IL-6 TNF-Alfa

Adipose tissue tumor necrosis factor and interleukin-6expression in human obesity and insulin resistance

PHILIP A. KERN, SUBRAMANIAN RANGANATHAN, CHUNLING LI,LINDA WOOD, AND GOURI RANGANATHANDivision of Endocrinology, Department of Medicine, University of Arkansas for Medical Sciencesand the Central Arkansas Veterans Healthcare System, Little Rock, Arkansas 72205Received 1 August 2000; accepted in final form 23 January 2001

Kern, Philip A., Subramanian Ranganathan, Chun-ling Li, Linda Wood, and Gouri Ranganathan. Adiposetissue tumor necrosis factor and interleukin-6 expression inhuman obesity and insulin resistance. Am J Physiol Endo-crinol Metab 280: E745E751, 2001.Adipose tissue ex-presses tumor necrosis factor (TNF) and interleukin (IL)-6,which may cause obesity-related insulin resistance. We mea-sured TNF and IL-6 expression in the adipose tissue of 50lean and obese subjects without diabetes. Insulin sensitivity(SI) was determined by an intravenous glucose tolerance testwith minimal-model analysis. When lean [body mass index(BMI) ,25 kg/m2] and obese (BMI 3040 kg/m2) subjectswere compared, there was a 7.5-fold increase in TNF secre-tion (P , 0.05) from adipose tissue, and the TNF secretionwas inversely related to SI (r 5 20.42, P , 0.02). IL-6 wasabundantly expressed by adipose tissue. In contrast to TNF,plasma (rather than adipose) IL-6 demonstrated the stron-gest relationship with obesity and insulin resistance. PlasmaIL-6 was significantly higher in obese subjects and demon-strated a highly significant inverse relationship with SI (r 520.71, P , 0.001). To separate the effects of BMI from SI,subjects who were discordant for SI were matched for BMI,age, and gender. By use of this approach, subjects with low SIdemonstrated a 3.0-fold increased level of TNF secretionfrom adipose tissue and a 2.3-fold higher plasma IL-6 level(P , 0.05) compared with matched subjects with a high SI.Plasma IL-6 was significantly associated with plasma nones-terified fatty acid levels (r 5 0.49, P , 0.002). Thus the localexpression of TNF and plasma IL-6 are higher in subjectswith obesity-related insulin resistance.

type 2 diabetes

OBESITY HAS BECOME a national epidemic with enormouspublic health implications (25), and recent studieshave demonstrated a further 6% increase in the inci-dence of obesity [body mass index (BMI) .30 kg/m2]over a 7-yr period (30). There is a strong correlationbetween obesity and insulin resistance in both diabeticand nondiabetic subjects (27), and the risk of diabetesincreases 11-fold as the BMI increases from 20 to 30(8). Although insulin resistance accompanies all pa-tients who become obese, the degree of insulin resis-tance varies considerably, and the relationships be-

tween obesity, insulin resistance, and type 2 diabetesare not well understood.

Obesity represents an expansion of adipose tissuemass, and one explanation for obesity-related insulinresistance is the production of factors by adipose tissuethat render some subjects more insulin resistant thanothers. Numerous adipocyte secretory products haverecently been described that play a role in carbohy-drate and lipid metabolism (14, 21, 23). One suchadipocyte secretory product is tumor necrosis factor(TNF)-a. A new role for TNF was proposed in 1993 withthe description of TNF expression by adipose tissueand the elevated expression of TNF in obese, insulin-resistant rodents and humans (17, 20, 24). Although itis unclear how adipose TNF expression may causeinsulin resistance (36), TNF is known to impair insulinreceptor signaling (18). TNF also inhibits lipoproteinlipase (LPL) and stimulates lipolysis in adipocytes (34),and the resulting increase in circulating nonesterifiedfatty acids (NEFA) would be expected to contribute toinsulin resistance (7).

Another adipocyte secretory product that may beinvolved in insulin resistance is interleukin (IL)-6,which is a cytokine secreted by many cells, includingadipocytes and adipose stromal cells (11, 15). LikeTNF, IL-6 inhibits the expression of LPL, but, unlikeTNF, IL-6 does not stimulate lipolysis (13, 16). IL-6secretion is increased in the adipocytes of obesesubjects (29) and may be important either as a cir-culating hormone or as a local regulator of insulinaction.

Although many studies have examined the role ofTNF in insulin resistance, relatively few of these havebeen in humans, and none has examined cytokineexpression in detail along with the measurement ofinsulin resistance. In this study, we examined theexpression of TNF and IL-6 in human adipose tissuefrom nondiabetic subjects with varying degrees of obe-sity and insulin resistance. We found that TNF secre-tion from human adipose tissue and circulating plasmaIL-6 were both highly associated with obesity-associ-ated insulin resistance.

