1st Page

Journal of Animal Science - Article



This article in

  1. Vol. 85 No. 7, p. 1712-1718
    Received: Dec 22, 2006
    Accepted: Mar 13, 2007
    Published: December 8, 2014

    2 Corresponding author(s):


Oral N-acetyl-l-cysteine is a safe and effective precursor of cysteine1

  1. R. N. Dilger and
  2. D. H. Baker2
  1. Department of Animal Sciences and Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana 61801


Relative bioavailability and toxicity of N-acetyl-l-Cys (NAC) were evaluated in 9-d chick growth assays. The bioavailability of NAC relative to Cys was determined by feeding young chicks a highly purified crystalline AA diet singly deficient in Cys. Bio-availability estimates were obtained using standard slope-ratio methodology. N-Acetyl-l-cysteine was shown to be as effective as Cys in supporting chick growth, and was assigned a relative bioavailability value of 100%. To assess toxicity, a nutritionally adequate corn-soybean meal diet was supplemented with graded concentrations of NAC (isomolar to 10, 20, 30, or 40 g/kg of Cys, as-fed). When NAC supplied 10 or 20 g/kg of Cys, chick growth performance was unaffected, but NAC supplying 30 or 40 g/kg of Cys reduced (P < 0.05) BW gain by 13 and 34%, respectively, relative to the unsupplemented control diet. Only plasma-free NAC was substantially increased (P < 0.05) because of excess dietary NAC; plasma-free Cys was unaltered. We concluded that dietary NAC is efficacious in supplying Cys in support of chick growth, and only large excesses of NAC are growth depressing. Hence, the human clinical benefits of oral NAC likely result from its ability to deliver Cys safely and effectively to the portal circulation.


The use of crystalline AA as dietary supplements in human and animal nutrition has gained popularity in the United States (Frost and Sullivan, 2006). This includes the sulfur AA (SAA) Cys, which participates in a variety of metabolic pathways involving Met, glutathione, CoA, and taurine (Stipanuk, 2004). Studies suggest that Cys should be considered a conditionally indispensable AA (Shoveller et al., 2006), contingent on Met status of the animal. However, dietary supplementation of Cys is complicated because its free sulfhydryl group is capable of spontaneous oxidation, resulting in product instability. Moreover, the oxidized form, cystine (Cys-Cys), is relatively insoluble, making both forms poor candidates for enteral or parenteral nutritional formulas. Recent studies suggest that excess dietary Cys may be deleterious to animal growth and feed consumption in various species (Dilger et al., 2007). Therefore, safe and functional alternative Cys precursor products have been sought.

N-Acetyl-l-cysteine (NAC) serves as a precursor for Cys because the α-amino nitrogen is protected by an acetyl group, and most tissues possess the ability to remove this acetyl group and thus furnish Cys. For this reason, NAC has been studied extensively as both a therapeutic agent and a direct Cys precursor. Oral administration of NAC remains a primary treatment for acetaminophen overdose (Vale and Proudfoot, 1995), and NAC has also been shown to positively affect overall antioxidant status (Ahola et al., 1999). In this regard, NAC serves as a direct antioxidant per se and as a precursor for Cys, the availability of which is rate limiting for hepatic glutathione synthesis. However, early pharmacokinetic studies suggested that oral NAC bioavailability was low (Borgström et al., 1986; Olsson et al., 1988; De Caro et al., 1989), between 6 and 10%, because of low blood concentrations of NAC. These studies did not account for the deacetylation of NAC to form Cys during normal absorptive processes in the gut. Additionally, a paucity of information exists regarding the toxicity of dietary NAC. Such evidence is increasingly necessary considering the growing interest in dietary supplemental AA in preventive healthcare (Frost and Sullivan, 2006).

The primary objectives of the studies herein were to 1) determine the bioavailability of NAC relative to Cys in support of growth and 2) evaluate the toxicity of NAC when supplied in excess of the dietary Cys requirement.


