The Entrance of Glucose and Other Sugars into Joints By MARIANW. ROPES,ALEX F. MULLERAND WALTERBAUER Entrance of glucose into joints was determined in calves and in patients with joint disease. In calves under two weeks, entrance was rapid; delayed in older calves. In patients with joint disease there was delay in appearance and slow subsequent entrance, more marked with greater severity of disease. The rate with galactose was the same; that with pentoses and methylated glucoses usually more rapid. The nature of the mechanism of transport is not apparent. Esseva determinate le entrata de glucosa in le articulationes de vitellos e de patientes con morbos articular. In vitellos de minus que duo septimanas de etate, le entrata esseva rapide; in vitellos plus avantiate ill0 esseva retardate. In patientes con morbo articular un retardate apparition de glucosa e un lente entrata subsequente de ill0 esseva notate, e isto esseva plus marcate con plus sever formas del morbo. In studios con galactosa le mesme comportamento esseva constatate; con le pentosas e le glucosas methylate le valores esseva generalmente plus rapide. Le natura del mechanismo de transport0 non es apparente. T HE BARRIER between blood and synovial fluid remains poorly characterized and information concerning the transfer of various substances has accumulated s1owly.l Understanding of the mechanism of exchange of glucose across the blood-joint barrier and the effect of disease thereon is of special importance in view of the fact that cartilage is an avascular tissue and depends for its nutrition largely on synovial fluid. Recent evidence indicates that substances can reach cartilage from the blood vessels of subchondral bone,2-5but that the major source of nutrition is the joint fluid. The glucose concentration in synovial fluid obtained from normal animals and humans with no evidence of joint disease or with relatively mild inflammation (as in traumatic joint disease and in mild cases of rheumatoid or infectious arthritis) has been shown to be approximately the same as that of the With more severe inflammation, however, as in rheumatoid or infectious arthritis, the sugar concentration in the fluid is reduced, occasionally to zero in very severely involved joints. The present investigations were undertaken in order to learn more of the mechanism of the exchange of glucose between blood and fluid in normal and diseased joints. MATERIALS AND METHODS Glucose entrance was determined in 47 joints of 19 calves ranging in age from 6 days to 4 months and in weight from 100 to 350 lbs. The animals were being held for slaughter From the Department of Medicine, Haroard Medical School and the Medical Seruices of the Massachwetts General Hospital. This is publication no. 280 of the Robert W. Looett Memorial for the Study of Cjippling Disease. Grants in support of these investigations have been received from the Commonwealth Fund, New York City, the National Institute of Arthritis and Metabolic Diseases, Grani #-2123, National Institutes of Health, United States Public Health SerTuice, and the La& Research hboratoria, Elf LiUv Co. 496 ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS 497 and had been fasting for at least 14 hours. After the calf had been restrained on his side, specimens of jugular blood and of synovial fluid from the astragalotibial or carpometacarpal joints were obtained. Sixty to 120 cc. of 25 per cent glucose in water (in 5 cases containing also 3 per cent sodium thiocyanate) were injected into the jugular vein with a syringe, and subsequently a constant drip of 25 per cent glucose, by means of a polyethylene catheter, was maintained for varying periods of time. The fluid specimens were withdrawn from the astragalotibial and carpometacarpal joints at regular intervals until the joint cavities were empty. It was necessary to tap the joints for each specimen. Definitely hemorrhagic fluids were discarded. Following the cessation of the intravenous drip, blood specimens were withdrawn from the jugular vein at regular intervals.* Phloridzin, (1 cc. containing 2.25 to 10 p M ) and dinitrophenol (0.5 to 1.5 cc. containing 2 to 10 p M per cc.) were injected into the selected joints, ( 5 and 18 respectively) usually from one to 15 minutes before the administration of glucose. In one animsl the intra-articular injection was made 18 hours before and on one occasion 40 minutes after the injection of the glucose. In order to compare the effect of other substances that would cause comparable synovial irritation saline was injected into 5 joints, concentrated turpentine into 1, and orthonitrophenol in the same concentration as dinitrophenol into 3 joints. The entrance of glucose into the human knee joint was studied in 20 patients with rheumatoid arthritis (24 tests), one patient with probable rheumatoid arthritis, one patient with gonorrheal arthritis, and three patients with Charcot knees ( 8 tests). In one of the patients with rheumatoid arthritis and one with Charcot joints, studies were made before and during treatment with ACTH or cortisone. The patients varied in age from 17 to 67; the duration of disease and effusion and the severity of disease are noted in table 1. When classified in accordance with the American Rheumatism Association diagnostic criteria,' 13 had classical, 7 definite and 1 probable rheumatoid arthritis. After the patients had fasted for at least 14 hours, specimens of venous blood and joint fluid were obtained. The fluid was aspirated with an 18-gauge lumbar puncture needle which was left in place throughout the experiment. Five per cent glucose in normal saline solution or water was then injected intravenously by the drip method, 500 cc. being given in 20 to 25 minutes and the remaining 600 to 1000 cc. more slowly. Blood specimens were taken from the opposite arm and specimens of synovial fluid from the knee at regular intervals until the cavity was empty. A separate syringe was used for each sample of