Quantification of hyaluronan (HA) using a simplified fluorophore-assisted carbohydrate electrophoresis (FACE) procedure

Tags: Methods in Cell Biology, UNCORRECTED PROOF, disaccharides, room temperature, PK, procedure, mL aliquots, collagen, ammonium acetate, temperature, disaccharide, unsaturated, Hascall, standards, gel electrophoresis, Analytical Biochemistry, FACE analysis, enzyme activity, gel solution, stock solution, cell cultures, National Institutes of Health, UNCORRECTED PROOF Methods in Cell Biology, chondroitin, chondroitin sulfate, FACE detection, FACE protocol, tissue samples, background values, supernatant, background correction, Lerner Research Institute, aliquot, Anthony Calabro, Mark Lauer, mL aliquot, centrifuge tubes, centrifuge tube, homogeneous solution, Vincent C. Hascall, Mark E. Lauer, Ethanol Precipitation, carbohydrate chemistry, Gel Preparation, Acrylamide Gel, Cell Biology, Ronald J. Midura1, Electrophoresis Equipment, Cleveland Clinic, GAG
Content: UNCORRECTED PROOF
Quantification of hyaluronan (HA) using a simplified fluorophore-assisted carbohydrate electrophoresis (FACE) procedure?
Ronald J. Midura1, Valbona Cali, Mark E. Lauer, Anthony Calabro, Vincent C. Hascall Lerner Research Institute, Cleveland Clinic, Cleveland, OH, United States Writing this chapter was made possible by support from the Program of Excellence in Glycosciences funded by Grant P01 HL107147 from the National Institutes of Health, National Heart, Lung and Blood Institute. We dedicate it to the memory of Mark Lauer who, sadly, died October 15, 2015. He did most of the early work in redesigning the protocols for the FACE procedures described in this chapter. 1 Corresponding author: Email address: [email protected] (R.J. Midura)
Contents
1. Introduction
2
2. Principle and Theory
5
3. Equipment
6
3.1. Electrophoresis Equipment
6
3.2. Imaging Equipment
8
3.3. Gel Analysis Software
9
3.4. Other Equipment
10
4. Materials (High Quality, Reagent Grade)
11
5. Solutions
12
5.1. 100 mM Ammonium Acetate, pH 7.0
13
5.2. Tris­Acetate Solution
13
5.3. AMAC/NaBH3CN Solution
13
5.4. Proteinase K Stock Solution
13
5.5. Chondroitinase ABC Enzyme Solution
14
5.6. Acrylamide Gel Solution
14
6. AMAC Labeling of HA and CS Disaccharides
15
Methods in Cell Biology, Volume xx ISSN 0091-679, https://doi.org/10.1016/bs.mcb.2017.08.017
© 2017. 1
2
Methods in Cell Biology
UNCORRECTED PROOF
6.1. Proteinase K Digestion
15
6.1.1. cultured cells
15
6.1.2. Conditioned media
15
6.1.3. Tissues
15
6.2. First Ethanol Precipitation
16
6.3. Digestion of HA and CS
17
6.4. Second Ethanol Precipitation
17
6.5. Sample Labeling With AMAC
18
6.6. Preparation of FACE Standards
19
7. Gel Preparation
19
7.1. Gel Casting
19
7.2. Chilling the Gels
19
7.3. Loading the Samples
20
7.4. Electrophoresis
21
8. Image Acquisition
21
9. Data Normalization
23
References
24
Abstract Hyaluronan (HA) exhibits numerous important roles in physiology and pathologies, and these facts necessitate an ability to accurately and reproducibly measure its quantities in tissues and cell cultures. Our group previously reported a rigorous and analytical procedure to quantify HA (and chondroitin sulfate, CS) using a reductive amination chemistry and separation of the fluorophore-conjugated, unsaturated disaccharides unique to HA and CS on high concentration acrylamide gels. This procedure is known as fluorophore-assisted carbohydrate electrophoresis (FACE) and has been adapted for the detection and quantification of all glycosaminoglycan types. While this previous FACE procedure is relatively straightforward to implement by carbohydrate research investigators, many nonglycoscience laboratories now studying HA biology might have difficulties establishing this prior FACE procedure as a routine assay for HA. To address this need, we have greatly simplified our prior FACE procedure for accurate and reproducible assessment of HA in tissues and cell cultures. This chapter describes in detail this simplified FACE procedure and, because it uses an enzyme that degrades both HA and CS, investigators will also gain additional insight into the quantities of CS in the same samples dedicated for HA analysis.
1. INTRODUCTION Hyaluronan (HA) is a glycosaminoglycan (GAG) normally consisting of thousands of disaccharides [-glucuronate-(1,3)-N-acetylglucosamine-(1,4)-]n forming hydrophilic, viscous macromolecules in the
Methods in Cell Biology
3
UNCORRECTED PROOF
5­10 MDa range (Toole, 2004). In contrast to the other GAGs (chondroitin/dermatan sulfate, CS/DS; heparin/heparan sulfate, HS; keratan sulfate, KS), HA is not synthesized as a proteoglycan in the Golgi. The HA synthases (HASs) have to be transported from the endoplasmic reticulum to the plasma membrane, where they can be activated and add alternately the UDP-sugar substrates onto the UDP-anchored reducing end, thereby elongating the polymer that is being extruded through the plasma membrane into the extracellular space (Weigel, 2015). Evolution has used HA's simple structure in a myriad of ways that are still challenging basic and clinical researchers today. The following list covers most, but not all of HA's physiology and pathophysiology: (1) its macromolecular Physical Properties are responsible for the viscosity of the eye's vitreous body and in the synovial fluids of diarthrotic joints (Necas, Bartosikova, Brauner, & Kolar, 2008); (2) its continuous synthesis and catabolism in the pericellular matrix of many, likely most, cells are essential to balance intracellular cytosolic UDP-N-acetylglucosamine concentration and maintain normal cytosolic functions (Hascall et al., 2014); (3) it is activated and synthesized into intracellular compartments, including the endoplasmic reticulum, during cell division when the extracellular level of glucose is more than ~ 2.5 Ч normal (diabetic hyperglycemia) due to increased cytosolic UDP-sugars inducing autophagy and impaired function after completing cell division (Hascall et al., 2014; Moretto et al., 2015); (4) it is modified in the extracellular matrix (ECM) by the HA-binding protein, TNF-induced protein 6 gene (TSG6), that transfers heavy chain proteins (HCs) from serum II to form HC­HA matrices, which is required for successful oocyte ovulation and therefore necessary for human life (Salustri, Camaioni, Di Giacomo, Fulop, & Hascall, 1999); (5) the production of an HC­HA matrix is also directly involved in most inflammatory processes through interactions with Inflammatory cells recruited to the site of injury (Jiang, Liang, & Noble, 2007; Lauer, Loftis, de la Motte, & Hascall, 2015; Wang, de la Motte, Lauer, & Hascall, 2011); and last, but not least (6) an aberrant interaction of HA with variants of CD44, the HA-binding cell surface protein on most cells, promotes migration, antiapoptotic intracellular pathways, and metastasis in several cancers (Toole, 2004). The increasing breadth and depth of HA research initiated intracellular HA meetings that expanded in the early 2000s. They were formal
