Nutrition and immunology

Decreased Resistance and Immune Response toEscherichia coli Infection in Chicks with Low orHigh Intakes of Vitamin A1

AHARON FRIEDMAN, AMIT MEIDOVSKY,GABRIEL LETTNER AND DAVID SKLAN

Department of Animal Sciences, Faculty of Agriculture,Hebrew University of Jerusalem, Rehouot 76100, Israel

ABSTRACT The effects of vitamin A excess or insufficiency on resistance to Escherichia coll infection andsubsequent anti-E. coli immune responses were examined in chicks. Chicks receiving depleted (0 jig/kg),sufficient (0.85 mg/kg) or excess (1000 mg/kg) levelsof vitamin A in their feed were inoculated by a subcutaneous injection of pathogenic E. coll (1 x IO9 and 2x IO9 cfu per chick). Susceptibility to E. coll was determined by mortality, morbidity and immune responses(antibody production and T lymphocyte proliferation).Excess or insufficient vitamin A led to increased susceptibility of chicks to E. co/i infection; this was accompanied by depressed immune responses. Chicksreceiving excess vitamin A were more sensitive to E. cotíthan vitamin A-depleted chicks. This was reflected inhigher mortality and morbidity rates and in severelydepressed immune responses. In contrast to chicksreceiving excess vitamin A, T lymphocyte responses(though not antibody responses) of vitamin A-depletedchicks achieved levels similar to those of vitamin A-sufficient birds with a lag period of 6 to 10 d. Therefore,reduction in resistance to E. coli infection, resulting fromvitamin A excess or deficiency, probably was compounded by a delayed immune response. J. Nutr. 121:395-400, 1991.

INDEXING KEY WORDS:

•chicks •vitamin A•E. coll •Infection •Immune response

Vitamin A deficiency and excess is associated withdepressed immunity (1-3). Vitamin A deficiency isaccompanied by low levels of serum im-munoglobulins (4—6),impaired irnmunoglobulin (Ig)G(7, 8) and IgA responses (6), reduction in delayed-typehypersensitivity reactions (9-11 ), depressed responsesto mitogens (12, 13) and reduced natural killer cellactivity (14). We previously showed that vitamin Ainsufficiency depressed in vitro T lymphocyteresponses and in vivo antibody production to defined

protein antigens (1). Excess vitamin A decreasedimmune responsiveness in chicks to defined proteinantigens (15), and high intakes of vitamin A werereported to decrease cellular immune functions inmice (16).

Whether or not a decreased immune, response, resulting from deficient vitamin A, increases mortalityand morbidity as a result of infections has not beenestablished. Studies have shown that children deficient in vitamin A are more susceptible to diseasethan those receiving an adequate amount of the vitamin (17-19). The relationship between vitamin Aand infection in chickens has been investigated(20-22). These studies described reduced resistance toinfection associated with marginal vitamin A intake,but they did not discuss possible related immunefunctions or the effects of excess vitamin A.

The purpose of our study was to examine the effects of vitamin A excess or insufficiency on thecapacity of chicks to resist a pathogenic Escherichiacoh challenge and to evaluate the respective immuneresponse.

MATERIALS AND METHODS

Animals. New Hampshire x Leghorn chicks wereobtained from a local hatchery and maintained fromthe day of hatching in temperature-controlledbrooders with free access to water and the experimental diet (15), which was supplemented with 0,0.85 or 1000 mg/kg of retinol equivalent as retinylacetate. Animals were weighed at weekly intervals.

Determination of retinal. Homogenized liver wassaponified with ethanolic KOH and extracted into

'Funded by the Israeli Poultry Marketing Board.

0022-3166/91 $3.00 ©1991 American Institute of Nutrition. Received 4 April 1990. Accepted 5 September 1990.

395

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from

396 FRIEDMAN ET AL.

petroleum ether as described by Vaisman et al. (23).Retinol was determined by reverse-phase high performance liquid chromatography on a C18 column(Varian Instruments, Palo Alto, CA) using retinylacetate as the internal standard (24).

