Disaccharide Analysis of Glycosaminoglycans From Twenty‐Four Organs of Young and Aged Mice

Abstract

ABSTRACT

Glycosaminoglycans (GAGs) are abundant negatively charged polysaccharides ubiquitously expressed in mammalian organs, having essential biological functions in development and homeostasis. It has been scarcely reported that GAG structures are changed in aged and diseased human organs; however, an overall landscape of GAGs in individual organs of an animal is missing. Here, we performed an in‐depth analysis of disaccharide composition of chondroitin/dermatan sulfate (CS/DS) and heparan sulfate (HS) from 24 organs of young and aged mice. Quantification of the disaccharide species showed that CS/DS dominates almost all organs, but HS dominates in five organs. As expected, the sulfation pattern of CS/DS and HS varied in different organs, which likely correlates with their biological functions in a given organ. Nevertheless, the age‐dependent alteration is more prominent in the disaccharides of CS/DS, which may suggest CS/DS has a more important role in aging. This first report of a comprehensive analysis of GAGs amount and structure should be highly relevant in understanding how GAGs affect diseases and aging.

Introduction

Glycosaminoglycans (GAGs) refer to a group of polysaccharides, including chondroitin sulfate (CS), dermatan sulfate (DS), heparan sulfate (HS), heparin and hyaluronic acid (HA). Unlike other natural carbohydrate polymers, GAGs are linear polysaccharides composed of repeating disaccharides of amino and uronic acid sugars. CS/DS is composed of N-acetyl galactosamine (GalNAc) and uronic acid. While CS contains only glucuronic acid (GlcA), the term CS/DS is used because many chains contain both glucuronic and iduronic acid (IdoA), in variable proportions. HS and heparin share the same type of disaccharides, glucosamine (GlcN) and GlcA/IdoA. The sugar units of CS/DS and HS can be substituted by sulfate at various positions, which further increases the negative ionic property and structural diversity. HA is composed of N-acetyl glucosamine (GlcNAc) and GlcA lacking sulfate substitution. GAGs, but not heparin, are expressed ubiquitously in all tissues, having essential biological roles through interaction with diverse ligands [1]. Molecular and genetic studies have identified all the enzymes involved in GAGs biosynthesis which significantly advanced the understanding of the regulation of GAGs biosynthesis and biological functions. The important functions of the GAGs are most evident in development as illustrated through knockout of the enzymes responsible for biosynthesis in animals [2–5]. Due to the structural heterogenicity, CS/DS and HS display an enormous structural variety in different cells/tissues, which is believed to be associated with their biological functions. In addition to the indispensable roles in development and homeostasis, recent studies have increasingly implicated GAGs in pathological processes of various diseases, such as inflammation, cancer metastasis and age-related diseases [6, 7]. For instance a recent report shows that the level of HS-3-O-sulfotransferase 2 is higher in Alzheimer's disease brain [8]. CS/DS has been found altered in molecular size, composition and expression level in a mouse model of AD [9].

It is important to understand the structural properties of GAGs under normal, aging or disease conditions, similarly to the importance in structural and biological comprehension of nucleic acid and proteins. However, restricted by the lack of a robust method to determine GAGs sequences, so far, fingerprinting disaccharides as the building blocks of GAGs has been used to determine the composition of GAGs. Using this method, a systemic analysis of all species of GAGs from most animal organs has not been done.

To provide an overall landscape of GAGs in mice and to determine whether the GAGs change with aging, we collected 24 organs from young and aged mice for a systemic analysis of disaccharide composition. Our results of 192 samples show characteristic features of different GAGs in organs. Because it is known that the sulfation pattern of GAGs determines the interaction with protein ligands, the observed changes in sulfated disaccharides of GAGs in some organs from aged animals could be the cause of age-related features. The obtained abundant information should provide a valuable reference to study how GAGs function in specific organs and affect aging.

Materials and Methods

Mice

C57BL/6 mice, 14 weeks old (young) were maintained in the animal facility at Biomedical Center, Uppsala University. Experiments were done under permission 01751/1020 approved by the local ethics committee and in accordance with animal welfare regulations. Twenty-1-month-old C57BL/6 (aged) were housed at the animal facility of Swedish Veterinary Agency, Uppsala, Sweden and were surplus mice belonging to protocol 5.8.18-05195/2020 approved by the regional animal ethics committee. Mice had free access to water and food during their lifespan without any intervention. Organs were collected after the mice were euthanized and kept frozen until use. A total of 24 organs were collected from four young or aged male, except uterus and ovary that were derived from four young and four aged female mice (192 organs in total). The following are specifications of selected organs: brain = whole brain; cartilage = xiphoid process in sternum; intestine = 12-cm-long jejunum starting 6 cm below stomach; liver = lateral left liver lobe; lymph nodes = submandibular plus mesenteric lymph nodes; muscle = right quadriceps femoris skeletal muscle; ovary = ovary plus oviduct; seminal gland = seminal plus coagulation glands; skin = approx. 4 square cm shaved skin from dorsum below neck. If not specified, both bilateral organs were taken.

