Skeletal muscle protein mobilization during the. progression of lactation. Emma J. Clowes, Frank X. Aherne, and Vickie E. Baracos.

Transcription

1 Articles in PresS. Am J Physiol Endocrinol Metab (October 26, 2004). doi: /ajpendo i Skeletal muscle protein mobilization during the progression of lactation Emma J. Clowes, Frank X. Aherne, and Vickie E. Baracos Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Alberta, T6G 2P5 Abbreviated title: Mechanisms of lactating sow muscle protein loss Name and address for mailing proofs: Dr. Vickie E. Baracos, University of Alberta, Department of Oncology, University Avenue, Edmonton, Alberta, T6G 1Z2 Contact: telephone number: fax number: address: vickie.baracos@ualberta.ca Copyright 2004 by the American Physiological Society.

2 ii ABSTRACT To investigate changes in muscle metabolism during lactation, serial biopsy of the triceps brachii was conducted in first-parity sows subjected to three degrees of selective protein mobilization through restriction of dietary protein intake (see J Anim Sci 81: , 2003). Muscle biopsies were taken 7 days prior to parturition and at 12 and 23 days of lactation. The following changes occurred after parturition, were progressive, and were significantly magnified in animals under the greatest degree of dietary protein restriction, and hence of protein mobilization. Decreased RNA:DNA ratio (capacity for protein synthesis) was observed. The presence of increased expression of several elements of the ubiquitin-proteasome proteolytic pathway suggested a robust catabolic response. However, as lactation progressed, and especially under conditions of increased dietary protein restriction, protein mobilization increased, muscle RNA:DNA ratio fell further, protease gene expression continued to rise, tissue free glutamine levels rose dramatically, and essential amino acid levels, especially branch chain amino acids and threonine fell to below pre-partum levels. Keywords: ATP-ubiquitin proteosome proteolytic pathway; muscle free amino acids; muscle RNA, DNA, and protein concentration

3 1 INTRODUCTION Lactation imposes a unique challenge to mammalian protein metabolism, especially for species such as the pig, which support high growth rates in their 10 to 12 offspring. This state represents a physiological maximum in protein anabolism with high rates of net protein export to the mammary gland. Milk production is sufficient to support a 40 to 50 kg net increase in litter body weight (equivalent to 20 to 25% of the maternal body mass) in about three weeks (10, 11, 33). Because of the high rate of net protein export from the dam, in the form of milk, lactation can also represent a physiological maximum in protein mobilization in the dam, as dietary intakes are often inadequate. The maternal catabolic response in lactation can approach levels seen in hypercatabolic patients after severe burn injury, with daily negative N balance at peak lactation of the order of 0.26 g N kg body wt -1 day -1 (0.9 g N kg Body Wt day -1 ; 21). Lactation thus represents a highly orchestrated physiologic state featuring intense milk protein anabolism and net catabolism of body protein reserves, especially skeletal muscle. The capacity to mobilize body protein from skeletal muscle is an important resource for lactation (10, 11, 12, 30), which enables the dam to care for and nourish her offspring, and at least temporarily minimizes the maternal commitment to foraging for food. Muscle protein mobilization is an adaptive response that enables a high level of milk production under conditions of poor nutrient supply (2, 21). Litter growth is largely maintained, even when feed is restricted to 50% of ad libitum intake (36). Maternal protein mobilization is related to overall milk protein production, and is proportional to the number of offspring suckled (5). Protein mobilization is also finely tuned in proportion to the dietary protein supply. At constant litter size, larger discrepancies between dietary protein intake and the demand for milk synthesis result in a greater impetus for protein mobilization (11, 12, 21, 45). Maternal protein loss cannot

4 2 continue unabated, and is attenuated if total protein mobilization exceeds a certain proportion of total body protein mass (11, 12) to prevent excessive depletion. Finally, protein mobilization must be sharply curtailed when weaning of the offspring occurs and milk production ceases. Our research was motivated by a desire to understand the alterations in skeletal muscle protein metabolism that permit a highly controlled diversion of muscle protein reserves towards milk production. Protein synthesis in skeletal muscle is highly regulated, and a fall in the rate of this process is suggested to contribute to protein mobilization during starvation (24, 26) and in lactation (6, 8, 34, 39). However, the regulation of protein catabolism in lactating animals remains unclear, and the participation of skeletal muscle proteolysis is especially poorly understood. We hypothesized that at constant litter size in lactating pigs, greater protein restriction would result in a larger proteolytic response. To explore potential mechanisms by which loss of maternal protein might be orchestrated, we developed a biopsy technique to collect triceps muscle samples in late gestation (day 107), mid-lactation (day 9 to 12) and within 3 h of weaning from the same animals. Samples were taken at all three time points, from pigs that lost 7, 9, or 16% of their body protein in lactation under conditions of differential dietary protein supply (11). Muscle RNA, DNA, protein, the expression of genes involved in skeletal muscle proteolysis, and muscle free amino acids were measured. MATERIALS AND METHODS This study was conducted in accordance with the Canadian Council of Animal Care Guidelines, and was approved locally by the Institutional Animal Policy and Welfare Committee.