General Procedures

All experimental procedures were approved by the University of Illinois Animal Care and Use Committee. Three studies were conducted using male chicks (New Hampshire male × Columbian female) obtained from the University of Illinois Poultry Farm. Chicks were housed in thermostatically controlled starter batteries with raised-wire flooring, in an environmentally controlled room with continuous lighting. From hatch to d 7 posthatch, chicks were fed a typical corn-soybean meal starter diet that provided 230 g/kg of CP (as-fed basis) and was adequate in all dietary nutrients (NRC, 1994). After an overnight fast, the chicks were weighed, wing-banded, and randomized to dietary treatments on d 8, such that average initial pen weights and weight distributions were similar across treatments.

Three bioassays were conducted, and 4 replicate pens of 3 chicks were assigned to each dietary treatment in each of the assays. The experimental basal diets (Table 1) were fed for a 9-d period (d 8 to 17 posthatch) during each bioassay. Experimental diets and tap water were freely available to chicks at all times. Body weights of individual chicks and pen feed intakes were measured at the termination of each bio-assay. Body weight gain, feed intake, and efficiency of gain (G:F) were calculated for each replicate pen.

The purified basal diet used in assays 1 and 2 was analyzed for both Met and cyst(e)ine, where cyst(e)ine refers to Cys + Cys-Cys (total of reduced plus oxidized forms) contained in the diet. Duplicate diet samples were preoxidized with performic acid (300 g/L of hydrogen peroxide, 800 g/L of formic acid) and then subjected to a 22-h acid hydrolysis (6 mol/L of HCl) under nitrogen gas at 100°C. Hydrolysates were assayed for Met and cyst(e)ine by ion-exchange chromatography (Model 119 CL Amino Acid Autoanalyzer, Beckman Instruments, Palo Alto, CA), as described by Chung and Baker (1992). The diet contained 1.2 g/kg of Met and 0.5 g/kg of cyst(e)ine (as-fed basis), making it severely deficient in both Met and cyst(e)ine relative to dietary requirements of ∼3 g/kg of digestible Met and ∼3 g/kg of digestible Cys (NRC, 1994). Other AA in the purified diet were in excess of the true digestible AA requirements of young chicks (NRC, 1994). When fully fortified with SAA, the experimental diet used in assays 1 and 2 has been shown to allow growth rates similar to those obtained with a typical corn-soybean meal diet providing 230 g/kg of CP (unpublished data).

Assay 1.

The objective of this assay was to quantify the Met requirement of chicks fed a purified crystalline AA diet containing more than adequate dietary cyst(e)ine. Supplemental dietary Met concentrations of 0.5, 1.0, 1.5, 2.0, and 2.5 g/kg were added (as-fed basis) to the basal diet fortified with 3.5 g/kg of Cys-Cys, and the supplemental Met requirement was based on chick growth performance.

Assay 2.

The bioavailability of NAC was evaluated relative to Cys, as measured by chick growth performance. The purified basal diet (Table 1) was supplemented with 2.0 g/kg of Met, (i.e., the dietary Met concentration from assay 1 deemed to be slightly above the minimal Met requirement). Dietary Cys was supplemented at 0, 0.35, 0.70, 1.05, and 2.50 g/kg (as-fed basis) to the cyst(e)ine-deficient basal diet. The first 4 Cys doses were expected to produce a linear growth response, and the fifth (greatest) supplemental Cys concentration was designed to provide Cys at the estimated dietary requirement of 3 g/kg, thereby serving as a positive control. N-Acetyl-l-cysteine was supplemented at 0.47 and 0.95 g/kg (as-fed basis), isomolar to 0.35 and 0.70 g/kg of Cys, respectively, and evaluated for its ability to serve as a Cys source by standard slope-ratio methodology.

Assay 3.

The toxicity of excess dietary NAC was evaluated using a corn-soybean meal basal diet (Table 1) adequate in all nutrients for chicks of this age (NRC, 1994). Dietary NAC was supplemented at 0, 13.47, 26.94, 40.41, and 53.88 g/kg (as-fed basis) (isomolar to 0, 10, 20, 30, and 40 g/kg of Cys), and toxicity was evaluated after a 9-d feeding period by chick growth performance and plasma-free SAA concentrations. Blood was collected (cardiac puncture) from each of the chicks in 3 of the 4 replicate pens (selected randomly), pooled by pen replicate, and processed as described by Dilger et al. (2007). Simultaneous chromatographic separation of plasma-free SAA (Met, Cys, Cys-Cys, and NAC) was performed using a previously described HPLC procedure (Dilger et al., 2007).

