Human skeletal muscle ascorbate is highly responsive to changes in vitamin C intake and plasma concentrations

ckground:Vitamin C (ascorbate) is likely to be essential for skeletal muscle structure and function via its role as an enzyme cofactor for collagen and carnitine biosynthesis. Vitamin C may also protect these metabolically active cells from oxidative stress. Objective:We investigated the bioavailability of vitamin C to hu- man skeletal muscle in relation to dietary intake and plasma con- centrations and compared this relation with ascorbate uptake by leukocytes. Design:Thirty-six nonsmoking men were randomly assigned to re- ceive 6 wk of 0.5 or 2 kiwifruit/d, an outstanding dietary source of vitamin C. Fasting blood samples were drawn weekly, and 24-h urine and leukocyte samples were collected before intervention, after in- tervention, and after washout. Needle biopsies of skeletal muscle (vastus lateralis)were carried out before and after intervention. Results:Baseline vastus lateralis ascorbate concentrations were w16 nmol/g tissue. After intervention with 0.5 or 2 kiwifruit/d, these concentrations increasedw3.5-fold to 53 and 61 nmol/g, respectively. There was no significant difference between the re- sponses of the 2 groups. Mononuclear cell and neutrophil ascorbate concentrations increased onlyw1.5- andw2-fold, respectively. Muscle ascorbate concentrations were highly correlated (P, 0.001) with dietary intake (R= 0.61) and plasma concentrations (R= 0.75) in the range from 5 to 80mmol/L. Conclusions:Human skeletal muscle is highly responsive to vitamin C intake and plasma concentrations and exhibits a greater relative uptake of ascorbate than leukocytes. Thus, muscle appears to comprise a relatively labile pool of ascorbate and is likely to be prone to ascorbate depletion with inadequate dietary intake. This trial was registered at the Australian New Zealand Clinical Trials Registry (www.anzctr.org.au) as ACTRN12611000162910.Am J Clin Nutr2013;97:800–7. INTRODUCTION Vitamin C (ascorbate) is an essential micronutrient with many important biological functions. Ascorbate is a cofactor for a va- riety of metalloenzymes that are necessary for the biosynthesis of collagen, carnitine, neurotransmitters, and peptide hormones as well as the regulation of transcription factors such as hypoxia- inducible factor-1 (1–3). Ascorbate also

Trials Registry (www.anzctr.org.au) as ACTRN12611000162910.Am J Clin Nutr2013;97:800–7. INTRODUCTION Vitamin C (ascorbate) is an essential micronutrient with many important biological functions. Ascorbate is a cofactor for a va- riety of metalloenzymes that are necessary for the biosynthesis of collagen, carnitine, neurotransmitters, and peptide hormones as well as the regulation of transcription factors such as hypoxia- inducible factor-1 (1–3). Ascorbate also acts as a potent water- soluble antioxidant, with the ability to scavenge a wide variety of reactive oxygen and nitrogen species and regenerate other small molecule antioxidants from their respective radicals (4). Humans have lost the ability to synthesize ascorbate from glu- cose because of mutations in the gene encoding L-gulono- g-lactone oxidase, which is the terminal enzyme in the ascorbate biosynthetic pathway (5). Therefore, an adequate and regular di- etary intake is essential to prevent hypovitaminosis C and the po- tentially fatal deficiency disease scurvy (6). Clinical manifestations of scurvy include blood vessel fragility and bleeding, which result in petechial and other hemorrhages, skin changes that result in follicular hyperkeratosis, impaired wound healing, gum swelling and bleeding, joint pain and effusions, anemia, weakness, and fa- tigue (6–8). Many of these symptoms are attributed to the role of ascorbate in collagen and carnitine biosynthesis (9). Plasma ascorbate status reflects recent dietary intake, whereas leukocyte concentrations are thought to more closely reflect tissue stores (10, 11). However, whether leukocytes are an accurate model for the ascorbate status of other tissues is uncertain. Our previous study with vitamin C–deficient Gulo knockout mice indicated that the maximal tissue uptake within different organs occurred at varying doses of ascorbate (12). Thus, the relative uptake of ascorbate by leukocytes may not be representative for all tissues or organs. Although ascorbate has previously been measured in skeletal muscle obtained at autopsy (13), to our knowledge, no studies have investigated the bioavailability of vitamin C to human skeletal muscle. Skeletal muscle contains relatively low concentrations of ascorbate compared with in other organs (13), but because of the large amount of skeletal muscle present in the body, it has been estimated to comprise#67% of the total body vitamin

