Blackwell Science - Synergy

Oxidized galectin-1 promotes axonal regeneration in peripheral nerves but does not possess lectin properties
Yoshimasa Inagaki^1,², Yoshiaki Sohma¹, Hidenori Horie³, Risa Nozawa¹ andToshihiko Kadoya¹

Galectin-1 has recently been identified as a factor that regulates initial axonal growth in peripheral nerves after axotomy. Although galectin-1 is a well-known b-galactoside-binding lectin, its potential to promote axonal regeneration as a lectin has not been reported. It is essential that the process of initial repair in peripheral nerves after axotomy is well clarified. We therefore undertook to investigate the relation between the structure and axonal regeneration-promoting activity of galectin-1. Recombinant human galectin-1 secreted into the culture supernatant of transfected COS1 cells (rhGAL-1/COS1) was purified under nonreducing conditions and subjected to structural analysis. Mass spectrometric analysis of peptide fragments from rhGAL-1/COS1 revealed that the secreted protein exists as an oxidized form containing three intramolecular disulfide bonds (Cys2–Cys130, Cys16–Cys88 and Cys42–Cys60). Recombinant human galectin-1 (rhGAL-1) and a galectin-1 mutant in which all six cysteine residues were replaced by serine (CSGAL-1) were expressed in and purified from Escherichia coli for further analysis; the purified rhGAL-1 was subjected to oxidation, which induced the same pattern of disulfide linkages as that observed in rhGAL-1/COS1. Oxidized rhGAL-1 enhanced axonal regeneration from the transected nerve sites of adult rat dorsal root ganglion explants with associated nerve stumps (5.0–5000 pgámL^-1), but it lacked lectin activity. In contrast, CSGAL-1 induced hemagglutination of rabbit erythrocytes but lacked axonal regeneration-promoting activity. These results indicate that galectin-1 promotes axonal regeneration only in the oxidized form containing three intramolecular disulfide bonds, not in the reduced form which exhibits lectin activity.

Successful nerve regeneration requires the concerted interplay of non-neuronal cells, growth factors, cell adhesion molecules, extracellular matrix materials, regenerating axons and recruiting macrophages [1–3]. However, although various neurotrophins have been investigated to promote axonal regeneration in vivo, it is still unclear what factor initiates axonal regeneration response to nerve injury. We found one of these factors from the culture supernatant of COS1 cells in a previous study [4]. The factor enhanced axonal regeneration in the in vitro nerve regeneration model employed, allowing for initial axon outgrowth from the proximal nerve stump to be monitored, which is comparable with the initial stages of nerve repair. Analysis of the purified protein indicated that it was identical to galectin-1. Recombinant human galectin-1 (rhGAL-1) confirmed that the protein promotes axonal regeneration not only in the in vitro experiment, but also in two other types of in vivo acellular nerve regeneration model. Furthermore, antibodies to galectin-1 clearly inhibited axonal regeneration in vivo as well as in vitro. These observations suggest that galectin-1 might play an important physiological role as a promoter of neuronal regeneration. Because galectin-1 has been identified only recently as a factor in promoting axonal regeneration, elucidating the mechanism of its activity and the structural features of galectin-1 are essential to understanding the initial process of axonal regeneration.

Galectins constitute a family of lectins that bind b-galactoside and are present in a wide range of animal species, ranging from Caenorhabditis elegans to human [5–8]. Ten members of the galectin family have been identified to date. Galectin-1 is the most well-characterized member of the galectin family. The complete amino acid sequence has been determined for galectin-1 from human [9–11], cow [11a], rat [12], mouse [13,14], chicken [15,16] and electric eel [17]. Galectin-1 is a homodimer with a subunit molecular mass of 14.5 kDa, and contains six cysteine residues per subunit. The b-galactoside-binding activity of galectin-1 is evident only in the presence of a reducing agent [12,18–21].

Galectin-1 is expressed in many tissues, including skeletal and smooth muscle, liver, lung, heart, skin, spleen and intestine [8]. In vitro studies suggest that galectin-1 is involved in the regulation of cell adhesion, cell proliferation and apoptosis in various cell types including lymphocytes, thymocytes and vascular cells [8]. Most of these prior studies were performed under reducing conditions, and the effects of galectin-1 were inhibited by lactose. The biological effects of galectin-1 were therefore assumed to depend on its ability to bind b-galactoside. Galectin-1 has also been localized and proposed to function in the central and peripheral nervous systems [22–25]. Galectin-1 mRNA becomes abundant in sensory neurons and motoneurons in the spinal cord and brainstem soon after neuronal differentiation [23]. Galectin-1 is also expressed in the developing brain [26] and olfactory system [25,27]. It has also been proposed to function as a lectin in neuronal pathfinding [27], contributing to fasciculation or the guidance of axons by acting as a substrate.

In our previous study, however, we purified galectin-1 as an axonal regeneration-promoting factor from the culture supernatant of COS1 cells under nonreducing conditions. rhGAL-1 promoted axonal regeneration at a concentration of 50 pgámL^-1, which is more than two or three orders of magnitude lower than the concentration at which galectin-1 exhibits lectin activity. Similar activity was also apparent in vivo using a peripheral axotomy model under nonreducing conditions. Therefore, it is possible that the mechanism by which galectin-1 exhibits this activity is different to that of its lectin activity.