Purification of GAGs

The collected organs were dried by lyophilization for 3 days. Dried tissues were weighed and treated with 200 μg/ml pronase (from Streptomyces griseus, Roche) in 50 mM Tris/HCl, pH 8, 1 mM CaCl₂, and 1% Triton X-100 by incubation at 55°C with shaking to degrade proteins. After 24 h, fresh pronase was added and the digestion was continued for another 24 h. After heat inactivation of the protease, digests were centrifuged at 10,000 g for 10 min and supernatants were collected. Samples were treated with 20 Sigma units/ml benzonase (after addition of MgCl₂ to reach final 2 mM) and incubated for 2 h at 37°C to degrade DNA. After heat deactivation of the DNAase, NaCl was added to a final concentration of 100 mM. The tissue lysate was incubated with 50 µl DEAE-Sephacel anion exchange gel (Cytiva) overnight in a tube. The gel was transferred to disposable spin centrifuge tubes (Costar 8163) and washed sequentially with buffer 1 (50 mM Tris/HCl, pH 8, 100 mM NaCl, and 0.1% Triton X-100); buffer 2 (50 mM NaOAc, pH 4, 100 mM NaCl, and 0.1% Triton X-100), deionized water, and 100 mM NH₄HCO₃ to wash away unbound materials. GAGs bound to the gel were eluted by 2 M NH₄HCO₃. After lyophilization, GAGs were dissolved in water and the amount was provisionally estimated by carbazole reaction [10]. Then, 1–2 µg GAGs from each sample were dried and subjected to degradation by chondroitinase ABC or a mixture of heparinase I + II + III. Chondroitinase ABC (Sigma) digestion was carried out overnight at 37°C in 10 μl 50 mM NH₄OAc, pH 7.5, containing 10 mIU enzyme and resulted in quantitative disaccharide production of both CS/DS and HA. Heparinases [(in house preparation, purified from E. coli, transformed with the pET-15 vector containing heparinase I, or vector pET-19b containing heparinase II or III, as provided by Prof. Jian Liu (University of North Carolina, Chapel Hill, NC)] digestion was carried out overnight at 37°C in 10 μl 50 mM NH₄OAc, pH 7.5, containing 5 mIU of each enzyme.

Disaccharide Analysis

The resulting disaccharides were labeled with fluorophore by adding 10 μl of a solution containing 20 mM repurified 2-aminoacridone (AMAC, Sigma 06627) and 1 M NaBH₃CN and incubating at 45°C for 16 h [modified from [11]]. The repurified AMAC was obtained by reversed-phase chromatography on a Zorbax RX-C8 semipreparative column (9.4 × 250 mm, 5 μm, Agilent), using a 13-min isocratic separation with 65% NH₄OAc and 35% acetonitrile as eluents at 3 ml/min. The AMAC-labeled disaccharides were analyzed as described previously [11], with slight modifications. Briefly, 1-μl samples were injected onto an XBridge BEH Shield RP18 (2.1 × 100 mm, 2.5 μm). Disaccharides were separated using a 39-min gradient run at 0.35 ml/min (0–1 min: 98% A: NH₄OAc, 60 mM, pH 5.6, and 2% B: acetonitrile, 1–3 min: 98%–96% A, 3–26 min: 96%–85% A, 26–28 min: 85%–10% A, 28–32 min: 10% A, 32–34 min: 10%–98% A, 34–39 min: 98% A) on a Thermo Scientific UltiMate 3000 Quaternary Analytical system with an FLD-3400RS fluorescence detector (excitation λ = 428 nm and emission λ = 525 nm). The column was kept at 30°C to improve performance and reproducibility.

The disaccharide species of CS/DS and HS were identified and quantified using standard disaccharides (Iduron, UK). The proportion of each species was derived by the fluorescent area of a given peak divided by the area of the sum of all identified peaks. The quantity of CS/DS, HA, and HS (expressed as ng/mg dry tissue) was obtained by summing the amount of each identified disaccharide. The amount of each disaccharide was calculated comparing the fluorescence area of the peaks with the fluorescence area of the relative standard, whose amount was known. The standard mixture was mock-treated in the same buffers and enzymes as the samples in each series of runs.

Preparation and Chain Length Analysis of In Vivo Labeled CS/DS and HS

Two young and two aged male mice were subcutaneously injected with 1 mCi ³⁵S-sulfate (Biotech-IgG; 2 × 50 µl in two sites) for in vivo labeling. After 2 h the mice were killed and brain, heart, kidney, liver, lung, and spleen were collected. GAGs were purified by the same procedure as above with two differences: i) the organs were not lyophilized before pronase digestion and ii) after DEAE-Sephacel purification, GAGs were detached from the remains of core proteins by beta-elimination (0.5 M NaOH, 0.1 M sodium borohydride for 16 h at 4°C). Sodium borohydride was inactivated by acidification reached by addition of 10% acetic acid, and the pH was brought to neutral with 2 M sodium bicarbonate. The released free GAGs were further isolated by DEAE-Sephacel as described above.

HS chains were degraded by nitrous acid treatment at pH 1.5 [12] and CS/DS chains by chondroitinase ABC digestion, as described above, to allow the analysis of intact CS/DS and HS chains, respectively. Labeled chains were applied on a Superose 6 column (Cytiva) run in 0.2 M ammonium bicarbonate. The flow rate was 0.25 ml/min, and fractions were collected every 2 min. ³⁵S radioactivity was measured by liquid scintillation.

Statistical Analysis

All statistical analyses were done in GraphPad Prism 10.1.1 (GraphPad Software, Boston, Massachusetts USA). Young and aged samples were compared with unpaired t tests with Welch correction. Statistically significant differences are indicated with *=p < 0.0332; **=p < 0.0021; ***=p < 0.0002; ****=p < 0.0001. For each group of organ disaccharide data points, i.e. the group of biological replicates, outliers were detected using the ROUT method. If an outlier was detected, all related disaccharide values from the same biological sample were manually excluded from subsequent analysis.

Results

Distinctive Disaccharide Composition of CS/DS From Different Organs of Young Mice

A total of 192 samples were prepared from 24 organs of four young and four aged mice for analysis of disaccharides of CS/DS, HA and HS. To analyze disaccharide composition of CS/DS, purified GAGs were degraded exhaustively with chondroitinase ABC and the resultant disaccharides were fluorescently labeled with AMAC. Separation of the fluorescent products by HPLC enabled the detection of six structures derived from CS/DS, that is, three disulfated, one 6-O and one 4-O monosulfated, and the nonsulfated disaccharides, plus one HA-derived nonsulfated structure (Figure 1 A1 and 1 A2). The chondroitinase lyase treatment led to losing the epimeric status of glucuronic or iduronic acid which is present in the native chains, resulting in unsaturated hexuronic acid (UA). The peak area of each identified disaccharide species was quantified and divided by the total area of all six peaks of CS/DS for each sample. The relative molar proportion of each species is presented in Figure 2 and Supporting Table 2. Variability in the quantification of disaccharides was inversely proportional, as expected, to species abundance. The average of the coefficients of variation (CVs) of the 24 organs was 3% for the predominant structure UA-GalNAc4S, while CV was between 9% and 20% for most of the other disaccharides (data not shown). The least present structure UA2S-GalNAc4S had a CV of 34%. To verify the technical variability contribution to the total variation, GAGs extracted separately from four equal portions of one liver lysate were treated with chondroitinase ABC, and the resultant CS/DS disaccharides were labeled and analyzed in parallel (Supporting Table 1). The CV of the molar proportion of the disaccharides were less than 2%, except the minor component UA-GalNAc6S that had a CV of 7%. From these data we can conclude that the observed variability in the presented data is mainly contributed by biological variations of the mice.