4
Methods in Cell Biology
UNCORRECTED PROOF
ized by Endre Balazs who founded the International Society for Hyaluronan Sciences (ISHAS) in 2004. ISHAS now has International Conferences every 2 years. Perusal of the Program and Abstracts of the most recent meeting or of previous meetings on the ISHAS website (see https://ishas. org/) provides an overview of the ongoing development and progress of the HA research fields to date. As the HA research community expands, it is important to have assays for HA contents in tissues and cultures that are accurate and readily accessible in any laboratory environment. This chapter describes an assay for that purpose that has been used in our Program of Excellence in Glycosciences Resource Core. Many analytical assays to quantify HA utilize specific glycosidase enzymes that specifically degrade this GAG. These bacterial glycosyl eliminase's remove a water molecule from the 1­4 linkage between N-acetylglucosamine and glucuronic acid in the HA backbone structure thereby releasing a unique unsaturated disaccharide that constitutes the repeat building block of HA: 2-acetamido-2-deoxy-3-O-(-D-gluco-4-enepyranosyluronic acid)-D-glucose or Di-HA for short. Several analytical procedures exist to assess Di-HA quantity with high sensitivity and specificity, including high-pressure liquid chromatography (Karamanos, Syrokou, Vanky, Nurminen, & Hjerpe, 1994; Midura, Salustri, Calabro, Yanagishita, & Hascall, 1994; Zebrower, Kieras, & Heaney-Kieras, 1991) and capillary zonal electrophoresis (Al-Hakim & Linhardt, 1991; Chang, Yang, Zhao, & Linhardt, 2012; Kitagawa, Kinoshita, & Sugahara, 1995). This chapter will focus on just one of these assays to quantify Di-HA and this procedure is referred to as fluorophore­assisted carbohydrate electrophoresis (FACE). This focus should not be considered as an endorsement of this method being superior to other existing analytical methods, but rather viewed as an emphasis on FACE's simplicity for implementation within a wider scientific community. FACE was first reported for the separation and detection of monosaccharides (Jackson, 1994a, 1994b) and then later adapted for the separation of unsaturated disaccharide end products released from GAG chains using specific enzymatic digestion protocols (Calabro, Benavides, Tammi, Hascall, & Midura, 2000; Calabro, Hascall, & Midura, 2000). The original report for FACE detection of Di-HA was a rigorous analytical procedure designed for use in laboratories already
UNCORRECTED PROOF
Methods in Cell Biology
5
invested in Carbohydrate Chemistry research (Calabro, Benavides, et al., 2000). In these earlier studies, chondroitin ABC lyase (or chondroitinase ABC; EC 4.2.2.4) was used to depolymerize HA, and it should be noted that this enzyme also degrades all CS and DS GAG chains. The key to chondroitinase ABC yielding a complete degradation of HA is keeping the overall pH of the digestion solution at or below 7 (Calabro, Benavides, et al., 2000). Thus, for most samples digested with chondroitinase ABC and analyzed by FACE, the outcome data will include an assessment of HA and CS/DS. Finally, FACE technology was expanded to assess the quantities of other GAG types such as HS (Calabro et al., 2001) and KS (Plaas, West, & Midura, 2001) in tissues and cell cultures, though the enzymes that specifically degrade HS or KS usually do not yield a complete release of all unsaturated disaccharide structures limiting FACE's analytical outcomes for these GAG types. Therefore, this chapter will focus strictly on HA (and CS) assessment and readers are referred to prior literature regarding details for FACE assessments of HS or KS. After our initial publications, we recognized that the FACE procedure utilizes relatively straightforward chemistry, vertical gel electrophoresis equipment, and UV imaging systems common to most life science laboratories. The data produced by FACE are a gel-based output familiar to the broader cellular and molecular biology community making it more recognizable to nonglycoscientists. This gel-based format for FACE permits the simultaneous analysis of multiple samples in a single electrophoretic run providing fairly good sample analysis throughput. Altogether, these aspects motivated us to further simplify and streamline the FACE procedure for the detection and quantification of HA by any competent research laboratory. Thus, the purpose of this chapter is to describe in detail a greatly simplified FACE procedure to quantify HA that could become operational in most cell and molecular biology laboratories desiring to explore HA science, but not currently involved in carbohydrate chemistry.
2. PRINCIPLE AND THEORY The chemistry behind the FACE method utilizes a Schiff's base reaction between aldehyde and primary amine groups to form a metastable imine bond that is subsequently stabilized by a reducing
UNCORRECTED PROOF
6
Methods in Cell Biology
agent (reductive amination). In the FACE reaction (Fig. 1), the aldehyde group is provided by the reducing end of the HA (or CS) disaccharide, the primary amine is provided by 2-aminoacridone (AMAC), and the reducing agent is sodium cyanoborohydride (NaBH3CN). This stable chemical conjugation of a single AMAC molecule (425 nm excitation wavelength/ 520 nm emission wavelength) to a single HA (or CS) disaccharide molecule provides sensitive fluorescence-based detection on a molar basis. The separation of the AMAC-conjugated disaccharides into unique migration positions is realized by their electrophoretic migration in a high concentration acrylamide gel (Fig. 2). Detection of these unique AMAC-labeled disaccharide bands is accomplished with advanced digital gel imaging. Conventional densitometry software is used to assess the integrated optical density (IOD) for each AMAC-labeled band, and disaccharide quantities are calculated from densitometry of AMAC-labeled standards. Conjugating AMAC to varying unsaturated disaccharides of differing chemistry does not affect AMAC's fluorescence intensity making this approach suitable for the simultaneous assessments of HA and CS unsaturated disaccharides (Fig. 3).