In this study, rhGAL-1 was expressed in COS1 cells and purified under nonreducing conditions from the culture supernatant. The structure of the purified protein was then determined in order to investigate the mechanism by which it promotes axonal regeneration. Secreted rhGAL-1 was shown to exist in an oxidized form with three intramolecular disulfide bonds, and the axonal regeneration-promoting activity of galectin-1 was confirmed to be unrelated to its lectin properties.

Experimental procedures

Results Discussion Acknowledgments References

Molecular cloning of human galectin-1 and construction of the mammalian expression vector

Human galectin-1 cDNA [11] was isolated from a SuperScript human liver cDNA library (Gibco-BRL) by amplification with the nested PCR [28], as described previously [4]. The products of the PCR were rendered blunt-ended with T4 DNA polymerase, digested with NotI, and inserted into the mammalian expression vector pEF18S [29]. The identity of the resulting plasmid (pEFGal1) was confirmed by dideoxy sequencing [30] with an automated sequencer (ABI, model 377).

Construction of bacterial expression plasmids

A plasmid for expression of human galectin-1 in Escherichia coli (pETGal1) was constructed as described previously [4].

A plasmid for bacterial expression of a mutant human galectin-1 in which all cysteines (residues 2, 16, 42, 60, 88, and 130) were replaced by serine was also constructed. To replace Cys130 with Ser, the oligonucleotides HLEG21 (5¢-AATTCAAGTTCCCCAACCGCCTCAACCTGGAGGCCATCAA

CTACATGGCAGCTGACGGTGACTTCAAGATCAAAAGCGTG

GCCTTTGACTGATAAGC-3¢) and HLEG22 (5¢-GGCCGCT-TATCAGTCAAAGGCCACGCTTTTGATCTTGAAGTCACCGT

CAGCTGCCATGTAGTTGATGGCCTCCAGGTTGAGGCGGTT

GGGGAACTTG-3¢, containing an additional sequence to generate a NotI site at the 3¢-end), which were designed in accordance with the sequence from the internal EcoRI site to the 3¢-end of the coding region of human galectin-1 cDNA, were annealed and inserted into the pGEX-5X-2 vector (Amersham Pharmacia). To replace the other five cysteines with serine, we performed a series of PCR amplifications with AmpliTaq DNA polymerase (ABI) and the following primers: 5¢-GAGAGAGGATCCCCATGGCTAGCGGTCTGGTCG-3¢ (HLEG15; the underlined portion representing an additional sequence designed to generate a BamHI site at the 5¢-end), 5¢-ATGGCTAGCGGTCTGGTCGCCAGCAACCTGAATCTCAAAC

CTGGAGAGAGCCTTCG-3¢ (HLEG16), 5¢-ACCTGAGCCTGCACTTCAACCCTCGCTTCAACGCCCACGG

CGACGCCAACACCATCGTGAGCAAC-3¢ (HLEG17), 5¢-GT-TGCTCACGATGGTGTTGGCGTCGCCGTGGGCGTTGAAGCG

AGGGTTGAAGTGCAGGCTCAGGT-3¢ (HLEG18), 5¢-AACTTGAATTCGTATCCATCTG-3¢ (HLEG19) and 5¢-AACTTGAATTCGTATCCATCTGGCAGCTTGACGGTCAGGT

TGGCCTGGTCGAAGGTGATGCTCAC-3¢ (HLEG20). A PCR was performed with HLEG16 and HLEG18 as primers and pEFGal1 as the template. A similar PCR was performed with primers HLEG17 and HLEG20. These two reaction mixtures were then combined and subjected to amplification with primers HLEG15 and HLEG19. The final PCR products were digested with EcoRI and BamHI, and inserted into the pGEX-5X-2 vector containing the mutated 3¢-portion of human galectin-1 cDNA. The identity of the resulting plasmid (pGEXCSGal1) was confirmed by dideoxy sequencing. This vector allowed expression of the mutant galectin-1 (CSGAL-1) as a glutathione S-transferase fusion protein; digestion of the fusion protein with factor Xa yields CSGAL-1 with the additional sequence Gly-Ile-Pro-Met at its N-terminus.

Expression and purification of recombinant proteins

Expression and purification of rhGAL-1 were performed as described previously [4]. For expression of CSGAL-1, cultures of E. coli strains DH5 were transformed with the pGEXCSGal1 expression vector, and grown overnight at 37 ¡C in Luria–Bertani medium. After dilution (1 : 100) with fresh medium, the cultures were incubated with shaking at 37 ¡C until they achieved a D₆₀₀ of 0.5–0.6. Expression of CSGAL-1 was induced by the addition of isopropyl thio-b-D-galactoside to a final concentration of 0.1 mM and further incubation for 3 h at 37 ¡C.