skeletal muscle obtained at autopsy (13), to our knowledge, no studies have investigated the bioavailability of vitamin C to human skeletal muscle. Skeletal muscle contains relatively low concentrations of ascorbate compared with in other organs (13), but because of the large amount of skeletal muscle present in the body, it has been estimated to comprise#67% of the total body vitamin C (10) and, thus, represents the major pool of vitamin C in the body. The accumulation of ascorbate in muscle tissue is thought to protect these metabolically active cells against oxidative stress (14). Therefore, vitamin C is likely to be essential for both skeletal muscle structure and function because of its dual role as an antioxidant and as an enzyme cofactor for collagen and carnitine biosynthesis (15, 16). Overall, it is likely that muscle tissue has a high requirement for and turnover of vitamin C. Therefore, the aims of the current study were to investigate the bioavailability of vitamin C to 1 From the Centre for Free Radical Research, Department of Pathology (ACC, SMB, JMP, and MCMV), and the Department of Plastic and Recon- structive Surgery (JWS), University of Otago, Christchurch, Christchurch, New Zealand. 2 Supported by the University of Otago, Dunedin, New Zealand, and Zespri International Ltd, Mount Maunganui, New Zealand. 3 Address correspondence to A Carr, Centre for Free Radical Research, Department of Pathology, University of Otago, Christchurch, PO Box 4345, Christchurch 8140, New Zealand. E-mail: anitra.carr@otago.ac.nz. Received October 18, 2012. Acceptedfor publication January 25, 2013. First published online February 27, 2013; doi: 10.3945/ajcn.112.053207. 800 Am J Clin Nutr2013;97:800–7. Printed in USA. 2013 American Society for Nutrition

human skeletal muscle in relation to dietary intake and plasma concentrations and to compare this relation to the uptake of ascorbate by peripheral blood leukocytes. Zespri gold kiwifruit (Actinidia chinensis var. Hort 16A), which are an outstanding dietary source of vitamin C (17), were used as the intervention in this study. Our previous human study in which we in- vestigated the bioavailability of vitamin C from gold kiwifruit indicated that the consumption of 0.5 kiwifruit/d resulted in a significant increase of plasma ascorbate concentrations in marginally deficient individuals, whereas the consumption of 2 kiwifruit/d was required to saturate the plasma, as shown by a significant increase in the urinary ascorbate excretion at this dose (18). In the current study, we supplemented participants with 0.5 or 2 kiwifruit/d for 6 wk and measured ascorbate concentrations in plasma, urine, leukocytes, and skeletal muscle (vastus lateralis) before and after intervention. SUBJECTS AND METHODS Participants This study was conducted according to the guidelines of the Declaration of Helsinki, and all procedures that involved human participants were approved by the Upper South Regional Ethics Committee (URA/11/02/003). The study was registered at the Australian New Zealand Clinical Trials Registry (www.anzctr. org.au; ACTRN12611000162910). Nonsmoking men aged 18–35 y were recruited from local tertiary institutes, and 134 subjects underwent a screening in- terview to determine their eligibility for the study. Exclusion criteria included being a recent smoker (within the past 1 y), having an allergy or intolerance to kiwifruit, consumption of vitamin C–containing supplements (within the past 3 mo), tak- ing prescription medication (within the past 3 mo), excessive alcohol consumption (.21 standard drinks/wk), high fruit and vegetable consumption (.5 servings/d), having diabetes or bleeding disorders, and fainting because of a fear of needles. Anthropometric measures were carried out to determine BMI (in kg/m 2 ), and a fasting venous blood sample was drawn to de- termine plasma ascorbate concentrations. Sample-size calculations indicated that at 80% power and with a= 0.05, a sample size of 15 participants per intervention group would detect a minimum difference of 10mmol ascorbate/L as determined by using data derived from our vitamin C

