SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Objective

To determine whether the development of osteoarthritis (OA) in men over a 33-year period is related to lower sulfate levels in stored serum collected during that time interval.

Methods

Stored serum samples from participants in the Veterans Administration Normative Aging Study were assayed for sulfate by ion-exchange chromatography. Samples had been obtained every 3–5 years during part or all of a 33-year portion of the study. Sulfate levels were determined in serum from all participants who underwent knee replacement surgery and had evidence of radiographic hand OA, from some of the participants who had evidence of radiographic hand OA but had not undergone knee replacement surgery, from all participants who underwent knee replacement surgery but had no evidence of radiographic hand OA, and from age-matched participants who had no evidence of OA by history, physical examination, or hand radiography.

Results

Serum sulfate levels in participants, with or without radiographic hand OA and/or knee replacements, who were ages 34–72 years at the first examination, ranged from 0.21 mM to 0.51 mM over the course of a maximum of 33 years. Both the overall mean and median sulfate levels rose from 0.32 mM at age 40–50 years to 0.38 mM at age 70–80 years, and the overall mean and median for all ages was 0.36 mM. There were no significant differences in sulfate levels between subjects in any of the 4 groups.

Conclusion

There was no evidence of a relationship between these serum sulfate levels and the development of OA. However, all samples were collected after overnight fasting, and no participant was younger than age 34 years at the initiation of the study. It remains to be determined whether differences in the time of ingestion of daily dietary protein providing sulfate are related to the development of OA, or whether sulfate levels measured at an earlier age could be a factor.

Proteochondroitin sulfate, primarily aggrecan, plays a major role in the mechanical support of cartilage. In addition to assisting in the positioning and orientation of collagen, it determines salt and water distribution and controls a volume domain of water many times the volume of the proteoglycan itself. This provides a major degree of cushioning, so that water is expressed from the domain under pressure and then can return when the pressure is released. These functions are dependant on the high charge of the sulfate substituents; therefore, any decrease in sulfation might be expected to affect the structure and stability of the cartilage. In addition, undersulfation might increase susceptibility to animal chondroitin-degrading enzymes, because these enzymes degrade much more readily whenever sulfate is absent from the adjacent sugars of chondroitin (1, 2). Consequently, undersulfation could produce a functional insufficiency in cartilage that in turn could relate to the onset and severity of osteoarthritis (OA).

The metabolism of chondroitin sulfate has been examined extensively in our laboratory, with a particular emphasis on where, how, and when sulfation takes place in relation to chondroitin chain formation under normal conditions and under conditions of low sulfate availability. Using cultures of various animal cell types or tissue explants, we (3–6) and other investigators (7–10) have found that sulfation of chondroitin is usually lowered when sulfate concentrations in medium are <0.2 mM. Normal fasting sulfate levels in humans are lower than those in any animals that have been tested (11). In adults, normal fasting sulfate levels have generally been reported to average between 0.25 mM and 0.4 mM (11–13), with a few individuals reported to have levels as low as 0.1 mM (12), suggesting a potential for chondroitin undersulfation. Furthermore, it has been suggested that the possible benefit of glucosamine sulfate and chondroitin sulfate in the treatment of OA (14, 15) may be attributable to supplementation with sulfate rather than glucosamine or chondroitin.

In order to investigate whether lower sulfate levels are related to the development of OA, we used stored fasting serum samples obtained over a period of up to 33 years from participants of the Normative Aging Study (16) initiated by the Boston Veterans Administration Outpatient Clinic. Serum sulfate was measured in stored samples that had been obtained from all participants who underwent knee replacement surgery between 1967 and 2000 (those with and those without evidence of radiographic hand OA) and from some participants who had not undergone knee replacement surgery during that same period but had evidence of OA by history, physical examination, and radiography. Results were compared with the serum sulfate levels in participants who had no evidence of the development of OA, by history, physical examination, or radiography. The mean and median levels of sulfate were slightly higher than those previously reported (12), with a gradual increase with age. There were no differences in serum sulfate levels between normal subjects and those in whom evidence of OA developed.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Assay materials.

Ultrapure water (>18 Mohm/cm) containing no sulfate was obtained with a Simplicity 185 water treatment system (Millipore, Bedford, MA). Chemical-grade sodium carbonate, sodium bicarbonate, sodium chloride, potassium sulfate, and D-(+)-glucose were obtained from Sigma (St. Louis, MO). Chemical-grade sulfuric acid and acetonitrile were obtained from Fisher Scientific (Pittsburgh, PA). Sodium hydroxide, 5N, was obtained from VWR (Bridgeport, NJ).