To purify CSGAL-1, we centrifuged the cultures for 30 min at 10 000 g, and the resulting pellets were resuspended in 20 vol NaCl/P_i. The cells were lysed by sonication, and the lysate was centrifuged at 10 000 g for 30 min. The supernatant obtained after sonication and centrifugation was applied, at a flow rate of 1.0 mLámin^-1, to a gluthathione–Sepharose 4B column (3 × 5 cm; Amersham Pharmacia) that had been equilibrated with NaCl/P_i. The column was washed with 50 mM Tris/HCl (pH 8.0) containing 1 mM CaCl₂ and 100 mM NaCl, after which factor Xa was injected into the column and the column was maintained at room temperature overnight. Proteins were then eluted with 50 mM Tris/HCl (pH 8.0) containing 100 mM NaCl. The fractions containing CSGAL-1 were concentrated with a YM3 membrane (Amicon) and, after a change of buffer to 20 mM Tris/HCl (pH 8.0), were applied at a flow rate of 0.5 mLámin^-1 to a Shodex IEC DEAE-825 column (8 × 75 mm; Showa Denko) that had been equilibrated with the same solution. The bound proteins were eluted from the column by application over 60 min of a linear gradient of 0–300 mM NaCl in 20 mM Tris/HCl (pH 8.0). The fractions containing CSGAL-1 were then applied, at a flow rate of 0.5 mLámin^-1, to a TSKgel Phenyl-5PW RP column (4.6 × 75 mm; Tosoh). Bound proteins were eluted by application over 40 min of a linear gradient of 16–64% (v/v) acetonitrile in 0.1% trifluoroacetic acid (TFA).

The concentrations of the recombinant proteins were determined by amino acid analysis with the AccQ Tag method [31], and subsequent experiments were performed on the basis of these data.

Preparation of antibodies to human galectin-1

Rabbit anti-(human galectin-1) serum was obtained as described previously. It was determined that this antibody reacted with galectin-1 alone [4].

Mass spectrometry

Intact wild-type and mutant galectin-1 proteins as well as proteolytic fragments of the wild-type protein were analyzed by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry using an instrument (Voyager-DE STR; PerSeptive) that was equipped with a pulsed nitrogen laser with a wavelength of 337 nm. A portion (0.5 µL) of each protein preparation was mixed with 0.5 µL of matrix solution on stainless steel plates, and the mixtures were then air-dried at ambient temperature. The samples were desorbed and ionized by irradiation using the pulsed nitrogen laser, and the signals of positive ions were collected, digitized, and averaged using a digital oscilloscope (Tectronix) controlled by a laboratory computer. The matrix solutions comprised sinapinic acid (Aldrich) dissolved at a concentration of 10 mgámL^-1 in acetonitrile : water (3 : 7, v/v) containing 0.1% TFA for intact galectin-1 and its mutant, and a-cyano-4-hydroxy cinnamic acid (Aldrich) dissolved at a concentration of 10 mgámL^-1 in acetonitrile : water (1 : 1, v/v) containing 0.1% TFA for the proteolytic fragments.

Proteolytic digestion of galectin-1 and disulfide bond assignment

The purified rhGAL-1 from the culture supernatant of COS1 cells was digested with trypsin (modified, sequencing grade; Boehringer Mannheim) in 0.1 M Tris/HCl (pH 6.8) for 16 h at 37 ¡C at an enzyme/substrate ratio of 1 : 25. A second digestion was performed for one of the resulting disulfide-bonded fragments with Achromobacter protease I (API, lysylendopeptidase; Wako) under the same conditions as for the trypsin digestion. Each digestion mixture was applied at a flow rate of 0.25 mLámin^-1 to a Symmetry C18/3.5 µm column (2.1 × 50 mm; Waters), and elution was performed over 50 min with a gradient of 1–50% solvent B [solvent A, 0.05% TFA; solvent B, 0.02% TFA in 2-propanol : acetonitrile (7 : 3, v/v)]. The peptide fragments recovered from each digestion were analyzed by MALDI-TOF mass spectrometry. The peptide masses were compared with those of predicted digestion fragments of human galectin-1, and disulfide bonds were assigned.

Multiangle laser light-scattering analysis

The absolute molecular size of galectin-1 proteins was determined by on-line, size-exclusion HPLC and light scattering. Protein samples (90 µL) dissolved in NaCl/P_i in the absence or presence of 4 mM dithiothreitol were applied to a Shodex KW 803 column (8 × 30 mm; Showa Denko) that had been equilibrated with NaCl/P_i. Elution was performed at a flow rate of 0.25 mLámin^-1, and absolute molecular size was determined with a WYATT DAWN DSP-F light-scattering photometer (Wyatt) and an RI SE-6 refractive index detector (Showa Denko). The molecular size distribution was calculated by the mass method using the ASTRA software program (Wyatt).

Assay for lectin activity

Lectin activity of rhGAL-1 and CSGAL-1 was determined by measurement both of hemagglutination of rabbit erythrocytes and of binding to lactose–agarose. Agglutination assays were performed in 96-well microtiter plates with serial twofold dilutions (50–0.39 µgámL^-1) of the purified proteins in NaCl/P_i in the absence or presence of 5 mM dithiothreitol. Samples (50 µL) were gently mixed with a 2% (v/v) suspension of erythrocytes (50 µL) and incubated at room temperature for 1 h; in the case of the reducing condition, samples were pre-incubated for 2 h with 5 mM dithiothreitol before addition to the erythrocytes. Agglutination activity was determined on the basis of the sedimentary state of the erythrocytes.

To assay the binding to lactose–agarose, oxidized or dithiothreitol-treated rhGAL-1 in NaCl/P_i was applied at a flow rate of 0.5 mLámin^-1 to a lactose–agarose column (5 × 50 mm; Honen), and elution was performed over 30 min with a linear gradient of 0–0.1 M lactose in NaCl/P_i.