fluid. During the procedure the patient lay at rest with the knee immobile and in as full extension as was possible. The entrance of sodium thiocyanate was investigated in 10 tests in 8 cases of rheumatoid arthritis and in 7 tests in 2 patients with Charcot knees in whom the entrance of glucose had been studied. In these experiments 20 cc. of 5 per cent sodium thiocyanate were injected intravenously over a period of 2 minutes, usually just prior to the injection of glucose, but in two cases on separate days. In 12 experiments, blood specimens were taken under oil to prevent any shift of thiocyanate from cells to plasma. Galactose entrance was studied in 7 patients with rheumatoid arthritis, and one patient with Charcot joints. Studies were carried out with D-xylose in 4 patients with rheumatoid arthritis and 1 patient with Charcot joints ( 2 tests); fructose in 1 patient with Charcot joints; L-xylose in 1 patient with rheumatoid arthritis; L-arabinose in 1 patient with rheumatoid arthritis ( 2 tests); D-arabinose in 2 patients with rheumatoid arthritis; 6methyl glucose in 1 patient with rheumatoid arthritis; and 3-methyl glucose in 2 patients with rheumatoid arthritis and 1 patient with Charcot joints ( 2 tests). In 4 patients with rheumatoid arthritis glucose and galactose (each in concentration of 5 per cent) were given simultaneously. The technic was identical with that used in the case of glucose ~ *The differences between the initial fasting levels of glucose in serum and fluid, often higher than the differences found in normal human joints,e were generally associated with abnormally high levels of blood glucose. It was assumed that these elevations of blood sugar occurred during the process of restraining the animals. 498 ROPES, MULLER, BAUER except for the fact that usually only 500 to 6001 cc. were given when the pentoses and the methylated glucoses were used. Most of the data with the pentoses and the methylated glucoses were collected over a 5 month period on the same patient (case 10,table 1 ) . The activity of this patient’s disease was seemingly stationary throughout the study period. Unfortunately, 6-methyl glucose had to b e administered more slowly than the other sugars since the patient developed a few extrasystoles and complained of vague chest discomfort. The removal of glucose from the joint was studied under similar conditions in one patient with rheumatoid arthritis. The removal of galactose was determined also in the same patient. After obtaining fasting blood and synovial fluid specimens, 0.75 cc. of 25 per cent glucose or galactose in normal saline was injected into the joint cavity; mixing was insured by alternate aspiration and reinjection of synovial fluid. Glucose was determined by the methods of Folins and of Somogyi-Nelson9JO; galactose and the methylated glucoses by the method of Hagedorn and Jensenll after y e s t fermentation*; the pentoses either by the method of Hagedorn and Jensenll after yeast fermentation or by that of Roe and Ricelz; and thiocyanate by the method of Lavietes, Bourdillon and KIinghoff er.13 Six-methylglucose was synthesized by one of us (A. F. M.) according to Freudenberg and Hiill,l4 with the following modifications: B-acetyl-3,5-benzylidene-l,2-monoacetone glucose (1.4) was prepared from 6-acetyl-1, 2-monoacetone glucose ( 1 . 4 ) according to D. J. Bell.15 The 3-time recrystallized 6-methylglucose (m.p. 143-145 C.) was dissolved in water and filtered through a Seitz-filter previous to its administration. Three-methylglucose was prepared by Ayerst-McKenna and Harrison Ltd. to whom we are most grateful for their generous supply. L-xylose was purchased from Hoffman La Roche Pharm. Indust., Basel, Switzerland; the arabinose from Pfanstiehl Chemical Co., the D-xylose from Fisher Scientific Co. All these substances were made up in sterile 5 per cent solutions in water, filtered through a Seitz-filter previous to administration, and given intravenously in a constant drip over approximately 30 minutes. They all had been tested previously in rabbits for possible pyrogenic reactions. RESULTS I . Entrance of Glucose and Thiocyanate into Joints of Normal Calves Characteristic curves depicting the rate of entrance of glucose and thiocyanate into the joints of normal calves are shown in figure 1. The shape of the curves apparently varied with the age of the calves. In animals less than 150 Ibs. (under 2 weeks of age), glucose appeared in the synovial fluid of both the astragalotibial and carpometacarpal joints relatively rapidly after intravenous administration. In older animals, over 150 Ibs. in weight, the entrance of glucose into the astragalotibial joints remained rapid, whereas in the carpometacarpal joints there was a delay in almost all cases in the appearance of glucose in the fluid (fig. 1B). Whether the difference in the rate of entrance into the front and rear joints in older calves is related to the greater trauma that the front joints suffer, or to other factors, is difficult to say. In both the young and the older calves thiocyanate entered the joints without any delay and the concentration rose rapidly. (fig. 1C ) . *Previous to precipitation by sodium tungstate, the aliquots of plasma and joint fluid were incubated at room temperature for 20 minutes with a 10 per cent yeast solution. (Fresh Fleishman’s yeast was washed 10 to 12 times with water in order to remove non-fermentable reducing substances. This solution was made fresh for each experiment. ) 499 ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS 8 1 300 ZOO I / / 100 0 20 10 0 40 30 SO C 0 Lo M 3 0 10 to 30 40 5C Fig. 1.-Entrance of glucose and thiocyanate into joints of normal calves following intravenous administration. Abscissa = time in minutes. LR = left astragalotibial joint; RR = right astragalotibial joint; LF = left carpometacarpal joint; R F = right carpometacarpal joint. A: 100 Ib. calf; B and C: 300 Ib. calves. -0and -0- indicate concentration of, glucose in serum and in synovial fluid; and x concentration of thiocyanate in synovinl fluid. In experiment A only one serum glucose level was obtained so the shape of the curve was estimated from those of comparable experiments. O W 50 11. Entrance of Glucose into Joints i n Various Joint Diseases Rheumatoid arthritis.