E. coli preparation and inoculations. E. coliserotype 0:78;K:80 was used in this study. The bacteria were grown in nutrient broth (Difco, Detroit,MI) for 16 h at 37*C and harvested at 5000 x g for 10min at 4'C. The pellet was resuspended in sterile

saline, washed three times and resuspended in 10 mLof saline. The number of bacteria were determined bycomparing optical density of a saline suspension at540 nm (Spectronic 70, Bausch & Lomb, Rochester,NY) with a previously calibrated standard. A suspension of 1 x IO10bacteria per milliliter was diluted(1:1) in glycerol and stored at -70*C until used.

Thawed bacteria were seeded on McConkey agar(Difco, Detroit, MI) to establish the number of viablecolony forming units at the time of inoculation. Bacterial death following freezing and thawing was lessthan 5%. Chicks were inoculated by a subcutaneousinjection containing different levels of E. coli in 0.5mL of saline.

T lymphocyte proliferation assay. The in vitroassay for antigen-specific proliferation of T lymphocytes in the chick has been described (15, 25). In brief,peripheral blood leukocytes of chicks, separated byslow centrifugation (35 x g, 15 min), were suspendedin Dulbecco's modified Eagle's medium supplementedwith antibiotics, 1 mmol glutamina, 1% nonessentialamino acids, 1% Na-pyruvate stock solutions (allfrom Biological Industries, Beit Haemek, Israel), 1 xIO-5 mol 2-mercaptoethanol (Sigma, St. Louis, MO)and 1% normal chicken serum. Leukocytes were cultured in 96-well, flat-bottom cluster plates (Costar,Cambridge, MA) in 200 \iL of medium (1 x 10* cellsper well) in the presence or absence of heat-killed E.coli (several doses ranging between 1 x IO6and Ix IO8bacteria per well). The cultures were placed in a humidified incubator containing 7.5% CO2 in air at39'C. Following 92 h of culture, 37 kBq (1 \iCi] of[3H]thymidine [specific activity 370 GBq/mmol (10Ci/mmol); Nuclear Research Center, Dimona, Israel]was added to each well for 4 h. The cultures thenwere harvested onto filters by a multiharvester (Dy-natech, Wesbart, England) and counted in a liquidscintillation counter with channel ratio quench correction calculating results as disintegrations perminute (Packard, La Grange, IL). Counting efficiencywas 55-60%. The results of cells from individualanimals are the average of quadruplicate cultures andare expressed in disintegrations per minute. The average response of cells in the presence of mediumalone (background responses) was 1100 dpm (range,400-2100 dpm); the average response to £.coli ofcells from nonimmunized birds was 900 dpm (range,430-1500 dpm).

Assay of antibody production. Antibodies specificfor E. coli were detected in sera of chicks by anenzyme-linked immunosorbent assay (ELISA)(25). Inbrief, dilutions of chick sera were added to microtiterplates coated with sonicated E. coli (10 ug protein perwell). Dilutions of a negative serum were routinelyadded to all microtiter plates. After three washes toremove excess unbound antibody, the bound chickanti-£.coli antibodies were determined by using aperoxidase-rabbit anti-chicken IgG (heavy and lightchain) (Bio Makor, Rehovot, Israel). Nonspecificbinding was blocked by chick serum that previouslywas shown to have no specific antibody activity. Per-oxidase substrate ABTS, 1 component (KPL Inc.,Gaithersburg, MD) was used for color development.The assay was read in a Bio-Tek ELISA reader (Bio-Tek Instruments, Winooski, VT) at 405 nm after 7min of enzyme reaction at a constant temperature(21*C). The results of individual animals are the av

erage of triplicate measurements and are expressed asa positive/negative ratio at a serum dilution of 1:400.The ratio was calculated by dividing the ELISA absorption of the tested antiserum by that of thenegative serum, which previously was shown to haveno anti-£.coli binding (25). The absorption values ofthe negative serum always were less than 0.2, and thesame negative serum was used in all experiments.