out to determine BMI (in kg/m 2 ), and a fasting venous blood sample was drawn to de- termine plasma ascorbate concentrations. Sample-size calculations indicated that at 80% power and with a= 0.05, a sample size of 15 participants per intervention group would detect a minimum difference of 10mmol ascorbate/L as determined by using data derived from our vitamin C bio- availability study (18). To allow for potential withdrawal be- cause of the length of the study, 36 nonsmoking participants with below-average plasma ascorbate concentrations were en- rolled by the study coordinator and provided signed informed consent. Two healthy men volunteered to undergo muscle biopsies to act as a comparator group. Their anthropometric data were BMI of 24 and 27 and ages of 31 and 39 y. Study design The study used a parallel-arm design (Figure 1), and par- ticipants were randomly assigned by the study coordinator into a low-dose group (0.5 kiwifruit/d) or a high-dose group (2 ki- wifruit/d) by using a random-numbers chart. Blocking was used to account for the range in prestudy plasma ascorbate concen- trations. The study was carried out in a phase I clinical trials unit and comprised a lead-in phase of 5 wk, an intervention phase of 6 wk, and a washout phase of 4 wk. Initially, participants were encouraged to reduce their dietary vitamin C intake by avoiding the consumption of juice and other vitamin C–fortified bever- ages and by substituting high–vitamin C–containing foods with low–vitamin C–containing foods. Fasting venous blood samples were drawn weekly throughout the study to monitor participant plasma ascorbate concentrations derived from normal daily diets and the intervention. Twenty-four–hour urine and extra blood for leukocyte isolations were collected at baseline (week 5), after intervention (week 11), and after washout (week 15). Muscle biopsies were carried out before and after intervention. Participants also completed 7-d food and beverage records on 4 occasions (at the beginning of the study, before intervention, after intervention, and after washout) to monitor their dietary vitamin C intakes. Intervention Gold kiwifruit (Actinidia chinensis var. Hort. 16A) were provided by Zespri International Ltd and stored at#48C. Par-

(week 15). Muscle biopsies were carried out before and after intervention. Participants also completed 7-d food and beverage records on 4 occasions (at the beginning of the study, before intervention, after intervention, and after washout) to monitor their dietary vitamin C intakes. Intervention Gold kiwifruit (Actinidia chinensis var. Hort. 16A) were provided by Zespri International Ltd and stored at#48C. Par- ticipants were provided with sufficient kiwifruit each week to consume 0.5 or 2 kiwifruit /d. The ascorbate content of the ki- wifruit was determined from an extract of the flesh that was measured by using HPLC with electrochemical detection (18). This method indicated that the fruit contained 116610 mg ascorbate/100 g (n= 5). Participants were asked not to consume the skins, and on the basis of the amount of fruit ingested, the amount of vitamin C consumed was calculated to bew53 mg for one-half of a kiwifruit orw212 mg for 2 kiwifruit. Sample collection and processing Plasma and urine Peripheral blood (4 mL) was collected into evacuated tubes containing K 3-EDTA and kept on ice at all times (19). Samples were centrifuged at 48C to pellet cells, and the plasma was collected and kept on ice for the extraction of ascorbate. Urine was collected over 24 h into collection bottles containing K 2- EDTA (final concentration: 100mmol/L) (20). Plasma and urine samples were treated with an equal volume of ice-cold 0.54- mol/L HPLC-grade perchloric acid with diethylene triamine pentaacetic acid (DTPA) 4 (100mmol/L) to precipitate the pro- tein (21). The perchloric acid and DTPA extracts were centri- fuged, and the deproteinated supernatant fluids were stored at 2808C until HPLC analysis. Mononuclear leukocytes and neutrophils Peripheral blood was collected into BD Vacutainer Cell Preparation Tubes (Becton, Dickinson and Co) that contained sodium heparin and kept at room temperature. Tubes were centrifuged in a horizontal rotor for 30 min at 1800 relative centrifugal force without brake. After centrifugation, the layer above the gel that contained the mononuclear leukocytes was 4 Abbreviations used: DTPA, diethylene triamine pentaacetic acid; HBSS, Hanks balanced salt solution; SVCT2, sodium-dependent vitamin C trans- porter 2. VITAMIN C UPTAKE

sodium heparin and kept at room temperature. Tubes were centrifuged in a horizontal rotor for 30 min at 1800 relative centrifugal force without brake. After centrifugation, the layer above the gel that contained the mononuclear leukocytes was 4 Abbreviations used: DTPA, diethylene triamine pentaacetic acid; HBSS, Hanks balanced salt solution; SVCT2, sodium-dependent vitamin C trans- porter 2. VITAMIN C UPTAKE BY HUMAN SKELETAL MUSCLE 801

