Cells Homeostatic Proliferation of Memory CD8 T Protein Energy Malnutrition Impairs

Protein Energy Malnutrition Impairs Homeostatic Proliferation of Memory CD8 T Cells

This information is current as Smita S. Iyer, Janel Hart Chatraw, Wendy G. Tan, E. John of September 28, 2021. Wherry, Todd C. Becker, Rafi Ahmed and Zoher F. Kapasi J Immunol 2012; 188:77-84; Prepublished online 23 November 2011; doi: 10.4049/jimmunol.1004027

http://www.jimmunol.org/content/188/1/77 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2011/11/23/jimmunol.100402 Material 7.DC1 http://www.jimmunol.org/ References This article cites 33 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/188/1/77.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 28, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Protein Energy Malnutrition Impairs Homeostatic Proliferation of Memory CD8 T Cells

Smita S. Iyer,*,†,1 Janel Hart Chatraw,‡,x,1 Wendy G. Tan,*,† E. John Wherry,{ Todd C. Becker,‖ Raﬁ Ahmed,*,† and Zoher F. Kapasi‡,x

Nutrition is a critical but poorly understood determinant of immunity. There is abundant epidemiological evidence linking protein malnutrition to impaired vaccine efficacy and increased susceptibility to infections; yet, the role of dietary protein in immune memory homeostasis remains poorly understood. In this study, we show that protein-energy malnutrition induced in mice by low-protein (LP) feeding has a detrimental impact on CD8 memory. Relative to adequate protein (AP)-fed controls, LP feeding in lymphocytic choriomeningitis virus (LCMV)-immune mice resulted in a 2-fold decrease in LCMV-specific CD8 memory T cells. Adoptive transfer of memory cells, labeled with a division tracking dye, from AP mice into naive LP or AP mice demonstrated that protein-energy malnutrition caused profound defects in homeostatic proliferation. Remarkably, this defect occurred despite the lymphopenic en- Downloaded from vironment in LP hosts. Whereas Ag-specific memory cells in LP and AP hosts were phenotypically similar, memory cells in LP hosts were markedly less responsive to polyinosinic-polycytidylic acid-induced acute proliferative signals. Furthermore, upon recall, memory cells in LP hosts displayed reduced proliferation and protection from challenge with LCMV-clone 13, resulting in impaired viral clearance in the liver. The findings show a metabolic requirement of dietary protein in sustaining functional CD8 memory and suggest that interventions to optimize dietary protein intake may improve vaccine efficacy in malnourished individuals. The

