Biomacromolecules 2003, 4, 189-192
189
Hyaluronic Acid N-Deacetylase Assay in Whole Skin Marı´a O. Longas,*,† James D. Burden,† John Lesniak,† Ryan M. Booth,† James A. McPencow,† and Jung I. Park‡ Department of Chemistry and Physics, Purdue University Calumet, 2200 169th Street, Hammond, Indiana 46323, and St. Margaret Mercy Health Care Center, Dyer, Indiana 46311 Received November 1, 2002
Hyaluronic acid (HA) N-deacetylase(s) was quantified in whole skin, using a novel method that involved reaction of skin with exogenous HA as substrate. Acetyl (CH3CO-) moieties generated were converted chemically to MeOAc and quantified using gas chromatography/mass spectrometry. HA (1.7 mg) and skin (1.0 g) yielded 3.32 and 769.00 µg of MeOAc from the 69.0- and 76.5-year-old-patient samples, respectively. Without added HA, 194.00 µg of product was obtained from the 76.5-year-old-patient samples. With chondroitin as substrate, the yields were 2.89 and 818.04 µg of MeOAc from the 69.0- and 76.5-year-oldpatient samples, respectively. The K5 (capsular, Escherichia coli polysaccharide) substrate yielded no detectable product, except for 170.02 µg from the 76.5-year-old-patient samples. This highly sensitive method was used to demonstrate that human-skin-HA N-deacetylase(s) was first detectable at 69 years of age, highly active at 76.5 years of age, and specific for N-acetyl moieties of D-GlcNAc and D-GalNAc where C1 is β-linked as in HA and CH. Introduction
Materials and Methods
Hyaluronic acid (HA), an acidic glycosaminoglycan (GAG) ubiquitous in the intercellular space of connective tissue in a variety of species,1 is a polymer of [4)-β-D-GlcA(1f3)-β-D-GlcNAc-(1f]n arranged in unbranched chains of varying lengths with partly stiff and partly flexible 2D structure.1-3 This highly hydrophilic4 GAG has been degraded sequentially by enzymes known to cleave unsubstituted, nonreducing, terminal D-GlcA and D-GlcNAc, β-glucuronidase (EC 3.2.1.31), and β-N-acetylhexosaminidase (EC 3.2.1.52).5 In human female breast skin, HA loses its acetyl (CH3CO) moieties as a function of age6,7 to become almost completely N-deacetylated by the middle of the 7th decade of life.7 To test whether HA N-deacetylation originated from aberrations of HA biosynthesis or from an age-induced peptidase that removed N-acetyl moieties from a normally synthesized GAG, we cultured skin fibroblast in the presence of 3H-sugar nucleotides. The 3H-GAGs were purified and tested for N-acetylation. It was demonstrated that HA synthesized by skin fibroblasts of the 75-year-old donors was N-acetylated (M. O. Longas et al., unpublished work). This project was initiated to search for a peptidase capable of removing HA N-acetyl moieties in the skin of the aged. It reports a novel gas chromatography/mass spectrometry (GC/MS) method to assay for HA N-deacetylase(s) in whole skin.
Materials. Postsurgical female breast skin (obtained at area hospitals) was examined by the pathologist in service to ascertain its normal morphology. After surgical removal of the subcutaneous fat, skin was stored frozen at -70 °C 2-3 h after surgery. Skin pools from four different patients of the following ages were utilized: 24.7 ( 0.94, 69.0 ( 1.50, and 76.5 ( 1.12 years. Rooster-comb HA was a gift from Dr. Endre A. Balazs, Matrix Biology Institute, Ridgefield, NJ. Chondroitin (CH) was a gift from the late Dr. Karl Mayer (Columbia University College of Physicians and Surgeons, New York, NY). Capsular polysaccharide (K5) of Escherichia coli (E. coli) was a gift from Dr. Willie F. Vann of the NIH.8 Water purification involved filtration, organic and ionic removal in the corresponding cartridges (Millipore Corp., Bedford, MA), and distillation. The distilled H2O was collected in glass, stored under nitrogen, and used throughout. All other reagents were of the highest quality commercially available. Assay of N-Deacetylase(s) with and without Exogenous HA as Substrate. Preliminary experiments were carried out to get the best conditions. The entire assay involved experiments A, B, and C. A. N-Deacetylation. Frozen skin (14 g) was sheared with a knife and homogenized at 0 ( 2 °C in 0.05 M (NH4)2SO4 buffer, pH 7.86. This buffer was prepared by dissolving 3.30 g of (NH4)2SO4 salt in 480 mL of water, adjusting the pH to 7.86 utilizing 22% NH4OH (Optima from Fisher Scientific), and then adjusting the total buffer volume to 500 mL using water. Each skin homogenate was divided into two aliquots, 12.5 mL each; one aliquot was combined with 12.5 mg of rooster-
* To whom correspondence should be addressed. E-mail: mol@ calumet.purdue.edu. † Purdue University Calumet. ‡ St. Margaret Mercy Health Care Center.
