Proc. Nati. Acad. Sci. USAVol. 91, pp. 4402-4406, May 1994Applied Biological Sciences

Calorie restriction prevents the occlusive coronary vascular diseaseof autoimmune (NZW x BXSB)F1 mice

(reduced calorie Intake/autolmmunity/cardiolipln autoantibodies)

HAJIME MIZUTANI*, ROBERT W. ENGELMAN*t, KOKEN KINJOH*, YOSHIYUKI KURATA*, SUSUMU IKEHARA§,YUJI MATSUZAWA¶, AND ROBERT A. GOOD**Department of Pediatrics, University of South Florida, All Children's Hospital, 801 Sixth Street South, St. Petersburg, FL 33701; tDepartment of BloodTransfusion, Osaka University Hospital, 2-15 Yamadaoka, Suita, Osaka, 565 Japan; §First Department of Pathology, Kansai MedicalUniversity, 1 Fumizonocho, Moriguchi, Osaka, 570 Japan; and ISecond Department of Internal Medicine, Osaka UniversityMedical School, 2-2 Yamadaoka, Suita, Osaka, 565 Japan

Contributed by Robert A. Good, February 2, 1994

ABSTRACT Male (NZW x BXSB)F1 (W/BFj) mice de-velop systemic autoimmunit involving autoantibodies, throm-bocytopenia, lupus nephritis, and coronary vascular disease(CVD) with myocardlal infarction. To determine whether thismurine lupus-asoted CVD can be prevented by the reduc-tion of dietary calories, male W/BF, mice were separated intoflve experimental groups and fed either ad libitum (desitedgroup A, n = 50), fed 32% fewer calories of an otherwisecomparable diet (designated group B6, n = 20), or initially fedad libitum and then switched to reduced calorie intake (RCI)feeding at ages 14, 17, or 22 weeks (designated B14, n = 10; B17,n = 20; or B22, n = 20). Occlusive CVD was prevented by RC.Life-span was sgnificantly extended among the early onset RCIcohorts, B6 and B14 (P = 0.0001 and P = 0.005), compared togroup A mice. Mean anti-cardiolipin autoantibody titers andmean levels of circulating immune complexes were also loweredIn RCI mice when all RCI mice were compared to ad libitumfed group A mice. Histological grades of both coronary vas-cular and glomerular lesions were snificantly less than thoseof group A mice (P < 0.001). Immunopreclpitates indicative ofimmunoglobulin deposition within coronary or glomerularvascular walls were also substantially less than those of groupA mice. These fidings ate a possible causal role foranti-cardiolipin autoantibody in development of autoimmuneCVD in W/BF1 mice and suggest that regulating dietarycalories can influence the mechanism involved in pathogenesisof autoimmune-associated CVD development.

Occlusive coronary vascular disease (CVD) contributes tothe morbidity and mortality of patients with systemic lupuserythematosus (1, 2), perhaps as a consequence of endothe-Hal injury by circulating immune complexes (CICs) andsubintimal deposition of immune reactants (2) or the throm-bogenic consequences of elevated titers of anti-cardiolipinautoantibody (aCL) (3-10).Of the strains of mice that develop lupus-like syndromes,

male (NZW x BXSB)F1 (W/BF1) mice have an unusuallyhigh incidence (-80%) of CVD (11-15). Male W/BF1 micedevelop systemic autoimmunity beginning at 10 weeks of agethat involves multiple autoantibodies to single- and double-stranded DNA (ssDNA and dsDNA), cardiolipin, and plate-lets and is manifested as progressive thrombocytopenia,lupus nephritis, and CVD (11-16). Although the major epi-cardial coronary arteries of male W/BF1 mice remain largelyunaffected, multiple small arteries and intramyocardial arte-rioles develop endothelial and intimal proliferative thicken-ings, become stenotic, often develop occlusive thrombi, andare frequently surrounded by zones of myocardial necrosis

(13, 14). The CVD of male W/BF1 mice lacks a substantialexudative inflammatory component, but immune reactantsincluding IgM, IgG2, and IgG3 have been demonstratedwithin intramyocardial arteriole walls (15).

