key: cord-0887347-z9suttt5
authors: van der Meer, Y.; Lammers, A.; Jansman, A. J. M.; Rijnen, M. M. J. A.; Hendriks, W. H.; Gerrits, W. J. J.
title: Performance of pigs kept under different sanitary conditions affected by protein intake and amino acid supplementation(,)(2)
date: 2016-11-03
journal: J Anim Sci
DOI: 10.2527/jas.2016-0787
sha: 7cf5574a22ba1e26e1155f19264de7b63b9e1c0f
doc_id: 887347
cord_uid: z9suttt5

There is growing evidence that requirements for particular AA increase when pigs are kept under low sanitary conditions. The extent to which reduction in growth performance is related to these increased requirements is unclear. To evaluate this relationship, an experiment (2 × 2 × 2 factorial arrangement) was performed with 612 male pigs (9 per pen) kept under low sanitary conditions (LSC) or high sanitary conditions (HSC) and offered ad libitum access to either a normal CP concentration diet (NP; 17, 15, and 15% CP for the starter, grower, and finisher phase, respectively) or a low CP concentration diet (LP; 20% CP reduced relative to NP for each phase), each of which containing a basal AA profile (AA-B) or a supplemented AA profile (AA-S). The supplemented diet type contained 20% more Met, Thr, and Trp relative to Lys on an apparent ileal digestible basis compared with the basal diet type. Pigs were followed for a complete fattening period and slaughtered at a targeted pen weight of 110 kg. Haptoglobin concentrations in serum (0.92 g/L for LSC and 0.78 g/L for HSC) and IgG antibody titers against keyhole limpet hemocyanin (3.53 for LSC and 3.08 for HSC) collected in the starter, grower, and finisher phases and pleuritis scores at slaughter (0.51 for LSC and 0.20 for HSC) were greater for LSC pigs compared with HSC pigs (P ≤ 0.01), illustrating that sanitary conditions affected health conditions. The ADG and G:F were greater for HSC pigs compared with LSC pigs (P ≤ 0.01). The number of white blood cells (WBC) was higher in (AA-S)–fed pigs compared with (AA-B)–fed pigs when kept at LSC but not at HSC [SS (sanitary conditions) × AA interaction, P = 0.04]. Pigs fed NP had a lower number of WBC compared with pigs fed LP (P = 0.02). The number of platelets in pigs fed AA-S diets was higher compared with pigs fed AA-B diets (P ≤ 0.01). A 20% reduction in dietary supplementation of Met, Thr, and Trp relative to Lys decreased G:F more in LSC pigs than in HSC pigs (interaction, P = 0.03), illustrating that dietary requirements for these AA differ depending on sanitary conditions. This study, performed under practical conditions, shows that AA requirements are dependent on sanitary conditions. Furthermore, supplementation of diets with particular AA may improve performance, especially under poor hygienic conditions. Dietary protein concentration as well as Met, Thr, and Trp supplementation can modify immune status, which may influence resistance to subclinical and clinical diseases.

ABSTRACT: There is growing evidence that requirements for particular AA increase when pigs are kept under low sanitary conditions. The extent to which reduction in growth performance is related to these increased requirements is unclear. To evaluate this relationship, an experiment (2 × 2 × 2 factorial arrangement) was performed with 612 male pigs (9 per pen) kept under low sanitary conditions (LSC) or high sanitary conditions (HSC) and offered ad libitum access to either a normal CP concentration diet (NP; 17, 15, and 15% CP for the starter, grower, and finisher phase, respectively) or a low CP concentration diet (LP; 20% CP reduced relative to NP for each phase), each of which containing a basal AA profile (AA-B) or a supplemented AA profile (AA-S). The supplemented diet type contained 20% more Met, Thr, and Trp relative to Lys on an apparent ileal digestible basis compared with the basal diet type. Pigs were followed for a complete fattening period and slaughtered at a targeted pen weight of 110 kg. Haptoglobin concentrations in serum (0.92 g/L for LSC and 0.78 g/L for HSC) and IgG antibody titers against keyhole limpet hemocyanin (3.53 for LSC and 3.08 for HSC) collected in the starter, grower, and finisher phases and pleuritis scores at slaughter (0.51 for LSC and 0.20 for HSC) were greater for LSC pigs compared with HSC pigs (P ≤ 0.01), illustrating that sanitary conditions affected health conditions. The ADG and G:F were greater for HSC pigs compared with LSC pigs (P ≤ 0.01). The number of white blood cells (WBC) was higher in (AA-S)-fed pigs compared with (AA-B)-fed pigs when kept at LSC but not at HSC [SS (sanitary conditions) × AA interaction, P = 0.04]. Pigs fed NP had a lower number of WBC compared with pigs fed LP (P = 0.02). The number of platelets in pigs fed AA-S diets was higher compared with pigs fed AA-B diets (P ≤ 0.01). A 20% reduction in dietary supplementation of Met, Thr, and Trp relative to Lys decreased G:F more in LSC pigs than in HSC pigs (interaction, P = 0.03), illustrating that dietary requirements for these AA differ depending on sanitary conditions. This study, performed under practical conditions, shows that AA requirements are dependent on sanitary conditions. Furthermore, supplementation of diets with particular AA may improve performance, especially under poor hygienic conditions. Dietary protein concentration as well as Met, Thr, and Trp supplementation can modify immune status, which may influence resistance to subclinical and clinical diseases. trients, such as AA. If the enhanced requirements are not compensated for by greater intake, repartitioning of dietary protein or AA away from development and growth tissues occurs (Le Floc'h et al., 2009) toward use by the immune processes. This shift of nutrients, together with an often-observed reduction in feed intake in case of immune stimulation, reduces pig performance (Klasing and Johnstone, 1991; Pastorelli et al., 2012a) .