Journal of Immunology, 2012, 188: 77–84. http://www.jimmunol.org/

eneration of long-lived immune memory is the basis homeostasis are well studied; it is independent of Ag and de- of vaccination and infection-generated immunity (1). pendent on the concerted action of the g-chain cytokines IL-7, for G Memory T cells arise after Ag clearance from a small survival, and IL-15, for turnover (7, 8). Cell survival and turnover subset of effectors derived from the clonal expansion and differ- also depend on nutrient availability; very little, however, is known entiation of naive T cells responding to Ag (2, 3). Upon Ag re- about how nutrition impacts memory T cell homeostasis. Because encounter, Ag-specific memory T cells respond more rapidly and proliferation of memory cells imposes a metabolic demand on efficiently than naive T cells, conferring the host with long-term amino acid supply to support protein synthesis, dietary protein by guest on September 28, 2021 protection (4). This dynamic memory pool is maintained at a rel- may be critical in sustaining T cell memory. atively constant size for prolonged periods, even up to a lifetime, The present study was designed to determine whether homeo- by homeostatic proliferation (5). static proliferation of memory CD8 T cells is impaired in protein- Homeostatic proliferation is a process in which influx of newly energy malnourishment (PEM). PEM is a major form of malnu- generated memory cells, by slow turnover, is balanced by death of trition and defined as an imbalance between intake of protein and pre-existing memory cells (6). The signals needed for memory energy and the optimal requirement to ensure the most favorable body growth and function (9). Malnutrition is a major health concern among millions in the developing world, and accumulating *Emory Vaccine Center, Emory University, Atlanta, GA 30322; †Department of evidence suggests that malnutrition also afflicts HIV-infected Microbiology and Immunology, School of Medicine, Emory University, Atlanta, patients, individuals with chronic disease, and aged populations in GA 30322; ‡Division of Physical Therapy, Department of Rehabilitation Medicine, School of Medicine, Emory University, Atlanta, GA 30322; xGraduate Program in industrialized nations (10, 11). There is abundant epidemiological Nutrition Health Sciences, Emory University, Atlanta, GA 30322; {Department of ‖ evidence suggesting that malnutrition impairs vaccine efficacy and Immunology, University of Pennsylvania, Philadelphia, PA 19104; and Division of increases susceptibility to infections; yet, the underlying mecha- Dermatology, David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, CA 90095 nisms remain poorly understood (9, 12). A better understanding of 1S.S.I. and J.H.C. contributed equally to this work. the role of nutrition in memory homeostasis may open avenues for Received for publication December 15, 2010. Accepted for publication October 22, nutritional interventions to improve vaccine efficacy in malnour- 2011. ished individuals. This work was supported by National Institutes of Health Research Grant We assessed quantitative and qualitative aspects of CD8 T cell AI30048 (to R.A.). memory in lymphocytic choriomeningitis virus (LCMV)-immune Address correspondence and reprint requests to Dr. Zoher F. Kapasi, Division of mice (.40 d postinfection) fed either a low-protein (LP) diet to Physical Therapy, Department of Rehabilitation Medicine, Emory University School develop PEM or an adequate-protein (AP) control diet. After 4 wk of Medicine, 1441 Clifton Road Northeast, Atlanta, GA 30322. E-mail address: [email protected] of dietary intervention, LP mice demonstrated a 2-fold reduction in The online version of this article contains supplemental material. Ag-specific memory CD8 cells compared with AP-fed controls, Abbreviations used in this article: AP, adequate-protein; LCMV, lymphocytic cho- suggesting impaired proliferation due to PEM. Using adoptive riomeningitis virus; LP, low-protein; MFI, mean fluorescence intensity; PEM, transfer of CFSE-labeled memory cells from AP mice into naive LP protein-energy malnutrition; poly(I:C), polyinosinic-polycytidylic acid; VV, vaccinia or AP mice, we confirmed that PEM caused profound defects in virus. homeostatic proliferation. Interestingly, memory cells in AP or LP Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 hosts were phenotypically similar with respect to the g-chain cy- www.jimmunol.org/cgi/doi/10.4049/jimmunol.1004027 78 MEMORY T CELLS IN PROTEIN MALNUTRITION tokine receptors, but were markedly less responsive to polyinosinic- Results polycytidylic acid [poly(I:C)]-induced acute homeostatic prolifer- PEM decreases CD8 T cell memory ation. Additionally, on challenge with LCMV-clone 13, memory Increased incidence of infections in protein-malnourished indi- cells in LP hosts displayed markedly reduced proliferation resulting viduals suggests that dietary protein may be a critical determinant in impaired viral clearance in the liver. Together, the data show that of immune memory. But the role of nutrition in memory homeo- dietary protein is critical for sustaining a functional memory CD8 stasis is yet to be precisely defined. The objective of the current pool. study was to determine whether PEM impairs CD8 memory homeostasis. Because PEM compromises the CD8 effector response Materials and Methods to LCMV (14), which, in turn, could contribute to decreased CD8 Mice, virus, and infections memory, we designed the study to specifically examine the effect Four- to 6-wk-old female C57BL/6 mice were obtained from The Jackson of PEM on established CD8 memory. Laboratory (Bar Harbor, ME). Thy1.1+ P14-transgenic mice with CD8 Mice were infected with the Armstrong strain of LCMV, which T cells expressing the TCR specific for the Db/GP33-41 epitope of LCMV initiates a vigorous immune response resulting in clearance of were obtained from The Jackson Laboratory and backcrossed to B6 mice in our colony. Mice were housed in cages and maintained on a 12-h light/ virus and the subsequent generation of a robust memory pool 12-h dark cycle at the Division of Animal Resources at Emory University. (3). LCMV-immune mice (.40 d postinfection) were randomly All experiments were initiated during the light cycle. Mice were infected assigned to receive either an AP diet (18% dietary protein) or an with 2 3 105 PFUs LCMV Armstrong i.p. For secondary challenge LP diet (0.6% dietary protein) (Fig. 1A). The diets were formu- 3 6 experiments, 2 10 PFU LCMV-clone 13 was given i.v. Recombinant lated to be isocaloric; therefore, the LP diet provided excess en- vaccinia virus (VV)-GP33 expressing the GP33–41 epitope of LCMV was Downloaded from propagated and used as previously described (13). Mice were infected with ergy from carbohydrates (14). Because optimal amino acid supply 2 3 106 PFU VV-GP33. All animal protocols were reviewed and approved is critical for cell proliferation and function, the effect of LP diet by the Institutional Animal Care and Use Committee at Emory University. on CD8 T cells may be directly attributed to limited amino acid supply in the LP diet. Diets As previously described, after 4 wk on a LP diet, mice become Semipurified diets were custom-prepared (Harlan-Teklad, Madison, WI) to protein-energy malnourished (14). This experimental setup allows test the specific effects of protein energy malnutrition on memory CD8 us to quantitatively assess the effects of PEM on CD8 memory http://www.jimmunol.org/ T cell homeostasis (14). Protein levels were controlled at the desired experimental level by decreasing amounts of essential and nonessential homeostasis without the confounding effects of malnutrition on amino acids. The LP diet contained 0.6% (5 g/kg) of protein, and the AP memory precursors and/or viral control. diet contained 18% protein (210 g/kg) by weight. LP diets contained more As demonstrated previously, weight loss occurs within 1 wk of carbohydrate in the form of sucrose and cellulose compared with AP diets initiating the LP diet and continues throughout the dietary period but were otherwise isocaloric, and both diets contained equal and identical (14). After 4 wk, mice on the LP diet showed a significant weight quantities of energy, fat, and micronutrients. Animals had free access to water and food at all experimental time points. loss of 20% of body weight, whereas body weight in AP mice remained relatively stable (data not shown). We first examined Flow cytometry