5 3 The experimental design, housing, and management of gilts during gestation and lactation, and their reproductive and lactational performance are described in Clowes et al. (2003a; 11). In brief, upon parturition first-parity sows (Camborough, PIC, Acme, AB, Canada) were allocated to be fed divergent levels of total lysine (50, 35, and 24 ± 1.0 g/day) and dietary protein (878, 647 and 491 ± 19 g/day). These dietary treatments induced sows to lose approximately 7, 9, or 16% of their body protein, as a percentage of their protein mass at parturition, in lactation, and to be in a more negative lysine balance in the first 20 days of lactation (-16, -27, and -42 g/day). At least 7 sows were allocated to each treatment; 8 sows to the 7% body protein loss treatment, 7 sows to the 9% body protein loss treatment, and 10 sows to the 16% body protein loss treatment. Losses of backfat (1.3 ± 0.29mm) during lactation were small and not different among treatments because sows were fed a similar energy level (61 ± 2.0 MJ Metabolizable Energy/day). In late gestation (day 107) and mid-lactation (day 9 to 12) a muscle sample was collected from alternate sides of each sow by biopsy. On day 23 of lactation, within a few minutes of slaughter, a third muscle sample was collected 2 to 3 h after weaning. Samples were immediately trimmed of connective and adipose tissue, frozen in liquid nitrogen, and stored at -70 C until analyzed for RNA, DNA, protein, free amino acids, and mrna expression of several components of the ubiquitin-proteasome proteolytic system by Northern hybridization analysis. To ensure sows were fasted for 16 h prior to collection, the evening prior to surgery and slaughter, feed was removed from sows at 1800 h. The long head of the triceps brachii muscle was chosen because it is a mixed-fibre type muscle in the pig (27, 40), and thus represents the main source of mobilizable protein in the body (16).

6 4 Muscle Biopsy. A preliminary experiment was conducted, using four animals (196 kg body wt and 17mm backfat depth at parturition), to develop a muscle biopsy technique that could be undertaken in the animal unit rather than surgery room, to a) minimize the additional handling, transport, and environment change stresses on the sow; b) reduce the time the suckling litter and sow are apart; and c) determine the most appropriate time to collect samples. The surgery was conducted in the sow s individual farrowing pen, under general anaesthetic as described by Clowes et al. (1998, 10), with minor modifications. Samples (2 to 4 g) were surgically collected from the long head of the triceps brachii muscle, and this procedure took about 10 minutes. Alternate sides of the animal were used for sampling in late gestation and mid-lactation. The surgery proved a minimal disruption to the sow, and had little impact on the piglets. The piglets were only separated from the sow for the duration of the surgery and recovery period, which comprised one suckling bout (40 to 60 min), and immediately commenced suckling once placed back on the sow. Sow body temperatures were not elevated post-surgery, feed intake in lactation was only slightly reduced 1 to 3 days post-surgery, and body weight losses were small (- 5 kg) and in the range for first-parity sows in our herd not subjected to surgery. Respectable litter growth rates (2.36 ± 0.23 kg/day) were achieved. Analyses Before analysis, individual muscle samples were pulverized in a mortar and pestle in liquid nitrogen and stored at -70 C until analysis. Muscle free amino acids. About 25 mg of powdered muscle tissue was homogenized in 1 ml of 3% (w/v) trichloroacetic acid (Sigma Diagnostics, St. Louis, MO). The mixture was centrifuged at 2,800 x g and the supernatant fraction, containing tissue free amino acids, was

7 5 analyzed by HPLC using orthopthaldialdehyde derivatization and ethanolamine as the internal standard (38). Muscle RNA, DNA and protein. Muscle RNA, DNA and protein were quantified, in the same sample, using a modification of the Schmidt-Thannhauser procedure (31). In brief, 250 mg of frozen powdered muscle tissue was incubated on ice for 10 min, with 4 ml of 2% (w/v) perchloric acid (PCA), to precipitate out the protein and nucleotides. The precipitate was centrifuged for 15 min at 2,800 g and 4 C, and washed with 4 ml of 2% PCA at 4 C. The pellet was re-suspended in 4 ml of 0.3N NaOH, incubated at 37 C for 1 h, and the alkaline digest cooled on ice. One hundred micro litres of digest was removed, diluted with distilled de-ionized H 2 O to make a 0.1N NaOH alkaline digest, and stored at -20 C until BCA assay micro-titre plate protein quantification (Pierce, Rockford, IL). Ice-cold 12% PCA (2 ml) was added to the remaining alkaline digest, incubated on ice for 10 min, centrifuged for 10 min, and the RNA containing supernatant was collected. The pellet was washed in 4 ml of 2% PCA, centrifuged, and the washings and pellet collected separately. The acid-soluble supernatant plus washings were diluted with H 2 O to make a 1% PCA solution to spectrophotometrically quantify the RNA [3: RNA (µg/ml) = (32.9 A A 232 ) x dilution factor]. The acid-insoluble precipitate was incubation in 10% PCA at 70 C for 1 h (13), and the DNA was quantified using the diphenylamine method (7, 15). Northern hybridization. Analysis was conducted as described by Adegoke et al. (2003; 1), with minor modifications. Total RNA was extracted from 200 to 300 mg of muscle tissue using TRIzol TM (Gibco BRL / Life Technologies Inc., Frederick, MD; 10). Ethidium bromide (1.7 Og) was added to the sample buffer of each sample, and samples were made up to the same volume using sterile distilled