3. EQUIPMENT 3.1. Electrophoresis Equipment Most vertical gel electrophoresis equipment can be used for FACE. We recommend using a system comparable to the BioRad Mini-PROTEAN® Tetra Cell apparatus to cast and run FACE gels. Tetra Cell plate dimensions should be 73 mm Ч 101 mm Ч 1 mm for running FACE. The quality of the glass plates to cast the gels is important to achieve optimal detection of faint fluorescing bands. The glass must be UV transparent and should not contain occlusions or imperfections that would obscure band detection and quantification (Fig. 4). We currently are using UV-transparent borosilicate glass plates of the dimensions listed earlier obtained from United Silica Products and, if careful to avoid chipping and scratching, one can reuse these plates for several years. We recommend 0.75 mm spacers for optimal imaging; larger spacers increase the thickness of the gel, which may decrease fluorescent band separation. We recommend using 10-well combs for producing straighter and flatter banding patterns, but 15-well combs can be
UNCORRECTED PROOF Fig. 1. Reductive amination reaction utilized by FACE methodology. For HA's disaccharide structure, the "R" groups would be hydrogen atoms and the carbon-4 position of the amine sugar would be in the equatorial position (i.e., N-acetylglucosamine).
8
Methods in Cell Biology
UNCORRECTED PROOF
Fig. 2. Electrophoretic migration positions of FACE standards. Note that the CCD camera captures bright bands on a dark background (left image) and it is often helpful for the user to invert the image to having dark bands on a light background (right image) to enhance band visibility. Standard bands are (1) N-acetylgalactosamine, (2) maltotetrose, (3) maltotriose, (4) maltose, (5) glucose, (6) unsaturated HA disaccharide, (7) unsaturated unsulfated chondroitin disaccharide, (8) 6-sulfated N-acetylgalactosamine, (9) 4-sulfated N-acetylgalactosamine, (10) unsaturated 6-sulfated chondroitin disaccharide, (11) unsaturated 4-sulfated chondroitin disaccharide, (12) unsaturated 2-sulfated chondroitin disaccharide, (13) unsaturated 4,6-disulfated chondroitin disaccharide, (14) unsaturated 2,6-disulfated chondroitin and unsaturated 2,4-disulfated disaccharides, and (15) unsaturated 2,4,6-trisulfated chondroitin disaccharide. Note that band 14 contains two disulfated chondroitin disaccharides that do not separate into unique positions under a broad survey run as shown; they sometimes can be separated by shortening the electrophoresis run time. * denotes a contaminant usually present in commercial lots of AMAC. used for larger sample sets with the understanding that there is an increased chance for band widening. 3.2. Imaging Equipment For gel image acquisition, a high-quality digital gel imaging system equipped with a UV transilluminator capable of emitting UV light around 365 nm wavelength is important to obtain optimal sensitivity for FACE detection of AMAC-labeled HA and CS disaccharides. We currently are using a UVP ChemiDoc-It2 515 integrated imaging sys
Methods in Cell Biology
9
UNCORRECTED PROOF
Fig. 3. AMAC fluorescence response unaffected by differing structures of conjugated unsaturated disaccharides. All AMAC-conjugated unsaturated disaccharides from HA and CS aligned with the calculated linear regression (R = 0.99; P < 0.001). tem, though any high-quality research grade imaging apparatus of similar specifications will suffice. A key aspect for the imaging system is the charge-coupled device (CCD) camera having the following specifications: 4 megapixel research-grade CCD providing even signal-to-noise detection over the entire chip surface, hermetically sealed CCD cooled down to - 60°C that greatly reduces background noise, a quantum efficiency of ~ 50%, data acquisition at 16-bit depth allowing for a wide linear dynamic range and simultaneous detection of low and high intensity bands, and interfaced with a variable focus high-quality lens (12.5­75 mm f/1.2) allowing one to fill the camera's field of view with the entire gel image. Commercial digital imaging units are integrated systems usually providing multiple band-pass filter slots (an ethidium bromide filter is important to prevent the appearance of the UV transilluminator bulbs in the gel's image), multiple light source capabilities, gel analysis software packages, and a computer/monitor to control acquisition and analyze data. 3.3. Gel Analysis Software Commercial integrated digital imaging systems typically offer gel analysis software either as a standard feature or as an optional accessory. The UVP ChemiDoc-It2 515 system includes VisionWorksLS
10
Methods in Cell Biology
UNCORRECTED PROOF
Fig. 4. Comparison of different gel plates for FACE imaging. Shown are inverted digital images of identical standards run on different gels using different gel plates exhibiting varying UV transparency and background uniformity. All images were acquired under the same imaging conditions including image acquisition time exposure (0.6 s). The first lane in each image is an AMAC blank lane to identify contaminant bands. The second lane is a sample of HA oligosaccharides forming a ladder pattern. The third lane (bands from top to bottom) contains glucose, unsaturated HA disaccharide, unsaturated, unsulfated CS disaccharide, HA tetra- and hexasaccharides. The fourth lane contains all FACE standards as defined in Fig. 2. The gel image on the left side is considered optimal for FACE analysis. The gel image in the middle exhibits too much speckling from occlusions in the gel plates even though it demonstrates good UV transparency. The gel image on the right side exhibits low UV transparency and would not be appropriate for low level detection of AMAC-labeled glycans. software and operates in a similar manner to that of a discontinued commercial gel analysis program called Gel Pro Analyzer (Media Cybernetics, Inc., Rockville MD). It is recommended that the software package included with the imaging system be used for gel analysis whenever possible for optimal compatibility. If the digital imaging system does not have a data analysis module, then one can download ImageJ at no cost (https:// imagej.nih.gov/ij/) that has the capability of quantifying bands on a gel. 3.4. Other Equipment Filter units to sterilize solutions (0.2 m pore size). Microcentrifuge tubes (1.5­2 mL).
UNCORRECTED PROOF
Methods in Cell Biology
11
Microcentrifuge (up to 15,000 Ч g). Vortex mixer. Magnetic stirrer. Stirring magnets (assorted sizes). Micropipettor set (overall volume range from 0.5 to 1000 L). Micropipette tips. Pipettes. Pipette bulb or electronic pipette filler. Ice bucket. Gel casting stand for Tetra Cell. High voltage power supply (1000 V recommended since electrophoresis occurs at 500 V for up to 1 h).