Despite the growing evidence that particular AA requirements are dependent on immune system activation, it remains unclear to what extent reduced performance during (subclinical) infections is related to changes in AA requirements and whether dietary supplementation of these AA can reverse the performance loss. Moreover, most evidence for increased AA requirements is obtained in studies in which pigs were repeatedly challenged with lipopolysaccharide (lPS; e.g., Kim et al., 2012; de Ridder et al., 2012; Rakhshandeh et al., 2010) or Complete Freund's Adjuvant (Kampman-van de Hoek et al., 2015; Le Floc'h et al., 2008) , and it can be questioned to what extent these results can be extrapolated to pigs in commercial conditions (Pastorelli et al., 2012b) . Therefore, we studied, under practical conditions, if the performance and immune status of pigs kept under different sanitary conditions is influenced by protein intake and AA supplementation. We hypothesized that increased provision of Met, Thr, and Trp would increase performance of pigs, particularly when kept under low sanitary conditions and low dietary protein intake.

The experimental protocol was approved by the Animal Care and Use Committee of Wageningen University, the Netherlands.

In a 2 × 2 × 2 factorial arrangement, groups of pigs were allocated to either high sanitary conditions (HSc) or low sanitary conditions (lSc) and were offered ad libitum access to 2 different diets, a normal CP concentration diet (nP) or a low CP concentration diet (lP), each having either a basal dietary AA profile (AA-B) or supplemented dietary AA profile containing 20% more Met, Thr, and Trp compared with the basal profile (AA-S).

In total, 612 (Topigs 20 × Tempo; Topigs, Helvoirt, The Netherlands) newborn boar piglets were selected on a commercial nursery farm in the Netherlands and allocated to either the LSC or HSC treatment.

Per nursery room, half of the boar piglets were selected for LSC and the other half for HSC treatment. Only HSC piglets received vaccinations in the first 9 wk of age. The HSC piglets were vaccinated, at 1 to 2 wk of age, against Mycoplasma hyopneumoniae (Porcilis M Hyo; MSD Animal Health, Boxmeer, the Netherlands); at 4 to 5 wk of age against M. hyopneumoniae, porcine circovirus type 2 (PcV2), and porcine reproductive and respiratory syndrome (PrrS; Porcilis M Hyo, Porcilis Circo, and Porcilis PRRS, respectively; MSD Animal Health) and Lawsonia intracellularis (Enterisol Ileitis Boehringer Ingelheim B.V., Alkmaar, the Netherlands); at 6 to 7 wk of age against Actinobacillus pleuropneumoniae (APP; Porcilis APP; MSD Animal Health) and influenza A virus (Gripovac3; Merial B.V., Velserbroek, the Netherlands); and at 8 to 9 wk of age against APP and influenza A virus (Porcilis APP and Gripovac3, respectively) by subcutaneous injection in the neck or, in the case of Enterisol, by oral drench. Piglets of both LSC and HSC treatments were housed in the same rooms until weaning (±24 d of age). After weaning, LSC and HSC pigs were group housed in different rooms to prevent cross-vaccination by the 2 living vaccines used in the HSC piglets (Enterisol Ileitis and Porcilis PRRS).

The HSC and LSC pigs were separately transported to the experimental farm (Vlierbos V.O.F., Neerloon, the Netherlands). As it was not possible to obtain all measurements during the study on 612 pigs on a single day, the LSC and HSC groups were split into 2 batches of 324 (180 from the LSC treatment and 144 from the HSC treatment) and 288 pigs (144 from the LSC treatment and 144 from the HSC treatment) arriving 1 wk apart. Therefore, pigs of batch 1 and 2 arrived at the experimental farm at an age of 10 and 11 wk, respectively.

Upon arrival, all pigs were individually weighed and, within sanitary condition treatment and batch, allocated to their pen based on BW to minimize variation between pens and within pens (17.3 ± 0.06 kg for LSC batch 1, 18.1 ± 0.07 kg for LSC batch 2, 15.9 ± 0.07 kg for HSC batch 1, and 17.4 ± 0.07 kg for HSC batch 2). The LSC pigs of each batch were housed in 5 LSC rooms and the HSC pigs were housed in 4 HSC rooms located in the same building. Each room had separate manure pits and separate ventilation regulation and contained 8 pens with 9 pigs per pen (0.8 m 2 space/pig), except for 1 LSC room, where 4 out of 8 pens were left empty. In addition, the HSC and LSC rooms were separated by a wall in the central corridor.

High sanitary condition rooms were intensively cleaned in 4 steps before arrival of the pigs: twice with foam (MS Topfoam LC Alk; MS Schippers, Bladel, the Netherlands) and high pressure washing and then treated twice with a disinfectant (MS Megades and MS Oxydes; MS Schippers). In addition, a strict hy-giene protocol was adhered to when entering the HSC rooms, which included showering, change of clothes, and use of a hairnet and face mask. People were not allowed to have access to a pig farm 48 h before entering HSC rooms. High sanitary condition animals received a preventative antibiotic injection (Fenflor; AUV Veterinary Services B.V., Cuijk, the Netherlands; 1 mL/pig, submuscular at Day 1 and 3 of the experiment) and were dewormed every 5 wk during the experiment starting at arrival (Flutelmium 0.6% premix; AUV Veterinary Services B.V.; topdressing, 1.5 mg Flubendazol/kg BW for 5 subsequent days).