[IMAGE OMITTED. SEE PDF]

[IMAGE OMITTED. SEE PDF]

Analysis of the data collected from young organs shows that the three disulfated disaccharides, endowed with distinct and partially recognized biological activities, constituted a minority of CS/DS structures in most organs and showed a considerable variability. UA-GalNAc4S,6S (Figure 2D) was less than 2% in bladder, brain, eye, heart, muscle, ovary, skin, uterus, but more than 10% in adrenal gland and spleen. Noticeably, it was as high as 37% in the liver. Overall, the proportion of UA-GalNAc4S,6S in the 24 organs had a mean of 5.9%; median = 3.7%; range 0.4-37%. UA2S-GalNAc4S (Figure 2E) was present at low level in brain, cartilage, spleen, and uterus and at high level in muscle and skin (mean of all organs = 4.5%; median = 4.7%; range 0.8%–11%). UA2S-GalNAc6S (Figure 2F) was a rare structure, comprised less than 1% in most of the organs (mean of all organs = 0.5%; median = 0.2%; range 0.1%–1.4%). The nonsulfated UA-GalNAc and the 6-O-monosulfated structures UA-GalNAc6S were present at an intermediate level in all organs, but again varied considerably. UA-GalNAc (Figure 2A) was 5%–10% in half of the organs, with lower level in bladder, cartilage, skin, spleen, uterus and higher level in adrenal gland, eye, harderian gland, heart, kidney, muscle, pancreas, salivary gland. The mean of all organs = 9.6%; median = 8.1%; range 3%–24%. UA-GalNAc6S (Figure 2B) represented 4%–10% in about half of the organs, with a lower amount in bladder, cartilage, liver, muscle, skin, uterus and higher amount in adrenal gland, eye, kidney, lung, lymph node, seminal gland, testicle. The mean of all organs = 7.3%; median = 5.3%; range 2%–17%. The 4-O-monosulfated UA-GalNAc4S (Figure 2C) was the predominant structure, varying the least (mean of all organs = 72%; median = 70%; range 50%–91%).

Examination of the distribution of the disaccharides with different sulfation degree showed that nonsulfated and disulfated species were minority structures regardless of organ (Figure 3A). Twelve organs contained less than 10% of disulfated structures, and ten organs contained between 10% and 16%, with mean and median = 10%; range from 2% (uterus) to 40% (liver). Although the amount of the different sulfated structures varied considerably, the level of total sulfation in CS/DS was similar in all organs (Figure 3C). Almost all organs contained between 0.90 and 1.10 sulfates/disaccharide with mean and median of all organs =1.01; range 0.79 (eye)−1.34 (liver). Therefore, liver appeared distinctive due to an unusually high proportion of the disulfated UA-GalNAc4S,6S (37%) and so its overall sulfation density was much higher than the other organs.

[IMAGE OMITTED. SEE PDF]

Age-Related Alterations of CS/DS Disaccharides

Comparison of the data from young and aged organs presented in Figure 2 showed statistically significant changes of the fine structures in several organs with aging. The disulfated structure UA-GalNAc4S,6S was significantly increased in 14 aged organs but not decreased in any organ, whereas UA2S-GalNAc4S was increased in nine organs and decreased in 2. On the other hand, the low abundant UA2S-GalNAc6S was not detected in five aged organs. The nonsulfated UA-GalNAc was increased in four organs and decreased in five. UA-GalNAc6S was decreased in six organs and did not increase in any organ. The prominent UA-GalNAc4S varied significantly only slightly with aging but decreased in spleen. Collectively, the disulfated UA-GalNAc4S,6S and UA2S-GalNAc4S were increased with aging, which led to the overall increase of disulfated structures in 21 of the 24 organs (Figure 3A, B and Supporting Table 3). The total sulfation density was statistically increased in 14 organs, but decreased in brain and colon. Although structural changes were observed in most of the organs with aging, the organ-specific patterns of CS/DS structure were similar between the young and aged organs, supporting the concept that there is a restrict structure-function correlation in a given organ which is maintained throughout life.

Characteristic Disaccharide Composition of HS From Different Organs of Young Mice

For HS analysis, aliquots of purified GAGs were exhaustively degraded with a mixture of heparinases I, II, and III. The resultant disaccharides were analyzed by the same procedure as for CS/DS. Eight disaccharides were identified including one non-, three mono-, three di- and one trisulfated disaccharides (Figure 1 B1 and B2). As for the chondroitinase degradation of CS/DS, heparinase degradation of HS also led to loss of the epimeric status of glucuronic or iduronic which is present in native chains. Figure 4 and Supporting Table 4 show the relative molar proportion of each disaccharide calculated from peak area of a given species divided by the total area of all eight peaks.

[IMAGE OMITTED. SEE PDF]

In the context of young organs, the average of the CV of each organ was equal or less than 4% for the most abundant structures of UA-GlcNAc, UA-GlcNS, UA2S-GlcNS6S, UA-GlcNAc6S and UA2S-GlcNS; the CV for the structures UA-GlcNS6S and UA2S-GlcNAc was between 5% and 8%, and for the least predominant UA2S-GlcNAc6S disaccharide was 12% (data not shown). Analytical variability was assessed using the same procedure as for CS/DS, showing inverse correlation of variation with abundance (Supporting Table 1). CVs of the different disaccharides were similar to the CVs found in the analysis of the organs. Thus, we can conclude that a substantial degree of the registered variability in HS analysis was due to the analytical method.