Assay of axonal regeneration-promoting activity

Promotion of axonal regeneration was assayed with a three-dimensional dorsal root ganglion (DRG) explant culture as described previously [4,32]. All preparations were obtained from 3-month-old Wistar rats (Nihon SLC), which were anesthetized with ether before killing. DRGs of similar size (Th-2 to Th-10) were dissected carefully and removed with their associated nerve stumps. Those with associated nerve stumps 0.7–2.0 mm in length were embedded in a collagen gel and cultured at 37 ¡C under 5% CO₂ in Ham’s F12 medium supplemented with insulin (5 µgámL^-1), transferrin (5 µgámL^-1), 20 nM progesterone, 0.1 mM putrescine and 30 nM selenium (all supplements from Sigma). Samples were diluted with the above culture medium and added to the cultures at various concentrations. The nerve stumps associated with the DRG explant extended in opposite directions: one toward the spinal cord (central nerve stump), and the other toward the sensory organs (peripheral nerve stump). The total number of regenerating axons at each of these transected nerve sites was counted before branching at a region adjacent to the sites with the use of a phase-contrast microscope. Data are expressed as means ±SEM, and differences were evaluated by analysis of variance (ANOVA), followed by Student–Newman–Keuls’s test. A P value of <0.05 was considered statistically significant.

Results

Experimental procedures Discussion Acknowledgments References

Purification and characterization of human galectin-1 secreted into the culture supernatant of COS1 cells

We have previously shown that endogenous galectin-1 secreted into the culture supernatant of COS1 cells enhances regeneration of peripheral nerves after axotomy. The secreted protein was purified and its biological activity was assayed under nonreducing conditions. Because galectin-1 contains six cysteine residues per subunit, it was possible that the secreted and active protein existed as an oxidized form containing intramolecular disulfide bonds. To characterize secreted galectin-1, we have now expressed rhGAL-1 in COS1 cells and purified the recombinant protein (rhGAL-1/COS1) from the culture supernatant of these cells. Whereas previous methods of purification of galectin-1 have relied on lactose–agarose or asialofetuin–agarose affinity chromatography performed under reducing conditions, we needed to purify the secreted protein under nonreducing conditions. We chose a combination of immunoaffinity chromatography with antibodies to human galectin-1 and reverse-phase HPLC.

COS1 cells incubated in Iscove’s modified Eagle’s medium supplemented with 10% fetal bovine serum were transfected with pEFGal1 with the use of the Transfectam reagent (Promega). The cells were subsequently cultured in medium containing 10% serum for 2 days, after which the culture supernatant was collected. The supernatant was applied to a NaCl/P_i-equilibrated column of Sepharose conjugated with the immunoglobulin G fraction of rabbit antiserum to human galectin-1. The column was washed with NaCl/P_i, and the bound galectin-1 was eluted with 0.1 M glycine/HCl (pH 2.7). The affinity-purified galectin-1 was then applied to a YMC-Pack Protein RP column (4.6 × 150 mm; YMC) ( Fig. 1 ).

The rhGAL-1/COS1 present in the major peak eluted from the reverse-phase HPLC column was analyzed by SDS/PAGE [33] ( Fig. 2 ) and by MALDI-TOF mass spectrometry ( Fig. 3 ). Both analyses demonstrated that rhGAL-1/COS1 as purified differed from the reduced form of the protein. Thus, the electrophoretic mobility of rhGAL-1/COS1 under reducing conditions differed from that under nonreducing conditions. Mass spectrometry revealed that the measured average molecular mass of rhGAL-1/COS1 was 14 620.5 Da, which was 36 Da larger than the theoretical average molecular mass of the reduced form predicted from the amino acid sequence of human galectin-1 (14 584.4 Da). However, the N-terminal amino acid of human galectin-1 has been shown to be blocked by an acetyl group (42 Da) [7]. Therefore, the measured average molecular mass of rhGAL-1/COS1 (14 620.5 Da) was actually 6 Da smaller than the theoretical average molecular mass of the reduced and acetylated form of the protein (14 626.4 Da).

The structure of rhGAL-1/COS1 was analyzed by a combination of peptide mapping and MALDI-TOF mass spectrometry. Nine major peaks were obtained after reverse-phase HPLC of a tryptic digest of purified rhGAL-1/COS1 ( Fig. 4A ). Each peak was analyzed by mass spectrometry and assigned to a tryptic fragment predicted from the sequence of human galectin-1 ( Table 1 ). Peptides TP-6 and TP-9 were assigned as the disulfide bond-containing peptides. TP-6, with a measured mass/charge (m/z) ratio of 3017.9, was assigned as the peptide in which Cys42 and Cys60 were linked. TP-9, with a measured m/z ratio of 5263.0, was assigned as the peptide containing two disulfide bonds and an acetylated N-terminus; in this peptide, Cys2 was linked to Cys88 or Cys130, and Cys16 was linked to Cys88 or Cys130. To determine which form of TP-9 was correct, the peptide was digested with API. HPLC revealed two major digestion products ( Fig. 4B ): TPAP-1, with a measured m/z ratio of 1796.5, was assigned as the peptide in which Cys2 was linked to Cys130, and TPAP-2, with a measured m/z value of 3482.5, was assigned as the peptide in which Cys16 was linked to Cys88. From these results, it was determined that secreted rhGAL-1/COS1 is an oxidized form of galectin-1 containing three intramolecular disulfide bonds (Cys2–Cys130, Cys16–Cys88, and Cys42–Cys60).