whether PEM quantitatively impaired CD8 T cell memory. To this by guest on September 28, 2021 Single-cell suspensions were prepared from spleen, and 106 cells were end, lymphocytes were isolated from the spleen of AP and LP stained in PBS containing 2% FBS and 0.02% sodium azide (FACS buffer) mice and examined for memory cells specific for the LCMV for 30 min at 4˚C followed by three washes in FACS buffer. Cells were epitopes Db/GP276-286 and Db/GP33-41. Whereas no differences stained with anti–CD8a-PerCP (53-6.7), Thy1.1-PE (OX-7), CD122-PE in percentage of CD44hi (33% in both groups) or total Ag-specific (TM-b1), Bcl-2–FITC (3F11), IFN-g–FITC (XMG1.2), TNF-a–allophy- cocyanin (MP6-XT22), IL-2–PE (JES6-5H4), Ki-67–FITC (B56) (all from cells were observed between both groups (Fig. 1B), estimation of BD Biosciences), CD127–PE-Cy7 (A7R34; eBioscience, San Diego, CA), total numbers of Ag-specific T cells in the spleen revealed a 2-fold IL-15Ra–FITC (BAF551; R&D Systems, Minneapolis, MN), and CD44- decrease in LP mice compared with AP mice (p , 0.05; Fig. 1B). Pacific Blue (IM7; BioLegend, San Diego, CA). Intracellular staining for The size of the total CD44-high memory pool was decreased by Bcl-2 and Ki-67 was performed directly ex vivo. Cytokine staining for IFN-g, TNF-a, or IL-2 was done after a 5-h in vitro stimulation with 0.1 2.8-fold in malnourished mice. mg/ml gp33 peptide using the Cytofix/Cytoperm kit, according to the We next determined whether the functional capacity of Db/ manufacturer’s instructions (BD Biosciences). MHC class I peptide tet- GP276 memory cells was also impaired in LP mice. For this ramers were generated and used as described (15). Samples were acquired purpose, splenocytes were stimulated for 5 h with GP276-286 using either FACSCalibur or FACS Canto II instruments (BD Biosciences). peptide, and cytokine production was determined. As measured Data were analyzed using FlowJo software (Tree Star). by the mean fluorescence intensity (MFI), IFN-g, TNF-a, and IL-2 CFSE labeling and adoptive transfer production was not different between the two dietary groups, suggesting that cytokine production on a per-cell basis ex vivo was CFSE labeling was performed as described (16). Splenocytes were obtained from immune C57BL/6 mice or immune P14 chimeric mice. Fifty million notimpairedbyPEM(Fig.2A,2B). However, the total number CFSE-labeled splenocytes were adoptively transferred i.v. via tail vein of cytokine-producing cells was significantly reduced in LP mice injection into naive recipients that had been on the LP or AP diet for 2 wk. (Fig. 2C). Longitudinal monitoring of T cell proliferation was performed by saphe- nous vein bleeding into 4% sodium citrate under isofluorane anesthesia. PEM decreases acute homeostatic proliferation of CD8 Poly(I:C) treatment memory T cells Mice were treated with 150 mg poly(I:C) (Sigma-Aldrich, St. Louis, MO) Decreased CD8 memory in LP hosts could be due to defects in dissolved in 0.5 ml PBS or PBS alone by i.p. injection, as previously homeostatic proliferation; because the receptors for the g-chain described (17). Splenocytes and bone marrow cells were analyzed 3 d later cytokines are key signals for homeostatic proliferation (8, 18), we for cycling memory CD8 T cells by Ki-67 staining. first determined whether expression of IL-7R and IL-15R was Statistical analysis decreased on memory cells in LP mice. We found no difference Data are expressed as the mean + SEM. Statistical analysis was performed either in percentage of receptor-positive cells or MFI of Ag- by the two-tailed Student t test using Prism software (GraphPad, La Jolla, specific CD8s between both dietary groups (Supplemental Fig. CA). A p value #0.05 was used to determine significance. 1; solid histogram denotes expression in GP276 tetramer+ CD8 The Journal of Immunology 79

FIGURE 1. Effect of PEM on LCMV- specific memory CD8 T cells. A, Experi- mental outline: 6–8-wk-old B6 mice were infected i.p. with LCMV Armstrong and after 40 d postinfection, mice were assigned to receive either an AP (18% dietary protein) or an LP diet (0.6% dietary protein). Analysis of LCMV-specific memory CD8 T cells in the spleen was performed after 4 wk of dietary treatment. B, Db/GP276 and Db/GP33 tetramer staining of CD8 T cells from the spleen of AP- (top panel) and LP- fed (bottom panel) mice. Data are percent of tetramer-positive CD8 T cells. C, Abso- lute numbers of Ag-specific CD8 T cells in the spleen in AP- (black bars) and LP-fed Downloaded from (white bars) mice. One representative experiment of three is shown. Data are mean of four mice per group 6 SEM. *p , 0.05. http://www.jimmunol.org/

