Adiponectin does not bind to gelatin: a new and easy way to purify high-molecular-weight adiponectin from human plasma.

Human plasma contains three forms of adiponectin, a trimer, a hexamer, and a high-molecular-weight (HMW) multimer. We previously reported HMW adiponectin was a gelatin-binding protein of 28 kDa (GBP28), it having been purified due to its affinity to gelatin-Cellulofine (Nakano, Y., et al. Isolation and characterization of GBP28, a novel gelatin-binding protein purified from human plasma. J. Biochem. 1996. 120: 803–12). Although HMW adiponectin binds to gelatin-Cellulofine, it cannot bind to gelatin-Sepharose. Gelatin-Cellulofine was made of formyl-Cellulofine and gelatin, and we found that HMW adiponectin binds to reduced formyl-Cellulofine with similar affinity as to gelatin-Cellulofine. Through only two steps using reduced formyl-Cellulofine and DEAE-Sepharose, HMW adiponectin can be effectively purified from human plasma.

formyl-Cellulofi ne and gelatin and HMW adiponectin bound to it ( Fig. 1C ). At the same time, we prepared a control resin with formyl-Cellulofi ne without gelatin and, surprisingly, HMW adiponectin bound to this reduced formyl-Cellulofi ne ( Fig. 1D ). Next, we tried hydrolyzed NHS-activated Sepharose, which has a similar structure to reduced formyl-Cellulofi ne except that the resin is made of agarose. HMW adiponectin did not bind to the hydrolyzed NHS-activated Sepharose and the binding of HMW adiponectin to reduced formyl-Cellulofi ne was not inhibited by 0.1 M Gal (data not shown). We also checked Cellulofi ne itself but no binding was observed. We found that a major HMW adiponectin could be precipitated with 12% polyethylene glycol (PEG) 4000 from human plasma and the redissolved precipitate did not bind to gelatin-Cellulofi ne before depletion of PEG 4000. The binding of HMW adiponectin to reduced formyl-Cellulofi ne was also inhibited by PEG 4000 ( Fig. 1E ). As reported before, HMW adiponectin does not bind to sulfate-Cellulofi ne ( 6 ). There remains at least one possibility that HMW adiponectin binds to hyaluronan, which is a nonsulfated glycosaminoglycan. We made hyaluronan-Sepharose from hyaluronic acid and EAH-Sepharose but HMW adiponectin did not bind to it ( Fig. 1F ). Although the characteristics of the affi nity to reduced formyl-Cellulofi ne are not clear, HMW adiponectin binds to it very fi rmly and can be eluted with NaCl.

Purifi cation of HMW adiponectin
Human plasma containing 0.1 M NaCl and 0.1 mM PMSF was applied to reduced formyl-Cellulofi ne and eluted with a NaCl gradient ( Fig. 2A ). In this step, HMW adiponectin was separated from other major proteins effectively and after equilibration against 10 mM Tris-HCl, pH8.0, 0.3M NaCl, and 1 mM CaCl 2 , the HMW adiponectin was applied to DEAE-Sepharose ( Fig. 2B ). When bound HMW adiponectin was eluted with NaCl, the resulting peak was almost only HMW adiponectin. Each aliquot of the original plasma, the fl ow-through pool, and the pool of HMW adiponectin fractions from reduced formyl-Cellulofi ne and the HMW adiponectin pool from DEAE-Sepharose was analyzed on a Superdex 200 HR 10/30 column. As shown in Fig. 2C , the original plasma contained HMW adiponectin, which was detected as a single peak by an ELISA specifi c for HMW adiponectin in fractions from 15 to 23 and as a molecule of about 420 kDa. By immunoblotting, three peaks corresponding to HMW, the hexamer and trimer adiponectins of apparent molecular weights of 420, 280, and 180 kDa were detected in fractions 19, 21, and 23, as described before ( 6 ). The fl ow-through pool contained HMW adiponectin of about 1/10 of the original plasma in fractions from 16 to 20 and the hexamer adiponectin of about 280 kDa in fractions from 20 to 23 became apparent by an ELISA specifi c for HMW adiponectin, which could detect the hexamer adiponectin with much lower effi ciency than HMW adiponectin ( Fig. 2D ). When the pool of HMW adiponectin fractions eluted from reduced formyl-Cellulofi ne was analyzed, HMW adiponectin was shown to be separated from other major proteins effectively ( Fig. 2E ). And as shown in Fig. 2F , an analysis of 0.2 M NaHCO 3 , pH8.5, at 4°C for 72 h. Hyaluronan-Sepharose was prepared as described before ( 7 ). Briefl y, EAH-Sepharose (1 ml), hyaluronic acid (9 mg), and N-ethyl-N'-(3-dimethylaminopropyl)-carbodiimide (75 mg) were mixed in 3 ml distilled water and then the pH was adjusted to 4.5 with 0.1 N HCl. The mixture was rotated at room temperature overnight.