-The synovial fluid glucose curves obtained after intravenous injection of glucose in the 24 tests in 20 patients with rheumatoid arthritis and one with probable rheumatoid arthritis varied markedly as indicated in table 1. The shapes of the curves and the range of variation are shown by characteristic examples (figure 2 ) . The blood sugar concentration rose rapidly, and was maintained usually beween 200 and 300 mg./'100 cc. for two hours. The initial rate of entrance into the joint fluid was slow in all but four tests' with fluid levels increasing only 25 mg./100 cc. or less in 20 minutes. In "In one of these, the patient was receiving cortisone intramuscularly and in another cortisone had been injected intra-articularly 9 days previously. so0 ROPES, MULLER, BAUER ? 8 1 I 350 e L o , i 0 2U 40 60 Bo 0 100 120 0 20 40 60 80 100 120 20 40 60 80 100 120 TIME IN M I N U T E S Fig. 2.-Entrance of glucose into joints of patients with rheumatoid arthritis following intravenous administration.-0and -0- indicate concentration in serum and synovial fluid respectively. A, B and C represent the same patients as those in figure 3 (cases 12, 14 and 17 in table 1). two patients with severe disease there was no alteration for 30 and 50 minutes, respectively (e.g. fig. 2C and case 17, table 1). After this initial lag, sugar entered the joint fluid, and the subsequent rate of entrance seemed to show an inverse relationship with the initial difference, in the fasting state, between the fluid and serum sugar concentrations. The smaller this initial difference the faster the rate of entrance and the more clcsely the fluid level approached the serum at the end of two hours when a tendency toward the establishment of equilibrium was usually apparent. In general, the fasting difference between fluid and serum sugar and the corresponding lag and slow rate of entrance tended to vary with the severity of the disease (table 1).In mild cases, as seen in table 1, the fluid sugar concentration tended to be only slightly below the blood level, and similarly, the rate of entrance of glucose after the initial lag was relatively fast. In contrast, patients with high initial differences and relatively slow sugar entrances usually had severe arthritis of long duration. This delayed entrance and slow rise of glucose in the fluid can be partially reversed by reducing the severity of synovial tissue inflammation as was apparent in studies on the patient with rheumatoid arthritis treated with ACTH. The lag in entrance was slightly shorter and the subsequent rate of entrance considerably faster, seven days after the patient was started on ACTH (80 mg. per day). Other signs of decreased inflammation were the reduction in volume of fluid from 38 to 11 cc., the drop in leukocyte count from 12,800 to 2350, and in neutrophils from 89 to 24 per cent, the improvement in mucin precipitate, and the slightly increased rate of entrance of thiocyanate and radioactive sodium, NaZ4(given in sodium chloride) .* l6 In a patient with Charcot *The rate of entrance of radioactive sodium (Na24) was determined in one pationt with rheumatoid arthritis and one with Charcot joints before and during treatment with ACTH. In these experiments radioactive sodium ( N H ' ) ~in) sodium chloride (4 to 10 cc.) was injected intravenously. These studies were conducted by Dr. Ann Wight Phillips, then Atomic Energy Commission Fellow, presently Assistant in Surgery, Harvard Meclical School. 501 ENTRANCE OF GLUCOSE AND OTHEH SUGARS INTO JOINTS knees, on the other hand, in whom treatment with ACTH caused no clinical improvement and no alteration in synovial fluid except an increase in relative viscosity, the rate of entrance of glucose remained unaltered. Other joint diseases.-The entrance of glucose into synovial fluid in the three patients with Charcot joints and the patient with gonorrheal arthritis showed the same types of curves as in rheumatoid arthritis and the same relationship between initial fluid sugar levels and the rate of entrance. (table 1). Entrance of Thiocyanate into Joints in Joint Diseases Thiocyanate, a readily digusible, nonutilizable substance, entered the joints differently (fig. 3 ) . There was no initial lag and in all cases the fiuid levels rose steadily though at different rates. In 5 of the 8 patients with rheumatoid arthritis and in the 2 patients with Charcot joints the concentration had reached equilibrium in 70 to 110 minutes. The three other rheumatoid cases difkred only in the slower rate of entry. No absolute correlation was apparent between the rates of entrance of thiocyanate and glucose although there was a tendency for thiocyanate equilibrium to be attained sooner in those joints into which glucose entered more readily. In the cases in which glucose entrance was slow, the rate of thiocyanate entrance was also reduced, though remaining, however, always much more rapid than that of glucose. (figs. 2 and 3 ) . I I I . Entrance of Galactose into Joints in Various Joint Viseases The synovial fluid curves obtained after intravenous injection of galactose in 6 patients with rheumatoid arthritis and one with Charcot joints were almost identical with those of the glucose experiments (fig. 4). In all cases there was an initial lag in appearance of galactose in the fluid, similar to that with glucose, and the' final fluid-serum relationships were the same. Entrance of Fructose into Joints in Joint Diseuse The rapidity with which fructose disappeared from the blood stream made it difficult to reach high blood levels. The curves of entrance into the two 2 0 20 40 60 80 100 120 140 0 20 40 TIME 60 IN 80 I00 120 0 20 40 60 80 100 120 MINUTES Fig. 3.