Statistical analyses. Significance of differences (P <0.01) was determined by analysis of variance usingDuncan's multiple range test for analysis of differencebetween groups (26).

RESULTS

To establish the effect of vitamin A nutrition onresistance of chicks to £.coli infection, preliminaryexperiments were performed to establish the pathoge-nicity and lethality of the inoculated bacterial strain.Vitamin A-sufficient chicks were inoculated by asubcutaneous injection with increasing levels ofviable £.coli colony forming units. Ensuing mortalitywas recorded for 8 d and plotted against the respectiveinoculation (Fig. 1). In this experiment, mortalityreached a plateau when levels higher than 4 x IO9cfuwere injected; similar results were obtained in fouradditional experiments. We elected to examine thepathogenic effects of £.coli by using two concentrations of colony forming units that caused 20% and50% mortality rates (1 x 10' cfu per chick and 2 x IO9cfu per chick, respectively) in vitamin A-sufficientchicks.

Hepatic vitamin A status of experimental chickswas determined by 21 d of age and was 48 ±5, 7 ±3and 750 ±25 ng/g for vitamin A-sufficient,-depleted and -excess groups, respectively. The following experiments were initiated on d 21 and terminated within 10 to 20 d (31 to 41 d of age). Nophysiologic manifestations of vitamin A excess or

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from

VITAMIN A NUTRITION AND RESISTANCE TO E. COLI 397

100

80-

60-

S 40 ••o

20

E. COLI INJECTED (CFU x 10 )

FIGURE 1 Mortality rates in vitamin A-sufficient chicksfollowing inoculation with pathogenic £.coi'. Groups of21-d-old chicks (25 chicks per group) were inoculated withdifferent doses of pathogenic £.coi colony forming units.Mortality was scored for 8 d. Mortality, expressed as apercentage of group size, is the group mean,- bars are SEM.

depletion were observed within this time period(weight loss, eye lesions or bone fragility).

Twenty-one-day-old chicks receiving depleted (0ug/kg), sufficient (0.85 mg/kg) or excess (1000 mg/kg)levels of vitamin A in their food were inoculated by asubcutaneous injection of pathogenic E. coli (1 x IO9and 2 x IO9cfu per chick). Mortality was recorded for7 d, then the surviving chicks were inspected for E.coli infections of the liver and pericard (perihepatitisand pericarditis). Chicks with lesions were consideredmorbid. Each group contained 25 chicks, and theexperiment was repeated seven times. The averagemeans of these experiments (Fig. 2) shows that vitamin A-depleted and -excess birds were significantly (P < 0.01) less resistant than vitamin-A-sufficient birds to both doses of infective E. coli. Thiswas shown by increased mortality of the two experimental groups (Fig. 2, upper panel), by morbidity ofthe survivors (Fig. 2, lower panel) and by temporalmortality (Fig. 3 , one experiment representative ofseven). Birds fed excess vitamin A were most sensitive, particularly after receiving large lethal doses of£.coli (Fig. 2, 2 x IO9cfu); all succumbed to disease by96 h postinoculation (Fig. 3). Vitamin A-depletedbirds were less sensitive; fewer birds died, but allsurvivors receiving 2 x IO9 cfu and 85% of thosereceiving 1 x 10' cfu were morbid (compared with50% and 30% in the vitamin A-sufficient group,respectively). The rate of death in vitamin A-depletedchicks was similar to that of vitamin-sufficient birds,but it occurred in larger numbers (Figs. 2 and 3).