removed, and cells were washed with phosphate-buffered saline and finally suspended in Hanks balanced salt solution (HBSS). After the first centrifugation step previously described, neutro- phils and erythrocytes were extracted from below the gel layer. Erythrocytes were removed by using dextran sedimentation and hypotonic lysis (22), and neutrophils were suspended in HBSS. Isolated leukocytes were counted by using a hemocytometer, standardized for cell number, suspended in HBSS, and an equal volume of ice-cold 0.54-mol/L perchloric acid/DTPA solution was added to precipitate the protein (21). Deproteinated super- natant fluids were stored at2808C. Muscle tissue Needle biopsies were performed by an experienced plastic surgeon. Local anesthetic (2 mL 1% lignocaine with adrenaline 1:200,000) was injected into the subcutaneous fat of the antero- lateral midthigh. A Quick-Core biopsy needle (14 gauge; 6 cm long with a 20-mm throw; Cook Medical Inc) was inserted into the vastus lateralis muscle and retrieved a small piece of tissue (13.56 8.2 mg;n= 52). Fascia, if present, was removed, and the muscle tissue sample was placed into a preweighed Eppendorf tube (Eppendorf International) on ice, the weight of the tissue was determined, and the sample was frozen at2808C. Our previous studies have shown that ascorbate in intact frozen tissue remains stable for many months (12, 23). Immediately before HPLC analysis, the frozen muscle tissue sample was homogenized for 60 s in 50mL ice-cold phosphate-buffered saline with a Dounce ground glass pestle, and an equal volume of ice-cold 0.54-mol/L HPLC-grade perchloric acid/DTPA solution was added (21). Deproteinated supernatant fluids were stored at2808C. Analysis of ascorbate by HPLC The ascorbate content of the kiwifruit, plasma, urine, leuko- cytes, and muscle tissue was analyzed by using reverse-phase HPLC with a Synergi 4 micron Hydro-RP 80-A column (Phe- nomenex NZ Ltd) and an ESA coulochem II electrochemical detector (ESA Inc) as described previously (18). The plasma ascorbate content was expressed as micromoles per liter, the urinary ascorbate content was expressed as micromoles per 24 h, the leukocyte ascorbate content was expressed as nanomoles per 10 8 cells, and the muscle tissue ascorbate content was expressed as nanomoles per gram of wet weight.

an ESA coulochem II electrochemical detector (ESA Inc) as described previously (18). The plasma ascorbate content was expressed as micromoles per liter, the urinary ascorbate content was expressed as micromoles per 24 h, the leukocyte ascorbate content was expressed as nanomoles per 10 8 cells, and the muscle tissue ascorbate content was expressed as nanomoles per gram of wet weight. Analysis of food and beverage records The number of servings of fruit and vegetables consumed by each participant was estimated from 7-d food and beverage re- cords as described previously (18). The vitamin C content of the consumed foods and beverages were estimated with Diet Cruncher software (version 1.6; Way Down South Software) and the 2006 New Zealand FOODfiles Food Composition Database (New Zealand Institute for Plant and Food Research Ltd). Statistical analysis Data are represented as either means6SDs or means6 SEMs as indicated. Differences between paired and unpaired data were determined by using the 2-tailedttest, andP#0.05 was considered significant. Linear regression analysis (Pearson’s correlations) and ANOVA with Fisher’s pairwise multiple- comparison procedures were carried out with SigmaStat soft- ware (version 11; Systat Software Inc). FIGURE 1.Study design. Parallel arms comprised 0.5 or 2 kiwifruit/d for 6 wk. 802 CARR ET AL

RESULTS Screening phase A total of 134 young men were screened for this study. Thirty- six subjects, who had below-average plasma ascorbate concen- trations and also satisfied other inclusion exclusion criteria, were enrolled. These individuals were randomly assigned to either a low-dose group (0.5 kiwifruit/d;n= 18) or high-dose group (2 kiwifruit/d;n= 18). One of the participants who enrolled in the high-dose group withdrew early in the study and was not used in the data analysis. Characteristics of screened and enrolled individuals are shown inTable 1. There were no significant differences between the 2 intervention groups, although both intervention groups had significantly lower plasma ascorbate concentrations than the screened group (P,0.001). The aver- age6SD fasting plasma ascorbate concentration for screened individuals was 48616mmol/L, with a range of 3 to 92mmol/L, whereas average6SD fasting plasma ascorbate concentrations for enrolled groups (low and high dose) were 34610 and 356 7mmol/L, respectively, with a range from 15 to 45mmol/L. Dietary intake of vitamin C An analysis of food and beverage records at baseline indicated a mean intake of,3 servings fruit and vegetables/d and a mean intake of,30 mg vitamin C/d (Table 2). The addition of 0.5 kiwifruit to the daily diet of the low-intervention group did not alter their baseline fruit and vegetable intake. However, this addition significantly increased their daily vitamin C intake from 29 to 73 mg/d (Table 2). The addition of 2 kiwifruit to the daily diet of the high-intervention group significantly increased their fruit and vegetable intake from 3 to 5 servings/d and their daily vitamin C intake.7-fold to 214 mg/d (Table 2). After the 4-wk washout period, the vitamin C intake of participants had re- turned to baseline intakes. Ascorbate status of plasma and urine At baseline, the low-dose group had a mean plasma ascorbate concentration of 23mmol/L and this increased to 46mmol/L after 6 wk of intervention (Figure 2). The high-dose group had a baseline mean concentration of 25mmol/L, and after in- tervention, this concentration increased to 63mmol/L (Figure 2), which was close to saturating (ie, did not increase further after additional vitamin C intake) (18). Plasma ascorbate