-Entrance of thiocyanate into joints of patients with rheumatoid arthritis following intravenous administration. The symbols -X - and -X- indicate concentration in serum and in synovial fluid. A, 3 and C represent the same patients as those in figure 2 (cases 12, 14 and 17 in table 1). RA 8. 640228 25. 308128 24. 168891 23. 327100 22. 310162 joints Charcot joints Chamt joints Gonococcal arthritis RA RA RA RA RA Prob. RA Charcot RA RA RA RA RA RA RA RA RA RA 5. 889261 6. 691529 7. 232864 864163 720330 703880 609351 119894 8999 296175 810235 8996 20047 416026 745476 305443 38 RA 54415 4. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 3 1 BA RA 2. 745536 3. 499948 M M M 21 - - - M 60 - - F 3 2 2 1 3 3 2 3 3 2 3 2 4 2 3 3 2 2 2 2 2 2 1 2 2 -- 3 1 F F 3 F M M M 1 M 4 2 1 1 2 1 1 4 2 M F M M M 2 M M M F F F 1 2 1 F Class M 6 2 0 3 9 0 6 3 2 7 Stage Sex 67 33 62 61 4 59 40 56 34 47 5 6 5 54 26 4 4 2 5 42 RA 1. 635099 Age (yrs.) Disease Case # Severitv Table 1.-Entrance 7 4 3 1 2 3mu. 12 4 2% 1 2 timo. 6 3 mo. ti ?2 Duration &usion (yrs.) -1 wk. 15 15 1 wk. 9 mo. 15 8 6 59; 5% 7 6 11 11 5 4 wks. ( ? 4 wks.) P9mo. ?9mo. 7 13 16 15 2 16 4 1 3 1 5 1 7 6 1 12 Duntion disease (yrs.) 63,750 400 ( 2 ) 1.250 4,200 150 4,950 ti,9nn 2,650 b,7no 9,500 -_ 31.600 16,850 27,400 23,600 33,000 33,750 58,100 26 - (4) 96 10 38 n 0 22 fi8 61 - 62 64 86 81 71 88 fi2 73 40 70 93 -- so - 22 52 48 27 24 Neutm phils % 10,900 (3) 25,550 17,450 ( 2 wks. before) 4,925 7.200 (2) 2.350 11,400 8.m GLUCOSE 3,800 (1) 2 0 . ~ 2 5 Leukocytes percu.mm. Synovial fluid Initial glucose 85 89 105 04 99 85 96 92 77 74 95 88 99 01 103 84 71 76 77 $9 82 77 78 114 82 89 91 57 inn Difference between serum and fluid sugar at two hours (mn./100 ce.) Maximum serum sugar (mg./ 100 ec.) 19 78 78 6 84 55 75 75 82 43 47 47 38 14 18 3 6 8 4 43 in0 85 88 75 96 64 65 59 81 69 21 18 97 0 64 53 73 71 81 78 34 44 49 30 14 25 32 39 42 21: 12 15 14 n 5 9 11 10 3 0 1 4 7 4 1 18 n 0 0 0 0 0 0 110 8 0 n 5 64 13 11 0 5 0 3 11 2 10 5 5 6 23 0 8 1 25 n 10 1 200 13 12 25 10 25 2 6 6 35 10 8 67 35 24 10 15 210 2 12 168 32 34 --36 119 (90min) 65 25 74 15 132 (98min) 235 205 104 265 (90 min) 52 40 n 6 31 -26 18 40 (40min) 30 55 (90min) 25 275 270 170 360 291 225 165 387 400 264 255 235 200 290 283 250 300 230 345 270 280 313 216 166 260 300 225 235 195 Y i X *B sr z 'd "E K 3 Increase in fluid serum concentration 0 ui with Various Joint Diseases in 10 in.20 serum fluid difference min. min. (mg./100 ce.) (ms./lOO ec.) - of Sugars into Knee Joints of Patients 4 5 4 40 47 5 54 33 49 6 5 5 67 6 6 6 6 5 6 3 6 67 RA RA RA RA RA RA RA Charcot joints RA RA RA RA Charcot joints Charcot joints RA RA RA RA RA CA RA Charcot joints 23. 327100 23. 327100 10. 720330 10. 720330 10. 720330 26. 758916 10. 720330 2. 745536 10. 720330 23. 327100 640228 720330 609351 415026 2 1 1 1 6 1 1 7 1 9 0 6 0 3 3 M M M M F F F F F F F M M 1 2 2 2 2 2 2 - - - 2 1 2 1 F 3 3 - M :j 4 1 1 1 F F M F M M M - - -- 1 8 3 3 2 3 3 2 3 S 3 -- 3 2 2 2 2 2 4 4 17 1 4 4 2 4 4 17 15 17 I - - - - (I.) 150 81 78 -10 94 48 46 40 0 80 -- 0 -- fi 0 5 2 0 IF 3 4 9 53 5 11 25 2n 20 40 18 30 11 40 10 24 0.7 1.7 13 27 55 5 14 7 33 12 0 12 10 28 15 24 0 13 6 0.1 0.7 14 1 1 3 8 23 3 0 0 4 2 3 4 33 20 5 89 49 9 9 15 87 86 76 50 - 69 88 93 77 - 83 96 91 230 320 148 188 10 159 241 246 175 216 234 5 5 143 177 242 242 178 143 217 105 95 128 69 345 110 (5) 135 115 ( 5 ) 0 -5 38 -11 -28 57 (89 min) -7 -24 -23 0.7 1.0 ( l 0 0 m i n ) 42 26 12 107 85 -35 10 7 (90 min) -8 0 (85 min) 0 237 (90min) 30 (5) 90 (105 min) 48 (5) usual occupation o r self-care; IV. incapacitated, largely o r wholly; little o r no self-care. The patients a r e listed in order of increasing difference between the initial serum and fluid glucose levels since this represents one indication of increasing severity i n the joints aspirated. ( I ) After intra-articular cortisone: (2) on ACTH; (3) on cortisone; (4) repeated on same day; (5) glucose & galactose measured together. - __ - - L-ARABINOSE 3 8.800 8,200 D-ARABINOSE 3 7,950 2 17,500 62 6-METHYL GLUCOSE 3 10,500 4s 3-METHYL GLUCOSE 3 mo. 11,950 3 17 ( R ) 400 ( L ) 350 ( R ) 150 - 3 17 17 15 - 4 13 GO 4 93 38 83 - 90 90 G9 650 ( R ) 0 750 ( R ) 12 450 (1,) FRUCTOSE (R) (L) L-XYLOS E 7.550 6 2 7 - 3 2 4 GALACTOSE 23.150 16,250 8,550 39.400 6.800 11.250 12,100 360 (1 wk. before) D-XYLOSE 4,050 3,700 2% 11 ? 9 mo. 6 6 mo. 1 1 3 mo. 7 3 7 8 11 ? 9 mo. 18 1 1 1 T h e stage of disease is determined by the degree of presumably irreversible involvement rather than activity. Stage I is characterized only by subchondral hone atrophy; stage 11, by cartilaginous and cortical destruction; stage I11 by deformities such as ulnar deviation and subluxation; stage IV, hy fibrow or bony ankylosis. Nodulss and significant muscle atrophy must not be present i n stage I. The functional classes are: I, ability to carry on all usual duties without handicap; TI, adequate for normal activities. despite handicap of discomfort or limited motion in one or more joints; 111, limited only to few or none of the duties of 8. 10. 12. 19. 5. 889261 7. 232864 11. 703888 13. 119894 16. 810235 17. 8996 20. 745476 22. 310162 8 cn 3 M I 3 i 8 4 r 0 504 ROPES, MULLER, BAUER 0 20 40 60 100 80 o 120 20 40 60 eo ioo 120 TIME IN M l N U TES Figure 4.-Entrance of glucose and galactose into joints of patients with rheumatoid arthritis following intravenous administration. -0and -0- indicate concentration of glucose in serum and in synovial fluid; - - and -A- concentration of galactose in serum and in synovial fluid. A and I3 represent individual patients (cases 17 and 7 in table 1). a knees of a patient with Charcot joints (fig. 5 ) , suggested that there was a lag in the entrance of fructose similar to that observed with glucose and galactose, but the observed concentrations in fluid and serum were extremely low. IV. The Entrance of D-xylose, L-xylose, D-nmbinose, L-arnbinose, 3-methyl Glucose and &methyl Glucose into Joints In general, the curves of entrance of D-xylose, L-xylose, D-arabinose, L-arabinose, 3-methyl glucose and 6-methyl glucose were alike (figs. 6, 7 and 8 “t_ -t 0 0 10 -- --_- -_--- *r -; 20 A* ; > -* ! RK 30 40 50 60 TIME IN MINUTES Fig. 5.