4. MATERIALS (HIGH QUALITY, REAGENT GRADE) Ultrapure water (MilliQ or equivalent). Ammonium acetate. 40% acrylamide solution (37.5:1 acrylamide to bis-acrylamide ratio) (BioRad, 161-0148). Ammonium Persulfate (BioRad, 161-0700, 10 g); make a 10% (w/v) solution in ultrapure water and store in 75 L aliquots at - 80°C indefinitely. TEMED (N,N,N,N-Tetra-methylethylenediamine) (BioRad, 161-0801, 50 mL). 1 Ч Tris/Borate/EDTA (TBE) electrophoresis buffer (we obtain a 1 L of 10 Ч stock from our institute's in-house lab and make 10 L at a time, which is stored at 4°C indefinitely; other commercial sources should work). Trizma base (Sigma, TRIZMA BASE, T-6791, 500 g, MW 121.1). Glacial acetic acid. AMAC (Thermo Fisher Scientific, A6289, 25 mg, MW 246.7). Dimethylsulfoxide (DMSO). Sodium Cyanoborohydride (NaBH3CN; Sigma-Aldrich, 156,159). Proteinase K (Invitrogen #25530015 for 100 mg and #25530031 for 1G). Bovine Serum Albumin. Chondroitinase ABC (Sigma Aldrich--Proteus Vulgaris/BSA Free # C3667-5UN (or 10UN), EC Number 232-777-9).
UNCORRECTED PROOF
12
Methods in Cell Biology
While several of our prior publications mention the use of both chondroitinase ABC and hyaluronidase SD to digest HA and CS in samples, it should be noted that this hyaluronidase is no longer commercially available. We have determined that adequate amounts of chondroitinase ABC in a digestion buffer maintained at pH 7 will degrade HA in samples quantitatively when given sufficient time. For example, see Fig. 5 where the same sample containing abundant amounts of HA was digested to an equivalent level by chondroitinase ABC alone as compared to chondroitinase ABC and hyaluronidase SD combined.
5. SOLUTIONS All solutions should be prepared using personal protective equipment and suitable ventilation conditions used when appropriate.
Fig. 5. FACE analysis of human synovial fluid samples digested with chondroitinase ABC and hyaluronidase SD as compared to chondroitinase ABC digestion alone. Shown are inverted digital images with the major band being AMAC-labeled unsaturated HA disaccharide. Lane 1 contains all FACE standards as defined in Fig. 2. Lanes 2 & 3 are duplicate sample aliquots digested with both chondroitinase ABC and hyaluronidase SD together. Lanes 4 & 5 are duplicate sample aliquots digested with just chondroitinase ABC alone. The faint bands toward the bottom third in lanes 2­5 are AMAC contaminants.
Methods in Cell Biology
13
UNCORRECTED PROOF
5.1. 100 mM Ammonium Acetate, pH 7.0 Dissolve 1.927 g ammonium acetate in 250 mL of water; the pH will be 7.0. Make 50 mL aliquots, filter sterilize and store at 4°C for up to 1 year. 5.2. Tris­Acetate Solution Dissolve 4.84 g of TRIZMA base into 80 mL water. Bring pH to 7.0 with glacial acetic acid and adjust volume to 100 mL to achieve a 400 mM concentration; filter sterilize and store at 4°C for up to 1 year. 5.3. AMAC/NaBH3CN Solution Add 2 mL DMSO to a vial containing 25 mg of AMAC; occasional agitation for 30 min at room temperature will assure suspension. Transfer the solubilized AMAC to a 15-mL centrifuge tube. Rinse the vial with another 1 mL DMSO and transfer to the same centrifuge tube. Add another 3.89 mL DMSO to the centrifuge tube to achieve a volume of 6.89 mL. Invert this capped tube several times to result in a homogeneous solution. Make a solution of NaBH3CN by adding 770 mg to 9.8 mL ultrapure water in a 15 mL tube (care needs to be taken making this solution as borohydride compounds release hydrogen gas). Invert to mix this solution at room temperature for 20 min with occasional inversion. Now, add 1.216 mL glacial acetic acid to your 6.89 mL of AMAC. Mix by inversion a few times. Mix 8 mL of the AMAC solution and 8 mL of the NaBH3CN solution in a 50 mL tube. Mix by inversion a few times. Immediately aliquot into microcentrifuge tubes (either 120 or 480 L aliquot volumes) and store at - 80°C indefinitely. This solution will contain 6.25 mM AMAC and 625 mM NaBH3CN and when thawed is ready to directly add to your samples. These concentrations of AMAC and NaBH3CN are in great excess compared to the amount of free reducing aldehyde groups in typical samples. Dispose of, and do not refreeze, any remaining reagent not used in the current experiment. 5.4. Proteinase K Stock Solution Add 10 mL of 100 mM ammonium acetate pH 7 to 100 mg of Proteinase K (PK); allow to dissolve for 30 min at room temperature. Transfer to a 150-mL flask or beaker. Rinse vial with another 10 mL of
14
Methods in Cell Biology
UNCORRECTED PROOF
100 mM ammonium acetate pH 7 and transfer to the 150-mL flask/beaker. Add 80 mL of 100 mM ammonium acetate pH 7 to the solubilized PK (total of 100 mL). Add 10 mg of SDS (or 100 L of a 10% solution) and stir until soluble. The final concentration of this stock solution is 1 mg/ mL PK containing 0.01% SDS. Store 1.5 mL aliquots in 2-mL microcentrifuge tubes at - 80°C indefinitely. Note: (1) this PK stock solution should be prepared, stored in aliquots, and frozen as quickly as possible, and (2) once a stock aliquot is thawed do not refreeze. Failure to freeze the PK aliquots as quickly as possible after resuspension and repeated freeze­thaw cycles of the PK stock aliquots have been noted to decrease PK activity. 5.5. Chondroitinase ABC Enzyme Solution Add 400 L of 100 mM ammonium acetate pH 7 to the vial containing 10 U of enzyme (25 mU/L concentration). If using the BSA free form of enzyme then add 0.1% (w/v) BSA to the 100 mM ammonium acetate pH 7 buffer as a stabilizing carrier protein to preserve optimal enzyme activity. Vortex mix and centrifuge at low speed to assure all of the liquid is at the bottom of the vial. Store 10 L aliquots in PCR microtubes at - 80°C indefinitely. 5.6. Acrylamide Gel Solution FACE gels consist of 20% acrylamide (37.5:1) in a 40 mM Tris­Acetate pH 7.0 buffer containing 2.5% glycerol. Prepare this gel solution in a 1-L Erlenmeyer flask by adding 500 mL 40% acrylamide (37.5:1), 100 mL Tris­Acetate (400 mM, pH 7.0), 375 mL distilled water, and 25 mL of glycerol. Note that slight variations in glycerol content in the acrylamide gel solution when cast into gels can result in large variations in the migration positions of AMAC-labeled glycan bands. Thus, because of glycerol's high viscosity, care must be taken to dispense the entire 25 mL of glycerol into the acrylamide stock solution. After thorough mixing, place 10 mL aliquots into 15-mL centrifuge tubes and store at 4°C indefinitely. Each 10 mL aliquot makes two standard mini gels using 0.75 mm spacers; this 1 L stock solution will yield enough acrylamide for 200 gels.