Examination of participants.

The Normative Aging Study was initiated at the Boston Veterans Administration Outpatient Clinic in 1963, with eventual recruitment of 2,280 normal, healthy men (almost all of whom were veterans), in order to examine changes in various normal and pathologic processes associated with aging. Upon recruitment (between 1963 and 1970) and every 3–5 years thereafter, thorough histories were obtained and the participants underwent general physical examinations by internist physicians. At each visit a serum sample was obtained for assay of a variety of components; multiple aliquots of each serum sample were frozen and stored in several tubes for future use. Chest and hand radiographs were obtained and were interpreted by radiologists. All data and diagnoses were recorded for future use. For the current study, an aliquot of the stored serum was thawed and assayed for sulfate, chloride, and glucose. The diagnosis of radiographic hand OA was made by a radiologist, and a description was recorded at the time of each examination; the diagnosis was confirmed by physical examination and was recorded by the examining physician.

Sulfate analyses by ion-exchange chromatography.

Ion-exchange chromatography was performed using a Metrohm 761 Compact IC equipped with an ion-exchange Metrosep A Supp 5 (150 × 4.0 mm) analytic column (Metrohm-Peak, Houston, TX), a suppressor, and a 20-μl sample loop. The 761 Compact IC eluent was carbonate/bicarbonate (3.2 mM Na2CO3, 1.0 mM NaHCO3; degassed for 1 hour) and 4.5% acetonitrile buffer, used as the mobile phase at a flow rate of 0.7 ml/minute at room temperature. The acetonitrile was added to the eluent in order to eliminate a peak preceding sulfate and to optimize good separation and quantification of the sulfate peak. Water and 100 mM H2SO4 were used at a flow rate of 0.6 ml/minute to regenerate the suppressor. Baseline was usually ∼13 μS/cm. The glucose analysis for sample control used a Metrohm-Peak 817 Bioscan equipped with an ion-exchange Metrosep Carb 1 column (250 × 4.5 mm, 5-μm pore size), a pulsed amperometric detector at 35°C, and a 20-μl sample loop. The 817 Bioscan eluent was 50 mM NaOH (degassed for 10 minutes and opened to the atmosphere for 1 hour) used as the mobile phase at a flow rate of 1.0 ml/minute. Baseline was usually ∼850 nA.

Aliquots of sera (20 μl) were diluted 100-fold in ultrapure water and injected using a Metrohm-Peak 766 autosampler. Chloride and sulfate concentrations were measured using the 761 Compact IC, while glucose concentrations were measured simultaneously in the same sample using the 817 Bioscan. The serum samples were also assayed for sodium, chloride, and glucose at the Bedford Veterans Administration Hospital clinical laboratory in order to compare the levels in sera obtained at each examination. This provided a means to correct for potential evaporation in the stored samples that yielded higher levels, and to determine potential bacterial contamination if glucose levels were substantially lower than the levels obtained when the sera were originally assayed. In a few samples, sulfate levels were adjusted accordingly for evaporation. Samples with glucose levels <3 mM (54 mg/dl), which were assumed to be attributable to bacterial contamination, were not included. Calibration standard curves were linear, with a relative standard deviation of 0–5% for each run; new curves were generated each day to ensure reproducibility and correct quantification.

Alternative method of sulfate analysis.

As an alternative analytic method, aliquots (50 μl) of sera were precipitated with trichloracetic acid, and a benzidine method was used to analyze the supernatants for sulfate (17).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Sulfate levels in 316 stored samples obtained from 91 participants of the Normative Aging Study over a 33-year period (1967–2000) were measured using the Metrohm-Peak ion-exchange chromatography method. The age of participants at the time of assay ranged from 34 years to 86 years. Participants were categorized into 4 groups, as follows: those with radiographic hand OA who had undergone knee replacement surgery (26 participants, 145 assays); those with radiographic hand OA who had not undergone knee replacement surgery (21 participants, 42 assays); those without radiographic hand OA who had undergone knee replacement surgery (10 participants, 41 assays); and those with no evidence of OA (34 participants, 88 assays). The mean and median sulfate levels were calculated for each individual, and overall values were calculated for each category. In addition, 186 samples obtained from 43 of these participants were analyzed by the benzidine method.