The most common structure detected is nonsulfated UA-GlcNAc disaccharide (Figure 4A) which constitutes the nonsulfated domains in the native HS chains, ranging between 40% and 60% in all young organs, except for four organs (cartilage: 65%; kidney: 37%; skin: 31%; spleen: 39%). The mean and median of all young organs = 49%; range 31%–65%. UA2S-GlcNAc disaccharide (Figure 4B) was between 1% and 3% in all organs, except in lung (5%) and thymus (4%). The mean of all young organs = 2.3%; median = 2.1%; range 1.3%–5.3%. UA-GlcNAc6S (Figure 4C) was between 4% and 8% in 17 organs, whereas cartilage contained less and, adrenal gland, kidney, liver, salivary gland, spleen, and thymus contained more. The mean of all young organs = 7.0; median = 6.3; range 3.3%–16%. The UA2S-GlcNAc6S (Figure 4D) represented the least present species; 21 organs had ≤ 1.0% but was higher in liver, spleen, and thymus. The mean of all young organs = 0.8%; median = 0.7%; range 0.3%–2.7%. The UA-GlcNS structure (Figure 4E) ranged between 16% and 24% in all organs, but was lower in skin and spleen. The mean and median of all young organs = 20%; range 13%–24%. UA2S-GlcNS (Figure 4F) was between 6% and 13% in all organs but higher in brain, lung and testicles. The mean and median of all young organs = 11%; range 6.2%–17%). Almost all organs contained between 2% and 4% of UA-GlcNS6S (Figure 4G) but was higher in adrenal gland, brain, kidney, liver, salivary gland, skin and spleen. The mean of all young organs = 4.4%; median = 3.3; range 1.8%–12%. Lastly, the trisulfated UA2S-GlcNS6S (Figure 4H) was between 3% and 7% in almost all organs, but higher in kidney, liver (29%), skin, and stomach. The mean of all young organs = 6.5%; median = 4.7; range 3.2%–29%.

The proportions of the position of sulfate substitutions are highlighted in Figure 5 and Supporting Table 5. N-sulfation of the glucosamine residue occurred in 30%–49% of the structures (Figure 5A) in all young organs, except skin (64%). The mean and median of all young organs = 41%; range 30%–64%. O-sulfation in 2-O position of hexuronic acid (Figure 5B), either glucuronic or iduronic in the native chains, occurred in 13–27% of the disaccharides in all organs but skin (37%). The mean of all young organs = 20%; median = 19%; range 13%–37%. 6-O-sulfation of glucosamine (Figure 5C) amounted to 9–20% in most organs but was higher in the adrenal gland, kidney, liver, skin, spleen and thymus. The mean of all young organs = 19%; median = 16%; range 9%–45%. Notably, skin showed the highest average sulfation degree (1.5 sulfates/disaccharide), whereas almost all organs had a sulfation degree between 0.7 and 1.0 sulfates/disaccharide except cartilage (0.53 sulfates/disaccharide) and muscle, which had less (Figure 5D). The mean and median of all young organs = 0.80; range 0.53–1.50. Figure 5E reports the relative distribution of 2-O sulfate within the different disaccharides in young organs. 2-O sulfation predominated when at the nonreducing end of GlcN-sulfated ± 6-O sulfated structure and occurred to much less extent in GlcN-acetylated ± 6-O sulfated structures. UA2S-GlcNS predominated over UA2S-GlcNS6S in all organs, except the kidney, liver and skin. Figure 5G reports the relative distribution of 6-O sulfate groups in the different disaccharides. The structures UA2S-GlcNS6S and UA-GlcNAc6S were equally predominant in all organs, but in skin and stomach where UA2S-GlcNS6S was more predominant and in spleen where UA-GlcNAc6S was predominant. The structures UA-GlcNS6S and UA2S-GlcNAc6S were the least common in all organs.

[IMAGE OMITTED. SEE PDF]

Altogether, unlike CS/DS, the variations of HS disaccharides in different organs were less pronounced. Although the overall difference of disaccharide composition from different organs is moderate, the distinctly higher levels of sulfated structures in the kidney, liver, lung, spleen and skin should be noted. Especially the HS from skin has a significantly higher proportion of disulfated (UA-GlcNS6S) and trisulfated (UA2S-GlcNS6S) disaccharides and has the highest overall N-, 2-O, and 6-O sulfation degree (Figure 5 and Supporting Table 5), even higher than kidney, liver, and lung where the important biological roles of HS has been most studied.

Age-Associated Changes of HS Fine Structure

Comparison of the data from young and aged organs showed statistically significant changes (Figures 4 and 5 and Supporting Table 4 and 5). However, these changes were minor in most cases, in contrast with CS/DS. Collectively, the sum of 6-O sulfate-containing disaccharides were notably increased with aging in 10 organs and decreased in none; the 2-O-sulfated-containing disaccharides were increased with aging in five organs and decreased in one and N-sulfated disaccharides increased with aging in six organs and decreased in one (Figure 5A-C). As a result, total sulfation increased in 10 organs and decreased in two (Figure 5D). Regardless of the fine structural changes in some organs with aging, similarly to CS/DS, the organ-specific characteristics of HS structure were unchanged with aging.

Amount and Proportion of GAGs in Different Organs and Age-Related Alterations

Finally, we assessed the amount of each type of GAGs by comparing the peak area of the fluorescent disaccharides to the area of standard disaccharides of known amount and summing all disaccharide species of each GAG (Figure 6 and Supporting Table 6 and 7). The sensitivity of the method allowed reliable quantification of the disaccharide species from all collected organs, including the tiny organs. The coefficients of variation of the amount, as expressed as ng GAG/mg dry tissue, was less than 30% in most young organs (data not shown).