8 6 deionized H 2 O. Muscle RNA to be hybridized with the cdna probes [14-kDa ubiquitinconjugating enzyme (14-kDa E2) and the C9 subunit of the 20S proteasome] were transferred to a nitrocellulose membrane (Nitropure, Micro Separations Inc., Westborough, MA), and RNA to be hybridized with the riboprobe for ubiquitin was transferred to a nylon membrane (Zeta-Probe, Bio-Rad Laboratories Inc., Hercules CA), and fixed onto the membranes by baking, under vacuum, for 2 h at 80 C. All blots were exposed to X-ray film (Kodak BioMax MR, Eastman Kodak Company, Rochester, NY), and evaluated quantitatively using a GS-670 Imaging Densitometer (Bio-Rad Laboratories Inc., Hercules, CA). The densitometric scans were normalized to the fluorescence level of the 18S ribosomal RNA ethidium bromide band. The amount of 18S rrna loaded did not vary among treatments (P = 0.98) or time points (P = 0.62) and there was no treatment * time interaction (P = 0.76), establishing consistent RNA gel loading. After stripping, all membranes were reprobed with a 32 P-labeled cdna fragment encoding glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Concentration curves, of increasing amounts of total RNA from lactating sow muscle, were run to determine the amount of total RNA to load for each probe. For example, muscle mrna expression of both 14-kDa E2 transcripts increased linearly (r 2 > 0.95) from 5 to 20 Og of total RNA (Figure 1). Muscle mrna expression of both ubiquitin transcripts plateaued between 5 and 10 Og of total RNA. Therefore, 15 Og of total RNA was loaded for the 14-kDa E2 hybridization, and 3 Og for the ubiquitin hybridization. Muscle C9 mrna expression increased linearly (r 2 = 0.998) from 15 to 42 Og of total RNA (Figure 1), thus 30 Og of total RNA was loaded onto a 1.2% agarose-formaldehyde gel for this hybridization. The design of the main experiment entailed the comparison of 3 animal treatments and 3 time points per animal, for a total of 75 samples. To accommodate this sample number, five gels and blots were simultaneously run for

9 7 each mrna measured. Each of the three animal treatments was represented within each blot, and all three time points were tested on the same blot for each sow. Statistical Analyses. Analyses involving continuous variables were computed using the GLM procedures of SAS version 8.3 (SAS Institute, Cary, NC). Variables were tested for normality using the Shapiro-Wilk statistic. The effect of feeding level in gestation (2.8, 2.5, and 2.3 kg/day), the degree of body protein lost in lactation (7, 9, and 16% of the calculated body protein mass at parturition) and their interactions were analyzed over time (late gestation, midlactation, and weaning) on the various muscle parameters using repeated measures ANOVA. Membrane was included as a covariate in the analysis of mrna expression. Because gestational feeding regimen had no effect (P > 0.35) on any parameters measured in lactation, results for the gestation treatment are not presented. In the event of a significant (P < 0.05) time x lactation protein loss interaction, the differences among time within each lactation protein loss treatment were computed using a priori orthogonal contrasts. To test for differences among lactation treatments in the time x lactation protein loss interaction, the absolute or percentage difference between the two time periods was analyzed by ANOVA. If lactation treatment was significant in the repeated measures ANOVA for any parameter, differences among lactation treatments in late gestation were tested for. If no treatment difference was observed in late gestation the values within treatment over the three time periods were averaged, and an ANOVA was performed on the mean value. If significant treatment differences were detected (P < 0.05), then these differences were computed using Fisher s protected least significant difference test.

10 8 RESULTS Changes in muscle RNA, DNA and protein concentrations. The capacity for protein synthesis (RNA:DNA ratio) decreased by about 15% between late gestation and mid-lactation (P < 0.05) in all treatments, and continued decreasing in sows that lost the most protein, to 68% of pre-partum levels at weaning (Figure 2a). The protein:dna ratio also decreased (P < 0.001) between late gestation and weaning in all treatments (Figure 2b). The size of this decline reflected (P < 0.10) the degree of protein lost in lactation, and resulted in protein:dna ratios at weaning that were respectively 84, 77, and 69% of pre-partum levels in sows that lost 7, 9, and 16% of their protein mass at parturition. Muscle RNA concentrations decreased (P < 0.001) 10 to 15% between late gestation and mid-lactation in all treatments. By the day of weaning they increased to pre-partum levels in sows that lost 7 and 9% of their body protein mass, but remained low in sows that lost the most protein (Figure 2c). Muscle DNA concentrations increased (P < 0.001) by about 5% between late gestation and mid-lactation in all treatments, and then increased to weaning in a manner that reflected the degree of protein lost in lactation (Figure 2d). The increase was largest (P < 0.05) in sows that lost the most protein, resulting in the DNA concentration of these sows at weaning being 134% of pre-partum levels. Changes in muscle ubiquitin-proteasome proteolytic pathway gene expression. When equal amounts of total RNA from each animal and time point were studied, marked changes in GAPDH mrna expression were observed over the course of lactation and by treatment (Figure 3). GAPDH mrna fell by nearly 50% (P < 0.001) between late gestation and mid-lactation and remained low until weaning. The overall GAPDH mrna level was lowest (P < 0.01) in sows that lost the most protein in lactation. These differences precluded the use of GAPDH mrna as a "housekeeping" gene to normalize gene expression. Therefore, the measured mrna expression

11 9 was corrected for slight differences in loading using the relative fluorescence of the 18S rrna ethidium bromide stain. Because the level of muscle mrna, for all genes measured, did not differ among treatments in late gestation (pre-partum), gene expression data are presented as a percentage of pre-partum values. Muscle mrna levels of several key elements of the ATP-ubiquitin-proteasome dependent proteolytic pathway in muscle increased (P < 0.05) as lactation progressed (Figure 4). The 1.2 kb 14-kDa E2 transcript showed the most marked increase. Many of the changes in mrna expression between late gestation and lactation reflected the degree of protein mobilization elicited by the dietary treatments in lactation, such that at weaning expression was highest in sows that lost the most protein (Table 1). The 1.2 kb 14-kDa E2 transcript more than doubled between late gestation and mid-lactation in all treatments. Between mid-lactation and weaning the increase in expression was highest (P < 0.05) in sows that lost the most protein, and were respectively 0, 60 and 120% of mid-lactation mrna expression values in sows that lost 7, 9, and 16% of their protein mass at parturition (Figure 4a). The 1.8 kb 14-kDa E2 transcript did not change between late gestation and midlactation in any treatment. But, expression of this transcript almost doubled (P < 0.05) between mid-lactation and weaning in sows that lost the most protein, and did not change in the other two treatments (Figure 4b). Overall the 1.2 kb ubiquitin transcript expression was highest (P < 0.05) in sows that lost the most protein, and were respectively 0.28, 0.37, and 0.40 ± 0.03 arbitrary densitometric units in sows that lost a 7, 9, and 16% of their protein mass at parturition (Figure 4c). In all treatments, 2.6 kb ubiquitin transcript expression increased by 40 to 60% over lactation. Although there were no significant treatment effects, the expression tended to be lowest in sows that mobilized the least protein in lactation (Table 4d). Expression of the proteasome