We next evaluated whether or not reduced resistance to infection by pathogenic E. coli was accompanied by a parallel decrease in E. coif-specificimmune responses. Chicks were inoculated with 2 xIO9cfu pathogenic £.coif. They were bled every 2 to 3

1x109 CFU 2x10 CFU

100 T

K-~--s

ÉD

eto

K

FIGURE 2 Effects of vitamin A on resistance of chicks topathogenic E. coli infection. Chicks given insufficient (-),sufficient (C)or excess (+)vitamin A (21 d old, 25 chicks pergroup, seven experiments) were inoculated with 1 x IO9or 2x IO9 chi £.coli. Mortality was scored for 7 d. Mortality,expressed as a percentage of group size, is the average ofmeans of seven experiments; bars are SEM (upper panel).Surviving birds were autopsied and scanned for pericarditisand perihepatitis. Chicks with these conditions were Ustedas morbid (lower panel). Morbidity, expressed as a percentage of surviving chicks within each group, is the average of means of seven experiments,- bars are SEM.

days thereafter to determine the presence of anti-£.coli antibodies in their sera (Fig. 4) and to determinethe proliferative response of peripheral blood leukocytes to heat-killed E. coli (Fig. 5).

The antibody titer of vitamin A-depleted chicksincreased with time, but it was significantly lower (P< 0.01) than the titer of vitamin A—sufficientchicksstarting from d 8 postinoculation and throughout theexperiment. The titer peaked sooner (d 8 comparedwith d 11) and diminished more rapidly than therespective titers of the vitamin A-sufficient group.Similar results were obtained in five consecutive experiments, ha additional experiments, when fewerbacteria were injected (1 x IO9), antibody titers ofdepleted chicks were lower than those of vitaminA-sufficient chicks, but absolute titers weresomewhat higher (not shown). Antibody titers ofchicks receiving excess vitamin A and inoculated

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from

398 FRIEDMAN ET AL.

ìo

oce

246

DAYS AFTER E. COLI INOCULATION

FIGURE 3 Effects of vitamin A on mortality rate resulting from infection by pathogenic E. coli. Chicks giveninsufficient (•),sufficient (o) or excess (A) vitamin A (21 dold, 25 chicks per group) were inoculated with 1 x IO9 or 2 xIO9 cfu E. coli. Mortality was scored for 7d. Daily mortalityis expressed as the mean percentage of group size; bars areSEM.

with 2 x IO9cfu E. coli were not assayed for antibodyproduction because chicks died within 3 to 5 d. Antibody responses of chicks with excess vitamin Asurviving 1 x IO9cfu inoculation (less than 15%) weredramatically depressed; positive/negative ratios were2 ±1 and 4 ±1, compared with positive/negativeratios of 8 ±2 and 11 ±3 of vitamin A-sufficientbirds 6 and 10 d after inoculation, respectively (P <0.01 for both groups).

The T lymphocyte proliferative response of vitamin A-depleted chicks receiving 2 x 10' cfu wasseverely depressed (Fig. 5) until 6 d after inoculation(P < 0.01 for d 3, 4 and 6). By d 9 and d 13 responseswere restored to levels similar to those of vitaminA-sufficient chicks; from d 13 on the responses ofboth groups gradually decreased at the same rate (notshown). Similar results were observed in six additional experiments. Similar responses were obtainedfollowing the inoculation of 1 x IO9cfu, although thedegree of T lymphocyte depression was less severe(not shown). T lymphocyte responses of chicks givenexcess vitamin A and receiving 2 x IO9cfu were notevaluated for the reasons stated here. The responsesof survivors receiving 1 x IO9cfu were minimal untild 9 postinoculation (1500 ±350 dpm compared with66,000 ±3500 dpm for vitamin A-sufficient birds; P< 0.005) and increased slightly thereafter(6000 ±1200 dpm compared with 85,000 ±5500dpm for vitamin A-sufficient birds by d 15; P <0.005).

DISCUSSION

In this study, excess or insufficient vitamin A ledto increased susceptibility of chicks to pathogenic E.

OL.