[IMAGE OMITTED. SEE PDF]

The total GAGs in the young organs ranged widely from the highest 3026 ng/mg dry tissue in the cartilage to the lowest 236 ng/mg in the seminal gland (Figure 6A). Eight organs (bladder, brain, cartilage, eye, lung, skin, stomach and uterus) had more than 1000 ng/mg. CS/DS varied from the highest 2755 ng/mg in the cartilage to the lowest 172 ng/mg in the thymus (Figure 6B). HS varied between 100 and 300 ng/mg in 17 organs, with lower levels in cartilage, lymph node, muscle (as low as 50 ng/mg), seminal gland and higher levels in bladder, kidney, and lung (the highest 829 ng/mg) (Figure 6C). HA was the least present GAG, mostly ≤ 100 ng/mg, ranging from 3 ng/mg in the liver to 534 ng/mg in skin (Figure 6D). The relative proportion of the different GAGs in the organs of young mice is shown in Figure 6E. CS/DS constituted more than 50% of total GAGs in 13 young organs, with the highest proportion (91%) in the cartilage. The mean and median of all young organs = 53%; range 23%–91%. HS was more than CS/DS in adrenal gland, kidney (75%), lung (61%), pancreas. The mean of all young organs = 30%; median = 28%; range 2%–75%. HA varied the most, ranging from 1% in liver to 47% in muscle. Seventeen organs contained ≤ 20% of HA, but lymph nodes, muscle, pancreas, salivary gland, seminal gland, skin and stomach contained more. The mean and median of all young organs = 17%; range 1%–47%.

Aging dramatically increased GAGs in kidney and spleen due to solely increasing of HS in kidney and both CS/DS and HS in spleen (Figure 6A–D). Total GAGs increased significantly in aged colon, intestine, and testicles but decreased in brain, eye, and pancreas. In addition, HS level was decreased in the aged brain, eye, lung, and increased in colon, heart, and spleen. Interestingly, HA is reduced in six organs, with the most striking reduction in skin, and increased in spleen.

The relative proportions between CS/DS and HS in young and aged heart, kidney, liver, lung was similar when derived either from the in vivo ³⁵S-sulfate labeled GAGs (data calculated from Figure 7) or from AMAC-labeled disaccharides (Figure 6E-F and Supporting Table 7). On the contrary, in the brain HS was 43% (young) and 42% (aged) of the total ³⁵S-labeled GAGs (CS/DS plus HS), but only 10% (young) and 6% (aged) calculated from the disaccharides fingerprint. This may indicate a faster turnover of HS compared to CS/DS in brain.

[IMAGE OMITTED. SEE PDF]

CS/DS and HS Chain Length Varies in the Different Organs But Does Not Change With Aging

To determine whether GAG chain length changed with aging, we analyzed in vivo ³⁵S-sulfate labeled GAGs from selected organs of two young and two aged mice. The labeled total GAGs were divided into two portions, treated with nitrous acid at pH 1.5 or chondroitinase ABC for HS and CS/DS degradation, respectively. The resulting materials were analyzed on a Superose 6 gel size column. Figure 7A shows that the molecular size of CS/DS differed in the different organs. Comparison of the profiles revealed that the predominant chains are longer in liver and spleen (~50 kDa; spleen has a second ~20 kDa peak) and shorter in brain, heart, and lung (~20 kDa); CS/DS chains were the shortest in kidney (~15 kDa). HS chains (Figure 7B) were in general shorter than CS/DS. The kidney, liver and spleen had a predominant peak equivalent to ~20 kDa (liver had a second population of ~10 kDa); HS chains from brain, heart, and lung were ~15 kDa long. There was no difference in the chain length of either CS/DS or HS from the organs of aged vs young mice because the chromatograms of all samples analyzed are superimposable or extremely similar.

Discussion

Advances in genomic and proteomic analytical methods have significantly enhanced our knowledge of the biological implications of nucleic acids and proteins. However, our understanding on the structural properties of GAGs lags behind, due to lacking an equivalent robust methodology of sequencing. Currently, the information on GAGs' structure relies on the analysis of their compositions and fragmented sequence information [13]. So far, one commonly used and reliable method is to determine the disaccharide species in a given population of GAGs, by which scattered data have been reported for individual organs. To get a comprehensive picture, we analyzed 192 samples prepared from 24 murine young and aged organs, with regarding to the disaccharide composition and distribution of different species of GAGs.

Adding up the amount of each disaccharide from CS/DS, HA, and HS showed the proportions of each GAG species, and a distinctive distribution in the organs. CS/DS is predominant (over 40% of total GAGs) in most of the organs, and comprising about 90% in cartilage, amounting 0.3% of dry weight of the tissue. CS/DS comprised almost 80% of GAGs in the brain; HS, less than 10%. HS is much less than CS/DS in the cartilage (2% of GAGs), but 75% in the kidney. The young tissues with highest HS amount were lung and kidney. HS amount in our young mice was in excellent agreement with the results reported for ten mouse tissue (figure 2 in ref [5]). Lung and kidney are highly vascularized organs. Endothelial glycocalyx is a specialized structure whose alteration plays a major role in physiological and pathological conditions [14–17]. One of the major components of endothelial glycocalyx are HS proteoglycans and it is therefore tempting to speculate that the high HS content found in lung and kidney reflects the glycocalyx abundance. In comparison, HA comprised an overall small portion of GAGs (less than 20%) in most organs and showed the most organ to organ variability: only 1–3% in the liver, kidney, spleen and testicle but almost 50% in the muscle, in agreement with [18]. In muscle HA influences muscle pain and stiffness both via its mechanical and signaling properties [19]. It is known that skin contains high level of HA [20, 21] and our analysis confirmed this (about 500 ng/mg dry tissue, constituting 30% of total GAGs). The biology of HA in skin is well established [22] as are its pharmacological/carrier/cosmetic use [23]. Our approach is global, that is, whole organs were analyzed. With the exception of cartilage, which is almost pure extra cellular matrix (ECM) without blood vessels, all other organs have a mixture of ECMs (including basement membranes), blood vessels, nerves, fat and parenchyma (epithelial cells mostly). Part of the observed differences in GAG levels may be due to the relative distribution of such sub-organ structures. Brain constitutes a special case with an ECM basically consisting of CS/DS. Muscle has a lot of blood vessels but also lots of loose and structured connective tissue with collagen, HA, and CS/DS. Liver has only 3% of ECM [24] but functions as a large blood reservoir and HSPGs are abundant in the space of Disse that separate hepatocytes from the liver sinusoidal compartment.