12 10 subunit C9 mrna also increased (P < 0.001) between late gestation and weaning. This increase tended (P = 0.09) to be highest in sows that lost the most protein (Figure 4e), and was reflected in a higher (P = 0.002) C9 expression at weaning in these sows (Table 1). Changes in muscle free amino acid concentrations. The chromatographic procedure used resolved 19 amino acids. Because muscle free amino acid concentrations in late gestation did not differ among treatments the amino acid data are presented as a percentage of pre-partum values. The average concentration of individual muscle free amino acids in late gestation is presented in Table 2. Muscle free amino acids were divided into two groups based on their pattern of change between late gestation and weaning. The first group consisted of six essential amino acids (EAA; leucine, isoleucine, valine, phenylalanine, lysine, and threonine). Methionine and histidine were not included in this group because a value was not obtained for every sow at every time point. Arginine was not included because its concentrations did not change over lactation. The second group consisted of seven non-essential amino acids (NEAA; alanine, glycine, serine, aspartic acid, asparagine, and glutamic acid). Glutamine was excluded from this group because it behaved differently from the other NEAA. Most muscle free amino acid (NEAA, EAA and glutamine) concentrations increased (P < 0.001) by about 30% between late gestation and mid-lactation (Figure 5a). A divergence in the pattern of free EAA and NEAA concentrations then occurred between mid-lactation and weaning. At weaning, muscle free NEAA concentrations remained elevated above pre-partum levels, as their concentrations either did not change (alanine, glycine, and serine) or slightly declined (asparagine, aspartate, and glutamate) between mid-lactation and weaning. In contrast, muscle free glutamine concentrations increased (P < 0.001) over this time period, resulting in muscle free glutamine concentrations at weaning that were more than double pre-partum levels

13 11 (Figure 5a). Muscle free taurine and citrulline concentrations were similar in late gestation and mid-lactation and increased (P < 0.05) about 25% between mid-lactation and weaning. By contrast EAA concentrations declined (P < 0.001) between mid-lactation and weaning to be slightly above (lysine and histidine), at the same level (valine, isoleucine, and tryptophan), or below (leucine, phenylalanine, and threonine) pre-partum levels. Unlike the other EAA measured, methionine and threonine concentrations did not change between late gestation and mid-lactation, and then declined 30% to weaning. This decline was significant (P < 0.05) for threonine only. The increase in muscle free NEAA concentrations between late gestation and mid-lactation (Figure 5b), and in glutamine concentrations between late gestation and weaning (Figure 5c) was proportional to the degree of protein mobilized by the sow in lactation. Increases in these amino acids were highest (P < 0.05) in sows that lost the most protein. The NEAA concentrations in mid-lactation were respectively 122, 131, and 146% of pre-partum values, and the glutamine concentrations at weaning were respectively 205, 240 and 308% of pre-partum values in sows that lost a 7, 9, and 16% of their protein mass at parturition. The decrease in muscle free EAA concentrations between mid-lactation and weaning, such as isoleucine, was greatest (P < 0.05) in sows that lost the most protein (Figure 5d). Similarly muscle free valine concentrations at weaning were lowest (P < 0.05) in sows that had mobilized the most protein. DISCUSSION Methodological considerations Use of indirect indices of muscle anabolism and catabolism

14 12 Direct measures of muscle protein synthesis and degradation were not made in our study. Classic techniques for measurement of these processes employing isotopically labeled amino acids would have been very difficult in the context studied here. Some of the limitations for research on protein turnover in large animals, and in swine in particular are: 1. cost prohibitive if sows and their progeny have to be killed, and disposed of suitably, for tissue sampling or due to contamination with radioactivity rather than returned to the research herd; 2. studies using isotopes are especially costly due to large animal size; 3. tracer recycling precludes making repeated measures in the same animals over time (14, 25); 4. due to unique metabolic pathways, urinary 3-methyl histidine may not be used to follow muscle protein catabolism in this species. A further consideration is that the level of stress and disruption in the experimental setting may not be compatible with maintenance of normal maternal behavior, as the approaches for in situ muscle metabolism tend to be relatively invasive and may require venous and arterial cannulation/ perfusion/ biopsy. For all of these reasons we opted for a variety of indirect indices of muscle metabolism in our serial biopsies. RNA:DNA ratio (capacity for protein synthesis) has been extensively reported and related to determined rates of protein synthesis (29). Proteolytic processes in muscle appear to have a large degree of transcriptional regulation, and increases in mrna expression of genes that encode proteins in the ATP-dependent proteolytic pathway (the co-factor ubiquitin, ubiquitin conjugation system enzymes, and the proteasome complex) generally correlate with protein levels and directly measured protein degradation rates (23, 37, 42, 43). Changes in skeletal muscle s free amino acid pool reflect the net balance of several inputs (amino acid uptake; appearance from protein degradation) and outputs (protein synthesis; export) and may be considered as a crude barometer of the net balance of amino acid appearance and loss in the