Address for reprint requests and other correspondence: P. A. Kern,Central Arkansas Veterans Healthcare System, 598/151 LR 4300West 7th St., Little Rock, AR 72205 (E-mail: [email protected]).

The costs of publication of this article were defrayed in part by thepayment of page charges. The article must therefore be herebymarked advertisement in accordance with 18 U.S.C. Section 1734solely to indicate this fact.

Am J Physiol Endocrinol Metab280: E745E751, 2001.

http://www.ajpendo.org E745

METHODS

Subjects. Fifty subjects were recruited for these studies.This research was approved by the Institutional ReviewBoard, and all subjects gave informed consent. All subjectswere weight stable at the time of the study. Subjects initiallyunderwent an oral glucose tolerance test using 75 g of glu-cose, and blood glucose was measured fasting and at 2 h.Subjects with diabetes (fasting blood sugar .126 mg/dl, 2-hglucose .200 mg/dl) were excluded. Of the 50 subjects, 15had impaired glucose tolerance based on a 2-h glucose of140200 mg/dl, and three of these subjects had impairedfasting glucose based on a fasting glucose of 110126 mg/dl.Subjects then underwent a frequently sampled intravenousglucose tolerance test (FSIVGTT) and an adipose tissue bi-opsy. The FSIVGTT and the biopsy were performed at least 3days apart.

Characteristics of the subjects that comprised this studyare shown in Table 1. Blood lipids were measured usingstandard clinical assays, and plasma NEFA were measuredusing a colorimetric assay (Waco Chemical, Richmond, VA).Of the 50 subjects studied, 39 were women and 8 wereAfrican-American. The subjects ranged from lean to veryobese, and insulin sensitivity (SI; using the SI index from theFSIVGTT] varied considerably. Some subjects demonstratedmoderate dyslipidemia, but no subject demonstrated fastingtriglycerides .400 mg/dl. Body composition was determinedusing bioelectric impedance (38).

SI measurements. The measurement of in vivo SI wasperformed in the fasting state with the minimal-model anal-ysis of the FSIVGTT (4, 5). We used the classic tolbutamide-modified test, which has been validated against the euglyce-mic clamp in humans (6, 41). In brief, catheters were placedfor glucose injection and blood sampling. Four basal bloodsamples were obtained, and the patient was given an intra-venous glucose bolus (11.4 g/m2) at time 0. At 20 min after theglucose injection, patients were given an injection of tolbut-amide (125 mg/m2), again followed by frequent blood sam-pling, according to the standard protocol. Together, 4 basaland 27 postglucose blood samples were taken, the last one at240 min. Glucose was measured in a glucose analyzer by useof the glucose oxidase method, and insulin was measuredusing radioimmunoassay. These measurements were per-formed in the Endocrinology Laboratory of the Indiana Uni-versity School of Medicine (Indianapolis, IN). The SI wascalculated using the MINMOD program (4) and was ex-pressed in microunits per milliliter per minute.

Adipose tissue biopsy. Abdominal subcutaneous adiposetissue (;10 g) was removed from each patient by incision,which avoids trauma to fat cells and minimizes the amount ofblood in contact with the fat cells. Some of the tissue wasimmediately frozen in liquid N2 for later RNA extraction,whereas the rest of the tissue was placed into cold DMEM forother assays.

Adipose tissue cytokine secretion. TNF and IL-6 may func-tion in an autocrine or paracrine manner; hence, we wishedto measure the local secretion of these cytokines into the

medium. Immediately after the biopsy, adipose tissue piecesof ;500 mg were minced and placed into serum-free DMEM(pH 7.4, 10 mM HEPES) at 37C for varying times. Figure 1illustrates the secretion of TNF and IL-6 into the medium ofthree subjects. There was little secretion of either cytokineinto the medium for the first 60 min, followed by an increasein secretion over the next 60 min. Medium cytokine levelscontinued to increase for up to 24 h. To compare TNF andIL-6 secretion among different subjects, we measured cyto-kine levels in the medium after 2 h at 37C. All data werenormalized to adipose DNA content to control for differencesin fat cell size. In general, IL-6 secretion from adipose tissuewas much higher than TNF. In all subjects studied, the TNFlevel in the medium at 2 h was 0.78 6 0.14 pg/mg DNA, andthe IL-6 level in the medium was 9.8 6 1.8 pg/mg DNA.

Measurement of TNF and IL-6. Adipose tissue TNF proteinwas measured using an ELISA (R&D Systems, Minneapolis,MN). This assay demonstrates an 8% intra-assay and a 15%interassay variation. This ELISA method was used to mea-sure TNF in fasting plasma as well as TNF secretion byadipose tissue (see Relationship between TNF and obesity).TNF mRNA levels were measured by competitive RT-PCR,as described by us previously (24). IL-6 was measured infasting plasma and secreted from adipose tissue using anELISA assay (R&D Systems). This assay demonstrates intra-and interassay variations of ,5%.