The effect of aging per se, that is without any overtly visible concomitant disease, on GAG structure and composition is scantly reported and is addressed in our study. Although the global organ structures of GAGs remain unaltered with aging, one overall change is increased sulfation degree of the GAGs in most of the aged organs. This may represent a compensation for differences in amount and/or distribution of the GAGs binding proteins because the binding of majority protein ligand to GAGs requires sufficient and organized sulfation [25]. For instance, HS 6-O-sulfation increased in ten organs in our analysis. With the same trend, the trisulfated UA2S-GlcNS6S structure was found to positively correlate with aging in human aorta HS [26]. Further, HS 6-O-sulfation was found to increase in aged rat myocardium [27].

Aging in mice is associated with increased body weight, particularly in kidney, liver, and spleen, increased fat deposition and fibrosis in certain organs, and decreased weight in some areas of the brain [28, 29]. Clearly, these changes could affect GAG content and composition. We could observe in our organ dissection increased adiposity of salivary glands and seminal glands. In addition, one of the eight aged kidneys was pale, a sign of a pathological process possibly ending with kidney failure. Aging affects all organs and tissues to varying extents. Single-cell analyses have showed aging-associated alterations in gene expression and the proteome, for example, upregulation of senescence-associated genes [30]. For instance, from a cellular/histological point of view, in the murine kidney the relative abundance of loop of Henle epithelial cells, mesangial cells and capillary endothelial cells decreases, whereas the glomerular basement membrane thickens, the glomerular capillaries dilate and vacuoles accumulate in epithelial cells under the renal cortex [31]. We detected a remarkable increase of HS in aged kidneys, which could be the result of histological changes.

GAGs constitute a considerable part in the brain extracellular matrix (ECM) having critical roles in maintaining the structure and integrity of diffuse ECM and perineural nets, affecting activities such as plasticity, memory, and axonal regeneration [32]. The prevalence of CS/DS in the brain (79%–73% of total GAGs in adult and aged brains, respectively) was a confirmation of several reports [33, 34]. Surprisingly, both CS/DS and HS reduced in the aged brain. It is known that brain atrophy is an age-related phenomenon, so the age-dependent reduction of the GAGs level may be responsible for the degeneration of the brain parenchyma with aging [35]. Our analysis was done at the whole organ level. When CS/DS from various regions of mouse brain was analyzed, and at different developmental stages, amount and structure differed considerably [36]. Reportedly, CS/DS 6-O sulfated structures decreased in aged rat brain [37], which aligns in our analysis with decreased disulfated UA2S-GalNAc6S. Scanty data report quantitative differences of HS in aging brain. For instance, HS amount decreased considerably in aged rat striatum, with no structural differences [38]. These observations open several questions, for instance if CS/DS decrease with aging is due to reduced synthesis or elevated catabolism.

Immunosenescence is an organism-wide aging process which leads to chronic inflammation and subsequent tissue damage and dysfunction [39]. Reportedly, neutrophils of aged mice that adhere to vascular endothelium in inflamed tissue exhibit a high grade of reverse transendothelial migration [40]. Instead of exerting their function at the inflamed site, activated neutrophils migrated to the lungs and caused vascular leakage and tissue damage [40]. The reverse transendothelial migration was attributed to tissue alterations and not to the neutrophils. This leads us to speculate whether the increased sulfation and/or altered disaccharide proportions of HS and CD/DS often observed in organs of aged mice may play a role in immunosenescence. GAGs present on endothelial cells and in the extracellular matrix interact with chemokines leading to generation of concentration gradients guiding leucocytes during their migration process [41]. If the changes in GAGs negatively affects the chemokine-GAG interactions, this could ultimately lead to alterations of leucocyte migration. Future studies are thus needed to shed light on the role of GAGs in immune dysfunction associated with aging.

GAG disaccharide fingerprint is obtained by different methodologies. One of the most popular is chain digestion by lyases to disaccharides followed by HPLC (our report) or capillary electrophoresis. Mass spectrometry detection is also used [42]. Results variability stems from different extraction, enzymatic digestion, disaccharide separation and detection protocols. Not surprisingly the values reported in the literature varies to a certain extent. The introduction of reference GAG samples could facilitate the comparison of the different methods and the obtained results. Nevertheless, our comparison of different samples analyzed using the same method in a single set of experiments holds firmly.

In general, our reported CS/DS and HS structures of murine organs compare well, but not perfectly, with the ones reported in literature. Brain CS/DS (figure 6 in [36]) and skin CS/DS (table II in [43]) structures were reported after differential splitting with chondroitinases AC and ABC, followed by HPLC separation of fluorescently labeled disaccharides. HS and CS structures of mouse heart using mass spectrometry of fluorescently labeled disaccharides were studied (figure 7 in [44]). The structure of HS from ten organs was studied using a methodology similar, but not identical, to ours (see figure 3 and table I in [5]).

Our GAG data complement the available expression data of HS and CS/DS biosynthetic enzymes. For instance RNA-seq and RT-qPCR data were presented for 20 HS biosynthetic enzymes and 20 HSPG in 21 mouse tissues [45]. Further, the expression data available in GTEx database of HS and CS/DS biosynthetic enzymes in 37 human tissues were reported [46].