The serum sulfate levels (1–10 sera per participant) among individuals in all 4 groups, as determined by ion-exchange chromatography, varied considerably. Characteristics (age range, range of sulfate levels, and mean sulfate level) of the participants for whom ≥3 analyses were performed are shown in Table 1. The mean and median sulfate levels for all participants in each group, including those who underwent only 1 or 2 analyses, are shown in Table 2. The serum sulfate levels for all groups of participants, according to the age at which serum sampling occurred, can be seen in Figure 1. In all groups, a slight increase in sulfate levels was observed as participants aged, but no differences between groups were apparent. Results using the benzidine method were slightly lower than those obtained using Metrohm-Peak ion-exchange chromatography but were less consistent and less reliable; these results also showed no significant differences between groups.

Table 1. Characteristics of participants with 3 or more analyses*
Participant no.Age range, yearsNo. of seraMean (range) sulfate level, mM
  • *

    The age range represents the ages during which the subject was enrolled in the study. OA = osteoarthritis.

Radiographic hand OA, knee replacement   
 010456–7150.35 (0.29–0.37)
 021246–78100.42 (0.37–0.49)
 026037–6780.36 (0.29–0.41)
 057147–7370.38 (0.32–0.46)
 061653–7460.34 (0.23–0.43)
 061744–7290.37 (0.29–0.45)
 090872–8230.38 (0.37–0.40)
 091362–8060.45 (0.41–0.48)
 108454–7360.41 (0.37–0.44)
 120755–7770.36 (0.30–0.41)
 126543–6860.38 (0.31–0.45)
 133762–7830.36 (0.32–0.38)
 134056–7250.42 (0.37–0.47)
 142544–6750.36 (0.31–0.41)
 157955–7770.37 (0.26–0.48)
 160458–7140.30 (0.26–0.38)
 160549–6330.33 (0.26–0.38)
 171737–6950.25 (0.21–0.32)
 174638–6550.30 (0.26–0.33)
 194945–7180.36 (0.27–0.39)
 500254–7460.36 (0.29–0.42)
 506442–7390.28 (0.23–0.34)
 519449–7280.36 (0.28–0.47)
Radiographic hand OA, no knee replacement   
 085155–7030.35 (0.29–0.42)
 088045–6740.33 (0.32–0.36)
 111142–7140.34 (0.28–0.38)
 116270–7730.32 (0.26–0.44)
 130657–6530.29 (0.25–0.31)
 151147–7030.40 (0.38–0.43)
No radiographic hand OA, knee replacement   
 010658–6430.43 (0.42–0.46)
 015243–6050.39 (0.33–0.43)
 034355–7660.35 (0.25–0.49)
 044764–8630.40 (0.36–0.44)
 060360–6630.49 (0.46–0.50)
 091847–7590.33 (0.25–0.39)
 140345–7160.37 (0.29–0.41)
 198446–7830.33 (0.28–0.37)
No radiographic hand OA, no knee replacement   
 010554–6330.43 (0.38–0.49)
 025954–7140.35 (0.28–0.49)
 063555–7650.33 (0.29–0.36)
 085445–5640.29 (0.21–0.33)
 112536–6040.29 (0.26–0.34)
 117747–7560.29 (0.22–0.36)
 145147–6930.42 (0.36–0.50)
 160836–6330.32 (0.23–0.39)
 161941–7360.41 (0.31–0.46)
 190856–7460.42 (0.33–0.49)
 191751–6830.37 (0.33–0.42)
 194057–7230.36 (0.34–0.37)
 197746–6770.36 (0.25–0.45)
 509457–7240.37 (0.32–0.44)
Table 2. Sulfate levels in all groups of participants*
GroupNo. of subjectsNo. of sera assayedMean SO4 level, mMMedian SO4 level, mM
  • *

    OA = osteoarthritis.

Radiographic hand OA, knee replacement261450.360.37
Radiographic hand OA, no knee replacement21420.350.34
Knee replacement, no radiographic hand OA10410.370.38
No knee replacement, no radiographic hand OA34880.350.34
thumbnail image

Figure 1. Sulfate levels in stored serum samples. Diagonal lines were generated with Microsoft Excel to show the linear trend/regression over time.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In this study, the absence of differences in the fasting serum sulfate levels between normal subjects and those with OA suggests that sulfate levels may not be related to OA. However, this study was limited to examination of sulfate in serum that had been obtained from fasting participants who, thus, had ingested no protein for at least 12 hours. In general, the major supply of sulfate in humans is provided by metabolism of cysteine and methionine, and consequently intake of a diet deficient in protein might be expected to result in decreased sulfate levels. The serum sulfate level in humans has been shown to increase dramatically 3–8 hours after ingestion of a single large amount of protein (12); in rats, sulfate levels have been shown to drop 30–50% after an overnight fast (11). Feeding rats a diet deficient in protein for 4 days has been shown to reduce sulfate levels by 50% (18).