Statistics. All data are expressed as means 6 SE. Toanalyze data between groups, a one-way ANOVA was per-formed, and secondary analysis was performed with theStudents t-test with Bonferroni correction. Analysis oftrends was performed using linear regression after log trans-formation. The Wilcoxon matched-pair sign-rank test wasused for the paired data in Table 2.

Table 1. Characteristics of the study subjects

nAge,yr

BMI,kg/m2

Fat,%

TG,mg/dl

LDL,mg/dl

HDL,mg/dl

FBG,mg/dl

Women 39 4061.6 3661.4 4361.2 118612 12565 5362 9362Men 11 4263.3 3363.8 2764.9 143624 124611 4561.8 10063

Data are expressed as means 6 SEM. BMI, body mass index; TG, triglyceride; LDL and HDL, low- and high-density lipoprotein,respectively; FBG, fasting blood glucose.

Fig. 1. Tumor necrosis factor (TNF)-a and interleukin (IL)-6 secre-tion from adipose tissue. TNF and IL-6 are secreted into the mediumover time from adipose tissue. Data represent means and SE of 3representative subjects. Both cytokines are expressed as pg/mg DNA.

E746 ADIPOSE EXPRESSION OF TNF AND IL-6

RESULTS

Relationship between TNF and obesity. To betterdefine the effects of obesity on TNF expression, wemeasured TNF mRNA in the adipose tissue from eachsubject, along with plasma TNF and TNF secretionfrom the adipose tissue. Subjects were divided into fourBMI groups representing lean (BMI ,25, n 5 9), over-weight (BMI 2530, n 5 9), moderately obese (BMI3040, n 5 17), and very obese subjects (BMI .40, n 515). Figure 2B shows TNF mRNA levels from subjectswith increasing BMI. There was considerable variabil-ity among the obese subject groups with respect to TNFmRNA levels, such that the differences between nor-mal lean subjects (BMI ,25 kg/m2) and obese subjectswere not statistically significant (NS; Fig. 2B). TNFprotein was also measured in these subjects; however,as shown in Fig. 2A, there was no relationship betweenplasma TNF and BMI. However, TNF secretion fromthe adipose tissue was higher in obese subjects. MeanTNF secretion was 0.16 6 0.06 pg/mg DNA in leansubjects (BMI ,25 kg/m2), and 1.21 6 0.36 pg/mg DNA

in subjects with a BMI between 30 and 40 kg/m2 (P ,0.05). Subjects with a BMI .45 kg/m2 demonstratedslightly lower TNF secretion (0.90 6 0.21 pg/mg DNA),but this was not significantly decreased compared withsubjects with a BMI of 3040 kg/m2. This effect of BMIon TNF secretion was still present when women andCaucasians were each considered separately and whensubjects with impaired glucose tolerance were elimi-nated. TNF secretion from adipose tissue was also lowin subjects with low body fat. TNF secretion in subjectswith ,30%, 3045%, and .45% body fat was 0.16 60.07 (n 5 10), 0.76 6 0.16 (n 5 14), and 1.1 6 0.28pg/mg DNA (n 5 18, P , 0.05 vs. ,30% group).

TNF expression and insulin sensitivity. As expected,there was a significant relationship between obesityand insulin sensitivity. As described previously by oth-ers (22), the relationship between BMI and SI is curvi-linear and best represented by a log/log transforma-tion, and in our subjects, BMI and SI were significantlyrelated (r 5 20.65, P , 0.001). Because SI variesconsiderably among nonobese subjects with normalglucose tolerance, we did not divide SI into subgroupsbut instead examined TNF expression over the spec-trum of SI. There was no significant relationship be-tween either plasma TNF or TNF mRNA levels and SI(data not shown). However, there was a significantdecrease in TNF secretion with increasing SI (Fig. 3),such that most of the insulin-sensitive subjects (SI .5)had lower levels of TNF secretion, and most of theinsulin-resistant subjects (SI ,2) had the highest lev-els of TNF secretion.

IL-6 expression with obesity and insulin resistance.The adipose tissue fragments secreted relatively highlevels of IL-6. When IL-6 expression was examined inthe same BMI groups, as described in the precedingsection for TNF, there was a tendency for an increasein IL-6 secretion from adipose tissue with increasingBMI and increasing body fat (Fig. 4B); however, thesechanges were not statistically significant. Plasma IL-6,however, was strongly associated with increasing obe-sity (Fig. 4A). In lean subjects (BMI ,25), plasma IL-6was 0.73 6 0.23 pg/ml and increased about fourfold to2.86 6 0.61 pg/ml in the most obese subjects (BMI .40,

Fig. 3. Relationship between TNF expression and insulin sensitivity(SI). TNF secretion into the medium was measured in subjects alongwith measurements of SI. TNF secretion is expressed as pg/mg DNAof adipose tissue, and the data are log transformed.