group had a mean plasma ascorbate concentration of 23mmol/L and this increased to 46mmol/L after 6 wk of intervention (Figure 2). The high-dose group had a baseline mean concentration of 25mmol/L, and after in- tervention, this concentration increased to 63mmol/L (Figure 2), which was close to saturating (ie, did not increase further after additional vitamin C intake) (18). Plasma ascorbate concentra- tions of the 2 groups were statistically different from each other within the first week of intervention, and this difference was maintained for the entire 6 wk of intervention (P,0.001). At the end of the 4-wk washout period, plasma ascorbate concen- trations for both high- and low-dose groups had decreased and were not different from baseline concentrations (Figure 2). After 6 wk of intervention, the urinary excretion of ascorbate had increased 2-fold in the low–kiwifruit-dose group and 15-fold in the high-dose group (Table 3), which suggested that the higher intake of kiwifruit resulted in plasma concentrations close to saturation (18). Ascorbate status of leukocytes and muscle tissue Baseline ascorbate concentrations of mononuclear cells and neutrophils are shown in Table 3. After 6 wk of intervention, mononuclear cell ascorbate concentrations increasedw1.5-fold for both the low–kiwifruit-dose group (P= 0.004) and the high– kiwifruit-dose group ( P= 0.004), whereas neutrophil ascorbate concentrations increased.2-fold (low-dose group:P= 0.0002; high-dose group:P= 0.003; Table 3). There were no significant differences in leukocyte ascorbate concentrations between the 2 intervention groups, which indicated that leukocytes were sat- urating with the low kiwifruit intake. Baseline mean muscle tissue ascorbate concentrations were w16 nmol/g tissue (Table 3), with a range from 1.0 to 43.2 nmol/g tissue. After 6 wk of intervention, there was aw3.5-fold in- crease in mean muscle tissue ascorbate concentrations to 53 and 61 nmol/g tissue in the low- and high-dose groups, respectively (P,0.0001; Table 3). Thus, muscle tissue appeared to have a sig- nificantly greater relative uptake of ascorbate than both mono- nuclear cells and neutrophils (Figure 3). There was no difference in muscle tissue ascorbate concentrations between the 2 in- tervention groups, which suggested that muscle tissue ascorbate concentrations were saturating with the low kiwifruit intake.

low- and high-dose groups, respectively (P,0.0001; Table 3). Thus, muscle tissue appeared to have a sig- nificantly greater relative uptake of ascorbate than both mono- nuclear cells and neutrophils (Figure 3). There was no difference in muscle tissue ascorbate concentrations between the 2 in- tervention groups, which suggested that muscle tissue ascorbate concentrations were saturating with the low kiwifruit intake. TABLE 1 Characteristics of the individuals screened and enrolled in the study 1 Screened group (n= 134) 0.5-kiwifruit/d group (n= 18) 2-kiwifruit/d group (n= 17) Age (y) 21 6322 6422 63 Weight (kg) 81 616 89 623 81 615 Height (cm) 182 67 181 67 181 67 BMI (kg/m 2 )24 6427 66* 25 64 Ascorbate (mmol/L) 48 616 34 610** 35 67** 1 All values are means6SDs. There were no significant differences between the 2 intervention groups. * , **For intervention groups compared with screened group (unpairedttest): *P,0.05, **P,0.001. TABLE 2 Dietary intake of vitamin C in 0.5- and 2-kiwifruit/d groups 1 0.5 kiwifruit/d 2 kiwifruit/d Between-group interventionPBaseline Intervention Washout Baseline Intervention Washout Fruit and vegetables (servings/d) 2.960.3 (18) 3.460.4 (17) 2.660.3 (15) 2.760.3 (17) 4.860.3* (17) 2.760.3 (16) 0.001 Vitamin C (mg/d) 28.663.1 (18) 73.464.2* (17) 31.465.1 (15) 29.163.1 (17) 21464.5* (17) 28.864.3 (16),0.0001 1 All values are means6SEMs;nin parentheses. For 0.5- compared with 2-kiwifruit/d groups after intervention,Pvalues were determined by using the unpairedttest. *For intervention compared with baseline,P,0.0001 (pairedttest). VITAMIN C UPTAKE BY HUMAN SKELETAL MUSCLE 803

The study examines the relationship between vitamin C intake and its bioavailability in human skeletal muscle.