LJGO

oCl-

ce>-oom

8 1 1 14 17 20

DAYS AFTER E. COLI CHALLENGE

FIGURE 4 Effects of vitamin A depletion on anti-E. coliantibody production following a challenge of pathogenic E.coli. Vitamin A-depleted (•)and -sufficient (o) chicks (21 dold, 30 chicks per group) received an inoculation containing2 x IO9 cfu E. coli. Serum was prepared every 2 to 3 dthereafter. The presence of anti-E. coli antibodies in sera ofsurviving birds was determined by ELISA. Antibodyresponses, expressed as the ratio between the tested anti-serum and a negative control serum (positive/negative [P/N]) at a serum dilution of 1:400, are the means of at leastfive chicks per data point; bars are SEM.

coli infection,- this was accompanied by depressedimmune responses. These results confirm and extendour previous studies showing depressed T lymphocyteand antibody responses to nonpathogenic protein antigens following excess or depleted vitamin A in thediet (15).

90000

4 6 8 10 12

DAYS AFTER E. COLI CHALLENGE

14

FIGURE 5 Effects of vitamin A depletion on anti-E. coliT lymphocyte proliferation following a challenge of pathogenic E. coli. Vitamin A-depleted (•)and -sufficient (o)chicks (21 d old, 30 chicks per group) received an inoculation containing 2 x IO9cfu E. cóli.T lymphocyte prolifer-ative responses to heat-killed E. coli were evaluated every 2to 3 d thereafter in surviving chicks. Peripheral blood leukocytes were cultured in the presence of heat-killed E. coli for92 h, then pulsed for 4 h with 37 kBq (1 p.Ci)[^Jthymidine.Proliferative responses, expressed in disintegrations perminute (dpm), are the means of at least four chicks per datapoint; bars are SEM.

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from

VITAMIN A NUTRITION AND RESISTANCE TO E. COU 399

Increased resistance to E. coli infection in chicksreceiving a short-term supplementation of vitamin Ato an adequate diet has been reported (20); however,that study did not address the effects of vitamin Aexcess or depletion on resistance to infection or onimmunoresponsiveness. Marginal vitamin A status inchicks was found to increase the severity of NewCastle virus infections (22), but an examination ofrelevant immune responses was not attempted.

Susceptibility to pathogenic £.coli, determined inour study by mortality and morbidity, was accompanied by depressed anti-£. coli specific immuneresponses, as reflected by reduced antibody production and T lymphocyte proliferation. The degree

1 of susceptibility and immune-response impairmentwas proportional to the dose of pathogenic bacteriainoculated. Note that all immune responses were assessed in chicks that survived the pathogenic challenge. This might have skewed results becauseanimals with the most highly depressed immunefunctions probably succumbed to £.coli infection.

Chicks receiving excess vitamin A were more sensitive to pathogenic E. coli than vitamin A-depletedbirds. This was reflected in higher mortality and morbidity and severely depressed immune responses asexpressed by anti-£.coli antibody production and Tlymphocyte proliferation. The deleterious effects ofexcess vitamin A on the immune response confirmand extend our previous findings (15) in which weemployed a soluble, nonpathogenic protein antigenbovine serum albumin. The mechanism by whichexcess vitamin A depresses the immun«-,response isnot known. We previously suggested that this mightresult from the down-regulation of nuclear receptorsfor vitamin A or, alternatively, might be the result oftoxic elevation of retinyl esters in the blood (15).

In contrast to chicks receiving excess vitamin A, Tlymphocyte responses (though not antibodyresponses) of vitamin A-depleted chicks achievedlevels similar to those of vitamin A-sufficient birdswith a lag period of 6 to 10 d. Therefore, reduction inresistance to £.coli infection probably was compounded by a delayed immune response (1). Previousreports have shown that vitamin A increased in vitroblastogenesis of thymocytes by up-regulating receptors for interleukin-2 (27, 28). Therefore, vitaminA depletion might reduce the rate of in vivo T lymphocyte division and differentiation following anti-genie stimulation. A similar case could be presentedfor B lymphocyte maturation, although antibodylevels of depleted chicks did not achieve those ofvitamin A-sufficient birds. Vitamin A has beenshown to have direct effects on the differentiation ofB lymphocyte hybridomas (29), but it also has beenshown to have negligible effects on IgM production,which is considered to be independent of helper Tlymphocytes (8). Therefore, our data, although nottested directly, could imply that IgM production was

intact and IgG production was defective because production of IgG requires the involvement of helper Tlymphocytes (30).