15 13 tissue. Decreases in muscle free amino acid levels are often larger and more numerous than those in the plasma pool under conditions of muscle atrophy (17, 32, 41). Changes in muscle variables between late gestation and lactation RNA and gene expression We provided evidence that up-regulation of the ubiquitin-atp-dependent proteolytic pathway occurs at the level of gene expression in the muscle of lactating sows that mobilize protein. The fall in total RNA and in GAPDH mrna make the increases in mrna for ubiquitin, 14-kDa E2, and the C9 subunit of the 20S proteasome all that more striking and selective. The marked increase in the 1.2 kb 14-kDa E2 transcript agrees with other evidence that this transcript is tightly up-regulated in various conditions of muscle atrophy in rats (reviewed by 4, 22). This is the first time that this proteolytic pathway has been studied in the muscle of lactating animals. Up-regulation of this pathway occurs in numerous conditions involving protein catabolism including fasting, cancer, sepsis, diabetes, acidosis, burn injury and muscle denervation and in all of these cases protein degradation and mrna are increased (4, 22, 23). The increase in mrna expression of genes that encode proteins in the ubiquitin-atp-dependent proteasome proteolytic pathway suggest that the rate of muscle protein degradation was higher in our sows during lactation than gestation. This inference is supported by several studies in the lactating first-parity sow. In early lactation (day 1 and 4) myofibrillar protein breakdown was at least 15% higher, and protein degradation was higher in sows fed < 36 vs 55 g lysine/day (20, 45). These differences were maintained until the end of lactation (day 15 and 20). In several studies in the lactating rat (34, 35) and dairy goat (39), where only protein synthesis was measured, a fall in protein synthesis was either not observed during lactation or was too small to account by itself for the net muscle protein loss. From these studies it can also be inferred that

16 14 protein degradation rates rose, and Pine et al. (1994ab; 34, 35) suggested that the increase in protein degradation was quantitatively the more important effect. We observed a decrease in the capacity for protein synthesis in muscle between late gestation and mid-lactation. By the time the sow was weaned this variable had changed further in a manner related to the degree of protein mobilized. This agrees with our previous observation of lower (~-20%) triceps muscle RNA:DNA ratio on day 25 of lactation, in sows that had lost > 10% of their body protein (10). These reductions in the capacity for protein synthesis may underlie a decrease in the rate of muscle protein synthesis. This inference is supported by data in the rat and dairy goat in which the rate of muscle protein synthesis was measured by either a flooding or tracer dose of [ 3 H] phenylalanine or L-[ 35 S] methionine. Compared to dry goats, no reduction in the fractional rate of muscle protein synthesis was observed in early-lactation in goats that had mobilized a moderate amount of protein (~28 g/day; 6). But a 30% reduction was observed in goats, of a similar weight, that had mobilized more protein (~57 g/day; 8). Similarly, the rate of muscle protein synthesis only decreased in rats that lost a large proportion of their body weight (20 vs 5%) during lactation (34); a 25% reduction in the fractional, and over a 45% reduction in the absolute, rate of muscle protein synthesis was observed. Changes in the rate of muscle protein synthesis may help provide control and sensitivity to the degree of protein mobilized in lactation. The large reduction in muscle GAPDH mrna between late-gestation and lactation in our sows suggests that the glycolytic pathway in muscle was down-regulated, at least at the level of gene expression. This is consistent with our recent observation that the common pattern of gene expression in diverse forms of muscle atrophy includes down regulation of multiple enzymes participating in muscle energy metabolism (23). The data could also suggest that the size of the

17 15 glycolytic muscle fibres were reduced relative to the oxidative fibres. Glycolytic muscle fibres appear to be targeted for degradation in muscle undergoing protein mobilization (18) and this was reflected in a decrease in the activity of glycolytic enzymes such as GAPDH (19, 27). Muscle free amino acids. The muscle free amino acid pool increased during early lactation, for most amino acids. The pattern observed later in lactation was strikingly different, with increased levels of glutamine and a fall in the levels of threonine, leucine, phenylalanine to below pre-partum levels. These changes (accumulation or loss) reflect an imbalance between the inputs into the muscle free amino acid pool (rate of amino acid appearance from myofibrillar protein breakdown and uptake from blood) and its outputs (rates of protein synthesis and amino acid export out of the tissue into blood). We observed a sharp rise in muscle free glutamine levels in sows towards the end of lactation, in a manner related to the degree of protein loss. These data suggest a relative oversupply of glutamine, and this is consistent with the observation that plasma glutamine concentrations also increase in later lactation (28), and that free glutamine concentrations increased by almost 4-fold (8% to 31% of free amino acids) in sow milk over the course of lactation (44). Similarly, the relative depletion of several EAAs in muscle suggests an overall inadequacy of EAA supply in the free amino acid pool in muscle in late lactation. Changes in muscle metabolism upon loss of divergent degrees of body protein By the end of lactation, sows on our study were calculated to have lost ~16, 20, and 36% of their muscle mass present at parturition (9). However, the demand for milk continually increases as the litter grows, even after feed intake has reached a maximum, so that rates of muscle protein mobilization increase as lactation progresses. This would occur to the greatest degree when

18 16 dietary protein was most limited. All of the changes described above (RNA:DNA ratio, protease gene expression, and amino acid pools) were magnified in the animals on the highest dietary protein restriction. This makes sense, considering that a law of diminishing returns applies to mobilization of a fixed resource. To release the same total amount of amino acids per day, the fractional mobilization rate must increase progressively to compensate as muscle mass shrinks. Our data is consistent with a continuous attempt to keep raising muscle mobilization rate, so that during later lactation and in animals provided the lowest dietary protein intake, all of the observed effects were magnified (i.e. further up-regulation of the ubiquitin-atp-dependent proteasome proteolytic pathway and decrease in the capacity for protein synthesis). The situation of diminishing returns described above may provide a mechanism whereby lactation becomes self-limiting. The fractional rate of muscle protein mobilization may simply have a maximal rate. Once this is attained the overall daily rate of amino acid delivery from muscle will shrink with the size of the muscle protein mass to a point when milk production will decline, for lack of substrate. This would be consistent with the fall in milk protein concentration seen in late lactation, which was most prominent in animals with the most restricted protein intake (11, 12). Maternal N balance became less negative, indicating a reduction in protein mobilization, and growth of the litter of piglets slowed. The highest degree of protein loss was also accompanied by a decline in ovarian function. These physiologic changes to delay or prevent a further cycle of reproduction would also help preserve the remaining maternal protein mass.