T cells, and black lines denote expression levels in CD44-low marrow, a major site for homeostatic proliferation (17), and di- naive CD8 T cells). These data suggested that memory cells in vision of memory cells was assessed by staining lymphocytes with the LP environment were poised to receive proliferative signals the cell proliferation marker Ki-67 (Fig. 3A). As shown in Fig. 3B, from IL-15. To directly assess responsiveness of memory CD8 frequency of Ki-67–positive cells (of total CD44-high CD8 T cells) T cells to acute proliferative signals, we treated AP and LP was lower in the spleen and bone marrow in LP mice, suggesting LCMV-immune mice with poly(I:C). Poly(I:C), an RNA analog that PEM caused defects in acute homeostatic proliferation. Fre- that induces IFN-g, which in turn upregulates IL-15 production, quency of LCMV-speciﬁc cells (gated on GP276 tetramer+ cells) by guest on September 28, 2021 leading to proliferation of memory CD8 T cells in an IL-15– responding to poly (1:C) treatment was also markedly decreased dependent manner (17, 19). Three days after poly(I:C) treat- (5% in LP mice compared with 21% in AP controls; Supplemental ment, lymphocytes were isolated from the spleen and bone Fig. 2).

FIGURE 2. Effect of PEM on cytokine production by LCMV-speciﬁc memory CD8 T cells. A, Intracellular cytokine staining for IFN-g, TNF-a, andIL-2after5hofGP276peptidestimulation. Data are percent of CD8 T cells that produced cytokine. B, MFI of cytokine expression after GP276 stimulation. C, Total cytokine-producing cells in the spleen after GP276 in AP- (black bars) and LP-fed (white bars) mice. One representative experiment of three is shown. Data are mean of four mice per group 6 SEM. *p , 0.05. 80 MEMORY T CELLS IN PROTEIN MALNUTRITION

FIGURE 3. Acute homeostatic proliferation of memory CD8 T cells in PEM. A,AP-andLP- fed Armstrong immune mice were injected i.p. with poly(I:C), and division of CD44hi CD8 T cells was determined 3 d later. B,Percentageof CD44hi CD8 T cells that responded to poly(I:C) in the spleen and bone marrow was assessed by intracellular staining with Ki-67. Histograms show median ﬂuorescence intensity of Ki-67 in CD44hi CD8 T cells; gray dashed lines show Ki-67 staining in CD44lo CD8 T cells. One representative mouse of four from two experimental repli- cates is shown. Downloaded from http://www.jimmunol.org/

PEM decreases basal homeostatic proliferation of CD8 an AP environment were labeled with the division tracking dye memory T cells CFSE and adoptively transferred into naive recipients that had In addition to defects in acute homeostatic proliferation, lower been maintained either on AP or LP diets for 2 wk (Fig. 4A). numbers of LCMV-speciﬁc memory CD8 T cells in LP mice Dietary regimes were continued for 4 wk after cell transfer. At this suggested that PEM may also impair basal homeostatic prolifer- time, lymphocytes were isolated from the spleen to compare di- ation. To determine if this was the case, memory cells generated in vision of CFSE-labeled cells between AP and LP mice. Inspection by guest on September 28, 2021

FIGURE 4. Basal homeostatic proliferation of memory CD8 T cells in PEM. A, Splenocytes were prepared from LCMV-immune mice, labeled with CFSE, and adoptively transferred into naive B6 recipients receiving either an AP or LP diet. Diets were continued for 4 wk after cell transfer. B, Homeostatic proliferation of donor cells was assessed in the spleen by ex- amining CFSE dilution. Histogram shows percent of transferred cells that divided. Bar graph shows total numbers of CD44hi T cells per spleen. In C, donor cells were obtained from Armstrong-immune P14 chimeric mice. Histogram shows percent of divided donor cells. Kinetic analysis of divided donor cells in blood is shown in the graph. One representative experiment of three is shown. Data are mean of three mice per group 6 SEM. *p , 0.05. The Journal of Immunology 81