Limiting dietary calories to levels 25-35% less than thoseconsumed ad libitum, while ensuring adequate consumptionof all dietary essentials, delays or abrogates the developmentof systemic autoimmunity in autoimmune-prone strains ofmice (17), including the (NZB x NZW)F1 (18-20), NZB (21,22), kd/kd (23), MRL/Mp-lpr/lpr (24, 25), and BXSB (26)strains. Reduced calorie intake (RCI) extends the healthfullongevity of mice of each of these autoimmunity-pronestrains and prevents strain-specific pathology, including se-vere glomerulonephritis and vasculitis (18, 19). Whether RCIcan prevent the development and reduce the severity ofCVDin W/BF1 mice has not been addressed.RCI suppresses the production of anti-DNA autoantibod-

ies and the formation of CICs (20, 25) and delays or preventsthe appearance of CD-5/Lyt-1+ B lymphocytes, a subpop-ulation linked to autoantibody formation (22). RCI has alsobeen shown to lower maintenance rates of cellular prolifer-ation within the intestine, skin, thymus, spleen, and lymphnodes (17, 21, 27), while preserving or even enhancingadaptive cellular proliferation such as in the regenerativehyperplasia of hepatocytes following partial hepatectomy(28) or as in the responses of lymphoid cells to antigens orphytomitogens (17).

In the present report, we fed dietary cohorts of maleW/BF1 mice two semipurified diets of comparable compo-sition so that calorie intake differed by 32%. To determinewhether the onset ofCVD in W/BF1 mice can be blocked andthe severity of vascular lesions that do develop can beminimized by RCI, dietary calories were restricted beginningwhen mice were either 6, 14, 17, or 22 weeks old. Mice wereevaluated for development of anticardiolipin or anti-dsDNAautoantibodies, for levels of CICs, for manifestations anddegree of coronary vascular and glomerular lesions, for theincidence of CVD, and for length of survival.

MATERIAL AND METHODSAnimals. Male W/BF1 mice and female BALB/c mice (The

Jackson Laboratory) were maintained in American Associ-ation for Accreditation of Laboratory Animal Care-accred-ited, specific pathogen-free conditions in accordance withref. 29.

Abbreviations: W/BF1, (NZW x BXSB)FI; CVD, coronary vascu-lar disease; RCI, reduced calorie intake; aCL, anti-cardiolipin au-toantibody; CIC, circulating immune complex; dsDNA, double-stranded DNA; ALP, alkaline phosphatase; PAS, periodic acid/Schiff reagent.tTo whom reprint requests should be addressed.

4402

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

Proc. Natl. Acad. Sci. USA 91 (1994) 4403

Semipurified Diets. Mice fed semipurified diet ad libitumthroughout life consumed w10.5 kcal/day (1 kcal = 4.18 kJ)and were designated group A. Four other groups ofmice wererestricted in dietary calories by 32% and consumed =7.1kcal/day. RCI mice were fed ad libitum prior to the initiationof RCI, which was imposed gradually with a 10% reductionin calories beginning when mice were either 6, 14, 17, or 22weeks old. Groups of RCI mice were designated groups B6,B14, B17, and B22 to indicate age at initiation of RCI. Com-position of the two semipurified diets has been presented indetail elsewhere (28). All dietary constituents were obtainedfrom ICN Biochemicals. Diets were low in dietary fat (-6%of total calories), differed in the level of total caloriesavailable, but were otherwise similar. Amounts of essentialdietary constituents in each diet were determined with regardto the gram and calorie consumption of mice fed ad libitum.The diet fed RCI mice was further enriched so that althoughless food in grams and fewer calories were consumed by RCImice, equivalent amounts of vitamins, minerals, essentialfatty acids, and 30% of calories as protein were consumed byall mice. Approximately 600% of total calories in both dietswere derived from carbohydrates. All mice were fed twiceweekly and weighed weekly. Food consumption was deter-mined for individually housed RCI mice and on a cage basisfor pair-housed ad libitum fed mice by weighing the foodoffered and reweighing the remnant food at the end of thefeeding interval.

Circulating Autoantibodies to dsDNA. Calf thymus dsDNA(Sigma) was used to coat each well ofa microtiter plate. Afterblocking with bovine serum albumin, 50 Ad of serum diluted1:80 was incubated in DNA-coated wells for 1 hr at 370C.Plates were washed, alkaline phosphatase (ALP)-conjugatedgoat anti-mouse IgG was added, the ALP reaction wasdeveloped, and the optical absorbance was measured at 405nm, plotted on a standard curve, and recorded as values from1 to 100 units (30).