19 17 Acknowledgements We thank Dr. Simon S. Wing (McGill University, Montreal, Canada) for provision of plasmids containing cdna sequences encoding rat 14-kDa ubiquitin-conjugating enzyme, Dr. Keiji Tanaka (Tokyo Institute for Medical Research, Tokyo, Japan) for kindly providing plasmids encoding the rat proteasome subunit C9, and Dr. J. Walker (University of Wisconsin-Madison) for the riboprobe for ubiquitin. The cdna probe for GAPDH was generated by RT PCR, kindly provided by Gordon Murdoch (University of Alberta), using primers designed according to Yelich et al. (1997; 46). The authors gratefully acknowledge the Alberta Agricultural Research Institute, Farming for the Future Matching Funding Program and Alberta Pork for their financial support. Pig Improvement Company (Canada) Ltd. provided the experimental animals, and the staff of the Swine Research Unit, especially Janes Goller and Jay Willis, cared for the animals and assisted with experimental procedures. The surgical expertise of Dr. Artur Cegielski, and the assistance of Charlane Gorsak and Brenda Tchir, the technical assistance of Renata Meuser, A Dunichand- Hoedl, Joan Turchinsky, Rose O Donoghue, Shirley Shostak, Gary Sedgewick, and other staff in the Department of Agricultural, Food and Nutritional Science, and the statistical assistance of Dr. R. T. Hardin, are gratefully acknowledged. Current address for Emma Clowes: Alberta Agriculture, Food and Rural Development, #204, St., Edmonton, AB, T6H 5T6. emma.clowes@gov.ab.ca

20 18 REFERENCES 1. Adegoke, OAJ, McBurney MI, Samuels SE, and Baracos VE. Modulation of intestinal protein synthesis and protease mrna by luminal and systemic nutrients. Am J Physiol Gastrointest Liver Physiol 284: G1017-G1026, Allison JB, and Wannemacher RW. The concept and significance of labile and over-all protein reserves of the body. Am J Clin Nutr 16: , Ashford AJ, and Pain VM. Effect of diabetes on the rates of synthesis and degradation of ribosomes in rat muscle and liver in vivo. J Biol Chem 261: , Attaix D, Aurousseau E, Combaret L, Kee A, Larbaud D, Rallière C, Souweine B, Taillandier D, Tilignac T. Ubiquitin-proteasome-dependent proteolysis in skeletal muscle. Reprod Nutr Dev 38: , Auldist DE, Carlson D, Morrish L, Wakeford CM, King RH. The influence of suckling interval on milk production of sows. J Anim Sci 78: , Baracos VE, Brun-Bellut J, and Marie M. Tissue protein synthesis in lactating and dry goats. Br J Nutr 66: , Burton KA. A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem J 62: , Champredon C, Debras E, Mirand PP, Arnal M. Methionine flux and tissue protein synthesis in lactating and dry goats. J Nutr 120: , Clowes EJ. Lactational protein loss and sow performance. PhD thesis. University of Alberta, Edmonton, AB, Canada, Clowes EJ, Williams IH, Baracos VE, Pluske JR, Cegielski AC, Zak LJ, and Aherne FX. Feeding lactating primiparous sows to establish three divergent metabolic states: II Effect

21 19 of nitrogen partitioning and skeletal muscle composition. J Anim Sci 76: , Clowes EJ, Aherne FX, Foxcroft GR, and Baracos VE. Selective protein loss in lactating sows is associated with reduced litter growth and ovarian function. J Anim Sci 81: , 2003a. 12. Clowes EJ, Aherne FX, Schaefer AL, Foxcroft GR, and Baracos VE. Parturition body size and body protein loss during lactation influence performance during lactation and ovarian function at weaning in first-parity sows. J Anim Sci 81: , 2003b. 13. Forsberg AM, Nilsson E, Werneman J, Bergstrom J, and Hultman E. Muscle composition in relation to age and sex. Clin Sci (Colch) 81: , Garlick PJ, McNurlan MA, and Preedy VR. A rapid and convenient technique for measuring the rate of protein synthesis in tissues by injection of [ 3 H]phenylalanine. Biochem J 192: , Giles KW, and Myers A. An improved diphenylamine method for the estimation of, deoxyribonucleic acid. Nature 207: 93, Goodman MN, and Ruderman NB. Starvation in the rat. I. Effect of age and obesity on organ weight, RNA, DNA, and protein. Am J Physiol Endocrinol Metab 239: E269-E276, Hammarqvist F, Wernerman J, Rustom A, von der Decken A, and Vinnars E. Addition of glutamine for total parenteral nutrition after elective abdominal surgery spares free glutamine in muscle, counteracts the fall in muscle protein synthesis, and improves nitrogen balance. Ann Surg 209: , 1989.