Rooms for the LSC pigs were not cleaned after a previous batch of commercial finisher pigs left the facility 2 d before, and no hygiene protocol was applied. Starting at 5 wk after arrival, fresh manure of another commercial pig farm was spread in the LSC pens every 2 wk until end of the experiment to enhance antigenic pressure. Low sanitary condition pigs did not receive any medication or preventive treatment.

The experimental period lasted from December 11, 2013, until April 16, 2014. Animals were monitored for the complete fattening period, divided in 3 phases, that is, starter (0-34 d), grower (35-49 d), and finisher phases (from Day 50 until a target average pen weight of 110 kg BW). At the end of each phase, pigs were individually weighed.

Pigs were allocated to 2 diets, NP (17, 15, and 15% CP for the starter, grower, and finisher phases, respectively) or LP (20% CP reduced relative to NP for each phase), each of which contained a basal or a supplemented AA profile. This resulted in 4 dietary treatments: Low protein -basal amino acid diet, low protein-supplemented amino acid diet, normal protein -basal amino acid die, and normal protein -supplemented amino acid diet. Each diet was fed to the pigs in both sanitary regimes resulting in 8 treatment groups.

The apparent ileal digestible (AId) Lys to NE ratio of the diets was reduced in each subsequent phase of the experiment to follow a 3-phase feeding system. For the NP, the ratio was based on the Lys to NE requirements for boars according to the NRC (2012). The values for Lys to NE requirements were multiplied by 0.95 to make sure that the dietary energy concentration was not limiting the growth performance of the pigs. This resulted in diets for the starter, grower, and finisher phases containing 0.90, 0.81, and 0.75 g AID Lys/MJ of NE. For the LP, the inclusion level of all protein-containing ingredients was decreased by 20% relative to the NP and replaced by maize starch and Opticell (Agromed Austria GmbH, Kremsmünster, Austria), resulting in 0.72, 0.65, and 0.60 g AID Lys/ MJ of NE.

The basal AA profile (AA-B) was designed based on a factorial approach to cover the requirements for body protein deposition based on results from Bikker et al. (1994) , Le Bellego and Noblet (2002) , and the NRC (2012) and to cover losses associated with basal endogenous AA in ileal digesta based on results from Jansman et al. (2002) and the NRC (2012), related to losses of AA in skin and hair based on resluts from the NRC (2012), and AA losses related to cell turnover based on resluts from Moughan (1998) . All values were expressed in the same units for a pig of 50 kg BW with an assumed protein deposition of 138 g/d. The Met + Cys (45% of AID Lys) and Trp (15% of AID Lys) concentrations in AA-B diets, obtained in this manner, were adjusted to 51% Met + Cys and 18% Trp based on results from Knowles et al. (1998) and Jansman et al. (2010) , as we considered these to be far below the requirement values (CVB, 2011; NRC, 2012) . The supplemented AA profile (AA-S) was derived from the AA-B profile by increasing the Met, Thr, and Trp ratio relative to Lys by 20%. These AA were increased in particular as they are believed to be important as building blocks for proteins, for example, acute-phase proteins, synthesized in case of immune system activation Le Floc'h et al., 2008 , 2012 Rakhshandeh et al., 2010) , because of their function as precursors for important immune related components and antioxidants, and also because of their effects on several immune processes. Methionine is known to be an important methyl donor (Burke et al., 1951) and antioxidant (Wu, 2009) , Thr plays an important role in mucus synthesis for gut integrity and immune function (Wu, 2009) , and Trp is known as a precursor of melatonin and serotonin, both known to inhibit inflammatory cytokines (Wu, 2009) .

The ingredient and nutrient composition of the diets is shown in Tables 1, 2, and 3. All diets were isocaloric on a NE basis and contained TiO 2 as an indigestible marker. Diets were analyzed for AA composition by acid hydrolysis at 110°C for 23 h and ion-exchange chromatography with postcolumn derivatization with ninhydrin (ISO13903; ISO, 2005a) and Trp by alkaline hydrolysis at 110°C for 20 h ion-exchange chromatography with fluorescence detection (MOD.0094 version G; ISO 13904; ISO, 2005c) .

Per pen (9 pigs), 1 feeder was used and feed and water were offered ad libitum. The feed was provided as pellets via a computerized automatic system (Fancom Multiphase; Fancom B.V., Panningen, the Netherlands), which registered the mass of feed delivered per pen per day. At the end of each phase (starter, grower, and finisher), remainders of the diet per pen were collected and weighed to determine the feed intake per pen per phase. The computerized feeding system was calibrated before the trial started and after each phase.

At wk 13, 18, and 24 of age, pigs in 4 pens per room and 4 rooms per sanitary treatment were sampled for feces by rectal stimulation. In 3 subsequent 2 AA-B = basal dietary AA profile; AA-S = supplemented dietary AA profile containing 20% more Met, Thr, and Trp compared with the basal profile.