Circulating aCLs. Anticardiolipin antibodies were detectedby solid-phase enzyme immunosorbent assay as described(31). Briefly, 50 pI of a cardiolipin solution, 10 pg of cardio-lipin per ml (Sigma) in 0.01 M phosphate-buffered saline (pH7.4), was incubated in each well of a microtiter plate for 12 hrat 4°C. Serum samples diluted 1:80 (50 I4) were added andincubated for 1 hr, ALP-conjugated goat anti-mouse IgG (100IlI) was added, and the ALP reaction was developed and readas described above (30).CICs. Determination of CIC levels was made using an

anti-mouse C3 F(ab')2 (Cappel Laboratories) solid-phaseimmunoassay as described (32). Anti-mouse C3 F(ab')2 (10,ug/ml, 200 Au per well) was added to microtiter wells andincubated at 4°C overnight. To determine CIC levels, a 1:80dilution ofserum was added and incubated at 37°C for 60 min.ALP-conjugated rabbit anti-mouse IgG (200 IL, 1:1000) wasadded, and the ALP reaction was developed and read asdescribed above (30).

Histological Analysis. Formalin-fixed tissues were sec-tioned at 2-4 ,um and stained with either hematoxylin/eosinor periodic acid/Schiff reagent (PAS). Each heart and kidneywere cut transversely into three blocks, and from each blockserial sections were prepared. The degree of CVD andmyocardial infarction was classified from 0 to 3+, accordingto the method used by Berden et al. (15). Grade 1 indicatesminimal PAS-positive deposits along and within one coro-nary vascular wall; grade 2 indicates PAS-positive depositswith narrowing ofthe lumen in two or three coronary vessels;grade 3 indicates four or more affected coronary vessels inthe section. Myocardial infarctions were often associatedwith CVD grades 2 and 3. The degree of glomerulonephritiswas classified from 0 to 3+, using a grading scale slightlymodified from methods previously reported (15). Grade 1indicates thickening and increased cellularity of the mesan-

Table 1. Influence of calorie intake on W/BF1 meanbody weights

Age, weeks

Group 6 10 15 20 25 35A 20.9 25.8 27.7 30.4 31.3 32.2B6 16.2 20.2* 21.4* 21.6* 23.5* 22.2*B14 22.2 27.0 26.8 24.4* 26.7* 26.6*B17 24.3 27.7 29.8 24.5* 25.0* 24.8*Bn 22.1 24.8 27.7 28.8 26.3* 25.7**P < 0.001 compared to body weight of age-matched group A mice.

gium; grade 2 indicates increased mesangial plus increase ofglomerular cellularity, inflammatory infiltrates, and presenceof capsular adhesions; grade 3 indicates severe glomerulardestruction with tubular cast formation.Immunohistochemistry and Immunofluorescence. Heart

sections were evaluated for the deposition ofimmunoglobulinwithin microvasculature using either biotinylated goat anti-mouse IgG or IgM, an avidin-biotin-peroxidase complex, anddiaminobenzidine as substrate in a Vectostain kit (VectorLaboratories) (33). Kidney sections were evaluated for glo-merular IgG or IgM deposition using standard methods ofimmunofluorescence and fluorescein-conjugated anti-mouseIgG or IgM.

Statistical Analysis. Development of systemic autoimmu-nity was compared by Kaplan-Meier analysis, using a Bon-ferroni adjustment. Rather than using a nominal level of 0.05to indicate marginal significance, a more conservative Pvalue of 0.005 was used to identify important mortalitydifferences. Statistical analysis of autoantibody levels wasperformed using paired t tests. Histological grades werecompared using the Mann-Whitney U test.

RESULTSPhysical Parameters. Mice fed semipurified diet at either

calorie intake level grew and gained body weight (Table 1).Mice restricted in calories consumed 7.1 kcal/day and wereleaner and lighter in body weight than ad libitum fed controls,

80 0

0-

-iI-

0So

60

40

20

0 10 20 30 40AGE (WKS)

FIG. 1. Kaplan-Meier analysis of mortality attributable to sys-temic autoimmunity of W/BF1 mice fed ad libitum (A) (n = 50), fedad libitum until 14, 17, or 22 weeks of age [B14 (n = 10), B17 (n = 20),or B22 (n = 20)] and then reduced in calorie intake (B) or fed RCIlevels from 6 weeks of age B6 (n = 20). *, P < 0.0001; t, P < 0.005.