FIGURE 5. Protective immunity of memory CD8 T cells in PEM. A, To assess protective immunity of memory CD8 T cells in PEM, memory cells from AP- fed P14 chimeric mice were labeled with CFSE and transferred to naive AP- or LP-fed recipients. Recip- ients were challenged with LCMV-clone 13, and recall responses were assessed 7 d after transfer. B shows a representative flow plot of percentage of Thy 1.1+Db/GP33+ tetramer-specific CD8 T cells in AP- and LP-fed hosts. Bar graph shows total Db/GP33 tetramer-specific CD8 T cells per spleen. In C, intracellular cytokine staining for IFN-g and IL-2 after 5 h of GP33 peptide stimulation is shown. Data are percent of CD8 T cells that produced both IFN-g and IL-2. In D, viral titers in liver and lung are shown. Downloaded from *p , 0.05. http://www.jimmunol.org/ of CFSE profiles revealed that the vast majority of memory cells in PEM impairs recall responses of memory CD8 T cells LP recipients had failed to divide compared with efficient ho- Because rapid recall response to secondary infection is a charac- meostatic proliferation in AP recipients (Fig. 4B). Overall, the teristic feature of memory cells (3), we next investigated whether number of CD44hi, CFSE-labeled CD8 T cells was 8-fold lower in PEM also compromised the ability of memory CD8 T cells to re- LP mice compared with AP mice (0.3 6 0.2 3 105 in LP mice 6 3 5 , spond to a secondary infection. For this purpose, memory cells from versus 2.5 0.6 10 in AP mice; p 0.001). + In a separate experiment, CFSE-labeled memory cells from im- Thy 1.1 P14 chimeric-immune mice were adoptively transferred by guest on September 28, 2021 mune mice containing Thy 1.1 P14-transgenic memory cells specific into naive recipients that had been maintained either on AP or for the Db/ GP33-41 LCMV epitope were transferred to naive re- LP diets for 2 wk. Mice were infected with the clone 13 strain of cipients that had been on an AP or LP diet for 2 wk. Dietary regimes LCMV at 1 d posttransfer, and the secondary memory response were continued 4 wk after cell transfer. Homeostatic proliferation of was assessed at day 7 posttransfer (Fig. 5A). transferred cells was tracked by longitudinal bleeding each week for At 1 d posttransfer, similar numbers of Thy1.1+ P14 cells were a total of 4 wk. The data showed that transferred cells in AP mice present in adoptive AP and LP hosts (data not shown). However, began to divide, whereas transferred cells in LP mice divided at examination of secondary effector proliferation at day 7 revealed a a much slower rate such that the percentage of divided Thy1.1+ P14 10-fold decrease in secondary effectors in LP hosts compared with memory cells was significantly higher in the AP mice at days 22 and AP hosts (Fig. 5B; AP mice, 30.3 3 105 6 8.2; LP mice, 3.1 3 105 6 29 (Fig. 4C). A representative flow plot of donor cells in the blood at 0.8; p , 0.05). A representative flow cytometric analysis of cells is day 22 is shown. Thus, the data clearly demonstrate that PEM im- shown. We stained for several surface activation markers such as paired homeostatic proliferation of CD8 memory cells. CD27, CD127, CD69, CD62L, programmed death-1, CD122, and

FIGURE 6. In a separate experiment (A), malnourished memory cells from LP-fed mice were transferred to naive AP- or LP-fed recipients subse- quently challenged with VV-GP33. Recall responses were assessed 7 d after transfer. B shows a representative ﬂow plot of percentage of Thy 1.1+Db/GP33+ tetramer-speciﬁc CD8 T cells in AP and LP hosts. The total number of tetramer-positive cells per spleen is shown in C. One representative experiment of three is shown. Data are mean of three mice per group 6 SEM. 82 MEMORY T CELLS IN PROTEIN MALNUTRITION

FIGURE 7. Protein supplementation rescues malnutrition-induced defect in CD8 T cell memory maintenance. A, Experimental outline: 6–8-wk-old P14- chimeric mice were infected i.p. with LCMV Armstrong, and 40 d postinfection, mice were assigned to receive either an AP (18% dietary protein) or LP diet (0.6% dietary protein). A third group of mice received LP diet for 4 wk followed by switch to AP diet. Analysis of LCMV-speciﬁc memory CD8 T cells in the spleen was performed after 8 wk of dietary treatment. B, Percentage, absolute numbers of GP33-speciﬁc donor CD8 T cells, and total CD44-high Downloaded from memory CD8 T cells in the spleen in AP (black bars), LP (white bars), and LP-rescue (striped bars) mice. One representative experiment of two is shown. Data are mean of four mice per group + SEM. *p , 0.05.