Reduced resistance to pathogenic £.coli also mayinvolve other leukocytes. Decreased resistance to £.coli infection has been shown to be accompanied byreduced clearance of the bacteria from peripheralblood (Leitner, G. and Heller, E. D., personal communication). In addition, vitamin A has been shownto directly affect monocyte functions (31-34). Thus,the amount of vitamin A in the diet might affect theactivity of several immune system cells and may notbe limited to the malfunction of one cell type. Because the physiologic functions of these cells are different, a general metabolic regulatory pathway couldbe expected rather than a constitutive function ofvitamin A; this theory is supported by the rapidrecovery of immune functions following replacementof vitamin A in depleted rats and chicks (1, 30).

LITERATURE CITED

1. Friedman, A. &. Sklan, D. (1989) Impaired T lymphocyteimmune response in vitamin A depleted rats and chicks. Br. J.Nutr. 62: 439-449.

2. Sklan, D. (1987) Vitamin A in human nutrition. Prog. FoodNutr. Sci. 11: 39-55.

3. Shapiro, P. E. &.Edelson, R. L. (1985) Effects of retinoids on theimmune system. In: Retinoids: New Trends in Research andTherapy. Rctinoid Symposium, Geneva (Saurai, A. H., éd.),pp.225-235, Karger, Basel, Switzerland.

4. Leutskaya, Z. K. &. Fais, D. (1977) Antibody synthesis stimulation by vitamin A in chickens. Biochim. Biophys. Acta 475:207-216.

5. Harmon, B. G., Miller, E. R., Hoefer, J. A., Ullrey, D. E. A.Luecke, R. W. (1963) Relationship of specific nutrient deficiencies to antibody production in swine. I. Vitamin A J.Nutr. 79: 263-268.

6. Sirisinha, S., Darip, M. D., Moongkamdi, P., Ongsakul, M. &Lamb, A J. (1980) Impaired local immune response in vitaminA deficient rats. Clin. Exp. Immunol. 40: 127-135.

7. Bamett, J. B. (1983) Immunomodulating effects of 13-cis-rctinoic acid on the IgG and IgM response of BALB/c mice. Int.Arch. Allergy Appi. Immunol. 72: 227-233.

8. Smith, S. M. &.Hayes, C. E. (1987) Contrasting impairments inIgM and IgG responses of vitamin A deficient mice. Proc. Nati.Acad. Sci. USA 84: 5878-5882.

9. Uhr, }. W., Weissmann, G. &. Thomas, L. (1963) Acute hyper-vitanninosi s A in guinea pigs. Et,Effects on dekyed type hyper-sensitivity. Proc. Soc. Exp. Biol. Med. 112: 287-291.

10. Athanassiades, T. J. (1981) Adjuvant effect of vitamin A pal-mitate and analogues on cell mediated immunity J. Nati.Cancer Inst. 67: 1153-1156.

11. Smith, S. M., Levy, N. S. &. Hayes, C. E. (1987) Impairedimmunity in vitamin A-deficient mice. J. Nutr. 117: 857-865.

12. Nauss, K. M., Mark, D. A &. Suskind, R. M. (1979) The effectof vitamin A deficiency on the in-vitro cellular response ofrats. J. Nutr. 109: 1815-1823.

13. Davis, C. Y. &. Sell, J. L. (1983) Effect of all-trans retinol andretinoic acid nutriture on the immune system of chicks. J.Nutr. 113: 1914-1919.

14. Nauss, K. M. & Newberne, P. M. (1985) Local and regionalimmune function of vitamin A deficient rats with ocularherpes simplex virus (HSV) infections. J. Nutr. 115: 1316-1324.

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from

400 FRIEDMAN ET AL.

15. Friedman, A. & Sklan, D. (1989) Antigen-specific immuneresponse impairment in the chick as influenced by dietaryvitamin A. J. Nutr. 119: 790-795.