-A

i-

I---B22

-.B_,

Dl *

! ''---- 6

Applied Biological Sciences: Mizutani et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

4404 Applied Biological Sciences: Mizutani et al.

F 1( 2._. I cidappeurJ (itI /2{c hOMIf ( )t 'Ilh ( })PP().f}S i'( /1b£a~! .'

Proc. Natl. Acad. Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

Proc. Natl. Acad. Sci. USA 91 (1994) 4405

wcc00C)-J

0-J0(I)Ir

3 KIDNEY

2 4000=0 0

1 - co _s*o

A B6 B14 B17 B22

3 _ HEART

2 -c%Bo 0 0

1 _c _ " 0

A B6 B14 B17 B22

FIG. 3. Individual histological scores ofCVD lesions in hearts orof glomerulonephritis in kidneys from mice of each dietary cohort.Hearts and kidneys from group A mice were collected, evaluated,and graded upon autoimmune-based death of each mouse when15-30 weeks old. Hearts and kidneys of group B622 mice werecollected at euthanasia of mice when 40 weeks old. Criteria forscoring are given in the text.

which consumed 10.5 kcal/day. Body weights ofRCI B6, B14,B17, and B22 mice were significantly lower than those ofgroupA mice when the mice were 10, 15, 20, 25, and 35 weeks ofage, respectively (P < 0.001).

Incidence of Fatal Autoimmunity. A significant reduction indeath attributable to systemic autoimmunity occurred amonggroup B6 mice compared to all other groups (P = 0.0001against group A, P = 0.006 against B14, P = 0.002 against B17,P = 0.001 against B22) (Fig. 1). Reducing calories beginningwhen W/BF1 mice were 14 weeks old reduced their mortalitycompared to that ofgroup A mice (P = 0.005). Ninety percentofgroup B6 mice and 60% ofgroup B14 mice survived beyondthe age of 40 weeks. Delaying the initiation of RCI increasedthe incidence of fatal systemic autoimmunity among W/BF1mice. Mortality of group B17 mice was only marginally lessthan group A mice (P = 0.06). Mice of group B22 developedfatal systemic autoimmunity at a rate comparable to that ofad libitum fed group A controls (P = 0.34). Fatal systemicautoimmunity developed in 50% ofmice in groups A, B17, andB22 when mice were 20, 22, and 19 weeks old, respectively.

Incidence and Histopathology of CVD and Glomerulone-phritis. Hearts and kidneys from mice were serially sectionedand examined. Immunologically based lesions within themicrovasculature of the kidney and coronary vessels wereevident. Lesions of murine lupus-associated CVD occurredtypically within ventricular walls and consisted of endothelialcell proliferation, PAS-positive deposition within the coro-nary arterioles, frequent thrombosis, and myocardial necro-sis (Fig. 2). The mean grade ofCVD lesions for all RCI mice,groups B622, was 0.52 + 0.66 and was significantly less thanthat of ad libitum fed group A mice, which had a mean scoreof 2.00 ± 1.00 (P < 0.001; mean ± SD) (Fig. 3). A subintimalbrown precipitate within the coronary vascular walls, indic-ative of IgG and IgM subintimal deposition, was evident in

GROUP aCLA

B6

B14

B17

B22BALB/c

CIC dsDNA

ELISA UNITS

FIG. 4. Mean + SD titers, expressed in ELISA units, of circu-lating autoantibodies to dsDNA or aCLs or ofCICs in sera from miceof each dietary cohort.

immunohistological preparations of heart sections from micewith CVD. This precipitate was more abundant in heartsections from group A mice compared to those of RCI mice(Fig. 2).