CD127 and found no difference in expression of these molecules impaired in malnourishment due to defects in homeostatic pro- between AP and LP hosts (data not shown). Furthermore, Thy1.1+ liferation; and second, that memory cells in malnourished hosts P14 cells in AP and LP hosts were functionally competent and could have defects in recall responses, leading to impaired viral clear- produce IFN-g,TNF-a, and IL-2 upon GP33 peptide stimulation ex ance upon reinfection. These findings have important implications http://www.jimmunol.org/ vivo (Fig. 5C). The observation that LP hosts were unable to support for understanding the basis of increased infections in malnour- robust expansion of secondary effectors would suggest that LP hosts ished individuals in the world’s poorest countries. The data also would be less effective at controlling virus compared with AP hosts. suggest that malnourishment that frequently accompanies aging Indeed, assay of viral titers in the liver revealed significantly higher and chronic diseases may contribute to impaired immunity ob- titers in LP mice compared with AP mice (LP mice, 5.0 6 0.07 served in these states. In all, the data show that a robust memory log10/mg of tissue; AP mice, 4.2 6 0.4 log10/mg of tissue; p , 0.05; response generated under nutritionally optimal conditions is im- Fig. 5D). No difference in viral titers was observed in the lung. In paired by poor nutrition. a separate experiment, we assessed recall proliferation of malnour- The complex and reciprocal relationship between nutrition and by guest on September 28, 2021 ished memory cells in AP versus LP hosts (Fig. 6A). Representative infections has long been recognized. Studies in humans and ex- flow plots of LP donor cells in AP and LP hosts are shown in Fig. perimental animals show that infections place a significant met- 6B. The data show that recall proliferation of LP memory cells in abolic demand on the host; for instance, increased mobilization of AP hosts was superior to that in LP hosts (Fig. 6C), suggesting that amino acids from peripheral tissues, largely skeletal muscle, malnutrition-induced defects in proliferation may be rescued by AP supports gluconeogenesis in the liver to meet the heightened supplementation. glucose needs of rapidly dividing immune cells (12, 23, 24). Cell- culture studies show that activated T cells increase amino acid Impaired memory maintenance in malnourishment is rescued uptake by ∼6-fold compared with resting cells (25). Collectively, by dietary protein supplementation these data indicate that activated T cells have increased amino Finally, we wanted to assess whether malnourishment-induced impairment in memory maintenance could be rescued by dietary intervention with an AP diet. For this purpose, Armstrong immune P14 chimeric mice were assigned to one of three dietary regimens as outlined in Fig. 7A: 1) AP for 8 wk; 2) AP for 4 wk followed by switch to LP for 4 wk; and 3) LP for 4 wk followed by a switch to AP for 4 wk (LP-rescue). After 8 wk of each respective dietary treatment, lymphocytes were isolated from the spleen of AP, LP, and LP-rescue mice and examined for Thy 1.1+ memory cells specific for the LCMV epitope Db/GP33-41 (Fig. 7B). Estimation of total numbers of Ag-specific T cells in the spleen revealed that AP supplementation of LP hosts resulted in a complete quantitative rescue of memory CD8 T (Fig. 7C). Together, these data underscore the importance of dietary amino acid availability in optimal CD8 memory homeostasis (Fig. 8). FIGURE 8. PEM impairs CD8 memory homeostasis. Malnourished Discussion memory CD8 T cells demonstrate markedly reduced homeostatic proliferation despite no measurable decrease in cell-surface expression of the Although nutrition is recognized as being a critical determinant of g-chain cytokine receptors. Malnourished memory CD8 T also conferred immunity, an understanding of the mechanisms by which poor reduced protective immunity due to impaired recall proliferation. The data nutrition impairs immune function remains poorly characterized show that dietary protein is critical for sustaining a long-lived and func- (20–22). The current study bridges this gap by presenting two tional memory CD8 T cell pool and that this defect may be rescued by main findings: first, that maintenance of memory CD8 T cells is protein supplementation. The Journal of Immunology 83 acid requirements to support at least two fundamental processes, Indeed, replenishing protein in malnourished hosts rescued the increased bioenergetic demands and increased protein synthesis. proliferative defects due to malnourishment (Figs. 7, 8). Further- It therefore comes as no surprise that cell-mediated immune more, recall proliferation was superior during adoptive transfer of responses are compromised in protein-malnourished individuals. LP memory cells into AP hosts compared with LP hosts (Fig. 6). Salimonu et al. (26) found significantly lower T lymphocyte count Although this result points to the critical role of environmental/ in malnourished compared with well-fed children at baseline and host factors in regulating memory homeostasis, it does not exclude at 3, 10, and 21 d postimmunization with measles virus. Several the possibly that any T cell-intrinsic defect resulting from mal- studies have also shown decreased responses to tuberculin test in nutrition was also corrected as a consequence of protein supple- infants and children with PEM (21, 27). More recently, we dem- mentation. It is likely, therefore, that a combination of factors, onstrated that PEM compromises the effector response to LCMV both environmental and T cell intrinsic, contribute to impaired infection in mice (14). Decrease in effector responses led to a 3 memory homeostasis in malnourishment. Experimental targeting log-fold increase in serum viral titers in PEM mice at day 8, of pathways involved in amino acid transport and metabolism in whereas virus was completely cleared in control mice. Thus, there T cells and subsequent response in well-nourished and malnour- is abundant evidence indicating that malnutrition impairs effector ished mice would be required to thoroughly dissect the role of CD8 responses to infection. However, relatively little is known T cell-intrinsic versus host/environmental factors in malnutrition. about whether malnutrition impairs memory maintenance. It is noteworthy that in malnourished children, an improvement In the current study, we demonstrate that PEM induced a 2-fold in immune function, as measured by Ab titers, after protein sup- decrease in the memory CD8 compartment. Total lymphocyte plementation has been reported (31). In contrast, however, nu- numbers were also decreased, suggesting that mice on an LP diet merous studies have failed to find significant benefit of protein Downloaded from were lymphopenic. Studies have shown that lymphopenia, resulting supplementation in the severely malnourished (32, 33). It is pos- from severe infection, chemotherapy, irradiation, or malnourish- sible that there may be a narrow window for successful inter- ment, induces proliferation of naive cells to memory cells that vention, and the ability to rescue immune function may depend on are phenotypically indistinguishable from true Ag-experienced a number of factors including severity of malnourishment, type of memory cells (28). Lymphopenia-induced proliferation of naive infection, and age of the host. Further studies are needed to de-