Preparation and physicochemical analysis of oxidized rhGAL-1 and CSGAL-1

The oxidized form of human galectin-1 was prepared in order to investigate its physicochemical properties and the relation between lectin activity and axonal regeneration-promoting activity. Oxidized galectin-1 was obtained from bacterially expressed rhGAL-1 by the air oxidation method with CuSO₄ as a catalyst. DEAE-purified rhGAL-1 was diluted 20-fold with 20 mM Tris/HCl (pH 8.0), CuSO₄ was added to a final concentration of 0.0001% (w/v), and the mixture was maintained overnight at 4 ¡C to allow disulfide bond formation. The sample was subsequently concentrated with a YM3 membrane (Amicon) and applied to a YMC-Pack Protein RP column (1 × 25 cm; YMC). Fractionation was performed over 60 min at a flow rate of 2.0 mLámin^-1 with a linear gradient of 32–40% acetonitrile in 0.1% TFA. Oxidized rhGAL-1 with same disulfide linkages as those in rhGAL-1/COS1 eluted as a major peak. The structure of oxidized rhGAL-1 was determined by a combination of peptide mapping and MALDI-TOF mass spectrometry (data not shown). Reverse-phase HPLC analysis revealed that the retention time of RP-purified oxidized rhGAL-1 was markedly shorter than that of DEAE-purified rhGAL-1 ( Fig. 5 ). SDS/PAGE analysis ( Fig. 2 ) and MALDI-TOF mass spectrometry ( Fig. 3 ) showed that the oxidized rhGAL-1 contained three intramolecular disulfide bonds but did not contain an acetylated N-terminus.

The mutant CSGAL-1 was also expressed in and purified from E. coli ( Fig. 2 ). The measured average mass of CSGAL-1 was 14886.5 Da ( Fig. 3 ), indicating that the purified protein contained the designed amino acid sequence without any further modification.

To assess the monomeric–dimeric structure of oxidized or reduced rhGAL-1 in a physiological solution (NaCl/P_i), absolute molecular mass was determined by a combination of size-exclusion HPLC and multiangle laser light scattering ( Fig. 6 ). The retention volume of oxidized rhGAL-1 was larger than that of reduced rhGAL-1. The absolute molecular mass of the oxidized protein was calculated as 14.5 kDa, likely corresponding to a monomeric structure, whereas that of the reduced protein was 30 kDa, likely corresponding to a dimeric structure.

Differences in secondary structural features between oxidized and reduced rhGAL-1 were probed by measuring CD from 250 to 200 nm. The CD spectrum of the oxidized form exhibited a small negative signal around 205–210 nm, whereas that of the reduced form showed a broad negative peak around 215–220 nm. This shift in the CD spectrum is indicative of a substantial change in secondary structure induced by the formation of disulfide bonds.

Lectin activity of oxidized rhGAL-1 and CSGAL-1

Lectin activity of oxidized rhGAL-1 and CSGAL-1 was determined by assaying both hemagglutination of rabbit erythrocytes and binding to lactose–agarose. For the hemagglutination assay, NaCl/P_i was used as a negative control and concanavalin A as a positive control. Oxidized rhGAL-1 did not induce agglutination even at a concentration of 50 µgámL^-1; however, when the assay was performed under reducing conditions, rhGAL-1 induced agglutination at concentrations of Ò6.25 µgámL^-1 ( Fig. 7 ). In contrast to oxidized rhGAL-1, CSGAL-1 induced hemagglutination even under nonreducing conditions at a concentration of 6.25 µgámL^-1.

b-Galactoside-binding activity was assayed with the use of a lactose–agarose column. Oxidized rhGAL-1 did not bind to lactose–agarose, whereas the dithiothreitol-treated protein did ( Fig. 8 ). SDS/PAGE analysis confirmed the presence of oxidized rhGAL-1 in the flow-through fraction of the column, whereas the reduced protein was detected in the bound fraction (data not shown). Thus, oxidized rhGAL-1 was not able to bind to b-galactoside.

Axonal regeneration-promoting activity of oxidized rhGAL-1 and CSGAL-1

To evaluate axonal regeneration-promoting activity, we employed an in vitro model that consists of DRG explants with associated nerve stumps. This in vitro model preserves the three-dimensional structure of cells, and those factors applied for examination are thought to work on the sites of spinal nerve stumps. The target sites in this model are the same as the proximal sites of in vivo axotomized nerve stumps. In fact, our previous experiments revealed that the effects of galectin-1 on axonal regeneration using this in vitro model were consistent with those of two other in vivo neural regeneration models [4].