The liver has generally been considered the primary, if not the sole, location in which sulfate is formed from cysteine and methionine. However, investigators in our laboratory have previously demonstrated that other cells (19), including human skin fibroblasts from some individuals (5), can supply sulfate by this mechanism, but we have also shown that neither mouse chondrocytes (20) nor human chondrocytes (21) can do so. Recently, it was reported that chondroitin sulfate from bovine patellar OA cartilage is fully sulfated (22). However, a cow has been reported to have a serum sulfate level of 1.8 mM (11), 5-fold that of humans; therefore, any undersulfation related to sulfate levels in these animals would be highly unlikely. Moreover, bovine OA may not be reflective of OA in humans.

From this information, it can be assumed that a longer fast or a daily diet that did not contain protein until later in the day would result in lower sulfate levels. Thus, there could well be periods of time during which sulfate levels are low enough to result in undersulfation of proteochondroitin. Because the half-life of cartilage proteoglycan may be as long as 10 months (23), areas susceptible to damage could be formed intermittently and would remain for an extended period of time. In addition, the capability to produce sulfate from cysteine and methionine could vary and be lower in some individuals, as observed with human skin fibroblasts (5), thus providing a further potential for undersulfation. It is noteworthy that several drugs, including aspirin and acetaminophen, are metabolized for excretion by conjugation with sulfate. Use of these agents has been shown to lower sulfate levels in humans (13, 24–26) and in other animals as well (18, 27–29). For these reasons, it remains possible that sulfate levels in some people under specific circumstances of protein consumption and/or drug ingestion could affect OA, even though results of the current study have shown no such effect.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Hoffman P, Linker A, Meyer K. The acid mucopolysaccharides of connective tissues. II. Further experiments on chondroitin sulfate B. Arch Biochem Biophys 1957; 69: 43540.
  • 2
    Hoffman P, Linker A, Meyer K. The acid mucopolysaccharides of connective tissue. III. The sulfate linkage. Biochim Biophys Acta 1958; 30: 1845.
  • 3
    Humphries DE, Silbert CK, Silbert JE. Glycosaminoglycan production by bovine aortic endothelial cells cultured in sulfate-depleted medium. J Biol Chem 1986; 261: 91227.
  • 4
    Silbert JE, Palmer ME, Humphries DE, Silbert CK. Formation of dermatan sulfate by cultured human skin fibroblasts: effects of sulfate concentration on proportions of dermatan/chondroitin. J Biol Chem 1986; 261: 13397400.
  • 5
    Silbert CK, Humphries DE, Palmer ME, Silbert JE. Effects of sulfate deprivation on the production of chondroitin/dermatan sulfate by cultures of skin fibroblasts from normal and diabetic individuals. Arch Biochem Biophys 1991; 285: 13741.
  • 6
    Silbert JE, Sugumaran G, Cogburn JN. Sulphation of proteochondroitin and 4-methylumbelliferyl β-D-xyloside-chondroitin formed by mouse mastocytoma cells cultured in sulphate-deficient medium. Biochem J 1993; 296: 11926.
  • 7
    Sobue J, Takeuchi J, Ito K, Kimata K, Suzuki S. Effect of environmental sulfate concentration on the synthesis of low and high sulfated chondroitin sulfates by chick embryo cartilage. J Biol Chem 1978; 253: 61906.
  • 8
    Ito K, Kimata K, Sobue M, Suzuki S. Altered proteoglycan synthesis by epiphyseal cartilages in culture at low SOmath image concentration. J Biol Chem 1982; 257: 91723.
  • 9
    Van der Kraan PM, de Vries BJ, Vitters EL, van den Berg WB, van de Putte LB. The effect of low sulfate concentrations on the glycosaminoglycan synthesis in anatomically intact articular cartilage of the mouse. J Orthop Res 1989; 7: 64553.
  • 10