3 Supplied the following per kilogram of diet: 3.0 mg riboflavin, 20 mg niacin, 10 mg d-pantothenic acid, 150 mg choline chloride, 0.015 mg cyanocobalamin, 40 mg dl-α-tocopheryl acetate, 1.5 mg menadione, 6,000 IU retinyl acetate, 1,200 IU cholecalciferol, 0.2 folic acid, 1.0 thiamin, 1.0 pyridoxine HCl, 50 m manganese oxide, 267 mg iron SO4·H2O, 60 mg copper SO4·5H2O, 140 mg zinc SO4·H2O, 0.44 mg disodium selenium trioxide, and 1.0 potassium iodate.

4 Opticell (Agromed Austria GmbH, Kremsmünster, Austria).

days, 1 sample per pig was collected and samples were pooled per pen and stored at −20°C. Sampling pens were equally distributed over dietary treatments.

Frozen feces samples were dried at 103°C in an oven for 24 h to determine DM content (method 930.15; AOAC; ISO, 1999) and were analyzed for N by the Kjeldahl method (ISO 5983; ISO, 2005b) . Before Ti analysis (Short et al., 1996; Myers et al., 2004) , samples were freeze-dried and ground to pass a 1-mm screen using a Retsch ZM 100 mill (Retsch GmbH, Haan, Germany). Apparent total tract digestibility (Attd) for DM and N was calculated using TiO 2 as an indigestible marker (Kotb and Luckey, 1972) .

At the start of the experiment, 2 pigs with an average weight per pen were selected for blood sampling at 13, 18, and 24 wk of age from the vena cava. Selected pigs were sampled during each of the 3 phases. Per sampling moment, two 9-mL tubes per animal were filled: 1 EDTA tube for blood cell counts (Vacuette; Greiner Bio-One, Kremsmünster, Austria) and 1 serum tube for acute-phase protein and natural antibody (nab) analysis (Vacuette). Blood samples collected in EDTA tubes were immediately stored on ice and transported to the lab where blood cell counts were performed using a Microcell counter (Sysmex pocH-iV Diff; Toa Medical Electronics Co., Ltd., Kobe, Japan). Blood samples in serum tubes were allowed to clot for 1 h at room temperature, after which serum was collected after centrifugation for 10 min at 5,251 × g at room temperature and stored at −20°C pending analysis of haptoglobin (Tridelta Phase Haptoglobin Assay, catalog number TP-801; Tridelta Development, Ltd., Maynooth, Ireland), pig major acute-phase protein (Cusabio Pig-MAP, ELISA, catalog number CSB-E13425p; Cusabio Biotech Co, Ltd., Wuhan, Hubei Province, China), and Nab titers against keyhole limpet hemocyanin (KlH) types IgG and IgM using ELISA.

In each room with 8 pens of pigs, 2 pens with the normal protein -basal amino acid diet were selected for oral fluid sampling by using a swine oral fluid test kit (Tego oral fluids kit; ITL Corporation, Melbourne, Australia). At wk 14, 20, and 24 of age (1 time point per phase), a clean rope was hanged in the pen at pigs' shoulder height and securely tied. Pigs were allowed to chew on the rope for 30 min. Subsequently, the rope was removed from the pen (by wearing gloves) and placed in a clean pouch bag. The bag was closed and the fluid was extracted from the rope by squeezing the rope through the bag. Oral fluid was collected from the 

Antibody titers were determined as described by de Koning et al. (2015) with the minor modification that a 4-step dilution (40, 160, 640, 2,560 times diluted) of the sera was made instead of a 3-step dilution.

All pigs per pen were slaughtered in the week in which the average BW of the pigs in that pen was close to the target weight of 110 kg. In the slaughterhouse, lungs were collected, examined, and scored by a pathologist for pleuritis (0 to 2 scale, in which 0 = absence of pleuritis, 1 = adhesion of lung tissue with film-like tissue, and 2 = lung tissue completely grown together) and pneumonia lesions (0 to 3 scale, in which 0 = absence of pneumonia, 1 = one spot of pneumonia, 2 = a few spots of pneumonia, and 3 = diffuse deviation of pneumonia spots). Carcass weight, backfat thickness, and muscle thickness were measured in the slaughterhouse. Both fat and muscle thickness were measured at 6 cm from the back midline between the third and fourth last rib. Lean meat and dressing percentages were calculated per pig from the parameters obtained at slaughter using the following formulas: lean meat (%) = 66.86 − 0.6549 × (fat, mm) + 0.0207 × (muscle, mm) (Engel et al., 2012) and dressing (%) = (carcass weight/live weight) × 100% (Watkins et al., 1990) .

Data were analyzed as a 2 × 2 × 2 factorial arrangement using the GLM procedure for parameters measured at slaughter and the Mixed Model procedure for other performance parameters and serum and blood data (SAS 9.3; SAS Inst. Inc., Cary, NC) with pen as experimental unit for all parameters. For all data, the normality of the distribution of Studentized residuals was assessed by the Shapiro-Wilk statistic. If required, transformation of data was performed to obtain normal distribution of residuals. Values are presented as least squares means ± SEM, and effects were table 4. Performance parameters of growing pigs kept under different sanitary conditions and fed diets differing in protein content and AA supplementation, (bold P-values are significant and underlined P-values are considered as tendency)

No. of pens 6 9 9 9 9 8 8 8 8 and batch) was used as covariate in the model in the first statistical evaluations but finally omitted because of absence of statistical significance.