Oxidized rhGAL-1 enhanced axonal regeneration from transected nerve sites of adult rat DRG explants with associated nerve stumps ( Figs 9 and 10 ). At both central and peripheral sites, at a concentration of 5.0, 50, 500 or 5000 pgámL^-1, oxidized rhGAL-1 significantly enhanced axonal regeneration in comparison with the control cultures (P < 0.05, ANOVA; Fig. 10 ). Even at a concentration of 5.0 pgámL^-1, oxidized rhGAL-1 induced a significant increase in the average number of regenerating axons at both central [control, 26.6 ± 7.3 (n = 26); oxidized rhGAL-1, 42.3 ± 6.8 (n = 26)] and peripheral [control, 52.2 ± 12.7 (n = 25); oxidized rhGAL-1, 72.5 ± 18.6 (n = 25)] sites (P < 0.05, ANOVA). Furthermore, at a concentration of 500 pgámL^-1, axonal regeneration at both central and peripheral sites was enhanced significantly compared with the 5.0 pgámL^-1 group [average number of regenerating axons: 51.6 ± 7.8 (n = 12) and 92.1 ± 23.9 (n = 10) for central and peripheral sites, respectively; P < 0.05, ANOVA]. Thus, whereas oxidized rhGAL-1 did not demonstrate lectin activity, it exhibited marked axonal regeneration-promoting activity in a dose-dependent manner. In contrast, CSGAL-1, which induced hemagglutination, had no effect on axonal regeneration even at a concentration of 5.0 ngámL^-1[average number of regenerating axons: 28.5 ± 5.7 (n = 12) and 51.2 ± 6.4 (n = 10) for central and peripheral sites, respectively]. These results clearly show that the axonal regeneration-promoting activity of oxidized galectin-1 is unrelated to its lectin properties.

Discussion

Experimental procedures Results Acknowledgments References

Recently, we reported that galectin-1 is a factor regulating initial axonal growth in peripheral nerves after axotomy [4]. Galectin-1 is a well-known animal lectin that exhibits b-galactoside-binding activity only in its reduced form. In our previous study [4], however, galectin-1 was purified from the culture supernatant of COS1 cells under oxidative conditions. Furthermore, galectin-1 promoted axonal regeneration at low concentrations, which are two orders of magnitude lower than the concentrations at which reduced galectin-1 exhibits lectin activity. The present study investigates the relation between structure and axonal regeneration-promoting activity of galectin-1. We have shown that rhGAL-1 secreted from COS1 cells exists as an oxidized form with three intramolecular disulfide bonds, and that galectin-1 promotes axonal regeneration in this oxidized form, but not in the lectin form.

Purification and analysis under nonreducing conditions revealed that rhGAL-1/COS1 contains intramolecular disulfide linkages between Cys2 and Cys130, Cys16 and Cys88, and Cys42 and Cys60. Throughout the purification and analysis procedures, we used methods to minimized any possible occurrence of disulfide exchange reaction. The pattern of intramolecular disulfide linkages of the purified protein should therefore not have changed. This pattern of disulfide linkages is identical to that determined previously for oxidized bovine galectin-1 which lacks lectin activity [34]. Thus, this pattern appears to be conserved among different species and to represent the most stable conformation of oxidized galectin-1.

Oxidized rhGAL-1 and CSGAL-1 were expressed in and purified from E. coli in order to investigate whether the oxidized conformation is essential for the axonal regeneration-promoting activity of galectin-1. Oxidized rhGAL-1, which was shown to contain the same disulfide linkages as rhGAL-1/COS1, enhanced axonal regeneration from transected nerve sites of adult rat DRG explants with associated nerve stumps at low concentrations (pgámL^-1), whereas it lacked both hemagglutination-inducing and b-galactoside-binding activities. In contrast, CSGAL-1, a galectin-1 mutant in which all cysteines were replaced by serine, did not promote axonal regeneration, but induced marked hemagglutination even under nonreducing conditions. Although CSGAL-1 has additional amino acids at the N-terminus, it functioned precisely like lectin under physiological conditions. We also expressed rhGAL-1 with the histidine tag sequence (MHHHHHHIGYPYDVPDYAGVEF) at the N-terminus, which promotes axonal regeneration using our in vitro assay (data not shown). This result indicates that the additional amino acids at the N-terminus of CSGAL-1 did not effect axonal regeneration-promoting activity. Furthermore we have already shown that galectin-3, which is another member of the galectin family, does not promote axonal regeneration in either our in vitro or in vivo model [4]. However, it has been reported that galectin-3 possesses typical lectin activity in solution under nonreducing conditions, and promotes neural cell adhesion and neurite growth [35,36]. Thus, the oxidized conformation of galectin-1 appears necessary for axonal regeneration-promoting activity, and this activity does not seem to be related to its lectin activity. Multiangle laser light-scattering analysis revealed that oxidized rhGAL-1 exists as a monomer in physiological solution, and CD analysis showed that the secondary structure of oxidized rhGAL-1 differs from that of reduced rhGAL-1. These results indicate that the disulfide bond formation alters the structure of galectin-1 in such a way as to confer the ability to promote axonal regeneration.

Galectins are members of a family of b-galactoside-binding animal lectins with significant sequence similarity in the carbohydrate-binding site [5]. In this respect, oxidized galectin-1 with three intramolecular disulfide bonds is not a galectin, because it lacks b-galactoside-binding ability.

Numerous studies on the biological effects of galectin-1, depending on its ability to bind b-galactoside, have been described. In contrast, effects of galectin-1 that are not blocked by the application of lactose have also been described [14,37–39]. Thus, mouse galectin-1 has been shown to act as a cytostatic factor and regulator of cell growth, and the growth-inhibitory effect is not inhibited by lactose. Our present results indicate that oxidized galectin-1, whose disulfide bonds may have been formed during culturing, is one possible explanation for these previously shown biological effects.