10.1021/bm025718g CCC: $25.00 © 2003 American Chemical Society Published on Web 12/14/2002
190
Biomacromolecules, Vol. 4, No. 1, 2003
comb HA dissolved in 12.5 mL of (NH4)2SO4 buffer. The other aliquot was combined with 12.5 mL of the same buffer and used as a control. A second control contained HA (12.5 mg) in 25.0 mL of buffer; a third control had 25.0 mL of the same (NH4)2SO4 buffer only. These experiments were repeated using skin that had been boiled in a minimum volume of water for 10 min to destroy any enzymatic activity. Each reaction mixture was placed in a dialysis membrane of 12 000-14 000 MW cutoff and dialyzed in 180 mL of the same (NH4)2SO4 buffer under toluene at 37 ( 1.0 °C for 24 h. Toluene was utilized to prevent bacterial growth. The dialysates were lyophilized to dryness and used to prepare CH3OAc (experiment B). To corroborate the yields of NH4OAc in the dialysates described above, the following control was included. Ammonium sulfate buffer (25.0 mL) was combined with 2.30 mg of standard NH4OAc, placed in the dialysis membrane, and incubated in 180 mL of the same (NH4)2SO4 buffer under the conditions of the N-deacetylation reaction. At the end of the reaction period, the dialysate was lyophilized to dryness. The yield of NH4OAc was determined by weight. Its identity was confirmed by comparing its Fourier transform infrared (FTIR) and Raman spectroscopy with those obtained before dialysis. This experiment was repeated 4 times in triplicate. A second set of experiments using unboiled skin was conducted under exactly the same conditions to test for depolymerization of HA used as substrate (experiment D). The N-deacetylation experiment was also repeated with unboiled skin using CH or K58 as substrates. B. Esterification. Glacial CH3COOH (0.5 mL) was combined with dry MeOH (2.0 mL), and 30 drops of concentrated H2SO4 was added. The reaction mixture was refluxed at 50-52 °C for 50 min and cooled to room temperature in a H2O bath. The product, MeOAc, was then extracted with CCl4 (500 µL), and 1 µL was analyzed immediately using GC/MS. Prior to GC/MS analysis, the product was kept sealed on ice to prevent evaporation. The method was >90% efficient to convert CH3COOH to MeOAc. Namely, 0.5 mL of CH3COOH (8.8 µmol), the limiting reagent, produced between 8.0 and 8.2 µmol of MeOAc. The yield was quantified using a standard curve of MeOAc developed as indicated (experiment C). This method was utilized to esterify CH3COOH formed by reaction of acetyl (CH3CO-) moieties generated during the enzymatic N-deacetylation reaction (experiment A) with H2O present in the reaction mixture. The lyophilized product obtained in experiment A (0.3-0.7 g) was suspended in 2.0-3.0 mL of dry MeOH and subjected to the reaction conditions described in the preceding paragraph. The final product was extracted with 500 µL of CCl4. To remove unwanted, water-soluble impurities, this extract was mixed quickly with 250 µL of H2O and centrifuged in a clinical centrifuge at room temperature for 30 s. The organic layer was transferred to a test tube containing 0.10 mg of CaCl2, mixed quickly, and centrifuged as indicated. The liquid phase was kept sealed on ice prior to GC/MS analysis (experiment C).
Longas et al.