Statistical Analysis

All data were subjected to ANOVA using the GLM procedure (SAS Inst. Inc., Cary, NC). Data were analyzed using pen means, with procedures appropriate for a completely randomized design. Data are presented as mean values with pooled SEM. In all cases, separation of means was carried out using the LSD multiple-comparison procedure of SAS, assuming an α level of 0.05 unless otherwise stated. Linear and quadratic responses were also evaluated using single df contrasts.

In assay 1, BW gain data were fitted to a 1-slope broken-line regression model (Robbins et al., 1979; Robbins et al., 2006) to estimate the minimal Met requirement of chicks fed a purified diet containing surfeit dietary Cys. In assay 2, relative bioavailability of NAC was evaluated using standard slope-ratio methodology (Sasse and Baker, 1973). This evaluation did not include data from chicks that received 2.5 g/kg of supplemental Cys because this treatment served to validate the maximal growth response to Cys in the purified diet used herein. Body weight gain (dependent variable) was regressed on supplemental Cys intake (independent variable) from Cys (X1) or NAC (X2) in a multiple linear regression analysis using the GLM procedure of SAS. Relative bioavailability of NAC was evaluated using the ratio of slopes [(X2/X1) × 100] obtained from the multiple linear regression analysis.


Assay 1

Chick BW gain increased (P < 0.05) because of Cys-Cys supplementation, indicating the basal diet was first limiting in Cys (Table 2). In the presence of excess dietary cyst(e)ine, Met supplementation resulted in a quadratic (P < 0.01) response in BW gain, feed intake, and G:F. The fitted 1-slope broken-line regression of either chick BW gain (Figure 1) or G:F over the 9-d assay suggested a minimal supplemental Met requirement of 1.2 g/kg for chicks fed the cyst(e)ine-adequate purified diet used here. Nonetheless, for assay 2, a supplemental concentration of 2.0 g/kg of Met was selected to provide more than adequate Met without furnishing significant excess Met from which Cys could be produced via the transsulfuration pathway.

Assay 2

Supplementation of the basal diet with Cys (diets 1 to 4) or NAC (diets 1, 6, and 7) resulted in a linear increase (P < 0.05) in BW gain, feed intake, and G:F (Table 3), suggesting the basal diet for this assay (containing 2.0 g/kg of supplemental Met) was markedly deficient in cyst(e)ine. Efficiency of gain, but not BW gain, was improved (P < 0.05) by 2.5 g/kg of supplemental Cys compared with Cys at 1.05 g/kg. Slope-ratio methodology provided a NAC bioavailability estimate of 106% relative to Cys when BW gain was regressed on supplemental Cys intake.

Assay 3

No decreases in BW gain, feed intake, or G:F were observed with supplementation of NAC up to 26.94 g/kg (isomolar to 20 g/kg of Cys; Table 4). However, ingestion of 40.41 and 53.88 g/kg of NAC resulted in decreased (P < 0.05) growth performance. Overall, quadratic (P < 0.01) decreases in BW gain, feed intake, and G:F occurred with increased concentrations of supplemental NAC. No chick mortality was observed during this assay.

Plasma-free SAA concentrations resulting from ingestion of excess levels of NAC are shown in Table 4. Plasma Met and Cys-Cys increased (P < 0.05) because of excess NAC, but the effect showed no consistent pattern. Although not unexpected, excess NAC had no effect on plasma-free Cys. However, graded excesses of NAC did increase (P < 0.05) plasma-free NAC from an undetectable concentration to 240 μmol/L.