Two pigs selected for blood sampling died during the grower phase and. as such. the data of these pigs are missing for the finisher phase. Data of another pig selected for blood sampling was omitted from the data set as this pig was treated with antibiotics in the starter phase due to lung problems. No other clinical signs of illness were observed during the experiment. All results are presented in Tables 4 through 8. For clarity, selected treatment interactions are represented in Fig.  1A through 1I.

Mean BW at start of the experiment was greater for LSC pigs (17.7 ± 0.1 kg) compared with HSC pigs (16.6 ± 0.1 kg; P ≤ 0.01; Table 4 ). The ADG was (52 g/d) lower for LSC pigs compared with HSC pigs for the starter phase (P ≤ 0.05) and (55 g/d) during the complete fattening period (P ≤ 0.01) but not during the grower phase (P > 0.10). The LP pigs had (56 g/d) lower ADG in the complete fattening period compared with NP pigs (all, P ≤ 0.05). The AA-B pigs tended to have (28 g/d) lower ADG compared with AA-S pigs in the starter phase when kept under HSC but not under LSC [sanitary conditions (Sc) × AA, P ≤ 0.10].

For ADFI, an interaction was present for SC × AA in the starter phase (P ≤ 0.01) and over the complete fattening period (P ≤ 0.05; Fig. 1A ) but not in the grower phase (P > 0.10). The AA-B pigs had (54 g/d in the starter phase and 69 g/d over the complete fattening period) lower ADFI compared with AA-S pigs when kept under HSC but not under LSC.

The G:F was (0.035 g/g) lower for pigs kept under LSC compared with HSC for the starter phase (P ≤ 0.01) and (0.013 g/g) for the entire grower-finisher period (P ≤ 0.01) but not for the grower phase (P > 0.10). The G:F was 0.038 g/g lower for LP pigs compared with NP pigs in all phases (all, P ≤ 0.01). The AA-B pigs had a (0.008 g/g) lower G:F compared with AA-S pigs in the entire grower-finisher period (P ≤ 0.01). The greater G:F for AA-B pigs compared with AA-S pigs for the entire experimental period was (0.025 g/g) greater for LSC pigs compared with HSC pigs (SC × AA, P ≤ 0.05; Fig. 1B) . A tendency for a similar interaction was found for the G:F in the grower phase (P ≤ 0.10).

In LSC pigs but not HSC pigs, reduction of dietary CP concentration reduced serum haptoglobin by 0.24 g/L (CP × SC, P ≤ 0.01; Table 5 ; Fig. 1I ). The LSC pigs showed lower haptoglobin concentrations over time whereas HSC pigs had lower concentrations during the grower phase compared with the starter phase and showed greater concentrations again in finisher phase (SC × phase, P ≤ 0.05; Fig. 1E ). The LSC pigs had (0.29 g/L) greater serum haptoglobin concentrations compared with the HSC pigs (P ≤ 0.01).

The LSC pigs had (0.03 g/L) lower PigMAP concentrations in the grower phase compared with the starter phase and (0.09 g/L) greater concentrations again in the finisher phase compared with the grower phase, whereas the HSC pigs had (0.09 g/L) lower concentrations in the grower phase compared with the starter phase (SC × phase, P ≤ 0.05; Fig. 1F ).

Keyhole limpet hemocyanin-specific IgM antibody titers in serum tended to be (0.04) lower for AA-S-fed pigs compared with AA-B-fed pigs (P ≤ 0.10). Keyhole limpet hemocyanin-specific IgG antibody titers were (0.45) greater for LSC pigs compared with HSC pigs (P ≤ 0.05).

The number of white blood cells (WBc) was (1.8 × 10 9 /L) greater (Table 6) in AA-S-fed pigs compared with AA-B-fed pigs when kept under LSC but not under HSC (SC × AA, P ≤ 0.05; Fig. 1C ). Over time, the concentration of WBC in pigs decreased (by 4.1 × 10 9 /L) but the number of red blood cells consistently increased (with 0.5 × 10 12 /L) in all treatment groups (P ≤ 0.01). Pigs fed the NP had a (3.1%) lower number of WBC compared with pigs fed the LP (P ≤ 0.05). Hemoglobin concentration was (0.1 mmol/L) lower in AA-S-fed pigs compared with AA-B-fed pigs, particularly under HSC (SC × AA, P ≤ 0.05, and P ≤ 0.05 for AA). Hemoglobin concentrations increased with age in all treatment groups (0.74 mmol/L); however, this increase was greater in HSC pigs compared with LSC pigs (SC × phase, P ≤ 0.05). Pigs fed NP had a (0.1 mmol/L) greater hemoglobin concentration than pigs fed LP (P ≤ 0.05). Mean cell volume was (0.7, 10 -15 L) greater for AA-S-fed pigs compared with AA-B-fed pigs in LSC but this was reversed in HSC (SC × AA, P ≤ 0.01; Fig. 1D ). The mean cell volume was (0.4, 10 -15 L) greater in pigs fed NP compared with pigs fed LP (P ≤ 0.05). The number of platelets decreased (by 258 × 10 9 /L) in pigs over time for all treatments, but the decrease was (60%) greater in HSC pigs compared with LSC pigs (interaction, P ≤ 0.05).

The concentration of platelets in pigs fed AA-S diets was (121 × 10 9 /L) greater compared with pigs fed AA-B diets (P ≤ 0.01).