Fig. 2. Effect of body mass index (BMI) on plasma TNF, adiposesecretion of TNF into the medium, and TNF mRNA levels. Subjectswere divided into BMI groups as described in the text, representinglean subjects (BMI ,25) and subjects with increasing degrees ofobesity. A: plasma TNF was measured and expressed as pg/ml, alongwith TNF secreted into the medium of the adipose tissue expressedas pg/mg DNA, as described in METHODS. B: TNF mRNA levels weremeasured using RT-PCR, as described in METHODS. *P , 0.05 vs. BMI3040 and BMI .40.

E747ADIPOSE EXPRESSION OF TNF AND IL-6

P , 0.05). In a similar manner, plasma IL-6 was lowerin subjects with low percent body fat. Plasma IL-6 was0.84 6 0.19 pg/ml (n 5 10) in subjects with ,30% bodyfat and was 2.05 6 0.38 (n 5 14) and 2.58 6 0.44 (n 518) pg/ml in subjects with 3045 and .45% fat, respec-tively (P , 0.05). The relationship between SI andplasma IL-6 was examined in the same manner asdescribed for TNF. In contrast to TNF, adipose-se-creted IL-6 demonstrated no significant relationshipwith SI (r 5 20.04, P 5 NS). However, there was ahighly significant relationship (r 5 20.71, n 5 38, P ,0.001) between plasma IL-6 and SI, as shown in Fig. 5.Plasma IL-6 was 3.0 6 0.53 pg/ml in the most insulin-resistant subjects (SI ,2) and was 0.82 6 0.19 pg/ml inthe most insulin-sensitive subjects (SI .5, P , 0.05).

One mechanism by which TNF may cause insulinresistance is through an increase in adipocyte lipolysis,leading to a rise in plasma NEFA. Hence, the relation-ship between cytokine expression and plasma NEFAwas examined. The only significant relationship withplasma NEFA levels was with plasma IL-6 and adipose

TNF secretion. As shown in Fig. 6, there were signifi-cant increases in plasma NEFA levels in subjects withhigher levels of plasma IL-6 (r 5 0.54, P , 0.001).There was also a significant association betweenplasma NEFA and TNF secretion (r 5 0.35, n 5 37, P ,0.05), although this association was less robust thanthe association with IL-6.

Cytokines and insulin resistance independent of obe-sity. Insulin resistance is exacerbated by obesity, lead-ing to a significant relationship between SI and BMI.Therefore, we examined the relationship between adi-pose cytokine expression and SI without the confound-ing effects of BMI. To factor out obesity, we identifiedsubjects who were of the same BMI but who werediscordant for SI. We compared the cytokine expressionof subjects with insulin resistance (SI ,2.0) with thatof subjects with less insulin resistance (SI .3.0) whowere matched for BMI (65 kg/m2), age (610 yr), andgender. Using these criteria, we were able to matchnine subjects with SI ,2.0 with nine subjects with SI.3.0. As shown in Table 2, these subjects were wellmatched for age and BMI, and there were significantdifferences in SI by virtue of subject selection. Nodifferences were noted between plasma TNF or adipose

Fig. 4. Effect of BMI on IL-6 expression. A: plasma IL-6 (pg/ml) wasmeasured in subjects from the same BMI groups as described in Fig.2. *P , 0.05 vs. BMI 3040 and BMI .40 groups. **P , 0.05 vs. BMI.40 group. B: adipose tissue IL-6 secretion (pg/mg DNA) was mea-sured in subjects as described in METHODS.

Fig. 5. Relationship between plasma IL-6 and SI. Plasma IL-6 wasmeasured as described in METHODS and expressed as pg/ml. Data arelog transformed.

Fig. 6. Relationship between plasma IL-6 and plasma nonesterifiedfatty acids (NEFA). Plasma IL-6 levels were plotted against theplasma NEFA level for each subject.


IL-6 expression. However, the insulin-resistant sub-jects had significantly higher levels of plasma IL-6 aswell as significantly higher levels of adipose TNF se-cretion (P , 0.05). In these matched subjects, TNFsecretion and plasma IL-6 were two- to threefoldhigher in the insulin-resistant subjects.

Previous studies have demonstrated that IL-6 andTNF interact with each other in both 3T3-L1 adipo-cytes and mice (3, 16). We examined TNF and IL-6expression from each subjects adipose tissue to deter-mine whether there was any relationship between IL-6and TNF expression. As shown in Fig. 7, there was astrong linear relationship between the secretions ofIL-6 and TNF from the adipose tissue (r 5 0.81, P ,0.0001). On the other hand, there was no significantrelationship between plasma IL-6 and plasma TNF(data not shown).