Methods in Cell Biology
15
UNCORRECTED PROOF
6. AMAC LABELING OF HA AND CS DISACCHARIDES 6.1. Proteinase K Digestion FACE analysis can be done on a wide variety of different tissues and cell types, though there are some important distinctions for cultured cells, conditioned media, and tissues. 6.1.1. Cultured cells Remove (and save) the conditioned media and wash the cells at least once with PBS. Dilute the 1 mg/mL PK stock solution to 0.1 mg/mL in 100 mM ammonium acetate pH 7 buffer and add 1 mL per 9.6 cm2 (i.e., 35 mm dish) of a confluent cell monolayer. 6.1.2. Conditioned media Add sufficient 1 mg/mL PK stock solution to the culture medium to yield a final concentration of 0.1 mg/mL PK. 6.1.3. Tissues Unlike cell cultures, tissues contain much greater amounts of ECM and need more aggressive digestion to release GAGs and DNA. We recommend directly adding 250 L of 1 mg/mL PK stock solution per 100 mg wet weight of tissue. Note: Keep accurate measures of tissue wet weights and approximate numbers of cells per culture surface as these values are used to estimate the amounts of chondroitinase ABC needed to achieve a complete digestion of HA (step 6.3) and the volume of AMAC solution needed for proper labeling of HA's unsaturated disaccharide (step 6.5). Incubate the samples at 60°C until tissues/cell layers are digested. In our experience a 2 h digestion should be sufficient to achieve an exhaustive digestion of the cell and ECM content containing up to 1.0 million cells. Cell layers containing 2­5 million cells usually require 3­4 h to assure an exhaustive digestion. We recommend the same digestion times for the culture medium samples as for their respective cell layer samples. For tissues, each tissue type usually requires differing digestion times to achieve an exhaustive digestion; vortex every 30 min
16
Methods in Cell Biology
UNCORRECTED PROOF
until the tissue is completely dissolved. For example, brain tissues dissolve in 3­4 h (depending upon age), yet liver tissues usually require 5­6 h to assure an exhaustive digestion outcome. In some extreme cases such as hyaline cartilage, tissue samples may require longer digestion times. After digestion is complete, we recommend saving a small aliquot from each sample for DNA analysis (see Section 9). For a 100 mg tissue wet weight sample digested in 250 L, we typically remove a 5 L aliquot for DNA analysis. For a cultured cell layer sample digested in 1 mL, we typically remove a 25 L aliquot for DNA analysis. If not proceeding to the first ethanol precipitation immediately (step 6.2), then the rest of the sample digest can be stored at - 20°C. 6.2. First Ethanol Precipitation Add four volumes of prechilled (- 20°C) 100% ethanol to each volume of sample, vortex mix and store at - 20°C overnight. For example, a 250 L PK digest of 100 mg tissue would be mixed with 1 mL of prechilled 100% ethanol to achieve a final ethanol concentration of around 80% ethanol. On the next day, centrifuge the samples at 14,000 Ч g for 10 min; note that this short centrifugation step can be done at room temperature since the sample starts at - 20°C and will still be cold at the end of the run. Discard the supernatants and recover the pellets containing precipitated GAGs and DNA. Wash the pellets by adding a second identical volume of prechilled (- 20°C) 75% ethanol and vortex. Centrifuge samples at 14,000 Ч g for 10 min as before. Discard the supernatant from this wash as before and tilt the tube with the pellet rotated to the top of the tube in order to remove residual ethanol using a pipette from the bottom of the tube assuring not to disturb the pellet. We have found that air drying the pellet at room temperature for 20 min to remove any remaining ethanol to be the most convenient method when dealing with a large number of samples and allows for an easier dissolving of the pellet in enzyme buffer. If you choose to use a vacuum dryer to remove the residual ethanol, then extra attention needs to be placed on determining the appropriate drying time so as not to overly desiccate the pellet, which will make dissolving the pellet more difficult. Add 35 L of 100 mM ammonium acetate pH 7 to each pellet and vortex mix to resuspend each GAG/DNA sample. We typically allow these resuspended samples to remain at room temperature for up to 20 min to fully resuspend the pelleted material before a second vortex
Methods in Cell Biology
17
UNCORRECTED PROOF
mixing. As a precaution to eliminate any residual PK activity in the resuspended pellets, the samples are heated at 100°C for 5 min and then chilled on ice for 5 min. If the samples are to be stored at this point, then we recommend storing them at - 20°C. If not, then proceed to step 6.3. 6.3. Digestion of HA and CS Add 1 L of chondroitinase ABC (25 mU) to each 35 L of sample, vortex mix and incubate at 37°C overnight (typically 18 h). This amount of enzyme over this digestion time period should be sufficient to digest HA and CS contents in most tissues having 100­200 mg wet weight or 10 million cells with their associated ECM. Note: Chondroitinase ABC's unit definition is the quantity of enzyme releasing 1 mol of unsaturated disaccharide from CS per minute at 37°C. Assuming no loss of enzyme activity during the 18-h time period, then 25 mU of enzyme under the stated conditions should digest ~ 10 mg of HA (or ~ 13 mg of CS) releasing unsaturated disaccharides for subsequent AMAC labeling. However, this stated enzyme-specific activity should be viewed as a theoretical maximum and actual sample digestions would likely exhibit lower overall digestion efficiencies. Nevertheless, this recommended amount of chondroitinase ABC assures a complete degradation of HA and CS into unsaturated disaccharide for most samples. It should be emphasized that chondroitinase ABC's ability to digest HA depends upon keeping the enzyme buffer pH in the range of 6­7 (as opposed to this enzyme's optimal pH range of 7­8 for the digestion of CS). We have selected a pH of 7 for our enzyme buffer as the best compromise to complete the digestion of both HA and CS over an 18-h time period. 6.4. Second Ethanol Precipitation At the end of the chondroitinase ABC digestion, some liquid will condense near the top of the tubes. Given the small volumes of these digestions, we recommend all samples should be centrifuged at low speed to make sure all of the liquid is at the bottom of the tube prior to the second ethanol precipitation step. Add 160 L of prechilled (- 20°C) 100% ethanol to each sample and vortex mix. Samples can be stored overnight at - 20°C if required or one can proceed through the following steps. Centrifuge samples at 14,000 Ч g for 10 min at room temper
18
Methods in Cell Biology
UNCORRECTED PROOF