-Entrance of fructose into joints of a patient with Charcot joints (case 23 in table 1). and * indicate concentration in serum and in synovial fluid. RK = right knee and LK = left knee. * 505 ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS ,-, B 300 300 z- \ I 0 20 40 60 80 100 0 120 \ 20 40 60 80 100 120 TIME I N MINUTES Fig. 6.-Entrance of pentoses into joint of a patient with rheumatoid arthritis following intravenous administration ( 5 separate experiments in case 10 in table 1). 0 and 0 indicate concentration of D-xylose in serum and in synovial fluid; and concentration of L-xylose in serum and in synovial fluid; and concentration of D-arabinose in serum and in synovial fluid; A and concentration of L-arabinose in serum and in synovial fluid (the solid lines represent one experiment, the dotted lines a second experiment two months later). . . I J 140 TIME IN MINUTES Figure 7.-Entrance of 3-methyl glucose, 6-methyl glucose and glucose into joint of a patient with rheumatoid arthritis following intravenous administration. (3 separate experiments in case 10 in table 1). 0 and indicate concentration of 3-methyl glucose in serum and in synovial fluid; 8 and o concentration of 6-methyl glucose in serum and in synovial fluid; 0 and o concentration of glucose in serum and in synovial fluid. ROPES, MULLER, SAUER 0 20 40 60 80 100 120 TIME 0 I I I I 20 40 60 80 I 100 I 120 I N MINUTES Fig. %-Entrance of glucose, 3-methyl glucose, D-xylose and thiocyanate into joints of a patient with Charcot joints following intravenous administration. (A indicates right knee and B left knee of case 23 in table 1). 0 and o indicate concentration of glucose in serum and in synovial fluid; 0 and 0 concentration of 3-methyl glucose in serum and in synovial fluid; and 0 concentration of D-xylose in serum and in synovial fluid; and x and X concentration of thiocyanate in serum and synovial fluid. 8). Following intravenous administration the sugars appeared relatively rapidly in the joint fluid. The rate of entrance varied from patient to patient and was relatively slow in the patient with Charcot joints in whom the entrance of thiocyanate also was relatively slow (fig. 8 ) . The reproducibility of the experiments is well shown in figure 6. The two entrance studies done on the same patient with L-arabinose are, for practical purposes, superimposable even though they were done at approximately a two month interval. There were some differences in the rates of entrance of the individual sugars that are of interest when compared with their transport across other barriers. For example, the entrance of L-arabinose was slightly less rapid than that of Darabinose. In the case of xylose, D-xylose entered the joint at essentially the same rate as L-xylose in 3 cases, but with a lag in 3 others. V. The Effect of Simultaneous Administration of Glucose and Galactose on Their Entrance into Joints in Rheumutoid Arthritis Patients In order to determine whether there was any evidence of competition of sugars entering the joint simultaneously, glucose and galactose were given intravenously at the same time. In the case of galactose there was almost no difference in the time of appearance or rate of entrance when it was given alone or with glucose (table 2, fig. 9). However, the entrance of glucose was delayed in 3 of 4 patients when glucose was given with galactose. In one of these patients, however, the subsequent rate of entrance of glucose was rapid. V l . The Effect of Enzyme lnhibitors on Entrance of Glucose into Joints in Nomull Calves Enzyme inhibitors that could be injected into joints, dinitrophenol and phloridzin, were utilized in the studies in calves to determine whether they ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS 507 1 - 0 0 4 G c, 350 t I Fig. 9.-Entrance of glucose and galactose into joint of a patient with rheumatoid arthritis following intravenous administration (case 16 in table 1). Abscissa = time in minutes. A: glucose injected alone; B: galactose injected alone; C: glucose and galactose injected simultaneously. 0 and 0 indicate concentration of glucose in serum and synovial fluid; and A and A concentration of galactose in serum and in synovial fluid. altered the entrance of glucose. Other substances that might cause comparable synovial irritation, orthonitrophenol, saline and concentrated turpentine, were used for comparison. In almost all the joints into which the various test substances had been injected the initial rate of entrance of glucose was more rapid than in the uninjected control joints. In many instances the shapes of the curves of entrance resembled in all other respects those in the control joints. However, in the majority of the dinitrophenol joints (13 out of 18) there was a definite reduction in the rate of entrance occurring between 5 and 35 minutes after injection of glucose (fig. lo).* A similar lessening in the rate of entrance oc'A reduction in rate of entrance was amsidered significant only when the slope of entrance decreased by 2.2% degrees. The measurements were made by an observer who had no knowledge as to which joints had been injected with the various substances). 508 ROPES, MULLER, BAUER Table Z.--Effect of Simultaneous Adminbtration of Glucose and Gakzctose on Entrance of Sugars into Synovial Fluid -___ Case No. -___ - .. Initial serumfluid glucose difference (mg./100 cc.) 745476 78 78 52 703880 5 39 8 810235 60 53 74 889261 - Sugar entrance Final at appearfluid sugar Appearance time ance time concentration (min.) (mg./100 cc.) (mg./100 cc.) Sugar galactose glucose galactose +glucose galactose glucose galactose +glucose galactose glucose galactose 55 40 55 85 20 20 30 20 20 30 20 85 +glucose galactow glucose galactose +glucose 12 23 -- 18 5 20 8 28 6 12 17 25 12 12 13 4 5 37 17 10 10 20 40 59 53 58 77 85 97 70 105 104 85 115 13 66 140 54 91 time (min.) (120) (105) (120) ( 105 ) (85) (85) (85) (85) ( 100) (100) (100) (103) (90) - (90) (100) (100) 500 400 2 300 0 200 $2 0 100 n 0 10 20 30 40 50 60 TiME I N 0 10 LO 30 40 50 60 70 MiNUTES Fig. 10.