DISCUSSION

Since the initial description of TNF expression byadipose tissue, several lines of evidence have suggestedthat TNF overproduction by adipose tissue may beinvolved in the pathogenesis of the insulin resistance ofobesity. TNF mRNA levels were high in obese, insulin-resistant rodents, and the infusion of a soluble TNFbinding protein into insulin-resistant fa/fa rats im-proved insulin sensitivity and improved the defect ininsulin receptor and insulin receptor substrate-1 auto-phosphorylation in fat and muscle (18, 20). Recentstudies using genetic manipulations resulting inknockout or depletion of TNF or TNF receptor haveconfirmed the importance of TNF in rodent insulin

resistance (9, 19, 40), although one such study (37)found no role for TNF or the TNF receptor in insulinresistance.

Relatively few studies have examined the relation-ship between TNF and insulin resistance in humans.Studies by us (24) and others (1, 17) demonstratedelevated levels of adipose TNF mRNA and protein inobese subjects and a decrease in TNF with weight loss.No study has examined the relationship between SIand TNF, although one study noted a significant cor-relation between TNF mRNA levels and fasting insulin(17), and several studies observed a decrease in TNFafter weight loss (12, 17, 24). High TNF secretion fromhuman adipose tissue was associated with decreased[3H]glucose incorporation into lipids (26).

It is not clear whether TNF functions locally orcirculates in a sufficiently high concentration to influ-ence distant targets. Plasma TNF has been measured,and several studies have observed increased plasmaTNF levels in obese subjects and in subjects withhyperinsulinemia or insulin resistance (10, 42, 43).Plasma TNF was elevated in male diabetic subjectscompared with male controls, but no such relationshipwas observed in women (35). In an attempt to bindplasma TNF and reverse insulin resistance in humans,diabetic or insulin-resistant subjects have been givenan injection of anti-TNF binding protein. In both stud-ies, there was no improvement in insulin resistance(31, 33).

The role of IL-6 in insulin resistance has been muchless studied. IL-6 is secreted by many cells, includingadipocytes and adipose stromal cells (11, 15) and isincreased after a meal (32). Like TNF, IL-6 inhibits theexpression of LPL, but unlike TNF, IL-6 does notstimulate lipolysis (13, 16). Linking IL-6 to insulinresistance are studies demonstrating increased IL-6secretion in the adipocytes of subjects with obesity (29)and diabetes (2).

In the studies described herein, we measured TNFand IL-6 gene expression at several levels from theadipose tissue of lean and obese subjects and relatedthis expression to SI, a reliable measure of insulinsensitivity. Both IL-6 and TNF were expressed andsecreted by human adipose tissue, although IL-6 levelswere much higher in both adipose tissue and plasma.The most consistent relationship between cytokine ex-pression and obesity-related insulin resistance in-volved increased TNF secretion from adipose tissueand increased plasma IL-6 levels. Elevated TNF andIL-6 expression was found in subjects who were onlymoderately obese (BMI .30) and increased progres-sively with decreasing SI. The relationship between

Table 2. Insulin-sensitive and insulin-resistant subjects matched for BMI and age

SI BMIAge,yr

Plasma TNF,pg/ml

Plasma IL-6,pg/ml

Adipose TNF Secretion,pg/mg DNA

Insulin sensitive 4.660.34 34.961.8 40.363.0 1.7260.26 1.4360.30 0.4760.15Insulin resistant 1.560.12* 34.862.4 39.162.8 1.8760.31 3.2160.79* 1.4260.47*

Values are means 6 SE. SI, insulin sensitivity; TNF, tumor necrosis factor; IL-6, interleukin-6. *P , 0.05 vs. insulin sensitive.

Fig. 7. Relationship between IL-6 and TNF secretion from adiposetissue. IL-6 and TNF (pg/mg DNA) were measured in the mediumfrom adipose tissue pieces cultured for 2 h, and the level of mediumIL-6 was plotted against the level of TNF.


plasma IL-6 and SI was very strong, with a highlysignificant inverse correlation and a fivefold differencebetween the most insulin-resistant and most insulin-sensitive subjects. Thus both TNF and IL-6 were asso-ciated with both obesity and insulin resistance; how-ever, it was the adipose-secreted form of TNF and theplasma level of IL-6 that displayed the strongest rela-tionships.

The subjects in this study were heterogeneous withregard to degree of obesity, gender, and race, and it ispossible that a study using a more focused group ofsubjects would yield different results. However, weobserved no consistent effect of gender or race on cyto-kine expression in these subjects. This study also reliedon plasma cytokine levels and cytokine secretion fromadipose tissue, and these measurements may not bereflective of cytokine biological effects at the tissuelevel.