N-Acetyl-l-cysteine was as effective as Cys in supporting chick growth performance and was therefore assigned a relative bioavailability value of 100%, suggesting the young chick is able to completely metabolize NAC to Cys when delivered orally. This finding was previously suggested using enteral administration in the rat (Baker and Han, 1993) and parenteral administration in the pig (Shoveller et al., 2006), but to our knowledge no studies have estimated the oral bioavailability of NAC using a standard slope-ratio growth assay. In addition, ingestion of dietary NAC isomolar to 40 g/kg of Cys resulted in no chick mortality and only a 34% reduction in BW gain. Therefore, excess dietary Cys provided by NAC resulted in relatively minor effects as compared with the pernicious outcome of feeding an isosulfurous concentration of excess dietary Cys (Dilger et al., 2007).

The first bioassay was conducted to establish the required level of Met in our purified diet that contained surfeit cyst(e)ine. We deemed this important for 2 reasons. First, by having Met at its required level, Cys would be singly limiting such that the responses to graded dosing of Cys or NAC in assay 2 would not be limited by Met deficiency anywhere on the response curve. Second, we wanted Met to be at, but not in great excess of, its minimal requirement because excess Met would furnish Cys via transsulfuration, and thereby possibly confound interpretation of the Cys and NAC responses. Establishing a supplemental Met requirement as such (Table 2; Figure 1) is an essential, and often overlooked, step in estimating the relative bio-availability of NAC. Although Met was included at 2.0 g/kg (in slight excess of the minimal Met requirement determined in assay 1), the data clearly showed a growth response to supplemental Cys (as NAC or Cys) in assay 2, confirming that our first 4 Cys doses were within the linear response range.

It is apparent that considerable confusion exists with regard to the meaning of the term “NAC bioavailability.” Previous pharmacokinetic studies (Borgström et al., 1986; Olsson et al., 1988; De Caro et al., 1989) have focused on the amount of an oral NAC dose that reaches the bloodstream as NAC itself. Little emphasis has been placed on the chemical modification of NAC (i.e., deacetylation) within the gut lumen and enterocyte during absorption. Therefore, pharmacokinetic studies suggesting that less than 10% of oral NAC is absorbed into portal blood as NAC per se are misleading. Our approach, however, was based on the assumption that beneficial effects of NAC do not result from NAC itself, but rather from NAC delivering Cys for in vivo functions (e.g., glutathione synthesis). Thus, the ability of NAC (the test precursor) to provide Cys relative to Cys itself (the standard nutrient) was assessed in assay 2. The results of this study (Table 3) clearly showed that oral NAC was 100% effective as a Cys precursor.

Deacetylation of NAC via the enzyme aminoacylase I (EC occurs in various tissues not limited to the liver, kidney, lung, and intestine (Sjödin et al., 1989; Yamauchi et al., 2002). The gut most likely plays a significant role in positively affecting the portal flux of Cys, especially when considering the large mass of intestine and the extensive first-pass Cys metabolism that likely occurs in intestinal tissue (Stipanuk and Rotter, 1984; Bos et al., 2003; Shoveller et al., 2005). Studies evaluating whether NAC influences circulating Cys concentrations (Ahola et al., 1999; Shoveller et al., 2006) have typically supplied NAC intravenously to simulate parenteral nutrient provision. Sjödin et al. (1989) clearly showed that dietary NAC was hydrolyzed to Cys in almost stoichiometric amounts in rat, mouse, and human tissues. However, although many tissues possess aminoacylase I activity, the importance of first-pass intestinal metabolism should not be overlooked. Observations in our chick model and in the pig (Shoveller et al., 2006) strongly support the efficient conversion of NAC into Cys. Therefore, the role of NAC in supplying Cys for metabolic and physiologic purposes should be considered a primary function.

On establishing the efficient utilization of NAC as a Cys precursor, we focused on the safety of dietary NAC when supplied to chicks in great excess of the dietary Cys requirement. Assay 3 clearly showed that oral provision of NAC supplying up to 20 g/kg of Cys was completely innocuous to chick growth performance. These toxicity data are in stark contrast to previous studies that compared the relative toxicities of Met, Cys, and Cys-Cys (Dilger et al., 2007). Excess dietary NAC providing 10 or 20 g/kg of Cys had no effect on growth, whereas NAC providing 30 or 40 g/kg of Cys depressed BW gain similar to that of excess Cys-Cys (Dilger et al., 2007). However, unlike Cys, excess dietary NAC caused no chick mortality after a 9-d feeding period.