-Entrance of glucose into joints of normal calves following intravenous administration. A. Dinitrophenol was injected into left carpometacarpal joint (LF) before administration of glucose. B. Dinitrophenol was injected into right carpometacarpal joint ( R F ) before administration of glucose. 0 and 0 indicate concentration of glucose in serum and synovial fluid. The vertical arrows indicate time at which intravenous glucose was started; horizontal arrows indicate duration of injection. so9 ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS curred in only 5 of 37 control joints, in 1 of 5 saline, in the 1 turpentinc, in 1 of 5 phloridzin and 1 of 3 orthonitrophenol joints. The occurrence of the marked change in the slope of entrance of glucose was significantly more frequent in the dinitrophenol joints than in the others. (using the chi-squared test p < .01) VII. Exit of Glucose and Other Sugars from Joints In order to aid in the interpretation of the lag in the entrance of gIucose and galactose into joints, these sugars were injected on separate occasions into the knee joint of one of the patients with rheumatoid arthritis in whom glucose and galactose entrances were exceptionally slow (fig. l l ) , table 1, Case 17). During 100 minutes after intra-articular injection of glucose the fluid level fell 95 mg./100 cc. whereas in 2 hours the galactose level decreased only approximately 15 mg./100 cc. Comparison of the breakdown of glucose and galactose in vitro by joint fluid cells was obtained by incubating uncentrifuged fluid containing approximately 9000 cells with glucose and with galactose at 37 C. In the case of glucose the content had fallen from 100 mg./100 cc. to 50 mg. after 24 hours. In the galactose experiment there was no decrease in the galactose concentralion after 22 hours’ incubation. In the present experiments the fluid glucose level did not begin to fall for 10 to 20 minutes after the blood level fell, the lag being of the same magnitude as the delay in entrance. Similar relationships between blood and fluid glucose curves, with the resulting phenomenon of fluid level above blood, occur in various joint effusions after the oral administration of glucose. Cajori, Crouter, and PembertonlT reported that in three of four patients investigated, the synovial fluid glucose concentration was above that of the blood between 35 and 85 minutes following the oral administration of 100 grams of glucose. In this laboratory,’ synovial fluid glucose levels ranging as high as 67 mg./100 cc. 2001 8 A I e0” ci I 2ool I $h F IS0 1 I I 20 40 60 0 0 20 40 60 80 100 I20 TIME 0 IN I I I 80 100 120 MINUTES Fig. 11.-Removal of glucose and galactose from joint of a patient with rheumatoid arthritis. A. glucose injected into joint. B. galactose injected into joint. 0 and 0 indicate concentration of glucose in serum and synovial fluid; indicates concentration of galactose in synovial fluid. 510 ROPES, MULLER, BAUER Table 3.-Concentration of Glucose in Serum and Synovial Fluid in Nonfasting Patients* Glucose (mg./100 cc.) Case Disease 342755 rheumatoid arthritis 6760 25615 347 182 42840 358373 374295 rheumatoid arthritis rheumatoid arthritis gonococcal arthritis rheumatic fever gonococcal arthritis rheumatoid arthritis (L) (R) Fluid Serum 102 103 101 139 99 143 124 150 81 81 96 117 67 113 96 83 'Serum and fluid were obtained 3 to 4 hours after patients had eaten. above the blood level have been observed on eight occasions when effusions of various types were aspirated 3 to 4 hours after the patients had eaten meals containing approximately 100 grams of carbohydrate (table 3 ) . In the case of the pentoses, also, the fluid level remained higher than the corresponding blood level when the latter was falling. With the methylated sugars the concentration in the joint fluid never stayed above that in the blood in the few experiments in which it was observed. DISCUSSION AND SUMMARY It is evident that in calves over 2 weeks old and in patients with various joint diseases there is a delay in the appearance of glucose in the joint fluid following intravenous administration in contrast to the rapid appearance of a readily diffusible substance such as thiocyanate. The rate of entrance of glucose after this initial lag is also slower than that of thiocyanate. In general, the more severe the joint inflammation, the greater the initial lag and the slower the subsequent entrance of glucose. There is a similar though less marked tendency to reduction in the rate of entrance with greater severity of disease in the case of thiocyanate. Whatever the mechanism of the slow entrance of glucose, it plays an important role in determining the joint fluid sugar level. The difference between blood and fluid sugar concentrations in the fasting state is relatively constant for any one joint. It is low in normal and traumatic joints and increased in more severely inflamed j0ints.l Whenever the equilibrium between entrance, utilization and removal is disturbed in any way, this blood-fluid sugar difference tends to be re-established. In normal and traumatic joints it is presumably restored fairly quickly, as indicated by the patient with Charcot knee (case 23 table 1) in whom the fluid and blood levels were equal two hours after intravenous injection of glucose. In many rheumatoid and infectious joints, on the other hand, the restoration is retarded and the greater the fasting difference between blood and fluid glucose, the slower is the restoration (case 17 table 1) . The cause of the difficulty in entrance of glucose is not obvious. It is not associated with a general decreased rate of entrance of other substances. On ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS 511 the contrary, the permeability to proteins is greatest in the severely inflamed joints into which entrance ok sugar is slowest. Furthermore, the slow rate of appearance of glucose in joint fluid is not merely a consequence of increased utilization in the joint as indicated by the similar rate of entrance of galactose. In the above experiments, galactose was shown not to disappear rapidly after intra-articular injection and was not readily fermented in vitro by synovial fluid cells. Further suggestion that the slow entrance of glucose was not determined solely by utilization was furnished by the findings in a patient (table 1, case 17) in whom synovectomy, with the removal of a large amount of proliferated tissue, had been performed following glucose entrance studies. Two years later, at which time there had been no reaccumulation of the-fluid and no clinically apparent proliferation of the synovial tissues, 1.5 cc. of fluid was withdrawn and found to contain only 8 mg. of glucose/100 cc., approximately the same concentration present before synovectomy 2 years previously. The absolute polymorphonuclear leukocyte counts in the two fluids were 36,490 cells before operation and 11,630 cells 2 years later. In order to determine whether structural specificity, particularly stereospecificity, plays a role in the transport of sugars from blood to synovial fluid, the studies with the stereoisomers of xylose and arabinose and with 3- and 6-methyl glucose were undertaken. The rate of appearance of these sugars was in general more rapid than that of glucose and galactose. The isomers were alike except in 3 of the tests with D-xylose in which there was a lag in entrance similar to that seen with glucose. The results are suggestive of the possibility of specificity in transfer, but more definite evidence is needed. The relative rates of transfer of various hexoses and pentoses across the intestinal, renal, red cell and joint barriers are strikingly similar, if one compares an increasing order of absorption in intestine, kidney and red cell with a decreasing order in the joint. D-galactose, D-glucose, and fructose are transferred across the intestine^,^"^^ red cells25-20and probably rena127-2e barriers more rapidly than the stereoisomers of xylose or arabinose or the methylated glucoses, with few exceptions, whereas at the joint barrier there is apparently delayed entrance of D-galactose, D-glucose and fructose as compared with the others. Recent investigators suggest that the C1conformation of the pyranose ring is essential in a monosaccharide if it is to react with the constituent in the red cell surface which is involved in the movement of sugars through the membrane.30 The specificity in the effect of insulin on the transfer of sugars into tiss u e ~ the , ~ ~sugars responsive to insulin being galactose, glucose, D-xylose and L-arabinose, is different than that suggested in the joint. Furthermore, the apparent variations in the rates of transfer of different sugars into joints are in contrast to the results of studies of the absorption of sugars from the peritoneal cavity32 and the passage of sugars from blood to liver33 in both of which the rate was the same for all sugars tested. Further suggestion of a possible specificity of the mechanism of transfer of glucose across the blood-joint barrier is given by the results obtained with Sl.2 ROPES, MULLER, RAVER simultaneous administration of glucose and galactose. The marked delay in appearance time of glucose in three of the joints when injected intravenously simultaneously with galactose, in contrast to the absence of any lag in entrance of galactose, suggests competition of galactose for the mechanism by which glucose enters the joint, but further ,evidence is necessary. Competition of different monosaccharides has been found in the intestine,'" kidney,34 and red cell.25 The significant decrease in the slope of the curve of glucose entrance in joints into which dinitrophenol had been injected suggests a possible specific effect of dinitrophenol. If this represents an inhibition of phosphorylation, the mechanism of the decreased transport could be either a decrease in the energy required for the transport, or injury of cells so that they can no longer transport sugar actively. The results of these experiments suggest the possibility that the mechanism for transfer of glucose from blood to joint fluid may be specific and selective, as in the intestine, renal tubule and red cell, but the evidence is not conclusive. The possibility of an enzymatically controlled transport, so clearly discussed in the case of the red cell membrane by Rosenberg and Wilbrandt,35 the activity of which is reduced in the presence of joint inflammation, or of an association-dissociation phenomenon, such as that hypothesized by Shannon and Fisher36 in the case of the renal tubule must be considered. In any case, one can agree with Rosenberg and Wilbrandt,"5 discussing entrance into the red cell, that . . . the complexity of the mechanism involved in glucose penetration may seem surprising. Apparently it has to be considered as an inherited remnant, indicating the general biological occurrence of the phenomena discussed here." " SUMMARY 1. In calves less than two weeks old, following glucose administration intravenously, glucose enters the astragalotibial and carpometacarpal joints rapidly, at essentially the same rate as thiocyanate. 2. In older calves, there is a delay in the appearance of glucose in the carpometacarpal joints. 3. In patients with rheumatoid arthritis and other joint diseases there is a delay in the appearance of glucose in the joint fluid and the subsequent rate of entrance is slow, in contrast to relatively rapid appearance and subsequent entrance of thiocyanate. 4. The degree of the initial lag and the reduction in subsequent rate of entrance increase with greater severity of disease. 