C. Quantitation of Methyl Acetate. A method to quantify standard MeOAC, using GC/MS, was developed. The instrument employed contained an HP-5 column (crosslinked 5% Ph Me silicone) that was 30 m × 0.25 mm × 0.25 µm film thickness connected to a Hewlett-Packard G1800A GCD system equipped with a 5890 gas chromatograph and electron ionization detector. The following chromatographic temperatures were used: inlet port, 200 °C; detector, 280 °C; oven, initial, 35 °C. The ionization detector was set to scan in the 10-425 m/z range. Standard MeOAc was diluted in CCl4 to concentrations ranging from 5 to 280 ng/µL, and 1 µL was injected to develop a standard curve. Helium, the carrier gas, was monitored at 0.7 mL/min. The run was allowed to proceed for 2.8 min, after it had been determined in preliminary experiments that this was the optimal time. To clean the column after each run, the oven temperature was programmed to increase 10 °C/min to 75 °C and kept at this temperature for 20 min. This method was used to identify and quantify the esterified products of the N-deacetylation reaction (experiments A and B). D. Test for Oligosaccharides Generated During the N-Deacetylation Reaction. The dialysates (180 mL each) of the second set of reactions of unboiled skin with HA as substrate (experiment A) were lyophilized, reconstituted separately in 1.0 mL of H2O, and desalted on Bio-Gel P2 columns (8 cm × 0.5 cm) equilibrated with H2O at a flow rate of 6.0 mL/h. The columns were precalibrated with a mixture of bovine serum albumin (BSA) (50 µg) and D-Glc (50 µg) in 250 µL of 0.15 M NaCl. Under these conditions, BSA eluted in the void volume and D-Glc eluted at 1.6 mL. Each lyophilized dialysate was suspended in 0.80 mL of H2O, applied to a column, and eluted with H2O. Fractions (0.30 mL each) were collected, lyophilized to dryness, redissolved in 100 µL of H2O, and assayed for D-GlcA, using the modified carbazole microassay reported previously.9,10 This experiment was repeated 3 times. Purification of HA From Skin Homogenate Used in the N-Deacetylation Reaction. HA was isolated mainly as described previously.5,7 Briefly, skin homogenate was digested exhaustively with papain. Unwanted impurities were removed by individual precipitations with (a) chloroformamyl alcohol, (b) cetylpyridinium chloride, and (c) calcium acetate.5,7 The GAGs in the final liquid phase were removed by sequential precipitations with absolute ethanol. HA precipitated with 45% (v/v) ethanol as described previously7. Results and Discussion Figure 1 shows the gas ion chromatogram of the product of N-deacetylation of HA by enzyme(s) in the skin of the 76.5-year-old patients. This product has 1.90-min retention time, which is the same as that of standard MeOAc. The intensity of this peak varied in the various samples that yielded MeOAc (Table 1). All of the other peaks shown in Figure 1 appeared in the gas ion chromatograms of the controls in which MeOAc was not detectable (Table 1). The product with 1.90-min retention time was quantified using a standard curve of MeOAc developed under the same
Hyaluronic Acid N-Deacetylase Assay
Biomacromolecules, Vol. 4, No. 1, 2003 191
Figure 1. Gas ion chromatogram of the product of HA N-deacetylation by endogenous enzyme(s) in skin of the 76.5-year-old patients. The ion with 1.90-min retention time is characteristic of standard MeOAc. Peak intensity ) ion count abundance. Table 1. Results of Assay of Hyaluronic Acid N-Deacetylase(s) in Whole Skin samples skin plus HA skin plus HA skin plus HA skin without HA, CH, or K5 skin without HA, CH, or K5 skin without HA, CH, or K5 skin plus CH skin plus CH skin plus CH skin plus K5c skin plus K5c skin plus K5c HA without skin chondroitin without skin K5 without skin preboiled skin in buffer preboiled skin + HA, CH, or K5 buffer only
age (years) 24.7 ( 0.94 69.0 ( 1.50 76.5 ( 1.12 24.7 ( 0.94 69.0 ( 1.50 76.5 ( 1.12 24.7 ( 0.94 69.0 ( 1.50 76.5 ( 1.12 24.7 ( 0.94 69.0 ( 1.50 76.5 ( 1.12
24.7, 69.0, or 76.5 24.7, 69.0, or 76.5
CH3COOCH3a (µg/g of wet skin) SDb ND 3.32 769.00 ND ND 194.00 ND 2.89 818.04 ND ND 170.02 ND ND ND ND ND
0.9 123.0
40.0 0.7 95.0
52.3
ND
a Quantified using a standard curve of MeOAc developed under the same conditions. Each of the skin pools from four different individuals of every age was subjected to experiments A-C. The means of four GC/ MS analyses conducted separately for each skin pool are reported here. b Standard deviation of the mean. c K5, capsular polysaccharide of E. coli.8 d ND ) not detectable.