    Van der Kraan PM, Vitters EL, de Vries BJ, van den Berg WB. High susceptibility of human articular cartilage glycosaminoglycan synthesis to changes in inorganic sulfate availability. J Orthop Res 1990; 8: 56571.
  • 11
    Krijgsheld KR, Scholtens E, Mulder GJ. Serum concentration of inorganic sulfate in mammals: species differences and circadian rhythm. Comp Biochem Physiol 1980; 67: 6836.
  • 12
    Tallgren LG. Inorganic sulphates in relation to the serum thyroxine level and in renal failure. Acta Med Scand 1980; S640: 3444.
  • 13
    Hoffman DA, Wallace SM, Verbeeck RK. Circadian rhythm of serum sulfate levels in man and acetaminophen pharmacokinetics. Eur J Clin Pharmacol 1990; 39: 1438.
  • 14
    Hoffer LJ, Kaplan LN, Hamadeh MJ, Grigoriu AC, Baron M. Sulfate could mediate the therapeutic effect of glucosamine sulfate. Metabolism 2001; 50: 76770.
  • 15
    Cordoba F, Nimni ME. Chondroitin sulfate and other sulfate containing chondroprotective agents may exhibit their effects by overcoming a deficiency of sulfur amino acids. Osteoarthritis Cartilage 2003; 10: 22830.
  • 16
    Bosse R, Ekerdt DJ, Silbert JE. The Veterans Administration Normative Aging Study. In: MednickSA, HarwayM, FinelloKM, editors. Handbook of longitudinal research: teenage and adult cohorts. Vol 2. New York: Praeger; 1984. p. 27395.
  • 17
    De Vries BJ, Vitters E, van den Berg WB, Schram D, van de Putte LB. Determination of small quantities of sulfate (0-12 nmol) in serum, urine, and cartilage of the mouse. Anal Biochem 1987; 163: 40817.
  • 18
    Krijgsheld KR, Scholtens E, Mulder GJ. An evaluation of methods to decrease the availability of inorganic sulphate for sulphate conjugation in the rat in vivo. Biochem Pharmacol 1981; 30: 19739.
  • 19
    Humphries DE, Silbert CK, Silbert JE. Sulphation by cultured cells: cysteine, cyteinesulphinic acid and sulphite as sources for proteoglycan sulphate. Biochem J 1988; 252: 3058.
  • 20
    Mroz PJ, Silbert JE. Effects of [3H]glucosamine concentration on [3H]chondroitin sulphate formation by cultured chondrocytes. Biochem J 2003; 376: 5115.
  • 21
    Mroz PJ, Silbert JE. Use of 3H-glucosamine and 35S-sulfate with cultured human chondrocytes to determine the effect of glucosamine concentration on formation of chondroitin sulfate. Arthritis Rheum 2004; 50: 35749.
  • 22
    Lammi MJ, Qu CJ, Laasanen MS, Saarakkala S, Rieppo J, Jurvelin JS, et al. Undersulfated chondroitin sulfate does not increase in osteoarthritic cartilage. J Rheumatol 2004; 31: 244953.
  • 23
    Maroudas A, Evan H. Sulphate diffusion and incorporation into human articular cartilage. Biochim Biophys Acta 1974; 338: 26579.
  • 24
    Morris ME, Levy G. Serum concentration and renal excretion by normal adults of inorganic sulfate after acetaminophen, ascorbic acid, or sodium sulfate. Clin Pharmacol Ther 1983; 33: 52936.
  • 25
    Van der Kraan PM, de Vries BJ, van den Berg WB, Vitters E, van de Putte LB. Effects of drug-mediated serum sulfate depletion on glycosaminoglycan synthesis. Agents Actions 1988; 23: 557.
  • 26
    Morris ME, Benincosa LJ. Sulfate homeostasis. II. Influence of chronic aspirin administration on inorganic sulfate in humans. Pharm Res 1990; 7: 71922.
  • 27
    De Vries BJ, van den Berg WB, van de Putte LB. Salicylate-induced depletion of endogenous inorganic sulfate: potential role in the suppression of sulfated glycosaminoglycan synthesis in murine articular cartilage. Arthritis Rheum 1985; 28: 9229.
  • 28
    Van der Kraan PM, de Vries BJ, Vitters EL, van den Berg WB, van de Putte LB. Inhibition of glycosaminoglycan synthesis in anatomically intact rat patellar cartilage by paracetamol-induced serum sulfate depletion. Biochem Pharmacol 1988; 37: 368390.
  • 29
    Van der Kraan PM, Vitters EL, de Vries BJ, van den Berg WB, van de Putte LB. The effect of chronic paracetamol administration to rats on the glycosaminoglycan content of patellar cartilage. Agents Actions 1990; 29: 21823.