cells to memory-phenotype cells would lead to an overestimation termine whether memory homeostasis can be rescued by amino http://www.jimmunol.org/ of the memory compartment in LP hosts. Thus, the true deficit acid supplementation in malnourished individuals. Indeed, rigor- induced by PEM may be underrepresented. However, it is also ous studies are also need to determine the impact of malnutrition possible that thymic output was decreased in LP mice, and/or on immune responses to vaccines with marginal efficacy, as this is circulating levels of IL-7 and IL-15 in LP mice were not suffi- of direct clinical and public health relevance. cient to induce proliferation of naive cells to memory-phenotype In summary, the data show that dietary protein intake is criti- cells (29). cal for the maintenance of a functional CD8 memory pool. The Impaired memory homeostasis in PEM could result from de- findings suggest that interventions to optimize dietary protein increased proliferation, increased cell death, or a combination of take may improve vaccine efficacy in malnourished individuals and both. Adoptive transfer experiments using CFSE-labeled cells decrease susceptibility to infections in chronically infected mal- by guest on September 28, 2021 showed that at 4 wk posttransfer, 77% of donor cells had an un- nourished patients. diluted CFSE profile in LP hosts compared with 25% in AP hosts. This suggests that defects in cell proliferation contributed to de- Acknowledgments creased memory pool in PEM. Further, comparable levels of the We thank Bogumila Koneiczny, Yelena Blinder, and Reben Rahman for ex- prosurvival protein Bcl-2 and Bcl-xL and the proapoptotic protein cellent technical assistance. Bim in memory cells from LP and AP mice argue against a proapoptotic memory phenotype in LP hosts (Fig. 3). Disclosures In addition to defects in basal homeostatic proliferation, memory The authors have no financial conflicts of interest. cells in LP hosts showed impaired responses to poly(I:C), which induces acute proliferation of memory cells in an IFN-g–, IL-15– dependent manner (19). This defect, which was also seen in the References bone marrow, one of the major sites of homeostatic proliferation, 1. Kaech, S. M., E. J. Wherry, and R. Ahmed. 2002. Effector and memory T-cell differentiation: implications for vaccine development. Nat. Rev. Immunol. 2: could be due to at least three reasons. First, cytokine levels in- 251–262. duced by poly(I:C) in LP hosts may not have been sufficient to 2. Masopust, D., S. M. Kaech, E. J. Wherry, and R. Ahmed. 2004. The role of drive increased proliferation. Indeed, amino acids such as gluta- programming in memory T-cell development. Curr. Opin. Immunol. 16: 217– 225. mine are critical for production of IFN-g in T cells (25). Second, 3. Wherry, E. J., and R. Ahmed. 2004. Memory CD8 T-cell differentiation during impaired activation of APC such as dendritic cells in LP hosts viral infection. J. Virol. 78: 5535–5545. could also be a contributing factor. Third, limited amino acid 4. Murali-Krishna, K., J. D. Altman, M. Suresh, D. J. Sourdive, A. J. Zajac, J. D. Miller, J. Slansky, and R. Ahmed. 1998. Counting antigen-specific CD8 availability in LP hosts could interfere with efficient cell-cycle T cells: a reevaluation of bystander activation during viral infection. Immunity 8: progression and turnover. In vitro studies by Angelini et al. (30) 177–187. 5. Ahmed, R., and D. Gray. 1996. Immunological memory and protective immu- showing that APC control T cell proliferation by controlling nity: understanding their relation. Science 272: 54–60. availability of cysteine in the extracellular fluid underscore the 6. Surh, C. D., and J. Sprent. 2005. Regulation of mature T cell homeostasis. Semin. importance of nutrient availability in T cell turnover. Immunol. 17: 183–191. 7. Boyman, O., J. F. Purton, C. D. Surh, and J. Sprent. 2007. Cytokines and T-cell Interestingly, despite demonstrating marked deficiencies in basal homeostasis. Curr. Opin. Immunol. 19: 320–326. and acute proliferation, memory cells in LP hosts did not appear to 8. Becker, T. C., E. J. Wherry, D. Boone, K. Murali-Krishna, R. Antia, A. Ma, and be phenotypically different from fully functional memory cells. R. Ahmed. 2002. Interleukin 15 is required for proliferative renewal of virus- specific memory CD8 T cells. J. Exp. Med. 195: 1541–1548. This observation raises the possibility that the effects of malnu- 9. Savy, M., K. Edmond, P. E. Fine, A. Hall, B. J. Hennig, S. E. Moore, K. Mulholland, trition may not be T cell intrinsic but rather may be environmental U. Schaible, and A. M. Prentice. 2009. Landscape analysis of interactions between nutrition and vaccine responses in children. J. Nutr. 139: 2154S–2218S. and suggest that impaired memory maintenance in malnourished 10. Chinen, J., and W. T. Shearer. 2010. Secondary immunodeficiencies, including individuals may be amenable to rescue by protein supplementation. HIV infection. J. Allergy Clin. Immunol. 125(Suppl 2): S195–S203. 84 MEMORY T CELLS IN PROTEIN MALNUTRITION