22 Henricksson J. The possible role of skeletal muscle in the adaptation to periods of energy deficiency. Eur J Clin Nutr 44 (Suppl 1): 55-64, Hood DA, and Pette D. Chronic long-term electrostimulation creates a unique metabolic enzyme profile in rabbit fast-twitch muscle. FEBS Lett 247: , Jones DB, and Stahly TS. Impact of amino acid nutrition during lactation on body nutrient mobilization and milk nutrient output in primiparous sows. J Anim Sci 77: , King RH, Toner MS, Dove H, Atwood CS, and Brown WG. Response of first-litter sows to dietary protein level during lactation. J Anim Sci 71: , Lecker SH, Solomon V, Mitch WE, and Goldberg AL. Muscle protein breakdown and the critical role of the ubiquitin-proteasome pathway in normal and disease states. J Nutr 129: 227S-237S, Lecker SH, Jagoe RT, Gilbert A, Gomes M, Baracos VE, Bailey J, Price SR, Mitch WE and Goldberg AL Multiple types of skeletal muscle atrophy involve a common program of changes in gene expression FASEB J 18: Li JB, and Wassner SJ. Effects of food deprivation and refeeding on total protein and actomyosin degradation. Am J Physiol Endocrinol Metab 246: E32-E37, Lobley GE, Harris PM, Skene PA, Brown D, Milne E, Calder AG, Anderson SE, Garlick PJ, Nevison I, and Connell A. Reponses in tissue protein synthesis to sub- and supramaintenance intake in young growing sheep: comparison of large-dose and continuousinfusion techniques. Br J Nutr 68: , 1992.

23 Lowell BB, Ruderman NB, and Goodman MN. Regulation of myofibrillar protein degradation in rat skeletal muscle during brief and prolonged starvation. Metabolism 35: , McAllister RM, Reiter BL, Amann JF, and Laughlin MH. Skeletal muscle biochemical adaptations to exercise training in miniature swine. J Appl Physiol 82: , McMillan DJ. Development of low protein diets for sows: Effect on performance and energy metabolism. MSc thesis, University of Alberta, Edmonton, Alberta, Canada, Millward DJ, Garlick PJ, James WPT, Nnan-Yelugo DO, and Ryatt JS. Relationship between protein synthesis and RNA content in skeletal muscle. Nature 241: , Mullan BP, and Williams IH. The chemical composition of sows during their first lactation. Anim Prod 51: , Munro HM, and Fleck A. Recent developments in the measurement of nucleic acids in biological materials. Analyst 91: 78-88, Munro HM. Free amino acid pools and their role in regulation. In: Mammalian Protein Metabolism, Vol 4, edited by H. N. Munro. New York: Academic Press, 1970, p Noblet J, and Etienne M. Estimation of sow milk nutrient output. J Anim Sci 67: , Pine AP, Jessop NS, and Allan GF. Maternal protein reserves and their influence on lactational performance in rats. 2. Effects of dietary protein restriction during gestation and lactation on tissue protein metabolism and Na +, K + ATPase (EC ) activity. Br J Nutr 72: , 1994a.

24 Pine AP, Jessop NS, Allan GF, and Oldham JD. Maternal protein reserves and their influence on lactational performance in rats. 4. Tissue protein synthesis and turnover associated with mobilization of maternal protein. Br J Nutr 72: , 1994b. 36. Pluske JR, Williams IH, Zak L, Clowes EJ, Cegielski AC, and Aherne FX. Feeding lactating primiparous sows to establish three divergent metabolic states. 3. Milk production and pig growth. J Anim Sci 76: , Sacheck JM, Ohtsuka A, McLary SC, and Goldberg AL. IGF-1 stimulates muscle growth by suppressing protein breakdown and expression of atrophy related ubiquitin ligases, atrogin-1 and MuRF1. Am J Physiol Endocrinol Metab 287:E , Sedgwick GW, Fenton TF, and Thompson JR. Effect of protein precipitating agents on the recovery of plasma free amino acids. Can J Anim Sci 71: , Tesseraud S, Grizard J, Debras E, Papet I, Bonnet Y, Bayle G, and Champredon C. Leucine metabolism in lactating and dry goats: effect of insulin and substrate availability. Am J Physiol Endocrinol Metab 265: E402-E413, Uhrin P and Liptaj T. Effect of training on fibre composition and phosphate metabolites in rest measured in vitro in muscles of young pigs. Comp Biochem Physiol B Biochem Mol Biol 102B: , Vinnars E, Bergstöm J, and Fürst P. Influence if the postoperative state on the intracellular free amino acids in human muscle tissue. Ann Surg 182: , Voisin L, Breuillé D, Combaret L, Pouyet C, Taillandier D, Aurousseau E, Obled C, and Attaix D. Muscle wasting in a rat model of long-lasting sepsis results from the activation of

25 23 lysosomal Ca 2+ -activated, and ubiquitin-proteasome proteolytic pathways. J Clin Invest 97: , Wing SS, and Banville D. 14-kDa ubiquitin-conjugating enzyme: structure of the rat gene and regulation upon fasting and by insulin. Am J Physiol Endocrinol Metab 267: E39- E48, Wu G, and Knabe DA. Free and protein-bound amino acids in sow s colostrum and milk. J Nutr 124: , Yang H, Pettigrew JE, Johnson LJ, Shurson GC, Wheaton JE, White ME, Koketsu Y, Sower SF, and Rathmacher JA. Effects of dietary lysine intake during lactation on blood metabolites, hormones, and reproductive performance in primiparous sows. J Anim Sci 78: , Yelich JV, Pomp D, and Geisert RD. Ontogeny of elongation and gene expression in the early developing porcine conceptus. Biol Reprod 57: , 1997.