Figure 2. Standard curve of MeOAc vs the area of the ion peak with 1.90-min retention time (Figure 1). Peak area is expressed as ion count abundance. Each point is an average of eight different determinations.
conditions (Figure 2). The method is linear in the 10-280 ng/µL range, using CCl4 as diluent for MeOAc; its sensitivity is 10 ng/µL. In addition to being identified and quantified as MeOAc (Figures 1 and 2), the product of HA N-deacetylation was
Figure 3. Ion mass spectrogram of the product of HA N-deacetylation by enzyme(s) contained in the skin of 76.5-year-old patients. The peak at m/z 74 corresponds to the product with 1.90-min retention time (Figure 1) and is also that of standard MeOAc. Peak intensity is expressed as ion abundance.
further confirmed using mass spectrometry where its m/z, 74, was the same as that of standard MeOAc. Figure 3 shows the ion mass spectrogram of the product of HA N-deacetylation by enzymes(s) in the skin of the 76.5-year-old patients (experiments A and B). The peak at m/z 74 was detectable at different intensities in all of the samples that yielded MeOAc but was not used to quantify the product. All other peaks in Figure 3 appeared in the controls in which MeOAc was not detectable. Table 1 shows that the only product of exogenous HA N-deacetylation with endogenous skin enzymes(s) originated from the 76.5-year-old-patient samples. Of special significance is the fact that under the same conditions only 3.32 µg of MeOAc was obtained from the skin of 69.0-year-old patients and none from the 24.7-year-old patients. These results suggest that the HA N-deacetylase(s) reported here is induced or activated as the skin ages. Another possibility is that such activity results from an age-mediated chemical modification(s) of a normal enzyme. With CH as substrate, the yields from the 76.5-year-oldpatient samples were apparently higher but within 1 standard deviation from the amount obtained with exogenous HA and skin of the same age (Table 1). These results suggest that CH is as good a substrate for skin-HA N-deacetylase(s) as HA. The N-deacetylase reported here did not work with K5, an unsulfated heparin precursor-like polysaccharide of E. coli.8 This was demonstrated by the yield of MeOAc that was slightly lower but within 1 standard deviation from the amount obtained when no substrate was added to the skin of the 76-year-old patients (Table 1). It is worth noting that the amount of product obtained with K5 as substrate suggests specificity of the N-deaceatylase under discussion for the glycosidic linkage(s) of the DGlcNAc from which it removes the N-acetyl group. Specifically, in HA, C3 and C1 of D-GlcNAc form β linkages, respectively, with C1 and C4 of adjacent D-GlcA’s. In K5, C1 and C4 of D-GlcNAc form R linkages, respectively, with C4 and C1 of adjacent uronic acids. Consequently, the glycosidic linkage configurations and secondary structures of HA and K5 are different. In CH, the glycosidic linkages and configurations are the same as those of HA. The conformations of CH and HA should therefore be similar, even though CH has D-GalNAc
192
Biomacromolecules, Vol. 4, No. 1, 2003
instead of D-GlcNAc. The N-deacetylase reported here cleaved the N-acetyl moieties of both HA and CH, demonstrating that the right configuration(s) at the glycosidic bond(s) is required. This observation is supported by the fact that DS35, a dermatan sulfate of human skin that precipitates with 35% (v/v) ethanol and has the glycosidic linkage(s) configuration(s) of CH and HA, is also N-deacetylated in the skin of the aged.7 Because skin alone (76.5-years), under the conditions of the N-deacetylation reaction, produced enough acetyl groups to get 194 µg of MeOAc, the product obtained with K5 as substrate under the same conditions should originate from N-deacetylation of endogenous HA (Table 1). These results confirm previous findings on the in vivo N-deacetylation of human female breast skin HA in the middle of the 7th decade of life.6,7 At pH 7.86, the pH of the N-deacetylation reaction medium, CH3COOH formed by reaction of acetyl (CH3CO-) ions (removed enzymatically) with H2O should exist largely as CH3COO- ions. In the presence of (NH4)2SO4 provided in the buffer, NH4+ ions and CH3COO- ions reacted to form CH3COONH4 salt that was