This study represents the first demonstration that oxidized galectin-1, which appears to be the form of the protein secreted from cells, promotes axonal regeneration, and that this activity is strictly dependent on the oxidized structure of the molecule. Furthermore, this activity by oxidized galectin-1 was observed at concentrations (pgámL^-1) substantially lower than those at which the lectin effects of galectin-1 are exhibited in neuronal cells in vitro (> ngámL^-1) [27]. Given that the physiological extracellular environment is oxidative, it appears likely that one important role of galectin-1 is to promote axonal regeneration in the peripheral nervous system of adult animals. Previous immunohistochemical analysis has demonstrated the localization of galectin-1 in regenerating sciatic nerves, as well as in both sensory neurons and motor neurons [4]. Thus, oxidized galectin-1 likely acts as an autocrine or paracrine factor to promote axonal regeneration, functioning more like a cytokine or chemokine than as a lectin.

Acknowledgments

Experimental procedures Results Discussion References

We thank Kazumi Fuju and Naomi Kubota for excellent technical support. We also thank Dr Tadashi Sudo for support.

References

Experimental procedures Results Discussion Acknowledgments

  1.   Fawcett, J.W. & Keynes, R.J. (1990) Peripheral nerve regeneration. Annu. Rev. Neurosci. 13, 43–60.

  2.   Bunge, R.P. (1993) Expanding roles for the Schwann cell: ensheathment, myelination, trophism and regeneration. Curr. Opin. Neurobiol. 3, 805–809.

  3.   Ide, C. (1996) Peripheral nerve regeneration. Neurosci. Res. 25, 101–121.

  4.   Horie, H., Inagaki, Y., Sohma, Y., Nozawa, R., Okawa, K., Hasegawa, M., Muramatsu, N., Kawano, H., Horie, M., Koyama, H., Sakai, I., Takeshita, K., Kowada, Y., Takano, M. & Kadoya, T. (1999) Galectin-1 regulates initial axonal growth in peripheral nerves after axotomy. J. Neurosci. 19, 9964–9974.

  5.   Barondes, S.H., Castronovo, V., Cooper, D.N.W., Cummings, R.D., Hirabayashi, J., Hughes, C., Kasai, K.-I., Leffler, H., Liu, F.-T., Lotan, R., Mercurio, A.M., Monsigny, M., Pillai, S., Polirer, F., Raz, A., Rigby, P.W.J. & Wang, J.L. (1994) Galectins: a family of animal b-galactoside-binding lectins. Cell 76, 597–598.

  6.   Harrison, F.L. (1991) Soluble vertebrate lectins: ubiquitous but inscrutable proteins. J. Cell Sci. 109, 9–14.

  7.   Kasai, K.-I. & Hirabayashi, J. (1996) Galectins: a family of animal lectins that decipher glycocodes. J. Biochem. (Tokyo) 119, 1–8.

  8.   Perillo, N.L., Marcus, M.E. & Baum, L.G. (1998) Galectins: versatile modulators of cell adhesion, cell proliferation, and cell death. J. Mol. Med. 76, 402–412.

  9.   Hirabayashi, J. & Kasai, K.-I. (1988) Complete amino acid sequence of a b-galactoside-binding lectin from human placenta. J. Biochem. (Tokyo) 104, 1–4.

  10.   Abbott, W.M. & Feizi, T. (1989) Evidence that the 14 kDa soluble beta-galactoside-binding lectin in man is encoded by a single gene. Biochem. J. 259, 291–294.

  11.   Couraud, P.-O., Casentini-Borocz, D., Bringman, T.S., Griffith, J., McGrogan, M. & Nedwin, G.E. (1989) Molecular cloning, characterization, and expression of a human 14-kDa lectin. J. Biol. Chem. 264, 1310–1316.

  11a.   Abbott, W.M., Mellor, A., Edwards, Y. & Feizi, T. (1989) Soluble bovine galactose-binding lectin. cDNA cloning reveals the complete amino acid sequence and an antigenic relationship with the major encephalitogenic domain of myelin basic protein. Biochem J. 259, 283–290.

  12.   Clerch, L.B., Whitney, P., Hass, M., Brew, K., Miller, T., Werner, R. & Massaro, D. (1988) Sequence of a full-length cDNA for rat lung beta-galactoside-binding protein: primary and secondary structure of the lectin. Biochemistry 27, 692–699.

  13.   Cooper, D.N. & Barondes, S.H. (1990) Evidence for export of a muscle lectin from cytosol to extracellular matrix and for a novel secretory mechanism. J. Cell Biol. 110, 1681–1691.

  14.   Wells, V. & Mallucci, L. (1991) Identification of an autocrine negative growth factor: mouse beta-galactoside-binding protein is a cytostatic factor and cell growth regulator. Cell 64, 91–97.

  15.   Ohyama, Y., Hirabayashi, J., Oda, Y., Ohno, S., Kawasaki, H., Suzuki, K. & Kasai, K.-I. (1986) Nucleotide sequence of chick 14 K beta-galactoside-binding lectin mRNA. Biochem. Biophys. Res. Commum. 134, 51–56.

  16.   Sakakura, Y., Hirabayashi, J., Oda, Y., Ohyama, Y. & Kasai, K.-I. (1990) Structure of chicken 16-kDa beta-galactoside-binding lectin. Complete amino acid sequence, cloning of cDNA, and production of recombinant lectin. J. Biol. Chem. 265, 21573–21579.