26 24 Fig. 1. Total RNA concentration curves for lactating sow muscle mrna expression for 14-kDa ubiquitin-conjugating enzyme (14-kDa E2) transcripts and the proteasome subunit C9. Total RNA indicated by membrane 18S rrna ethidium bromide stain from muscle hybridized against either a) the 1.8 and 1.2 kb 14-kDa E2 transcripts and b) the proteasome subunit C9. * mrna transcripts expressed in arbitary densitometric units. The mrna expression increased linearly with increasing levels of total RNA loaded for both 14-kDa E2 transcripts to 20 Og (1.8 kb: y = 0.845x 0.465, r 2 = 0.954; 1.2 kb: y = 0.861x 3.59, r 2 = 0.999), and for C9 (y = 0.486x ; r 2 = 0.998) to over 42 Og. Fig. 2. Changes in muscle a) RNA:DNA ratio, b) protein:dna ratio, c) RNA concentration, and d) DNA concentration, from day -7 to 23 of lactation, in first-parity sows that lost in lactation 7, 9, and 16% of their protein mass at parturition. Muscle RNA concentrations and RNA:DNA ratio differed (P < 0.05) with lactation treatment. All variables changed (P < 0.001) with time over lactation. Variables on day 23 without a common letter differ, xy (P < 0.05) and wx (P < 0.10). Changes in muscle variables between day 12 and 23 without a common letter (ab) differ (P < 0.05). Fig. 3. Muscle glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mrna expression during lactation, as a) Northern-blot hybridized against GAPDH from porcine triceps muscle (15 Og) taken immediately pre-partum (day 7), at mid-lactation (day 12), and at weaning (day 23); and b) quantification of GAPDH mrna, expressed in arbitrary densitometric units, from first-parity sows that lost in lactation 7, 9, and 16% of their protein mass. Level of total RNA loaded is indicated by the inverted 18S rrna ethidium bromide staining. Muscle GAPDH mrna

27 25 expression decreased (P < 0.001) between day -7 and 12, and differed (P < 0.05) among lactation treatments. Fig. 4. Muscle mrna expression of a) the 1.2 kb, and b) 1.8 kb transcripts of 14-kDa ubiquitinconjugating enzyme (14-kDa E2), c) the 1.2 kb, and d) 2.6 kb transcripts of ubiquitin (Ub), and e) the proteasome subunit C9, from day -7 (pre-partum) to 23 of lactation, in first-parity sows that lost 7, 9, and 16% of their protein mass in lactation. * mrna expression was normalized to 18S rrna, and expressed as a percentage of pre-partum values. All muscle variables increased (P < 0.001) over lactation. Overall ubiquitin 1.2 kb transcript mrna expression differed among treatments (P < 0.05) and was highest in sows that lost the most protein. The increase in mrna expression of both 14-kDa E2 transcripts, between mid-lactation and weaning, without a common letter (xy) differ (P < 0.05). The increase in C9 mrna expression, between late gestation and weaning, without a common letter (vw) tend to differ (P = 0.06). Fig. 5. Changes in free muscle: a) non-essential (NEAA) and essential (EAA) amino acid and glutamine (Gln) concentrations from day -7 (pre-partum) to 23 of lactation, as a percentage of pre-partum levels; treatment differences in b) NEAA, and c) glutamine concentrations on day 12 and 23 of lactation, as a percentage of pre-partum levels, in first-parity sows that lost in lactation 7, 9, and 16% of their protein mass; and d) EAA and isoleucine concentrations on day 23 of lactation, expressed as a percentage of day 12 concentrations. NEAA include: ala, gly, ser, asp, asn, and glu. EAA include: leu, ile, val, phe, lys and thr. Percentage amino acid levels without a common letter differ, xy (P < 0.05), and wx (P < 0.10).

28 26 Table 1. Muscle mrna expression of the ubiquitin-atp-dependent proteasome proteolytic pathway at weaning in sows with divergent protein losses Protein loss in lactation, % parturition mass P Ubiquitin, 2.6 kb 0.79 ± ± ± Ubiquitin, 1.2 kb 0.40 ± 0.04* 0.49 ± 0.04* 0.54 ± kDa E2, 1.8 kb 0.83 ± 0.11* 0.90 ± 0.11* 1.40 ± kDa E2, 1.2 kb 0.68 ± 0.16* 0.90 ± 0.18* 1.37 ± Proteasome subunit C ± 0.44* 10.3 ± 0.47* 12.9 ± Values are least square mean ± SE for triceps brachii mrna at weaning (day 23 of lactation), expressed in arbitary densitometric units normalized to 18S rrna, from first-parity sows that lost 7 (n = 8), 9, (n = 7), and 16 (n = 10) % of their protein mass in lactation. 14-kDa E2, 14-kDa ubiquitin-conjugating enzyme E2. * Means within a row with different symbols differ by the significance level in that row.

29 27 Table 2. Pre-partum triceps muscle free amino acid concentrations Amino Acid (5mol / mg wet tissue) SD Essential Amino Acids Arginine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Threonine Valine Non-Essential Amino Acids Alanine Aspartic Acid Asparagine Glutamic Acid Glutamine Glycine Ornithine Serine Taurine Values are means of 25 gilts taken in late gestation (day 107; pre-partum)

30 28 A mrna 14-kDa E2 14-kDa E2 (1.8 kb) kb 14-kDa E2 1.2 kb 14-kDa E B mrna C S rrna 14-kDa E2 (1.2 kb) C9 18 S rrna µg total RNA µg total RNA Figure 1

31 29 A 2.5 RNA:DNA a a b y y x B 0.9 protein:dna x wx C w RNA, ug/mg D 0.28 DNA, ug/mg a ab b x wx w Day of Lactation 7% 9% 16% Figure 2

32 30 A B 3.5 Day of Lactation GAPDH 18 S d -7 d 12 d 23 7% 9% 16% 7% 9% 16% 7% 9% 16% mrna GAPDH * * Protein Loss in Lactation Day of Lactation 7% 9% 16% Figure 3

33 31 A B kb 14-kDa E y xy x 1.8 kb 14-kDa E y x x C 300 D kb Ub kb Ub E Day of Lactation 160 w C9 140 v 7% 9% 16% 120 v Day of Lactation Figure 4

34 32 A B % pre-partum levels GLN NEAA NEAA, %prepartum w wx x C 80 EAA Day of Lactation D Day of Lactation 350 y 0 Gln, %prepartum xy x % mid-lactation levels y y x y y x Day of Lactation 7% 9% 16% Figure 5