ature and recover the supernatant in a separate 1.5 mL tube (Note: HA and CS unsaturated disaccharides are in this supernatant; the pellet will contain other intact GAGs and DNA). Wash the pellet with 100 L of prechilled (- 20°C) 75% ethanol, vortex mix and centrifuge at 14,000 Ч g for 10 min at room temperature once more. Recover this "pellet wash" supernatant and pool with the first supernatant for each sample. Lyophilize or vacuum dry these supernatant samples and immediately proceed to the AMAC labeling step listed below. The pellets should be air-dried for 20 min at room temperature. Resuspend these pellets in 20 L of 100 mM ammonium acetate pH 7, vortex mix and allow to dissolve at room temperature for 20 min. Vortex mix the resuspended pellet again and then briefly centrifuge at low-speed to assure that all sample liquid is at the bottom of the tube. Inactivate residual chondroitinase ABC activity by heating the samples at 100°C for 5 min, then placing them on ice for 5 min to achieve 4°C. Store these resuspended pellet samples at 4°C. 6.5. Sample Labeling With AMAC While stable in neutral pH buffer solutions, unsaturated disaccharides released from HA or CS may exhibit variable stability at room temperature in dry form. Thus, as a precaution, we recommend that the dried supernatant samples from the previous step should be immediately AMAC labeled to assure maximum recovery. The volume of labeling solution needed for each sample will vary depending upon the amount of your starting material. Optimal AMAC-labeling conditions should be determined empirically, but the following ratios of AMAC solution to sample amount are recommended as a starting point: (i) add 5 L of AMAC solution per 5­10 million cultured cells (and/or the GAGs secreted into their culture medium) and (ii) add 5 L of the AMAC solution for every 50 mg of tissue wet weight. For samples larger than those listed, these recommended starting ratios can be scaled up to accommodate a proper reaction. Since these AMAC reaction volumes are small, we recommend to briefly centrifuge the resuspended sample at low-speed to assure that all sample liquid is at the bottom of the tube. Incubate the samples at 37°C for 18 h in the dark to complete the AMAC-labeling reaction. Note: We have found that having a somewhat more concentrated AMAC-labeled sample, as opposed to one that is too dilute, is better
UNCORRECTED PROOF
Methods in Cell Biology
19
suited for FACE detection via digital gel imaging. Concentrated AMAC-labeled samples can be diluted with additional AMAC-labeling solution to achieve a better sample concentration. Samples that are too dilute, however, should not be processed further because FACE gels cannot accommodate more than 3 L sample volume per gel lane to achieve well-separated, flat bands (step 7.3). 6.6. Preparation of FACE Standards For investigators wanting to assess HA contents in their samples, the minimum number of standards required would be Di-HA, Di-0S, Di-4S, and Di-6S. As shown in Fig. 2, a larger set of standards can be generated when also assessing CS content in samples. A detailed protocol for the preparation of FACE standards is provided on our website (http://pegnac. sdsc.edu/cleveland-clinic/protocols/) and investigators are encouraged to download this detailed FACE protocol if they wish to prepare their own standards. Alternatively, our lab routinely prepares FACE standards and is willing to provide sufficient amounts of these standards at no cost other than a shipping fee.
7. GEL PREPARATION See Section 3.1 for recommendations on gel apparatus. 7.1. Gel Casting Each FACE gel (using 0.75 mm spacers) will need about 5 mL of the acrylamide solution (step 5.6) transferred to a clean tube and, for each 5 mL of acrylamide solution, 25 L of 10% (w/v) ammonium persulfate, and 5 L TEMED are added. Mix by inversion, transfer 5 mL to each preassembled casting plate unit and immediately place the sample comb at the top of the gel solution in order to form flat wells. Polymerization is usually complete within 10 min. 7.2. Chilling the Gels Place the casted gels into their gel apparatus. Quickly, add prechilled (4°C) 1 Ч Tris/Borate/EDTA (TBE) electrophoresis buffer to the inner and outer gel apparatus chambers to precool the gel. Remove the sample comb and rinse the gel wells with electrophoresis buffer to assure flat and even well surfaces resulting in flatter banding patterns during
20
Methods in Cell Biology
UNCORRECTED PROOF
FACE runs (failure to rinse the gel wells can result in an uneven well surface resulting in a distorted banding pattern due to residual acrylamide polymerizing unevenly in the well). The prechilled buffer levels in the outer chamber should be at or just below the visible gel wells in order to promote better cooling of the gels. If possible, place a stir bar in the bottom of the gel apparatus. Place the assembled gel box in a seamless, secondary plastic box container having the following dimensions: 14 length Ч 12 wide Ч 6 high of ~ 10 L capacity. We currently use 10 L Nalgene HDPE Pans distributed through Thermo Fisher Scientific (Cat. # 7120-0010). Surround the gel apparatus with ice up to where the gel box's top lid meets with the highest point of the buffer chamber. In a corner of the secondary plastic container, carefully pour water into the ice until the water level reaches about halfway up the gel apparatus to create an ice-water slurry bath. Place this entire assembly onto a stir plate to facilitate a more even heat dissipation during running of the gel. Cool the gel apparatus to 4°C for 1­2 h before running the gel (usually 1.5 h). Add fresh ice to the secondary container to replace any melted ice before running the gel and during the gel run. Maintaining gel temperature at 4­5°C prior to electrophoresis is extremely important for tight/flat bands and best separation of closely spaced bands (a nonconducting thermometer can be inserted into the buffer chamber on most commercial gel boxes to make temperature readings before and during the gel run). Cautionary note: Take care to make sure that all ice and water is well away from the electrical lines and their connectors to the power supply prior to introducing electrical current. 7.3. Loading the Samples Unlike protein gels using a stacking buffer where samples of 30 L volume loaded per lane is common, FACE does not employ a stacking phenomenon and can only accommodate much smaller sample volumes. We recommend loading 2­3 L of sample per lane for a 10-well gel. Add the AMAC-labeled samples directly to the well (a "loading buffer" is not necessary). Keep in mind that the greater the sample volume per lane, the thicker, and less resolved, the bands will appear. One lane in the gel should include an AMAC blank (no sample, just AMAC solution) that is used to identify and account for the trace impurities in the AMAC solution that migrate into the lane (see Fig. 2); AMAC itself
UNCORRECTED PROOF
Methods in Cell Biology
21
is uncharged and will not migrate in the gel. It is also important to include a lane of FACE standards (see step 6.6). The IOD of the Di-HA band of known quantity in the standard lane is used to calculate the amount of Di-HA in the sample lanes based on their measured IOD values. Note: Any unused gel lanes should be loaded with AMAC blank solution, which helps to keep banding patterns as flat and even as possible (e.g., reducing band curvature and widening) and helps to avoid undesired banding patterns. 7.4. Electrophoresis Once AMAC-labeled samples and standards are loaded, then the FACE gel should be run at 500 V constant voltage. The amperage at the beginning of FACE electrophoresis should be around 30 mA/gel and drop to around 15 mA/gel at the end of the run. For best separation of Di-HA from Di-0S, electrophoresis should be between 50 and 60 min run time (typically a 55 min run time gives good separation of these two closely spaced bands). Note: A run time of 45 min is recommended for optimum resolution of all CS bands including the disulfated disaccharides.