The AA-S-fed pigs had (0.7 × 10 9 /L) a greater number of blood lymphocytes compared with AA-Bfed pigs when provided a LP, but this was reversed (−0.5 × 10 9 /L) when pigs were provided a NP (CP × AA, P ≤ 0.05). The number of lymphocytes increased (0.9 × 10 9 /L) with age for all treatment groups (P ≤ 0.01). The number of monocytes (2.9 × 10 9 /L) decreased over time for all treatment groups, particularly for LSC pigs (SC × phase, P ≤ 0.01; Fig. 1H ). The AA-S-fed pigs had a (0.6 × 10 9 /L) greater monocyte number compared with AA-B-fed pigs (P ≤ 0.05). The NP-fed pigs had a (0.8 × 10 9 /L) lower monocytes number compared with LP-fed pigs (P ≤ 0.05). The concentration of granulocytes increased (by 0.01 × 10 9 /L) over time for LSC pigs and, in the HSC pigs, increased (0.02 × 10 9 /L) from the starter to the grower phases but decreased (0.03 × 10 9 /L) again in the finisher phase (SC × phase, P ≤ 0.05; Fig. 1G ). The concentration of granulocytes was (0.01 × 10 9 /L) greater for HSC pigs than for LSC pigs (P ≤ 0.01).

Results obtained at slaughter are presented in Table 7 . Carcass weight was (3.8 kg) lower for LP pigs compared with NP pigs but not dressing percentage (P ≤ 0.01 and P > 0.05, respectively). The LP pigs had (1.2 mm) greater backfat thickness (P ≤ 0.01) and (0.8%) lower lean meat percentage (P ≤ 0.01) compared with NP pigs. The AA-B pigs had (1.5 mm) lower muscle thickness (P ≤ 0.01; 0.5 mm), greater backfat thickness (P ≤ 0.01), and a (0.4%) lower lean meat percentage (P ≤ 0.01) compared with AA-S pigs.

Pleuritis scores were (0.3) greater for LSC pigs compared with HSC pigs (P ≤ 0.01). Percentage of lung surface with pleuritis was also (1.2%) greater for LSC pigs compared with HSC pigs (P ≤ 0.01). The AA-B pigs had a (1%) greater percentage of lung surface with pleuritis compared with AA-S pigs when kept in LSC, but for HSC pigs, this was reversed (SC × AA, P ≤ 0.05). The LP pigs had (0.2) greater pneumonia scores compared with NP pigs when kept under HSC; however, when kept under LSC, this was reversed (SC × CP, P ≤ 0.05). The AA-B pigs tended to have greater pneumonia scores compared with the AA-S pigs in all cases except in LSC pigs fed a LP (CP × AA, P ≤ 0.01). All animals were group housed in pens with 9 animals in total. Two animals per pen we selected for blood sampling.

The effect of experimental treatments on the ATTD of DM and N were consistent over the duration of the experiment and are presented averaged over all phases (Table 8 ). In general, treatment differences, albeit significant, were small. Apparent total tract digestibility for DM was (0.3%) lower for LSC pigs compared with HSC pigs during all phases (all, P ≤ 0.05). The AA-B pigs had (0.4%) lower ATTD of DM compared with AA-S pigs (P ≤ 0.01). Apparent total tract digestibility for N was (0.98%) greater for HSC pigs compared with LSC pigs during all phases (all, P ≤ 0.01), regardless the dietary CP content or AA profile. Apparent total tract digestibility for N increased Table 7 . Slaughter results of fattening boars kept under different sanitary conditions and fed diets with 2 CP levels and 2 AA profiles (bold P-values are significant and underlined P-values are considered as tendency)

No. of pens 6 9 9 9 9 8 8 8 8 5 SC = sanitary conditions. Considered significant when P ≤ 0.05 and considered a tendency when 0.05 < P ≤ 0.10. 6 A pen contained 9 pigs. 7 Body weight expressed 1 d before slaughter day; muscle, backfat, and lean meat is expressed as corrected for carcass weight, by including carcass weight as a covariate in the statistical model. 8 Pleurits was scored on a scale of 0 to 2 and pneumonia was scored on a scale of 0 to 3. 9 Percentage of lung surface affected by pleuritis or pneumonia. from 80.6 (starter phase) to 83.9 (grower phase) and 84.7% finisher phase (P ≤ 0.01). There were no interactions between phase and other independent variables.

Oral fluid samples of pigs in all sampled pens (18 pens in total) were positive for M. hyorhinis and negative for PRRS, M. hyopneumoniae, influenza A virus, and porcine cytomegalo virus in all phases. All pens were negative for porcine respiratory corona virus in the starter and grower phases, but in the finisher phase, 4 out of 9 LSC pens and 3 out of 9 HSC pens became positive. All pens were negative for PCV2 in the starter phase, but LSC pigs became positive in grower phase (7 out of 9 pens) and finisher phase (all, 9 pens). Only 1 out of 9 HSC pens became positive for PCV2 in the grower phase and was negative again in the finisher phase. The rest of the sampled HSC pigs were negative for PCV2 in all phases.

The main objective of the present experiment was to evaluate, under practical conditions, if diets with low or normal CP level and basal or supplemented Met, Thr, and Trp have differential effects on pig performance and immune status at different sanitary conditions.