Because obesity and insulin resistance are related toeach other, we wished to determine whether TNF andIL-6 expression were related to insulin resistance in-dependently of obesity. As described in Table 2, wepaired insulin-resistant subjects with more-insulin-sensitive subjects and matched them for BMI and age.By use of this analysis, high levels of TNF secretionand plasma IL-6 were both significantly associatedwith insulin resistance. Thus the expression of thesecytokines was associated with insulin resistance inde-pendently of obesity.

There are differences in the expression of TNF andIL-6 that may be important in understanding theirfunctions. IL-6 was secreted at high levels from adiposetissue, and there was a significant arteriovenous dif-ference in IL-6 across the adipose tissue bed, whereasthere was no arteriovenous difference with TNF (29).We found no relationship between plasma TNF andobesity or insulin resistance, although other studieshave noted increased plasma TNF with obesity (2, 10,42, 43). IL-6 and TNF may interact with each other, assuggested by the strong correlation between TNF se-cretion and IL-6 secretion in this study and by previousstudies that demonstrated increased IL-6 expression inresponse to TNF (3, 16). Together, these data suggestthat TNF functions locally at the level of the adipocytein a paracrine fashion, perhaps stimulating the secre-tion of NEFA, IL-6, or other circulating substances. Onthe other hand, plasma IL-6 circulates at high levelsand may be more important systemically and perhapsrepresents a hormonal factor that induces muscle in-sulin resistance.

It is noteworthy that two studies have tried, andfailed, to reverse insulin resistance with an injection ofanti-TNF binding proteins (31, 33). On the basis of thestudies described herein, we can speculate on severalpossible reasons for the failure of anti-TNF therapy inhumans. If TNF functions in a paracrine or autocrinefashion in adipose tissue, then the anti-TNF bindingproteins may not reach the microcirculation in suffi-cient concentration to prevent TNF-mediated effects.In addition, our data raise the possibility that IL-6 is

the major circulating component of obesity-related in-sulin resistance.

The development of insulin resistance with increas-ing adiposity suggests that an adipocyte product maybe important in insulin resistance. Both TNF and IL-6are adipocyte products that are overexpressed in obeseinsulin-resistant subjects, and we have shown that thesecretion of these cytokines is interrelated. Some ofthese cytokines may function systemically, others mayfunction locally, and still others may function to in-crease the secretion or synthesis of other adipocytefactors or to act as an adjuvant to the actions of otherinsulin resistance factors. One such insulin resistancefactor is NEFA, which are closely associated with in-sulin resistance (28, 39). TNF stimulates lipolysis inadipocytes (34); hence, it is possible that TNF functionsat the level of the adipocyte to stimulate lipolysis.Although IL-6 is not known to stimulate lipolysis (13,16), we found a significant relationship betweenplasma IL-6 and plasma NEFA levels, whereas therelationship between TNF expression and plasmaNEFA was much less robust.

These studies provide the first comprehensive anal-ysis of IL-6 expression in obese, insulin-resistant hu-mans and add to the data on TNF expression. To-gether, these studies suggest that obesity-relatedinsulin resistance represents a complex syndrome, me-diated by a number of adipocyte secretory products,which ultimately lead to defects in insulin action inother target organs.

We thank Dr. Richard Evans for statistical assistance, DeniseHargrove for assistance with subject recruitment, and the nurses ofthe General Clinical Research Center at the University of Arkansasfor Medical Sciences and Central Arkansas Veterans HealthcareSystem. We also thank Dr. Richard Bergman for supplying theMINMOD program, and Sarah Dunn for excellent secretarial assis-tance.

This study was supported by a Veterans Affairs Department MeritReview Grant M01-RR-14288 of the General Clinical Research Cen-ter, a Career Development Award from the American Diabetes As-sociation, and DK-39176 from the National Institutes of Health.

REFERENCES

1. Arner P. Obesity and insulin resistance in Swedish subjects.Diabet Med 13, Suppl 6: S85S86, 1996.

2. Bastard JP, Jardel C, Bruckert E, Blondy P, Capeau J,Laville M, Vidal H, and Hainque B. Elevated levels of inter-leukin 6 are reduced in serum and subcutaneous adipose tissueof obese women after weight loss. J Clin Endocrinol Metab 85:33383342, 2000.

3. Berg M, Fraker DL, and Alexander HR. Characterization ofdifferentiation factor/leukaemia inhibitory factor effect on li-poprotein lipase activity and mRNA in 3T3-L1 adipocytes. Cyto-kine 6: 425432, 1994.