8. IMAGE ACQUISITION Unlike protein gels, FACE gels should be imaged while still housed in their gel plates. We do not recommended to remove the gel from its gel plates because FACE gels are thin and fragile and almost always crack and break when the gel plates are removed. Also, unlike protein gels, FACE gels need to be imaged immediately after electrophoresis is completed. At the end of the electrophoretic run, the gel temperature should be between 6°C and 8°C and, upon removing the gel plate from the apparatus, the gel will rapidly warm up to room temperature. While this rapid warm-up of the gel is important to prevent condensation fogging during imaging, it should be emphasized that the AMAC-labeled unsaturated disaccharides are of low molecular weight and will be susceptible to diffusion over time at room temperature. Thus, if one waits 60 min at room temperature before imaging a FACE gel, then the banding patterns will widen and exhibit blurry edges. In effect, this band diffusion will diminish the ability to accurately distin
22
Methods in Cell Biology
UNCORRECTED PROOF
guish the Di-HA from the Di-0S bands (roughly 1-mm separation on typical FACE gels) for quantification purposes. As mentioned in step 3.2, FACE gels should be imaged using a research grade, high-quality cooled CCD camera system using an ethidium bromide filter to prevent an image of the UV transilluminator bulbs in the gel image. The camera system should have a narrow depth-of-field to provide optimal focusing of the fluorescing bands and avoid imaging of any dust particles that may settle on the top gel plate. Also, the camera system should have a variable focal length lens in order for the entire gel to fill the complete imaging field (this assures maximal image spatial resolution). Image acquisition should be done using a grayscale modality as color imaging will insert additional filters in front of the chipset. These filters used for color imaging will reduce the amount of light detection from the fluorescing bands thereby decreasing sensitivity and necessitating much longer exposure times. Image acquisition should be obtained at a bit-depth of either 12- or 16-bit depth. An 8-bit camera system acquires data in 256 grayscale bins. In comparison, a 16-bit depth camera system acquires data in 65,536 grayscale bins (see https://en.wikipedia.org/wiki/Grayscale for additional information). Thus, the linear dynamic range of a 16-bit chipset is increased by 256-fold over an 8-bit chipset. What this means for digital imaging of FACE gels is that a 16-bit depth will enable an investigator to quantify low intensity and high intensity bands accurately in a single acquired image. Most commercial software used to acquire digital images from cooled CCD imaging systems will provide a tool to visually assess the percentage of pixels in the image that are oversaturated (i.e., overexposed). These tools usually apply a mask layer above the gel image that color codes oversaturated pixels as a red color. Also, the software will provide a number value for the percentage of pixels that are oversaturated in the gel image. For accurate quantification of gel band intensities, it is recommended to keep the pixel saturation level to less than 1% of the total pixels in the image. This tool will assure that an inexperienced investigator will acquire images that are properly exposed and yield more accurate results. Also, such a tool will eliminate the need for digital postprocessing of the image that will "stretch" or "equalize" the grayscale of under exposed images (see https://en.wikipedia.org/wiki/
UNCORRECTED PROOF
Methods in Cell Biology
23
Histogram_equalization for a description of this postprocessing method). Most commercial high-quality, research-grade cooled CCD imaging systems provide remarkably even background values over the entire gel image. A lower quality CCD chipset may exhibit "hot spots" (brighter than average pixels) and "cold spots" (lower than average pixels) that may generate images of uneven background values. In these cases, an investigator may wish to use digital postprocessing approaches to provide a more even background correction in the digital gel images. If this postprocessing background correction is desired, then we strongly advise these individuals to become fully familiar regarding proper background correction filtering steps, which will help the investigator to avoid the pitfalls of introducing postprocessing artifacts that could potentially invalidate the rendered data (we refer investigators to https://imagej.net/Image_Intensity_ Processing#Background_correction for a nice description of proper background correction). As mentioned in step 7.3, at least one lane in the FACE gel should be loaded with the AMAC solution only. All commercial sources of AMAC exhibit varying amounts of trace contaminants and some of these contaminants carry negative charge and migrate into the FACE gel at distinct distances (see Figs. 2, 4, and 5). While these fluorescing contaminant bands do not migrate to the position of Di-HA and will not interfere with HA quantification, one of the fluorescing AMAC contaminants migrates to the position of Di-6S. If an investigator desires to accurately measure Di-6S, then assessing the IOD of the AMAC contaminant that comigrates to this location provides a correction factor that would need to be subtracted from the overall IOD at the Di-6S migration position.