Purifi cation of HMW adiponectin
NaCl and PMSF were added to human plasma to fi nal concentrations of 0.1 M and 0.1 mM, respectively, and then applied to reduced formyl-Cellulofi ne equilibrated with 10 mM Tris-HCl, pH7.4, and 0.1 M NaCl. After thorough washing, HMW adiponectin was eluted with a NaCl gradient from 0.1 to 2.0 M. The HMW adiponectin fractions were pooled and equilibrated against 10 mM Tris-HCl, pH8.0, 0.3 M NaCl, and 1 mM CaCl 2 by dialysis. Then, the HMW adiponectin pool was applied to DEAE-Sepharose equilibrated with 10 mM Tris-HCl, pH8.0, and 0.3 M NaCl, and after thorough washing with the same buffer, HMW adiponectin was eluted with a NaCl gradient from 0.3 to 2.0 M NaCl. If needed for further purifi cation, HMW adiponectin fractions were pooled and applied to a HiLoad 16/60 Superdex 200 prepgrade column after concentration and equilibration against 10 mM Tris-HCl, pH7.4, and 0.15 M NaCl. To determine its purity and size, an aliquot of the HMW adiponectin pool was applied to a Superdex 200 HR 10/30 column.

Osteoclast formation assay
RAW cells were kept in serum-free medium, CHO-S-SFM II containing 1% penicillin/streptomycin, and for subculturing, 0.25% (w/v) Trypsin-0.53 mM EDTA was used. For osteoclast formation assay, RAW cells were plated at a density of 2.5 × 10 3 cells/ well in 96-well plate in ␣ -MEM containing 10% FBS and 1% penicillin/streptomycin, and cultured for 2 days. Then, medium was changed to fresh medium containing RANKL (70 ng/ml) with or without HMW adiponectin (2.5 or 5.0 g/ml). After 2 days, the medium was changed to fresh medium containing RANKL (70 ng/ml) and the next day, tartrate-resistant acid phosphatase (TRAP) staining was performed.

TRAP staining
After cells were fi xed with 1% formaldehyde for 5 min and treated with acetone-ethanol (1:1) for 30 s, TRAP staining was performed at pH 5.0 in the presence of L(+)-tartaric acid using naphthol AS-MX phosphate in N, N-dimethyl formamide and Fast red violet LB salt.

General procedures
SDS-PAGE was performed by the method of Laemmli, followed by Coomassie staining or Western blotting. After treatment with anti-C and peroxidase-conjugated second antibody of a nitrocellulose membrane, bands were detected with Immobilon Western HRP Substrate according to the manufacturer's instructions. For protein size analysis, a prestained SDS-PAGE standard, Precision Plus Protein Standards Dual Color, was used.

HMW adiponectin binds to reduced formyl-Cellulofi ne
Recently, the sale of gelatin-Cellulofi ne was discontinued and so we purchased gelatin-Sepharose. Unexpectedly, HMW adiponectin did not bind to gelatin-Sepharose ( Fig. 1B ). Because gelatin-Cellulofi ne was made of formyl-Cellulofi ne and gelatin, we made gelatin-Cellulofi ne from purifi cation steps are better and the concentration step should be avoided if possible.

The activity of HMW adiponectin purifi ed by formyl-Cellulofi ne
Adiponectin inhibits osteoclastogenesis ( 9 ) and HMW adiponectin is most potent inhibitor among adiponectin molecules containing globular adiponectin by our experiments (data not shown). Using HMW adiponectin purifi ed by gelatin-Cellulofi ne and formyl-Cellulofi ne, their activities were assayed. HMW adiponectin purifi ed by formyl-Cellulofi ne inhibited osteoclastogenesis as shown in Fig. 3C and D, and the inhibition activity was similar, but slightly stronger than HMW adiponectin purifi ed by gelatin-Cellulofi ne (Fig. 3E, F).