Within the kidneys, PAS-positive thickening of glomerulartufts, mononuclear infiltrates, increased mesangial and glo-merular cellularity, and capsular adhesions were evident.The mean grade of glomerular lesions for all RCI mice,groups B6-22, was 0.83 ± 0.77 and was significantly less thanthat of ad libitum fed group A mice, which had a mean scoreof 2.37 ± 0.92 (P < 0.001; mean ± SD). Intense immunoflu-orescence, indicating deposition of IgG or IgM, was evidentin the affected glomeruli of group A mice, while immunoflu-orescence of glomeruli in kidneys of RCI mice was faint ornot observed (Fig. 2).Development of aCLs. The sera of mice were evaluated for

the presence of circulating autoantibodies to dsDNA orcardiolipin and for the presence of CICs (Fig. 4). Mean titersof autoantibodies that reacted with dsDNA (P < 0.05) orcardiolipin (P < 0.01) and levels of CICs (P < 0.01) amongRCI mice were lower than those of ad libitum fed group Amice.

DISCUSSIONThe potential therapeutic benefits of modifying dietary com-position or calorie intake on the progression of autoimmunedisease in humans have not been effectively addressed,typically evaluating dietary modulation ofrheumatoid arthri-tis or occasionally systemic lupus erythematosus (34-39).The potential of RCI in preventing CVD has not beeninvestigated.

Experimentally, limiting dietary calories delays the onsetand arrests the progression of systemic autoimmunity inshort-lived autoimmune-prone strains of mice (17-26). In thecurrent study, we determined that RCI could prevent CVD inW/BF1 mice, and, in those RCI mice where lesions devel-oped, the histological grade of those lesions was less severe.Immunohistological evidence of subintimal deposition ofimmunoglobulin in the coronary vessels of RCI mice wasreduced compared to substantial immunoprecipitates presentin the coronary vasculature of ad libitum fed mice. Thisabrogation of CVD development was accompanied by lower

FIG. 2 (on opposite page). (A) Macroscopic appearance of representative hearts from group A (Left) and group B6 (Right) mice showingsevere, multiple, necrotic myocardial infarctions in hearts from ad libitum fed group A mice and normal appearance of heart from RCI mice.(x3.5.) (B) Representative histopathologic section of CVD, grade 3, from a group A mouse. Note PAS-positive deposits within the media ofthe coronary arterial wall, presence of thrombus, and presence of minor inflammatory infiltrate. (x410; PAS.) (C) Representative histologicalsection of an unaffected heart from RCI group B6 mouse. (x410; PAS.) (D) Immunohistochemical preparation of a representative section ofheart from a group A mouse showing substantial subintimal brown precipitate in several small coronary arterioles, indicating the subintimaldeposition of IgG. (x260.) (E) Immunohistochemical preparation of a representative section of heart from a group B6 mouse showing scant orno subintimal brown precipitate. (x260.) (F and G) Immunofluorescent preparation of representative kidney sections from a group A mouseshowing substantial glomerular deposition of IgG (F) and absence of fluorescent signal in a glomerulus from a group B6 mouse (G). (x415.)

Applied Biological Sciences: Mizutani et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

4406 Applied Biological Sciences: Mizutani et al.

mean anti-cardiolipin and dsDNA autoantibody levels andreduced levels of CICs.Both coronary and glomerular autoimmune-based lesions

ofW/BF1 mice progress rapidly, early in life, so that delayingthe imposition ofRCI until mice are >17 weeks of age did notimpair the progression to more severe vascular pathologythat develops with age in mice of this strain. The mechanismby which early-onset RCI abrogates this aggressive vascularpathology has not been determined. Our previous studieshave shown that transplantation of T-cell-depleted bonemarrow (TCDM) from autoimmune-resistant strains of miceprevents CVD in W/BF1 recipients, showing that the au-

toimmunity of W/BF1 mice is attributable to genetic defectsthat reside within marrow stem cells (30). Conversely, trans-plantation of W/BF1 TCDM to autoimmune-resistant miceregularly induced systemic autoimmunity with CVD andglomerulonephritis in the recipients, and development ofCVD in recipient mice was accompanied by elevated titers ofaCL (30). Since, in the present report, reducing calorie intakewas associated with lower mean titers of aCL, and reducedincidence and severity of CVD, the protective abrogation ofCVD development by RCI in W/BF1 mice could be the resultof actions of the marrow stem cells that prevent the produc-tion of elevated aCL titers.The mechanism by which RCI increases healthful longev-