Ingestion of NAC caused no change in plasma-free Cys and only small but variable changes in plasma-free Met and Cys-Cys. We previously observed a similar outcome (Dilger et al., 2007) when feeding excesses of either Cys or Cys-Cys. Thus, it seems that plasma-free Cys remains fairly constant in the chick after acclimation to dietary SAA excesses for a period of 9 d. Shoveller et al. (2006), in newborn piglets, and Ahola et al. (1999), in preterm infants, infused NAC intravenously and observed no change in plasma cyst(e)ine. Additionally, Stabler et al. (2000), using a baboon model of severe prematurity, showed that postdelivery concentrations of plasma cyst(e)ine could not be maintained by direct infusion of Cys. Overall, these studies suggest that strict control of plasma cyst(e)ine concentrations exists in various species. Shoveller et al. (2006) suggested that NAC may be providing a reserve pool of Cys via glutathione biosynthesis, or perhaps by replacing Cys in mixed disulfides or by binding to plasma albumin.

No free NAC was detected in the plasma of chicks receiving unsupplemented diets, but plasma-free NAC reached a concentration of 240 μmol/L when chicks ingested 40 g/kg of NAC. However, our plasma NAC results do not allow a quantitative estimate of the quantity of Cys absorbed from the gut as NAC vs. Cys itself. Regardless, the free NAC that reaches the bloodstream is theoretically available to serve as a direct antioxidant, or it may be deacetylated to Cys for eventual incorporation into glutathione. Our results in the chick suggest a high tolerance for excess dietary NAC. A similar conclusion was reported previously (Baker, 2006), which suggests that NAC may be a safer over-the-counter nutritional supplement than Cys.

Previous research has suggested that Cys may act as an endogenous excitotoxin (Olney and Ho, 1970). In this regard, Cys has been shown to selectively activate N-methyl-d-aspartate receptors and cause brain damage similar to Glu (Olney et al., 1990; Puka-Sundvall et al., 1995). However, supplementation of excess Cys as NAC does not result in the pernicious effects observed by feeding Cys directly (Dilger et al., 2007). Explanation for why NAC is less toxic than Cys itself is perplexing but may result from 1) a slower rate of Cys absorption when provided as NAC or 2) preferential use of NAC over Cys for as yet undefined purposes within the enterocyte. Data in support of these hypotheses are lacking. Therefore, future investigations of NAC pharmacokinetics are warranted. Such studies should consider not only conversion processes of dietary NAC into Cys, but also the location where the free-sulfhydryl group of NAC is metabolized.

Work in various animal species suggests NAC is a safe and functional precursor of Cys. In this regard, NAC has several advantages over Cys (unstable) or Cys-Cys (insoluble) for use as a dietary supplement. If used as a dietary supplement, the amino group of NAC (but not of Cys or Cys-Cys) would be protected from Maillard destruction (Baker, 1979; Boebel and Baker, 1982). Moreover, unlike Cys-Cys, NAC is soluble and therefore suitable for liquid formulations. Our data suggest dietary NAC is 100% bioavailable compared with Cys in terms of supporting BW gain. Additionally, it is apparent that a considerable tolerance exists for NAC, because excess supplemental NAC concentrations (up to 26.94 g/kg, equivalent to 20 g/kg of Cys) caused no untoward effects. This finding is intriguing considering that the provision of 20 g/kg of Cys was 7-fold greater than the estimated dietary requirement for digestible Cys (∼3 g/kg) in the purified diet used herein (Table 4). For these reasons, it seems warranted to consider NAC as a safe and efficacious dietary precursor of Cys.