recovered in the dialysate (experiment A). This was demonstrated by the 85-90% yield of NH4OAc in the control that contained a known amount of this standard salt (experiment A). Hence, the percent recovery of acetyl moieties characterized as MeOAc should also fall on this range. Some product might have also been lost during other experimental procedures. Nevertheless, the sensitivity of the GC/MS method utilized (10 ng/µL) indicates that if any acetyl group was cleaved off when skin of the 24-year-old patients was used with any of the substrates (HA, CH, K5), its concentration was below the sensitivity limit. The same applies to reactions involving skin of the 69.0-year-old patients with K5 and to the controls containing preboiled or no skin (Table 1). To test for cleavage of the substrate backbone during the N-deacetylation reaction, HA was isolated from the skin homogenate at the end of the incubation period (experiment A) and subjected to electrophoresis on cellulose polyacetate membranes and to FTIR spectroscopy.6,7 Its mobility on cellulose polyacetate membranes was close to that of N-deacetylated HA reported previously6,7 (not shown). The FTIR spectrum of HA isolated from skin homogenate after the N-deacetylation reaction was also characteristic of the N-deacetylated HA reported previously6,7 (not shown). Specifically, the FTIR band at 1600-1650 cm-1 in the spectrum of N-deacetylated HA isolated from skin of the aged7 or from the skin homogenate after the N-deacetylation reaction displayed its main peak at 1560 cm-1, even though it extended between 1500 and 1800 cm-1. In GAGs, this
Longas et al.
band peaks at 1650-1600 cm-1 and originates from CdO stretching regardless of its source, D-GlcA or D-GlcNAc. It was shifted to 1560 cm-1 after the N-deacetylation reaction demonstrating that enough primary amines (-NH2) to be detectable using FTIR were created. This band is produced by C-N stretching with contributions from C-N-H in-plane bending. A further test for substrate-backbone cleavage during N-deacetylation was the carbazole assay9,10 performed using the lyophilized dialysates of experiment A after being desalted on Bio-Gel P2 (experiment D). Desalting was necessary because salts interfere with the assay. The results were negative, even though the sensitivity of the method was 0.12 µg. This suggests that if any HA-backbone cleavage took place, the fragment(s) had molecular weight(s) g12 00014 000, the MW cutoff of the dialysis membrane employed (experiment A). In conclusion, the highly sensitive and reproducible GC/ MS method reported here was used to assay HA Ndeacetylase(s) in whole female breast skin and to determine its specificity, which turned out to be N-acetyl moieties of D-GlcNAc and D-GalNAc where C1 is β-linked as in HA and CH. These data confirm previous findings on the in vivo HA N-deacetylation of female breast skin by an enzyme(s) the activity of which appears to be age-induced because it is detectable first at 69 years of age and becomes highly active at 76.5 years.6,7 An additional advantage of the method reported here is that it does not require purified GAGs as most known methods do.11 Acknowledgment. This work was supported by Purdue University Calumet Scholarly Research Release Award and NSF NS-FILIP DUE 9650826 grant to M. O. Longas. References and Notes (1) Meyer, K.; Davidson, E.; Linker, A.; Hoffman, P. Biochim. Biophys. Acta 1956, 21, 506-518. (2) Scott, J. E. In The Chemistry, Biology and Medical Applications of Hyaluronan and Its DeriVatiVes; Laurent, T. C., Ed.; Portland Press: London, 1998; pp 7-15. (3) Scott, J. E.; Heatley, F. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 48504855. (4) Hascall, V. C.; Hascall, G. K. In Cell Biology of Extracellular Matrix; Hay, E. D., Ed.; Plenum Press: New York, 1981; pp 39-63. (5) Longas, M. O.; Meyer, K. Biochem. J. 1981, 197, 275-282. (6) Longas, M. O.; Russell, C. S.; He, X.-Y. Biochim. Biophys. Acta 1986, 884, 265-269. (7) Longas, M. O.; Russell, C. S.; He, X.-Y. Carbohydr. Res. 1987, 159, 127-136. (8) Vann, W. F.; Schmidt, M. A.; Jann, B.; Jann, K. Eur. J. Biochem. 1981, 116, 359-364. (9) Dische, Z. J. Biol. Chem. 1947, 167, 189-198. (10) Bitter, T.; Muir, H. M. Anal. Biochem. 1962, 4, 330-334. (11) Longas, M. O. Anal. Biochem. 1990, 187, 355-358.
BM025718G