DISCUSSION
Although we reported the purifi cation of HMW adiponectin (GBP28) using gelatin-Cellulofi ne ( 4 ), and adiponectin was also reported to bind to collagen I, III, and HMW adiponectin pool from DEAE-Sepharose revealed that HMW adiponectin of about 420 kDa was successfully purifi ed. In the case, a little contamination is observed in HMW adiponectin pool; for further purifi cation a HiLoad 16/60 Superdex 200 prep-grade column can be used effectively.
The purifi ed HMW adiponectin was analyzed by SDS-PAGE ( Fig. 2G ). As reported before ( 4 ), characteristic bands were seen, i.e., when HMW adiponectin was boiled in Laemmli's loading buffer, it gave a single band corresponding to 28 kDa under reducing conditions, and without boiling, under reducing conditions it gave a band corresponding to 65 kDa, which represents the adiponectin trimer, with minor bands of 150 kDa and bigger at almost the top of the gel, which represent the adiponectin hexamer and HMW multimer, respectively. From 2.6 l of human plasma that contained 1.85 mg/L HMW adiponectin, using reduced formyl-Cellulofi ne, 50 ml, about 2.1 mg of HMW adiponectin could be effectively purifi ed and the yield was about 43%. The protein concentration was determined to be 1 mg/ml when the absorbance at 280nm was 1.34 ( 8 ). Because HMW adiponectin is easily adsorbed to surfaces of glassware and so on, fewer   Fig. 1 , and each fraction was also analyzed by 12.5% Laemmli's SDS-PAGE after being heat-denatured at 100°C for 3 min, under reducing conditions, followed by Coomassie blue staining. B: The HMW adiponectin pool from reduced formyl-Cellulofi ne in 10 mM Tris-HCl, pH 8.0, 0.3 M NaCl and 1 mM CaCl 2 was applied to DEAE-Sepharose, 10 ml. Each aliquot of the original human plasma (C), the fl ow-through pool from reduced formyl-Cellulofi ne (D), the pool of HMW adiponectin fractions eluted from reduced formyl-Cellulofi ne (E), and the HMW adiponectin pool from the DEAE-Sepharose (F) was applied to a Superdex 200 HR 10/30 column. Each fraction was analyzed by 12.5% Laemmli's SDS-PAGE after being heat-denatured at 100°C for 3 min, under reducing conditions, followed by immunoblotting with anti-C. Small triangles indicate elution peaks of thyroglobulin (669 kDa), ferritin (416 kDa), catalase (219 kDa), adolase (176 kDa), and BSA (67 kDa), respectively. G: 12.5% Laemmli's SDS-PAGE analysis of the purifi ed HMW adiponectin. Under reducing conditions: +, heat-denatured at 100°C for 3 min; and -, not heat-denatured. Protein bands were stained with Coomassie blue. very fi rmly and needs a NaCl concentration of more than 0.3 M to be released.
After all, we could establish a very simple and effective purifi cation method for HMW adiponectin involving reduced formyl-Cellulofi ne because most proteins do not show such strong affi nity for reduced formyl-Cellulofi ne, fl owing through it. Also, this short two-step method results in not only a good yield of the protein but also a protein exhibiting high activity. Until this method was developed, many steps were needed to purify HMW adiponectin and often resulted in a low yield and low activity, which is correlated with the V ( 5 ), we found it does not bind to gelatin. Moreover, surprisingly, it binds to reduced formyl-cellulofi ne but not to hydrolyzed NHS-activated Sepharose, whose structures should be very similar to each other. The structures of formyl-Cellulofi ne and NHS-activated Sepharose are shown in Fig. 4 . Because adiponectin has a similar structure to complement factor C1q and mannose-binding protein (MBP), there is a possibility that adiponectin recognizes a certain sugar structure or structures. As shown in Fig. 1F , HMW adiponectin does not bind to hyaluronan-Sepharose and, as reported before, it does not bind to sulfate-Cellulofi ne ( 4 ). Moreover, analysis involving glycoconjugate microarrays ( 10 ) did not reveal any interaction with many glycoconjugates containing heparin sulfate, chondoroitin sulfate, hyaluronic acid, mannan, LPS, and so on (data not shown). Adiponectin was reported to bind to LPS at pH 5 to 6 and this binding was inhibited by many sugars like lactose, fucose, and N-acetyl glucosamine ( 11 ). In fact, with HMW adiponectin, we could detect a weak interaction with LPS. There remains a possibility that under certain conditions, HMW adiponectin might bind to a certain sugar chain or chains specifi cally. However, this cannot be the case for HMW adiponectin and reduced formyl-Cellulofi ne because HMW adiponectin binds to reduced formyl-Cellulofi ne Fig. 3. The activity of HMW adiponectin purifi ed by formyl-Cellulofi ne. RAW cells were treated with medium containing RANKL (70 ng/ml) with or without HMW adiponectin for 2 days and with RANKL for one more day, and then stained for TRAP. Arrowheads and arrows indicate large and small osteoclasts, respectively. A: RANKL. B: Control. C: 2.5 g/ml, and D: 5.0 g/ml of HMW adiponectin purifi ed by formyl-Cellulofi ne. E: 2.5 g/ml and F: 5.0 g/ml of HMW adiponectin purifi ed by gelatin-Cellulofi ne.