In summary, the systemic analysis of disaccharides from as many as 24 organs of young and aged mice provides a global picture of GAG expression and disaccharide composition. The study has several relevant and important findings that warrant further investigation in human tissues.

Author Contributions

Emil Tykesson and Marco Maccarana conceived the study and analyzed the data. Emil Tykesson, Malin Eriksson, and Marco Maccarana performed experiments, Marco Maccarana supervised the overall study. All authors interpreted data and wrote the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

We are grateful to Kristofer Rubin for valuable comments.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

L. Kjellén and U. Lindahl, “Specificity of Glycosaminoglycan‐Protein Interactions,” Current Opinion in Structural Biology 50 (2018): 101–108.

J.‐P. Li, F. Gong, Å. Hagner‐McWhirter, et al., “Targeted Disruption of a Murine Glucuronyl C5‐Epimerase Gene Results in Heparan Sulfate Lacking L‐Iduronic Acid and in Neonatal Lethality,” Journal of Biological Chemistry 278, no. 31 (2003): 28363–28366.

S. Mizumoto and S. Yamada, “Congenital Disorders of Deficiency in Glycosaminoglycan Biosynthesis,” Frontiers in Genetics 12 (2021): [eLocator: 717535].

S. Mizumoto, S. Yamada, and K. Sugahara, “Human Genetic Disorders and Knockout Mice Deficient in Glycosaminoglycan,” BioMed Research International 2014 (2014): 1–24.

J. Ledin, W. Staatz, J. P. Li, et al., “Heparan Sulfate Structure in Mice With Genetically Modified Heparan Sulfate Production,” Journal of Biological Chemistry 279, no. 41 (2004): 42732–42741.

M. Ennemoser, A. Pum, and A. Kungl, “Disease‐Specific Glycosaminoglycan Patterns in the Extracellular Matrix of Human Lung and Brain,” Carbohydrate Research 511 (2022): [eLocator: 108480].

M. L. Escobar Galvis, J. Jia, X. Zhang, et al., “Transgenic or Tumor‐Induced Expression of Heparanase Upregulates Sulfation of Heparan Sulfate,” Nature Chemical Biology 3, no. 12 (2007): 773–778.

M. B. Huynh, N. Rebergue, H. Merrick, et al., “HS3ST2 Expression Induces the Cell Autonomous Aggregation of Tau,” Scientific Reports 12, no. 1 (2022): [eLocator: 10850].

M. B. Huynh, et al, “Glycosaminoglycans From Aged Human Hippocampus Have Altered Capacities to Regulate Trophic Factors Activities But not Abeta42 Peptide Toxicity,” Neurobiology of Aging 33, no. 5 (2012): [eLocator: 1005].

T. Bitter and H. M. Muir, “A Modified Uronic Acid Carbazole Reaction,” Analytical Biochemistry 4 (1962): 330–334.

N. Volpi, F. Galeotti, B. Yang, and R. J. Linhardt, “Analysis of Glycosaminoglycan‐Derived, Precolumn, 2‐aminoacridone‐labeled Disaccharides With LC‐Fluorescence and LC‐MS Detection,” Nature Protocols 9, no. 3 (2014): 541–558.

J. E. Shively and H. E. Conrad, “Nearest Neighbor Analysis of Heparin: Identification and Quantitation of the Products Formed by Selective Depolymerization Procedures,” Biochemistry 15 (1976): 3943–3950.

S. Perez, O. Makshakova, J. Angulo, et al., “Glycosaminoglycans: What Remains To Be Deciphered?,” JACS Au 3, no. 3 (2023): 628–656.

D. Pretorius, R. P. Richter, T. Anand, J. C. Cardenas, and J. R. Richter, “Alterations in Heparan Sulfate Proteoglycan Synthesis and Sulfation and the Impact on Vascular Endothelial Function,” Matrix Biology Plus 16 (2022): [eLocator: 100121].

S. Gaudette, D. Hughes, and M. Boller, “The Endothelial Glycocalyx: Structure and Function in Health and Critical Illness,” Journal of Veterinary Emergency and Critical Care 30, no. 2 (2020): 117–134.

F. E. Curry, “The Molecular Structure of the Endothelial Glycocalyx Layer (EGL) and Surface Layers (ESL) Modulation of Transvascular Exchange,” Advances in Experimental Medicine and Biology 1097 (2018): 29–49.

S. Reitsma, D. W. Slaaf, H. Vink, M. A. M. J. van Zandvoort, and M. G. A. oude Egbrink, “The Endothelial Glycocalyx: Composition, Functions, and Visualization,” Pflügers Archiv—European Journal of Physiology 454, no. 3 (2007): 345–359.

K. Watanabe, A. Oohira, I. Uramoto, and T. Totsuka, “Age‐Related Changes in the Content and Composition of Glycosaminoglycans Isolated From the Mouse Skeletal Muscle,” Journal of Biochemistry 100, no. 1 (1986): 167–173.

A. Amir, S. Kim, A. Stecco, M. P. Jankowski, and P. Raghavan, “Hyaluronan Homeostasis and Its Role in Pain and Muscle Stiffness,” PM&R 14, no. 12 (2022): 1490–1496.

J. R. E. Fraser, T. C. Laurent, and U. B. G. Laurent, “Hyaluronan: Its Nature, Distribution, Functions and Turnover,” Journal of Internal Medicine 242, no. 1 (1997): 27–33.

D. H. Lee, J. H. Oh, and J. H. Chung, “Glycosaminoglycan and Proteoglycan in Skin Aging,” Journal of Dermatological Science 83, no. 3 (2016): 174–181.

J. Muto, K. Sayama, R. L. Gallo, and K. Kimata, “Emerging Evidence for the Essential Role of Hyaluronan in Cutaneous Biology,” Journal of Dermatological Science 94, no. 1 (2019): 190–195.

K. Valachová, M. E. Hassan, and L. Šoltés, “Hyaluronan: Sources, Structure, Features and Applications,” Molecules 29, no. 3 (2024): [eLocator: 739].