4. Bergman RN, Finegood DT, and Ader M. Assessment ofinsulin sensitivity in vivo. Endocr Rev 6: 4586, 1985.

5. Bergman RN, Phillips LS, and Cobelli C. Physiologic evalu-ation of factors controlling glucose tolerance in man. Measure-ment of insulin sensitivity and beta-cell sensitivity from theresponse to intravenous glucose. J Clin Invest 68: 14561467,1981.

6. Bergman RN, Prager R, Volund A, and Olefsky JM. Equiv-alence of the insulin sensitivity index in man derived by theminimal model method and the euglycemic glucose clamp. J ClinInvest 79: 790800, 1987.


7. Boden G. Role of fatty acids in the pathogenesis of insulinresistance and NIDDM. Diabetes 46: 310, 1997.

8. Carey VJ, Walters EE, Colditz GA, Solomon CG, WillettWC, Rosner BA, Speizer FE, and Manson JE. Body fatdistribution and risk of noninsulin-dependent diabetes mellitusin women. The Nurses Health Study. Am J Epidemiol 145:614619, 1997.

9. Cheung AT, Ree D, Kolls JK, Fuselier J, Coy DH, andBryer-Ash M. An in vivo model for elucidation of the mecha-nism of tumor necrosis factor-alpha (TNF-alpha)-induced insulinresistance: evidence for differential regulation of insulin signal-ing by TNF-alpha. Endocrinology 139: 49284935, 1998.

10. Corica F, Allegra A, Corsonello A, Buemi M, Calapai G,Ruello A, Nicita Mauro V, and Ceruso D. Relationship be-tween plasma leptin levels and the tumor necrosis factor-alphasystem in obese subjects. Int J Obes 23: 355360, 1999.

11. Crichton MB, Nichols JE, Zhao Y, Bulun SE, and SimpsonER. Expression of transcripts of interleukin-6 and related cyto-kines by human breast tumors, breast cancer cells, and adiposestromal cells. Mol Cell Endocrinol 118: 215220, 1996.

12. Dandona P, Weinstock R, Thusu K, Abdel-Rahman E, Al-jada A, and Wadden T. Tumor necrosis factor-alpha in sera ofobese patients: fall with weight loss. J Clin Endocrinol Metab 83:29072910, 1998.

13. Feingold KR, Doerrler W, Dinarello CA, Fiers W, andGrunfeld C. Stimulation of lipolysis in cultured fat cells bytumor necrosis factor, interleukin-1, and the interferons isblocked by inhibition of prostaglandin synthesis. Endocrinology130: 1016, 1992.

14. Flier JS. The adipocyte: storage depot or node on the energyinformation superhighway? Cell 80: 1518, 1995.

15. Fried SK, Bunkin DA, and Greenberg AS. Omental andsubcutaneous adipose tissues of obese subjects release interleu-kin-6: depot difference and regulation by glucocorticoid. J ClinEndocrinol Metab 83: 847850, 1998.

16. Greenberg AS, Nordan RP, McIntosh J, Calvo JC, ScowRO, and Jablons D. Interleukin 6 reduces lipoprotein lipaseactivity in adipose tissue of mice in vivo and in 3T3-L1 adipo-cytes: a possible role for interleukin 6 in cancer cachexia. CancerRes 52: 41134116, 1992.

17. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, andSpiegelman BM. Increased adipose tissue expression of tumornecrosis factor-alpha in human obesity and insulin resistance.J Clin Invest 95: 24092415, 1995.

18. Hotamisligil GS, Budavari A, Murray D, and SpiegelmanBM. Reduced tyrosine kinase activity of the insulin receptor inobesity-diabetes. Central role of tumor necrosis factor-a. J ClinInvest 94: 15431549, 1994.

19. Hotamisligil GS, Johnson RS, Distel RJ, Ellis R, Papaio-annou VE, and Spiegelman BM. Uncoupling of obesity frominsulin resistance through a targeted mutation in aP2, the adi-pocyte fatty acid binding protein. Science 274: 13771379, 1996.

20. Hotamisligil GS, Shargill NS, and Spiegelman BM. Adiposeexpression of tumor necrosis factor-a: direct role in obesity-linked insulin resistance. Science 259: 8791, 1993.

21. Hotamisligil GS and Spiegelman BM. Tumor necrosis factora: a key component of the obesity-diabetes link. Diabetes 43:12711278, 1994.

22. Kahn SE, Prigeon RL, McCulloch DK, Boyko EJ, BergmanRN, Schwartz MW, Neifing JL, Ward WK, Beard JC, andPalmer JP. Quantification of the relationship between insulinsensitivity and b-cell function in human subjects. Evidence for ahyperbolic function. Diabetes 42: 16631672, 1993.

23. Kern PA. Potential role of TNFa and lipoprotein lipase ascandidate genes for obesity. J Nutr 127: 1917S1922S, 1997.

24. Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, andSimsolo RB. The expression of tumor necrosis factor in humanadipose tissue. Regulation by obesity, weight loss, and relation-ship to lipoprotein lipase. J Clin Invest 95: 21112119, 1995.