16. Dennert, G. &. Lotan, R. (1978) Effects of retinole acid on theimmune system: stimulation of T killer cell induction. Eur. J.Immunol. 8: 23-29.

17. Somer, A., Tarwotjo, I., Hussaini, G. & Susanto, G. (1983)Increased mortality in children with mild vitamin A deficiency. Lancet 2: 585-588.

18. Barclay, A.J.G., Foster, A &. Sommer, A. (1987) Vitamin Asupplements and mortality related to measles: a randomizedclinical trial. Br. Med. J. 294: 294-296.

19. Milton, R. C, Reddy, V. &.Naidu, A. N. (1987) Mild vitamin Adeficiency and childhood morbidity—an Indian experience.Am. J. Clin. Nutr. 46: 827-829.

20. Bang, B. G., Foard, M. A. & Bang, F. B. (1973) The effect ofvitamin A deficiency and Newcastle disease on lymphoid cellsystems in chickens. Ptoc. Soc. Fjtp. Biol. Med. 143:1140-1146.

21. Tengerdy, R. P. &. Nockels, C. F. (1975) Vitamin E or vitaminA protects chickens against E. coli infection. Poult. Sci. 54:1292-1296.

22. Sijtsma, S. R., West, C. E., Rombout, J.H.W.M. &. van derZijpp, A. J. (1989) The interaction between vitamin A statusand Newcastle disease virus infection in chickens. J. Nutr.119: 932-939.

23. Vaisman, N., MogUner, B. M. &. Sklan, D. (1985) Vitamin Aand E content of preterm and term milk. Nutr. Res. 5:931-935.

24. Sklan, D. & Halevy, O. (1984) Vitamin A metabolism in chickliver: some properties of the cytosolic lipid-protein aggregate.Br. J. Nutr. 52: 107-114.

25. Leitner, G., Heller, E. D. & Friedman, A. (1989) Sex related

differences in immune response and survival rate of broilerchickens. Vet. Immunol. Immunopathol. 21: 249-260.

26. Snedecor, G. W. &. Cochran, W. G. (1987) Statistical Methods,7th ed., Iowa State University Press, Ames, Iowa.

27. Sidell, N., Rieber, P. & Golub, S. (1984) Immunological aspectsof retinoids in humans. L Analysis of retinoic acid enhancement of thymocytc responses to PHA. Cell. Immunol.87: 118-125.

28. Sidell, N. & Ramsdell, F. (1988) Retinoic acid upregulatesinterleukin-2 receptors on activated human thymocytes. Cell.Immunol. 115: 299^309.

29. Sherr, E., Adelman, D. C., Saxon, A., Gilly, M., Wall, R. &.Sidell, N. (1988) Retinoic acid induces the differentiation of Bcell hybridomas from patients with common variable immunodeficiency. J. Exp. Med. 168: 55-71.

30. Pasatiempo, AM.G., Bowman, T. A., Taylor, C. E. &. Ross, A.C. (1989) Vitamin A depletion and repletion: effects on antibody response to the capsular polysaccharide of Streptococcus pneuminiae, type in (SSS-m). Am. J. Clin. Nutr. 49:501-510.

31. Goldman, R, (1985) Enhancement of colony stimulating factor-dependent growth of murine macrophage progenitors and theiractivity by retinoic acid. J. Cell. Physiol. 123: 288-296.

32. Rhodes, J. & Oliver, S. (1980) Retinoids as regulators of macrophage function. Immunology 40: 467-471.

33. Wirth, J. J. & Kierszenbaum, F. (1986) Stimulatory effects ofretinoic acid on macrophage interaction with blood forms ofTrypanosoma cruzi : involvement of transglutaminase activity.J. Immunol. 137: 3326-3331.

34. Nakamura, T. & Hemmi, H. (1988) Retinoic acid regulates IgGFc receptor expression in human myelomonocytic leukemiacells and normal peripheral monocytes. Eur. J. Haematol. 41:258-266.

by guest on June 12, 2015jn.nutrition.org

Dow

nloaded from