ity and impairs the development of autoimmune disease hasnot been elucidated, but it may involve neuroendocrine,immunologic, and metabolic systems (40) with effects onrates of cellular proliferation and responsiveness (17, 21),levels of free radical generation and capacity for detoxifica-tion (41, 42), and the efficiency of enzymatic repair ofDNA(43). To our knowledge, the prevention of CVD by limitingcalorie intake has not been reported previously. Early-onsetcalorie restriction not only abrogated the development ofCVD but also inhibited all aspects of W/BF1 autoimmunity.Further, the RCI W/BF1 mice thrived. The accessibility andmalleability of dietary variables, and the abundant experi-mental evidence for RCI's role in preventing and frequentlyreversing autoimmune disease, make it imperative that fur-ther evaluation ofthe role ofthe diet, especially calorie intakelevel, in the prevention ofautoimmune disease be elucidated.

We thank Jaime LeJohn, Una Owens, and Cheryl Thompson forthe excellent care, feeding, and monitoring of the animal collectionand C. Hendrix Brown (Department of Biostatistics and Epidemi-ology, University of South Florida) for assistance with statisticalanalyses. This research was supported by grants from the EleanorNaylor Dana Charitable Trust, National Institutes of Health(AGO5633 and CA41061), and Suncoast Cardiovascular Researchand Education Foundation.

1. Haider, Y. S. & Roberts, W. C. (1981) Am. J. Med. 70,775-781.

2. Bidani, A. K., Roberts, J. L., Schwartz, M. M. & Lewis, E. J.(1980) Am. J. Med. 69, 849-858.

3. Sturfelt, G., Nived, 0., Norberg, R., Thorstensson, R. &Krook, K. (1987) Arthritis Rheum. 30, 382-388.

4. Boey, M. L., Colaco, C. B., Ghanavi, A. E., Elkon, K. B.,Loizou, S. & Hughes, G. R. (1983) Br. Med. J. 287,1021-1023.

5. Harris, E. N., Gharavi, A. E., Boey, M. L., Patel, B. M.,Mackworth-Young, C. G., Loizou, S. & Hughes, G. R. (1983)Lancet 11, 1211-1214.

6. Thiagarajan, P., Shapiro, S. S. & DeMarco, L. (1980) J. Clin.Invest. 66, 397-405.

7. Shapiro, S. & Thiagarajan, P. (1985) Prog. Hemostasis.Thromb. 6, 263-285.

8. Harris, E. N., Asherson, R. A., Gharavi, A. E., Morgan,S. T., Derue, G. & Hughes, G. R. V. (1985) Br. J. Haematol.59, 227-230.

9. Carreras, L. 0., Defreyn, G., Machim, S. J., Vermylen, J.,

Deman, R., Spitz, B. & Van Assche, A. (1981) Lancet i,

244-246.

10. Ruiz-Arguelles, G. J., Ruiz-Arguelles, A., Deleze, M. & Alar-con-Segovia, D. (1989) J. RheumatoL. 16, 381-383.

11. Hang, L. M., Izui, S. & Dixon, F. J. (1981) J. Exp. Med. 154,216-221.

12. Oyaizu, N., Yasumizu, R., Miyama-Inaba, M., Nomura, S.,Yoshida, H., Miyawaki, S., Shibata, Y., Mitsuoka, S., Ya-sunaga, K., Morii, S., Good, R. A. & Ikehara, S. (1988) J. Exp.Med. 167, 2017-2022.

13. Yoshida, H., Fujiwara, H., Fujiwara, T., Ikehara, S. & Ha-mashima, Y. (1987) Am. J. Pathol. 129, 477-485.

14. Takemura, G., Fujiwara, H., Yoshida, H., Wu, Der-Jinn,Matsuda, M., Ishida, M., Kawamura, A., Fujiwara, T. &Kawai, C. (1989) Am. J. Pathol. 135, 989-999.

15. Berden, J. H. M., Hang, L., McConahey, P. J. & Dixon, F. J.(1983) J. Immunol. 130, 1699-1705.

16. Hashimoto, Y., Kawamura, M., Ichikawa, K., Suzuki, T.,Sumida, T., Yoshida, S., Matsuura, E., Ikehara, S. & Koike,T. (1992) J. Immunol. 149, 1063-1068.