25. Kuczmarski RJ, Flegal KM, Campbell SM, and JohnsonCL. Increasing prevalence of overweight among US adults: the

National Health and Nutrition Examination Surveys, 1960 to1991. JAMA 272: 205211, 1994.

26. Lofgren P, Van H, Reynisdottir VS, Naslund E, Ryden M,Rossner S, and Arner P. Secretion of tumor necrosis factor-ashows a strong relationship to insulin-stimulated glucose trans-port in human adipose tissue. Diabetes 49: 688692, 2000.

27. Ludvik B, Nolan JJ, Baloga J, Sacks D, and Olefsky J.Effect of obesity on insulin resistance in normal subjects andpatients with NIDDM. Diabetes 44: 11211125, 1995.

28. McGarry JD. What if Minkowski had been ageusic? An alter-native angle on diabetes. Science 258: 766770, 1992.

29. Mohamed-Ali V, Goodrick S, Rawesh A, Katz DR, MilesJM, Yudkin JS, Klein S, and Coppack SW. Subcutaneousadipose tissue releases interleukin-6, but not tumor necrosisfactor-alpha, in vivo. J Clin Endocrinol Metab 82: 41964200,1997.

30. Mokdad A, Serdula MK, Dietz WH, Bowman B, Marks J,and Koplan J. The spread of the obesity epidemic in the UnitedStates, 19911998. JAMA 282: 15191522, 1999.

31. Ofei F, Hurel S, Newkirk J, Sopwith M, and Taylor R.Effects of an engineered human anti-TNF-a antibody (CDP571)on insulin sensitivity and glycemic control in patients withNIDDM. Diabetes 45: 881885, 1996.

32. Orban Z, Remaley AT, Sampson M, Trajanoski Z, andChrousos GP. The differential effect of food intake and beta-adrenergic stimulation on adipose-derived hormones and cyto-kines in man. J Clin Endocrinol Metab 84: 21262133, 1999.

33. Paquot N, Castillo MJ, Lefebvre PJ, and Scheen AJ. Noincreased insulin sensitivity after a single intravenous adminis-tration of a recombinant human tumor necrosis factor receptor:Fc fusion protein in obese insulin-resistant patients. J ClinEndocrinol Metab 85: 13161319, 2000.

34. Patton JS, Shepard HM, Wilking H, Lewis G, AggarwalBB, Eessalu TE, Gavin LA, and Grunfeld C. Interferons andtumor necrosis factors have similar catabolic effects on 3T3-L1cells. Proc Natl Acad Sci USA 83: 83138317, 1986.

35. Pfeiffer A, Janott J, Mohlig M, Ristow M, Rochlitz H,Busch K, Schatz H, and Schifferdecker E. Circulating tumornecrosis factor alpha is elevated in male but not female patientswith type II diabetes mellitus. Horm Metab Res 29: 111114,1997.

36. Qi C and Pekala PH. Tumor necrosis factor-alpha-inducedinsulin resistance in adipocytes. Proc Soc Exp Biol Med 223:128135, 2000.

37. Schreyer, SC Chua SA Jr, and LeBoeuf RC. Obesity anddiabetes in TNF-a receptor-deficient mice. J Clin Invest 102:402411, 1998.

38. Segal KR, Gutin B, Presta E, Wang J, and Van Itallie TB.Estimation of human body composition by electrical impedancemethods: a comparative study. J Appl Physiol 58: 15651571,1985.

39. Unger RH. Lipotoxicity in the pathogenesis of obesity-depen-dent NIDDM. Genetic and clinical implications. Diabetes 44:863870, 1995.

40. Uysal KT, Wiesbrock SM, Marino MW, and HotamisligilGS. Protection from obesity-induced insulin resistance in micelacking TNF-alpha function. Nature 389: 610614, 1997.

41. Welch S, Gebhart SSP, Bergman RN, and Phillips LS.Minimal model analysis of intravenous glucose tolerance test-derived insulin sensitivity in diabetic subjects. J Clin EndocrinolMetab 71: 15081518, 1990.

42. Winkler G, Lakatos P, Salamon F, Nagy Z, Speer G, KovacsM, Harmos G, Dworak O, and Cseh K. Elevated serumTNF-alpha level as a link between endothelial dysfunction andinsulin resistance in normotensive obese patients. Diabetic Med16: 207211, 1999.

43. Zinman B, Hanley AJ, Harris SB, Kwan J, and Fantus IG.Circulating tumor necrosis factor-alpha concentrations in a na-tive Canadian population with high rates of type 2 diabetesmellitus. J Clin Endocrinol Metab 84: 272278, 1999.


Documents

IL-6 TNF-Alfa