MJ, To ¨ pfer- Petersen E & Calvete JJ (1997) The 2.4 A ˚ resolution crystal structure of boar seminal plasma PSP-I ⁄ PSP-II: a zona pellucida-binding glycoprotein heterodimer of the spermadhesin family built by a CUB domain archi- tecture. J Mol Biol 274, 635–649. 14 Centurio ´ nF,Va ´ zquez JM, Calvete JJ, Roca J, Sanz L, Parrilla I, Garcı ´ a EM & Martı ´ nez EA (2003) Influence of porcine spermadhesins on the susceptibility of boar spermatozoa to high dilution. Biol Reprod 69, 640–646. 15 Leshin LS, Raj SM, Smith CK, Kwok SC, Kraeling RR & Li WI (1998) Immunostimulatory effects of pig semi- nal proteins on pig lymphocytes. J Reprod Fertil 114, 77–84. 16 Yang WC, Kwok SCM, Leshin S, Bollo E & Li WI (1998) Purified porcine seminal plasma protein enhances in vitro immune activities of porcine peripheral lympho- cytes. Biol Reprod 59, 202–207. 17 Assreuy AMS, Calvete JJ, Alencar NMN, Cavada BS, Rocha-Filho DR, Melo SC, Cunha FQ & Ribeiro RA (2002) Spermadhesin PSP-I ⁄ PSP-II heterodimer and its isolated subunits induced neutrophil migration into the peritoneal cavity of rats. Biol Reprod 67, 1796–1803. 18 Rodriguez-Martinez H, Saravia F, Wallgren M, Tien- thai P, Johannisson A, Va ´ zquez JM, Martı ´ nez E, Roca J, Sanz L & Calvete JJ (2005) Boar spermatozoa in the oviduct. Theriogenology 63, 514–535. 19 Assreuy AMS, Alencar NMN, Cavada BS, Rocha-Filho DR, Feitosa RFG, Cunha FQ, Calvete JJ & Ribeiro RA (2003) Porcine spermadhesin PSP-I ⁄ PSP-II stimu- lates macrophages to release a neutrophil chemotactic substance: modulation by mast cells. Biol Reprod 68, 1836–1841. 20 Solı ´ s D, Romero A, Jime ´ nez M, Dı ´ az-Maurin ˜ oT& Calvete JJ (1998) Binding of mannose-6-phosphate and heparin by boar seminal plasma PSP-II, a member of the spermadhesin protein family. FEBS Lett 431, 273–278. 21 Kwok SCM, Soares MJ, McMurtry JP & Yurewicz EC (1993) Binding characteristics and immunolocalization of porcine seminal protein, PSP-I. Mol Reprod Dev 35, 244–250. 22 Jona ´ kova ´ V, Man ˇ a ´ skova ´ P, Graus M, Liberda J & Ticha ´ M (2000) Sperm surface proteins in mammalian fertilization. Mol Reprod Dev 56, 275–277. 23 Feinberg H, Uitdehaag JC, Davies JM, Wallis R, Drick- amer K & Weis WI (2003) Crystal structure of the CUB1-EGF-CUB2 region of mannose-binding protein associated serine protease-2. EMBO J 22, 2348–2359. 24 Gregory LA, Thielens NM, Matsushita M, Sorensen R, Arlaud GJ, Fontecilla-Camps JC & Gaboriaud C (2004) The X-ray structure of human mannan-binding lectin- associated protein 19 (MAp19) and its interaction site with mannan-binding lectin and 1-ficolin. J Biol Chem 279, 29391–29397. 25 Gregory LA, Thielens NM, Arlaud GJ, Fontecilla- Camps JC & Gaboriaud C (2003) X-ray structure of the Ca 2+ -binding interaction domain of C1s. Insights into the assembly of the C1 complex of complement. J Biol Chem 278, 32157–32164. 26 Sanz L, Calvete JJ, Mann K, Gabius HJ & To ¨ pfer- Petersen E (1993) Isolation and biochemical characteri- zation of heparin-binding proteins from boar seminal plasma: a dual role for spermadhesins in fertilization. Mol Reprod Dev 35, 37–43. 27 Minton AP (1994) Conservation of signal: a new algo- rithm for the elimination of the reference concentration as an independently variable parameter in the analysis of sedimentation equilibrium. In Modern Analytical Ultracentrifugation (Schuster TH & Laue TH, eds), pp. 81–93. Birkha ¨ user, Boston. Stability of PSP-I ⁄ PSP-II heterodimer M. A. Campanero-Rhodes et al. 5670 FEBS Journal 272 (2005) 5663–5670 ª 2005 FEBS . Analysis of the stability of the spermadhesin PSP-I ⁄ PSP-II heterodimer Effects of Zn 2+ and acidic pH Marı ´a Asuncio ´ n Campanero-Rhodes 1 ,. filtration chromatographic behaviour of PSP-I ⁄ PSP-II. Effects of Zn 2+ (B) and acidic pH (C). A 0.75 lgÆmL )1 solution of 125 I-labelled PSP-I ⁄ PSP-II alone