17. Engelman, R. W., Day, N. K. & Good, R. A. (1993) Proc. Soc.Exp. Biol. Med. 203, 13-17.

18. Fernanades, G., Friend, P., Yunis, E. J. & Good, R. A. (1978)Proc. Nat!. Acad. Sci. USA 75, 1500-1504.

19. Gajjar, A., Kubo, C., Johnson, B. C. & Good, R. A. (1987) J.Nutr. 117, 1136-1140.

20. Safai-Kutti, S., Fernandes, G., Wang, Y., Safai, B., Good,R. A. & Day, N. K. (1980) Clin. Immunol. Immunopathol. 15,293-300.

21. Ogura, M., Ogura, H., Ikehara, S., Dao, M. L. & Good, R. A.(1989) Proc. Nat!. Acad. Sci. USA 86, 5918-5922.

22. Ogura, M., Ogura, H., Ikehara, S. & Good, R. A. (1989) Proc.Nat!. Acad. Sci. USA 86, 4225-4229.

23. Fernandes, G., Yunis, E. J., Miranda, M., Smith, J. & Good,R. A. (1978) Proc. Natl. Acad. Sci. USA 75, 288-292.

24. Fernandes, G. & Good, R. A. (1984) Proc. Natl. Acad. Sci.USA 81, 6144-6148.

25. Kubo, C., Day, N. K. & Good, R. A. (1984) Proc. Natl. Acad.Sci. USA 81, 5831-5835.

26. Kubo, C., Gaijar, A., Johnson, B. C. & Good, R. A. (1992)Proc. Natl. Acad. Sci. USA 89, 3145-3149.

27. Lok, E., Scott, F. W., Mongeau, R., Nera, E. A., Malcolm, S.& Clayson, D. B. (1990) Cancer Lett. 51, 67-73.

28. Himeno, Y., Engelman, R. W. & Good, R. A. (1992) Proc.Natl. Acad. Sci. USA 89, 5497-5501.

29. U.S. Department of Health and Human Services (1985) Guidefor the Care and Use ofLaboratory Animals (U.S. Departmentof Health and Human Services, Washington, DC), PHS/NIHPubl. No. 86-23.

30. Mizutani, H., Engelman, R. W., Kinjoh, K., Kurata, Y.,Ikehara, S. & Good, R. A. (1993) Blood 82, 3091-3097.

31. Koike, T., Sueishi, M., Funaki, H., Tomioka, H. & Yoshida,S. (1984) Clin. Exp. Immunol. 56, 193-199.

32. Pereira, A. B., Theofilopoulos, A. N. & Dixon, F. J. (1980) J.Immunol. 125, 763-770.

33. Hsu, S. M., Raine, L. & Fanger, H. (1981) J. Histochem.Cytochem. 29, 577-582.

34. Shigemasa, C., Tanaka, T. & Mashiba, H. (1992) Lancet 339,1177.

35. Kjeldsen-Kragh, J., Haugen, M., Borchgrevink, C. F.,Laerum, E., Eek, M., Mowinkel, P., Hovi, K. & Forre, 0.(1991) Lancet 338, 899-902.

36. Beri, D., Malaviya, A. N., Shandilya, R. & Singh, R. R. (1988)Ann. Rheum. Dis. 47, 69-72.

37. Panush, R. S., Carter, R. L., Katz, P., Kowsari, B., Longley,S. & Finnie, S. (1983) Arthritis Rheum. 26, 462-471.

38. Skoldstam, L., Larsson, L. & Lindstrom, F. D. (1979) Scand.J. Rheum. 8, 249-255.

39. Darlington, L. G. & Ramsey, N. W. (1993) Br. J. Rheumatol.32 (Suppl. 6), 507-514.

40. Weindruch, R. & Walfor, R. L. (1988) The Retardation ofAging and Disease by Dietary Restriction (Thomas, Spring-field, IL), pp. 231-293.

41. Koizumi, A., Weindruch, R. & Walford, R. L. (1987) J. Nutr.117, 361-367.

42. Weindruch, R., Walford, R. L., Fligiel, S. & Guthrie, D. (1986)J. Nutr. 116, 641-654.

43. Licastro, F., Weindruch, R., Davis, L. J. & Walford, R. L.(1988) Age 11, 48-53.

Proc. Natl. Acad. Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0