5. The rate of entrance of galactose and probably fructose is the same as that of glucose. 6. The stereoisomers of xylose and arabinose, and glucose methylated in the 3 and in the 6 position, in most cases, enter into the joint fluid rapidly. 7 . The nature of the mechanism of transport is not apparent. ACKNOWLEDGMENTS The authors wish to express their appreciation to Dr. E. G. L. Bywaters (now at Canadian Red Cross Memorial Hospital, Taplow, Bucks, England) for the active part ENTRANCE OF GLUCOSE AND OTHER SUGARS INTO JOINTS 513 he played in the initial phase of the study; to the many other physicians in the Arthritis Group who have aided in these investigations; to Miss Elsie C. Rossmeisl, Mrs. R. Barbara Peabody, Miss Elizabeth Manning, Miss Katherine L. Brine, Miss Dorothy Woodhams and Miss Phoebe h e y , for their assistance. REFERENCES 1. Ropes, M. W., and Bauer, W.: Synovial Fluid Changes in Joint Disease. Cambridge, Harvard University Press, 1953. 2. Ingelmark, B. 0. E.: The nutritive supply and nutritional value of synovial fluid. Acta orthscandinav. 20~144, 1950. 3. Ekholm, R.: Articular cartilage nutrition. How radioactive gold reaches the cartilage in rabbit knee joints. Acta Anat. (Suppl. 15) pg. 1, 1951. 4. -: Nutrition of articular cartilage. Acta Anat. 24~323,1955. 5. Brodin, H.: Paths of nutrition in articular cartilage and intervertebral discs. Acta orthscandinav. 24:177, 1955. 6. Ropes, PVI. W., Kossmeisl, E. C., and Bauer, W.: The origin and nature of normal human synovial fluid. J . C h . Invest. 193795, 1940. 7. -, Bennett, G. A., Cobb, S., Jacox, R. and Jessar, R.: 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull. Rheumat. Dis. 9:175, 1958. 8. Folin, 0.:TWQrevised copper methods for blood sugar determination. J.Biol. Chem. 82:83, 1929. 9. Nelson, N.: A photometric adaptation of the Somogyi method for the determination of glucose. J.Biol.Chem. 153:375, 1944. 10. Somogyi, M.: Notes on sugar determination. J.Biol.Chem. 19519, 1952. 11. Hagedorn, H. C., and Jensen, B. N.: In Peters, J. P. and van Slyke, D. D.: Quantitative Clinical Chemistry. Vol. 11, Methods. Baltimore, Williams & Wilkins, 1932, p. 471. 12. Roe, J. H., and Rice, E. W.: h photometric method for the determination of free pentoses in animal tissues.. J. Bio1,Chem. 173307, 1948. 13. Lavietes, P. H., Bourdillon, J., and Klinghoffer, K. A.: The volume of the extracellular fluids of the body. J.Clin.Invest. 15261, 1936. 14. Freudenberg, K., and Hiill, G . : 2,6 Di Methyl Glucose. Berichte d.Dcutsch. Chem.Gesellsch. 74-B:237, 1941. 15. Bell, D. J.: Crystalline 6-methyl glucose. A new synthesis. J.Am.Chem. SOC.p. 859, 1936. 16. Ropes, M. W., and Phillips, A. W.: Unpublished data. 17. Cajori, F. A., Crouter, C . Y., and Pemberton, R.: The physiology of synovial fluid. A.M.A.Arch.Int.Med. 37~92, 1926. 18. Cori, C. F.: The fate of sugar in the animal body. I The rate of absorption of hexoses and pentoses from the intestinal tract. J.Biol.Chem. 66: 691, 1925. 19. Wilbrandt, W., and Laszt, L.: Untersuchungen uber die Ursachen der selecktiven Resorption der Zucker aus dem Darm. Biochem. Zeitsch. 259:398, 1933. 20. Verzar, F., and McDougall, E. J.: Absorption from the Intestine. New York, Longmans, Green & Co., 1936. 21. Westenbrink, H. G. K.: Ober die Spezifitat der Resorption einiger Monosen aus d t m Darme der Ratte und der Taube. Arch.Neere.Physio1. 21: 433, 1936. 22. CsBky, T. Z.: Ober die Rolle der Struktur des Glucosemolekuls bei der Resorption aus dem Dunndarm. Zeitsch. physiolog. Chemie. 2?7:47, 1942. 23. Campbell, P. N., and Davson, H.: Absorption of 3-methyl glucose from the small intestine of the rat and the cat. Biochem.J. 43:426, 1948. 24. Hele, M. P.: Phosphorylation and absorption of sugars in the rat. Nature 166:86, 1950. 25. Wilbrandt, W.: Die Wirkung des Phlorizins auf die Permeabilitiit der menschlichen Erythrocyten fur Glukose und Pentosen. Helvet. Physiol. Pharmacol. Acta. S:CG4, 1947. 26. Wilbrandt, W.: Die Permeabilitiit der roten Blutkorperchen fur einfache Zucker. Arch. gesamte Physiol. 241: 302, 1938. 27. Hober, R.: Uber die Ausscheidung 514 28. 29. 30. 31. ROPES, MULLER, BAUER von Zuckern durch die isolierte Froschniere. Pfliiger’s Arch. gesamte Physiol. 233181, 1934. Hamburger. H. J.: Weitere Untersuchungen iiber die Permeabilitat der Glomeriilusmembran fiir Stereoisomere Ziicker, mit besonderer Beriicksichtigung von Galactose. Biochem.Zeitsch. 128:185, 1922. Campbell, P. N., and Young, F. G.: Metabolic studies with 3-m~thyl glucose. 1. Its fate in the animal body. Bi0chem.J. 52:439, 1952. LeFevre, P. G., and hlarshall, J. K.: Conformational specificity in a biological transport system. Am.J.Physiol. 194333, 1958. Goldstein, M. S., Henry, W. L., Huddlestun, B., and Levine, R.: Action of insulin on transfer of sugars across cell barrirrs: common chemical con- 32. 33. 34. 35. 36. figuration of substances responsive to action of the hormone. Am.J. Physiol. 173:207, 1953. Cori, C. F., and Goltz, H. L.: Rate of absorption of hexoses and pentoses from peritoneal cavity. Proc.Soc. Euper.Biol.& Med. 623:122, 1925. Cori, C. F., and Goltz, H. L.: The permeability of liver and muscles for hexoses and pentoses. Proc.Soc.Exper. Biol.& Med. a, 23124, 1925. Shannon, J. -4.: The tubular reabsorption of xylose in the normal dog. Am. J. Physiol. 122:775, 1938. Rosenberg, T., and Wilbrandt, W.: Enzymatic processes in cell memlirane penetration. 1nt.Rev.Cytol. 1:65,1952. Shannon, J. A., nnd Fisher, S.: The renal tubular reabsorption of glucose in the normal dog. Am.J.Physio1. 122: 765, 1938. Walter Bauer, M.D., Chief of Medical Services, Mussachrcsetts General Hospital; lackson Professor of Clinicd Medicine, Harvarcl Medico1 School, Boston, Mass. Alex F . hluller, M.D., Mhdicin Adjuvant Clinique therapeutique, Universifb tle Gc-eva, Geneua, Switzcrlarul. Maruin W. Ropes, M.D.,Associate Physician in Mdicine, Massachusetts Generul Hospital; Assistant Clinical Professor of Medicine, Harvurtl Medical School, Boston, Mass.