9. DATA NORMALIZATION The quantitative data directly derived from FACE gels will need to be normalized in order to cross compare within a set of test groups run at the same time, or across different test groups run at different times. As mentioned in step 6.1, we strongly advise measuring the wet weights of tissue samples as accurately as possible; this is not generally practical for most cell culture samples. In some cases normalization of HA content from FACE analysis using these wet weight measures may
24
Methods in Cell Biology
UNCORRECTED PROOF
suffice. We should point out that many musculoskeletal tissue samples (e.g., tendon, ligament, or hyaline cartilage) contain ample amounts of collagen and an alternative normalization factor for HA content could be collagen content in the same sample. A surrogate measure for collagen content would be an assay for hydroxyproline as this amino acid is principally found in collagens (Cissell, Link, Hu, & Athanasiou, 2017; Hofman, Hall, Cleaver, & Marshall, 2011; Woessner, 1961). For cell culture samples, or tissue samples having ample cellular content, we recommend that a double-stranded DNA (dsDNA) assessment be used as a surrogate for assessing sample cell numbers. Currently, we are using the Quant-iTTM PicoGreen® dsDNA Assay Kit (ThermoFisher Scientific), but any other specific and sensitive assay would be appropriate. REFERENCES Al-Hakim, A., Linhardt, R.J., 1991. Capillary electrophoresis for the analysis of chondroitin sulfate- and dermatan sulfate-derived disaccharides. Analytical Biochemistry 195, 68­73. Calabro, A., Benavides, M., Tammi, M., Hascall, V.C., Midura, R.J., 2000. Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by fluorophore-assisted carbohydrate electrophoresis (FACE). Glycobiology 10, 273­281. Calabro, A., Hascall, V.C., Midura, R.J., 2000. Adaptation of FACE methodology for microanalysis of total hyaluronan and chondroitin sulfate composition from cartilage. Glycobiology 10, 283­293. Calabro, A., Midura, R., Wang, A., West, L., Plaas, A., Hascall, V.C., 2001. Fluorophore-assisted carbohydrate electrophoresis (FACE) of glycosaminoglycans. Osteoarthritis and Cartilage 9, S16­S22. Chang, Y., Yang, B., Zhao, X., Linhardt, R.J., 2012. Analysis of glycosaminoglycan-derived disaccharides by capillary electrophoresis using laser-induced fluorescence detection. Analytical Biochemistry 427, 91­98. Cissell, D.D., Link, J.M., Hu, J.C., Athanasiou, K.A., 2017. A modified hydroxyproline assay based on hydrochloric acid in Ehrlich's solution accurately measures tissue collagen content. Tissue Engineering Part C 23, 243­250. Hascall, V.C., Wang, A., Tammi, M., Oikari, S., Tammi, R., Passi, A., et al., 2014. The dynamic metabolism of hyaluronan regulates the cytosolic concentration of UDP-GlcNAc. Matrix Biology 35, 14­17. Hofman, K., Hall, B., Cleaver, H., Marshall, S., 2011. High-throughput quantification of hydroxyproline for determination of collagen. Analytical Biochemistry 417, 289­291. Jackson, P., 1994. The analysis of fluorophore-labeled glycans by high-resolution polyacrylamide gel electrophoresis. Analytical Biochemistry 216, 243­252. Jackson, P., 1994. High-resolution polyacrylamide gel electrophoresis of fluorophore-labeled reducing saccharides. Methods in Enzymology 230, 250­256.
Methods in Cell Biology
25
UNCORRECTED PROOF
Jiang, D., Liang, J., Noble, P.W., 2007. Hyaluronan in tissue injury and repair. Annual Review of Cell and Developmental Biology 23, 435­461. Karamanos, N.K., Syrokou, A., Vanky, P., Nurminen, M., Hjerpe, A., 1994. Determination of 24 variously sulfated galactosaminoglycan- and hyaluronan-derived disaccharides by high-performance liquid chromatography. Analytical Biochemistry 221, 189­199. Kitagawa, H., Kinoshita, A., Sugahara, K., 1995. Microanalysis of glycosaminoglycan-derived disaccharides labeled with the fluorophore 2-aminoacridone by capillary zone electrophoresis and high-performance liquid chromatography. Analytical Biochemistry 232, 114­121. Lauer, M.E., Loftis, J., de la Motte, C., Hascall, V.C., 2015. Analysis of the heavy-chain modification and TSG-6 activity in pathological hyaluronan matrices. Methods in Molecular Biology 1229, 543­548. Midura, R.J., Salustri, A., Calabro, A., Yanagishita, M., Hascall, V.C., 1994. High-resolution separation of disaccharide and oligosaccharide alditols from chondroitin sulphate, dermatan sulphate and hyaluronan using CarboPac PA1 chromatography. Glycobiology 4, 333­342. Moretto, P., Karousou, E., Viola, M., Caon, I., D'Angelo, M.L., De Luca, G., et al., 2015. Regulation of hyaluronan synthesis in vascular diseases and diabetes. Journal of Diabetes Research 2015, https://doi.org/10.1155/2015/167283. Necas, J., Bartosikova, L., Brauner, P., Kolar, J., 2008. Hyaluronic acid (hyaluronan): A review. Veterinбrnн Medicнna 53, 397­411. Plaas, A.H.K., West, L.A., Midura, R.J., 2001. Keratan sulfate disaccharide composition determined by FACE analysis of keratanase II and endo--galactosidase digestion products. Glycobiology 11, 779­790. Salustri, A., Camaioni, A., Di Giacomo, M., Fulop, C., Hascall, V.C., 1999. Hyaluronan and proteoglycans in ovarian follicles. Human Reproduction Update 5, 293­301. Toole, B.P., 2004. Hyaluronan: From extracellular glue to pericellular cue. Nature Reviews Cancer 4, 528­539. Wang, A., de la Motte, C., Lauer, M., Hascall, V.C., 2011. Hyaluronan matrices in pathobiological processes (minireview). FEBS Journal 278, 1412­1418. Weigel, P.H., 2015. Hyaluronan synthase: The mechanism of initiation at the reducing end and a pendulum model for polysaccharide translocation to the cell exterior. International Journal of Cell Biology 2015, https://doi.org/10.1155/2015/367579. Woessner, J.F., 1961. The determination of hydroxyproline in tissue and protein samples containing small proportions of this imino acid. Archives of Biochemistry and Biophysics 93, 440­447. Zebrower, M.E., Kieras, F.J., Heaney-Kieras, J., 1991. High pressure liquid chromatographic identification of hyaluronic acid and chondroitin sulfate disaccharides. Glycobiology 1, 271­276.

, pages, 0 Mb

Section C, 44 pages, 1.74 Mb

Jeff Wall, 13 pages, 0.07 Mb
Copyright © 2018 doc.uments.com