  17.   Paroutaud, P., Levi, G., Teichberg, V.I. & Strosberg, A.D. (1987) Extensive amino acid sequence homologies between animal lectins. Proc. Natl Acad. Sci. USA 84, 6345–6348.

  18.   Levi, G. & Teichberg, V.I. (1981) Isolation and physicochemical characterization of electrolectin, a beta-D-galactoside binding lectin from the electric organ of Electrophorus electricus. J. Biol. Chem. 256, 5735–5740.

  19.   Levi, G. & Teichberg, V.I. (1985) Isolation and characterization of chicken thymic electrolectin. Biochem. J. 226, 379–384.

  20.   Whitney, P.L., Powell, J.T. & Sandford, G.L. (1986) Oxidation and chemical modification of lung beta-galactoside-specific lectin. Biochem. J. 238, 683–689.

  21.   Abbott, W.M. & Feizi, T. (1991) Soluble 14-kDa beta-galactoside-specific bovine lectin. Evidence from mutagenesis and proteolysis that almost the complete polypeptide chain is necessary for integrity of the carbohydrate recognition domain. J. Biol. Chem. 266, 5552–5557.

  22.   Dodd, J. & Jessell, T.M. (1986) Cell surface glycoconjugates and carbohydrate-binding proteins: possible recognition signals in sensory neurone development. J. Exp. Biol. 124, 225–238.

  23.   Hynes, M.A., Gitt, M., Barondes, S.H., Jessell, T.M. & Buck, L.B. (1990) Selective expression of an endogenous lactose-binding lectin gene in subsets of central and peripheral neurons. J. Neurosci. 10, 1004–1013.

  24.   Mahanthappa, N.K., Cooper, D.N., Barondes, S.H. & Schwarting, G.A. (1994) Rat olfactory neurons can utilize the endogenous lectin, L-14, in a novel adhesion mechanism. Development 120, 1373–1384.

  25.   Puche, A.C. & Key, B. (1995) Identification of cells expressing galectin-1, a galactose-binding receptor, in the rat olfactory system. J. Comp. Neurol. 357, 513–523.

  26.   Joubert, R., Kuchler, S., Zanetta, J.P., Bladier, D., Avellana-Adalid, V., Caron, M., Doinel, C. & Vincendon, G. (1989) Immunohistochemical localization of a beta-galactoside-binding lectin in rat central nervous system. I. Light- and electron-microscopical studies on developing cerebral cortex and corpus callosum. Dev. Neurosci. 11, 397–413.

  27.   Puche, A.C., Poirier, F., Hair, M., Bartlett, P.F. & Key, B. (1996) Role of galectin-1 in the developing mouse olfactory system. Dev. Biol. 179, 274–287.

  28.   Newton, C.R. & Graham, A. (1994) PCR. Alden Press, Oxford, pp. 27–38.

  29.   Ohashi, H., Maruyama, K., Liu, Y.C. & Yoshimura, A. (1994) Ligand-induced activation of chimeric receptors between the erythropoietin receptor and receptor tyrosine kinases. Proc. Natl Acad. Sci. USA 91, 158–162.

  30.   Sanger, F., Nicklen, S. & Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl Acad. Sci. USA 74, 5463–5467.

  31.   Cohen, S.A. & Michaud, D.P. (1993) Synthesis of a fluorescent derivatizing reagent, 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate, and its application for the analysis of hydrolysate amino acids via high-performance liquid chromatography. Anal. Biochem. 211, 279–287.

  32.   Horie, H., Sakai, I., Akahori, Y. & Kadoya, T. (1997) IL-1 beta enhances neurite regeneration from transected-nerve terminals of adult rat DRG. Neuroreport 8, 1955–1959.

  33.   Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.

  34.   Tracey, B.M., Feizi, T., Abbott, W.M., Carruthers, R.A., Green, B.N. & Lawson, A.M. (1992) Subunit molecular mass assignment of 14,654 Da to the soluble beta-galactoside-binding lectin from bovine heart muscle and demonstration of intramolecular disulfide bonding associated with oxidative inactivation. J. Biol. Chem. 267, 10342–10347.

  35.   Kuwabara, I. & Liu, F.T. (1996) Galectin-3 promotes adhesion of human neutrophils to laminin. J. Immunol. 156, 3939–3944.

  36.   Pesheva, P., Kuklinski, S., Schmitz, B. & Probstmeier, R. (1998) Galectin-3 promotes neural cell adhesion and neurite growth. J. Neurosci. Res. 54, 639–654.

  37.   Wells, V. & Mallucci, L. (1992) Molecular expression of the negative growth factor murine beta-galactoside binding protein (mGBP). Biochim. Biophys. Acta 1121, 239–244.

  38.   Adams, L., Scott, G.K. & Weinberg, C.S. (1996) Biphasic modulation of cell growth by recombinant human galectin-1. Biochim. Biophys. Acta 1312, 137–144.

  39.   Yamaoka, K., Ingendoh, A., Tsubuki, S., Nagai, Y. & Sanai, Y. (1996) Structural and functional characterization of a novel tumor-derived rat galectin-1 having transforming growth factor (TGF) activity: the relationship between intramolecular disulfide bridges and TGF activity. J. Biochem. (Tokyo) 119, 878–886.s

Notes

Enzymes: Achromobacter protease I (EC 3.4.21.50); trypsin (EC 3.4.21.4).