11. Jewett, J. F., and F. M. Hecht. 1993. Preventive health care for adults with HIV Immunodeficiency associated with anorexia nervosa is secondary and improves infection. JAMA 269: 1144–1153. after refeeding. Immunology 94: 543–551. 12. Scrimshaw, N. S., and J. P. SanGiovanni. 1997. Synergism of nutrition, infection, 23. Beisel, W. R., W. D. Sawyer, E. D. Ryll, and D. Crozier. 1967. Metabolic effects and immunity: an overview. Am. J. Clin. Nutr. 66: 464S–477S. of intracellular infections in man. Ann. Intern. Med. 67: 744–779. 13. Kaech, S. M., S. Hemby, E. Kersh, and R. Ahmed. 2002. Molecular and func- 24. Beisel, W. R. 1972. Interrelated changes in host metabolism during generalized tional profiling of memory CD8 T cell differentiation. Cell 111: 837–851. infectious illness. Am. J. Clin. Nutr. 25: 1254–1260. 14. Chatraw, J. H., E. J. Wherry, R. Ahmed, and Z. F. Kapasi. 2008. Diminished 25. Carr, E. L., A. Kelman, G. S. Wu, R. Gopaul, E. Senkevitch, A. Aghvanyan, primary CD8 T cell response to viral infection during protein energy malnutri- A. M. Turay, and K. A. Frauwirth. 2010. Glutamine uptake and metabolism tion in mice is due to changes in microenvironment and low numbers of viral- are coordinately regulated by ERK/MAPK during T lymphocyte activation. J. specific CD8 T cell precursors. J. Nutr. 138: 806–812. Immunol. 185: 1037–1044. 15. Wherry, E. J., J. N. Blattman, K. Murali-Krishna, R. van der Most, and R. Ahmed. 26. Salimonu, L. S., A. O. Johnson, A. I. Williams, G. I. Adeleye, and B. O. Osunkoya. 2003. Viral persistence alters CD8 T-cell immunodominance and tissue distribution 1982. Lymphocyte subpopulations and antibody levels in immunized malnour- and results in distinct stages of functional impairment. J. Virol. 77: 4911–4927. ished children. Br.J.Nutr.48: 7–14. 16. Wherry, E. J., V. Teichgra¨ber, T. C. Becker, D. Masopust, S. M. Kaech, R. Antia, 27. Schlesinger, L., and A. Stekel. 1974. Impaired cellular immunity in marasmic U. H. von Andrian, and R. Ahmed. 2003. Lineage relationship and protective infants. Am. J. Clin. Nutr. 27: 615–620. immunity of memory CD8 T cell subsets. Nat. Immunol. 4: 225–234. 28. Murali-Krishna, K., and R. Ahmed. 2000. Cutting edge: naive T cells mas- 17. Becker, T. C., S. M. Coley, E. J. Wherry, and R. Ahmed. 2005. Bone marrow is querading as memory cells. J. Immunol. 165: 1733–1737. a preferred site for homeostatic proliferation of memory CD8 T cells. J. 29. Savino, W. 2002. The thymus gland is a target in malnutrition. Eur. J. Clin. Nutr. Immunol. 174: 1269–1273. 56(Suppl 3): S46–S49. 18. Schluns, K. S., K. Williams, A. Ma, X. X. Zheng, and L. Lefrançois. 2002. 30. Angelini, G., S. Gardella, M. Ardy, M. R. Ciriolo, G. Filomeni, G. Di Trapani, Cutting edge: requirement for IL-15 in the generation of primary and memory F. Clarke, R. Sitia, and A. Rubartelli. 2002. Antigen-presenting dendritic cells antigen-specific CD8 T cells. J. Immunol. 168: 4827–4831. provide the reducing extracellular microenvironment required for T lymphocyte 19. Tough, D. F., P. Borrow, and J. Sprent. 1996. Induction of bystander T cell activation. Proc. Natl. Acad. Sci. USA 99: 1491–1496. proliferation by viruses and type I interferon in vivo. Science 272: 1947– 31. Mathews, J. D., I. R. Mackay, S. Whittingham, and L. A. Malcolm. 1972. Protein 1950. supplementation and enhanced antibody-producing capacity in New Guinean 20. Sinha, D. P., and F. B. Bang. 1976. Protein and calorie malnutrition, cell- school-children. Lancet 2: 675–677. Downloaded from mediated immunity, and B.C.G. vaccination in children from rural West Ben- 32. Swaminathan, S., C. Padmapriyadarsini, L. Yoojin, B. Sukumar, S. Iliayas, gal. Lancet 2: 531–534. J. Karthipriya, R. Sakthivel, P. Gomathy, B. E. Thomas, M. Mathew, et al. 2010. 21. McMurray, D. N., S. A. Loomis, L. J. Casazza, H. Rey, and R. Miranda. 1981. Nutritional supplementation in HIV-infected individuals in South India: a pro- Development of impaired cell-mediated immunity in mild and moderate mal- spective interventional study. Clin. Infect. Dis. 51: 51–57. nutrition. Am. J. Clin. Nutr. 34: 68–77. 33. Neumann, C. G., G. J. Lawlor, Jr., E. R. Stiehm, M. E. Swenseid, C. Newton, 22. Allende, L. M., A. Corell, J. Manzanares, D. Madruga, A. Marcos, A. Madronõ, J. Herbert, A. J. Ammann, and M. Jacob. 1975. Immunologic responses in A. Lo´pez-Goyanes, M. A. Garcıá-Pe´rez, J. M. Moreno, M. Rodrigo, et al. 1998. malnourished children. Am. J. Clin. Nutr. 28: 89–104. http://www.jimmunol.org/ by guest on September 28, 2021