P. Bedossa and V. Paradis, “Liver Extracellular Matrix in Health and Disease,” Journal of Pathology 200, no. 4 (2003): 504–515.

U. Lindahl and J. P. Li, “Interactions Between Heparan Sulfate and Proteins‐Design and Functional Implications,” International Review of Cell and Molecular Biology 276 (2009): 105–159.

E. Feyzi, T. Saldeen, E. Larsson, U. Lindahl, and M. Salmivirta, “Age‐Dependent Modulation of Heparan Sulfate Structure and Function,” Journal of Biological Chemistry 273, no. 22 (1998): 13395–13398.

M. B. Huynh, C. Morin, G. Carpentier, et al., “Age‐Related Changes in Rat Myocardium Involve Altered Capacities of Glycosaminoglycans to Potentiate Growth Factor Functions and Heparan Sulfate‐Altered Sulfation,” Journal of Biological Chemistry 287, no. 14 (2012): 11363–11373.

J. W. Willows, M. Robinson, Z. Alshahal, et al., “Age‐Related Changes to Adipose Tissue and Peripheral Neuropathy in Genetically Diverse HET3 Mice Differ by Sex and Are Not Mitigated by Rapamycin Longevity Treatment,” Aging cell 22, no. 4 (2023): [eLocator: e13784].

M. Lessard‐Beaudoin, M. Laroche, M. J. Demers, G. Grenier, and R. K. Graham, “Characterization of Age‐Associated Changes in Peripheral Organ and Brain Region Weights in C57BL/6 Mice,” Experimental Gerontology 63 (2015): 27–34.

N. Almanzar, J. Antony, A. S. Baghel, et al., “A Single‐Cell Transcriptomic Atlas Characterizes Ageing Tissues in the Mouse,” Nature 583, no. 7817 (2020): 590–595.

J. Jin, X. Yang, H. Gong, and X. Li, “Time‐ and Gender‐Dependent Alterations in Mice During the Aging Process,” International Journal of Molecular Sciences 24, no. 16 (2023): [eLocator: 12790].

J. W. Fawcett and J. C. F. Kwok, “Proteoglycan Sulphation in the Function of the Mature Central Nervous System,” Frontiers in Integrative Neuroscience 16 (2022): [eLocator: 895493].

L. Djerbal, H. Lortat‐Jacob, and J. Kwok, “Chondroitin Sulfates and Their Binding Molecules in the Central Nervous System,” Glycoconjugate Journal 34, no. 3 (2017): 363–376.

S. S. Deepa, D. Carulli, C. Galtrey, et al., “Composition of Perineuronal Net Extracellular Matrix in Rat Brain,” Journal of Biological Chemistry 281, no. 26 (2006): 17789–17800.

M. Veldsman, “Brain Atrophy Estimated From Structural Magnetic Resonance Imaging as a Marker of Large‐Scale Network‐Based Neurodegeneration in Aging and Stroke,” Geriatrics 2, no. 4 (2017): [eLocator: 34].

C. Akatsu, S. Mizumoto, T. Kaneiwa, et al., “Dermatan Sulfate Epimerase 2 Is the Predominant Isozyme in the Formation of the Chondroitin Sulfate/Dermatan Sulfate Hybrid Structure in Postnatal Developing Mouse Brain,” Glycobiology 21, no. 5 (2011): 565–574.

S. Foscarin, R. Raha‐Chowdhury, J. W. Fawcett, and J. C. F. Kwok, “Brain Ageing Changes Proteoglycan Sulfation, Rendering Perineuronal Nets More Inhibitory,” Aging 9, no. 6 (2017): 1607–1622.

R. Raghunathan, N. K. Polinski, J. A. Klein, et al., “Glycomic and Proteomic Changes in Aging Brain Nigrostriatal Pathway,” Molecular & Cellular Proteomics 17, no. 9 (2018): 1778–1787.

J. Nikolich‐Žugich, “The Twilight of Immunity: Emerging Concepts in Aging of the Immune System,” Nature Immunology 19, no. 1 (2018): 10–19.

A. Barkaway, L. Rolas, R. Joulia, et al., “Age‐Related Changes in the Local Milieu of Inflamed Tissues Cause Aberrant Neutrophil Trafficking and Subsequent Remote Organ Damage,” Immunity 54, no. 7 (2021): 1494–1510.e7.

H. Crijns, V. Vanheule, and P. Proost, “Targeting Chemokine‐Glycosaminoglycan Interactions to Inhibit Inflammation,” Frontiers in Immunology 11 (2020): [eLocator: 483].

A. Zappe, R. L. Miller, W. B. Struwe, and K. Pagel, “State‐Of‐The‐Art Glycosaminoglycan Characterization,” Mass Spectrometry Reviews 41, no. 6 (2022): 1040–1071.

T. Hirose, S. Mizumoto, A. Hashimoto, et al., “Systematic Investigation of the Skin in Chst14‐/‐ Mice: A Model for Skin Fragility in Musculocontractural Ehlers‐Danlos Syndrome Caused by CHST14 Variants (mcEDS‐CHST14),” Glycobiology 31, no. 2 (2021): 137–150.

H. Luong, S. Singh, M. Patil, and P. Krishnamurthy, “Cardiac Glycosaminoglycans and Structural Alterations During Chronic Stress‐Induced Depression‐Like Behavior in Mice,” American Journal of Physiology‐Heart and Circulatory Physiology 320, no. 5 (2021): H2044–H2057.

S. Moon and Y. T. Zhao, “Spatial, Temporal and Cell‐Type‐Specific Expression Profiles of Genes Encoding Heparan Sulfate Biosynthesis Enzymes and Proteoglycan Core Proteins,” Glycobiology 31, no. 10 (2021): 1308–1318.

A. Basu, N. G. Patel, E. D. Nicholson, and R. J. Weiss, “Spatiotemporal Diversity and Regulation of Glycosaminoglycans in Cell Homeostasis and Human Disease,” American Journal of Physiology‐Cell Physiology 322, no. 5 (2022): C849–C864.