A contrast in sanitary conditions was generated by imposing a combination of differences in hygiene, antibiotic treatment, deworming, and a vaccination protocol, all applied to piglets originating from the same farm. Over the duration of the experiment, HSC pigs showed greater ADG (50 g/d) and G:F (0.013 kg/kg) than LSC pigs. Low sanitary condition pigs had greater serum haptoglobin levels and greater KLH-IgG titers than HSC pigs. This indicates that LSC pigs had a more active immune system compared with HSC pigs. In addition, at slaughter, higher pleuritis scores were observed in LSC pigs compared with HSC pigs. Taken together, the absence of clinical signs of illness of pigs in either treatment group and the lower ADG and G:F, elevated haptoglobin concentrations, increased KLH-specific IgG titers, greater pleuritis occurrence, and oral fluid samples positive for PCV2 in the grower and finisher phases in the LSC groups illustrate a difference in subclinical health status between the LSC and HSC pigs (Le Floc'h et al., 2006; van den Berg et al., 2007; Piñeiro et al., 2009) .

We expected that the sanitary regimes used in our study would affect PigMAP and haptoglobin concentrations in a similar way, as these are both considered to be positive acute-phase proteins and are expected to increase in response to immune stimulation (Murata et al., 2004) . Haptoglobin appeared more responsive to the difference in sanitary conditions than PigMAP, which is in line with Heegaard et al. (1998) and Kampman-van de Hoek et al. (2016) . In addition, PigMAP levels found here (0.14 g/L serum) were low compared with the values found by Piñeiro et al. (2009) and Kampman-van de Hoek et al. (2016; 0.9 and 1.4 g/L serum, respectively), whereas the haptoglobin concentrations reported by these authors were similar to values reported in the present study.

The KLH antigen is a relevant antigen to measure Nab levels as pigs had no previous exposure to KLH. Natural antibodies are defined as antigen-specific antibodies that are present in the absence of intentional immunization with specific antigens (KLH, in this case; Star et al., 2007) . The IgG titer against KLH was greater for LSC pigs than for HSC pigs. Because Nab play a role in the first line of defense against pathogens, the increase in Nab levels might be an adaptive response of the pigs to the higher infection pressure at LSC. Keyhole limpet hemocyanin-specific antibodies were shown to be cross-reactive with antigens of pathogens (Hamilton et al., 1999) . Therefore, it can not be excluded that antigenspecific Nab are products of adaptive immune responses and cross-reactive with structurally related antigens. Specific memory B cells were shown to be activated in a polyclonal but antigen-independent way (Lanzavecchia et al., 2006) . This way of activation is triggered by nonspecific microbial-associated molecular patterns, such as lipopolysaccharide or bacterial or viral nucleotide motifs. This B cell activation mechanism is likely responsible for the lifelong presence of circulating specific antibodies, which forms an important part of the first line of defense. It can easily be envisaged that memory B cells of pigs under LSC become more activated by microbial associated molecular patterns than under HSC, resulting in higher levels of KLH-specific IgG levels. For IgM titers against KLH, there were no differences between LSC and HSC pigs. In the study of Ploegaert et al. (2010) , a similar result in IgM and IgG Nab was found. Ploegaert et al. (2010) studied genetic and phenotypic correlation of Nab titers in dairy cattle and found greater estimates for environmental variation in the IgG than in the IgM isotype of Nab. Most probably, the LSC in the present study, as an environmental factor, stimulated KLH-specific IgG responses. Immunoglobulin M titers are believed to be influenced by genetic factors (Ploegaert et al., 2010) , which explains the absence of differences in IgM against KLH between sanitary regimes.

As the HSC pigs were vaccinated against several pathogens in their first 10 wk of life, vaccination and/ or antibiotic treatment might have modulated immune functions of these animals. Our results show that despite the vaccination of the HSC pigs, LSC pigs had significantly higher values for haptoglobin and IgG antibodies against KLH, showed reduced ADG and G:F, and had significant higher pleuritis scores, indicating that the effects of the LSC conditions outweighed the potential effects of the vaccinations or antibiotic treatment of the HSC pigs. However, the vaccinations of the HSC pigs may have reduced the immunological contrasts between the pigs kept under the different sanitary conditions. If this is the case, the observed interactions between diet and sanitary conditions could even be larger when vaccinations are omitted. In addition, vaccinations or antibiotic treatment may have affected other (undetermined) immune parameters as well.

At start of the experiment, the BW of LSC pigs was 1.1 kg greater compared with that of HSC pigs, likely related to the vaccination program of the HSC pigs prior to arrival at the experimental farm. During the starter phase, HSC pigs compensated for this lower starting weight, or the ADG of the LSC pigs was negatively influenced by the LSC.

A lower ADG of LSC pigs was also found in other studies evaluating a contrast in sanitary conditions (Williams et al., 1997; Le Floc'h et al., 2009; Pastorelli et al., 2012a) . This decreased ADG can be explained by the competition for use of nutrients between the immune system and for use in deposition in organs and body tissues. In general, it is perceived that use by the immune system has a high priority (Humphrey and Klasing, 2004; Le Floc'h et al., 2004) . From a metaanalysis, Pastorelli et al. (2012b) concluded that the major portion of the reduction in ADG (12.2% lower ADG of the total 16.3% lower ADG in pigs in poor housing conditions compared with unchallenged animals) was related to a decreased feed efficiency (in Pastorelli et al. [2012b] , indicated as "maintenance") rather than to a decrease in feed intake. This was also observed in the present experiment, where G:F was affected but ADFI was not influenced by sanitary conditions. The former indicates a greater maintenance requirement or a lower growth efficiency for LSC pigs as suggested by Pastorelli et al. (2012b) .