In conclusion, the use of crystalline AA as dietary supplements in human and animal nutrition has gained popularity in the United States. Ingestion of excess dietary l-Cys, l-cystine, or both was previously shown to cause pernicious effects, including mortality, in both rats and chicks. Therefore, safe and functional alternative l-Cys precursor products have been sought. N-Acetyl-l-cysteine serves as a precursor for l-Cys because the α-amino nitrogen is protected by an acetyl group, and most tissues in the animal body possess the ability to remove this acetyl group, thus furnishing l-Cys. In the assays described herein, NAC was shown to be as effective as l-Cys in supporting chick growth, and it was less toxic than l-Cys in terms of growth depression, absence of elevated plasma sulfur AA concentrations, and lack of mortality. Thus, NAC was deemed to be a safe and effective precursor for l-Cys.

View Full Table | Close Full ViewTable 1.

Composition of experimental basal diets, as-fed basis

Cornstarch 538.7 53.9
Corn (82 g/kg of CP) 431.9
Soybean meal (480 g/kg of CP) 421.7
Soy protein isolate (824 g/kg of CP) 40.0
Casein (848 g/kg of CP) 25.0
AA mixture3 198.6
Soybean oil 100.0 50.0
Solka-Floc4 30.0
Complete mineral mix5 53.7
Limestone (CaCO3) 13.0
NaCl 4.0
Dicalcium phosphate 18.0
Trace-mineral mix6 1.5
Vitamin mix 2.07 2.08
Choline chloride 2.0 2.0
NaHCO3 10.0
Ethoxyquin (125 mg/kg) +
dl- α-Tocopheryl acetate (20 mg/kg) +
dl-Met 1.5
Bacitracin premix9 0.5

View Full Table | Close Full ViewTable 2.

Methionine requirement of chicks fed diets containing excess dietary cyst(e)ine (assay 1)1

1. None 35e 122c 284e
2. As 1 + 3.5 g/kg of Cys-Cys 56d 127c 439d
3. As 2 + 0.5 g/kg of Met 120c 194b 621c
4. As 2 + 1.0 g/kg of Met 158b 219a 723b
5. As 2 + 1.5 g/kg of Met 176a 231a 764ab
6. As 2 + 2.0 g/kg of Met 187a 236a 795a
7. As 2 + 2.5 g/kg of Met 188a 236a 795a
Pooled SEM 5.4 7.2 15.7

View Full Table | Close Full ViewTable 3.

Bioavailability of N-acetyl-l-Cys (NAC) relative to Cys for chicks fed a cyst(e)ine-deficient purified diet (assay 2)1

1. None 135d 223b 606f
2. As 1 + 0.35 g/kg of Cys 161bc 245a 655de
3. As 1 + 0.70 g kg of Cys 171bc 247a 691c
4. As 1 + 1.05 g/kg of Cys 200a 266a 751b
5. As 1 + 2.50 g/kg of Cys 204a 248a 819a
6. As 1 + 0.47 g/kg of NAC 160c 245a 650e
7. As 1 + 0.95 g/kg of NAC 179b 261a 686cd

View Full Table | Close Full ViewTable 4.

Growth performance1 and plasma-free sulfur AA (SAA) concentrations2 of chicks fed excess N-acetyl-l-Cys (NAC; assay 3)

1. None 203a 306a 663a 99.2b 557.9 47.0b 0.0c
2. As 1 + 13.47 g/kg of NAC 208a 314a 662a 116.9ab 440.7 63.0ab 92.0b
3. As 1 + 26.94 g/kg of NAC 203a 298ab 684a 153.2a 411.5 67.9a 189.0a
4. As 1 + 40.41 g/kg of NAC 176b 280b 631b 129.4ab 483.8 53.0ab 212.1a
5. As 1 + 53.88 g/kg of NAC 133c 231c 574c 156.6a 472.3 55.1ab 239.6a
Pooled SEM 4.2 6.7 8.1 13.8 54.0 6.3 16.2
Figure 1.
Figure 1.

Body weight gain of young chicks (4 replicate pens of 3 chicks each) fed graded concentrations of supplemental Met in a purified crystalline AA diet containing more than adequate dietary cyst(e)ine (assay 1). A 1-slope broken-line regression was fitted to the data according to procedures described by Robbins et al. (2006); this model provided an estimated supplemental Met requirement of 1.21 ± 0.07 g/kg. XLR denotes X when less than the requirement or break point. Error bars represent pooled SEM.