Signs of respiratory problems, such as pleuritis or pneumonia, are related to reduced performance (Saco et al., 2011) . The LSC pigs had greater scores for pleuritis and greater percentage of lung surface with pleuritis at slaughter compared with HSC pigs, which might have negatively affected growth and G:F of LSC pigs. From the meta-analysis of Pastorelli et al. (2012b) , it appeared that a reduction in ADFI is the major contributor to reduced ADG in case of respiratory problems. In contrast, in our study, ADFI is not the major contributor to reduced ADG, indicating that respiratory problems are likely not the cause of the decrease in G:F in LSC pigs or that conclusions about the respiratory problems drawn by Pastorelli et al. (2012b) are not representative for our study. Pastorelli et al. (2012b) used many different challenge studies for the meta-analysis, differing from studies conducted under more practical conditions such as the present study.

The greater G:F of the HSC pigs might partly be due to a greater apparent total N digestion in these animals compared with LSC pigs. When assuming similar postabsorptive efficiencies for absorbed AA, the observed increase in N digestibility would typically explain approximately 20% of the observed increase in ADG, hence leaving 80% unaccounted for. Differences in ATTD for N correspond with results by Kampman-van de Hoek et al. (2016), who found a reduction in ATTD for N of 3.7% in LSC growing pigs compared with HSC pigs (P ≤ 0.01). The reduced ATTD for N might be due to intestinal infections, intestinal damage, or an increased digesta passage rate (Sandberg et al., 2006; Pastorelli et al., 2012b ). The small difference found in ATTD of DM might suggest that the ATTD of GE also differed to a similar extent.

Greater ADG and G:F were expected for NP-fed pigs compared with the LP-fed pigs. It is well known that additional AA under conditions where AA are limiting but sufficient energy is available lead to greater ADG and improved G:F (Noblet et al., 1987) . Supplementation of extra AA by increase in dietary CP (LP vs. NP) or by specifically supplementing Met, Thr, and Trp (AA-B vs. AA-S diet type) resulted in increased ADG, suggesting that one of these AA was limiting.

The increased G:F in AA-S pigs vs. AA-B pigs, particularly in LSC pigs (SC × AA interaction; Fig.  1B ), confirms our hypothesis that the dietary AA-S profile better matches the AA requirements of pigs under LSC. It should be noted, however, that this effect was not yet present in the starter phase. Our results are in accordance with other studies that show that immune stimulation by different challenges lead to increased requirements for specific AA compared with unchallenged pigs (Grimble and Grimble, 1998; Le Floc'h et al., 2004; Klasing, 2007; Rakhshandeh et al., 2014) . As a consequence, these animals require more AA for their immune system. The increased demand for these AA (Met, Thr, and Trp) for LSC pigs was expected to be greater for pigs fed the LP. This interaction, however, was absent. Therefore, the results of our study do not confirm our hypothesis that at low levels of protein intake, the requirements for Thr, Trp, and Met relative to Lys are increased. As illustrated by the meta-analysis by Pastorelli et al. (2012b) , the type of challenge has a major impact on the response of pigs. The contrast in sanitary conditions, as applied in our study, illustrates that also in absence of clinical disease, requirements for Met, Thr, and Trp are persistently affected over the entire weight range. This is in agreement with observations by Kim et al. (2012) for Met, using a repeated LPS model, but not in a study by de Ridder et al. (2012) for Trp, also using a repeated LPS model, in which the response to incremental intake of Trp was demonstrated to be transient.

Dietary protein concentration did not affect ADFI in our study. Apparently, the exchange of protein for starch did not affect the satiating potential of the diets. This is in agreement with Le Bellego and Noblet (2002) and Kerr et al. (2003) ; however, research in humans has indicated satiating effects of dietary proteins to exceed that of carbohydrates and fats (Andersen and Moore, 2004) .

A remarkable result is the interaction between SC and AA profile in ADFI, particularly observed during the starter phase. High sanitary condition AA-S pigs had greater ADFI compared with HSC AA-B pigs, whereas pigs in LSC ate the same amount, regardless of AA supplementation. The greater ADFI for HSC AA-S pigs resulted in a tendency for ADG in the same direction as the change in ADFI. This result illustrates that pigs in HSC were probably more limited in their growth by the AA-B profile than LSC pigs receiving the same profile, which was expected to be the other way around due to the effect of immune stimulation of pigs housed in LSC. We speculate that 1 of the 3 supplemented AA, probably Met, has restricted growth, particularly inhibiting HSC pigs to exploit their full ADG potential following a period of restricted growth before the start of the trial. Methionine is thought to be the limiting AA because the Met:Lys ratio deviates most from recommended values (e.g., CVB, 2011) compared with Thr and Trp. The growth restriction in HSC pigs before the start of the trial was possibly caused by the vaccination strategy and illustrated by a lower BW of HSC pigs at the onset of the trial. The AA profile of the basal diets may not have been sufficient to support compensatory growth following a period of growth restriction.

Increasing the dietary protein concentration increased serum haptoglobin concentrations in LSC pigs but not in HSC pigs. The effect of protein scarcity on total serum protein concentrations has been demonstrated in mice (Cooper et al., 1974) . Houdijk et al. (2007) found a reduced C-reactive protein and haptoglobin response for infected pigs fed a low protein diet. This may reflect a sensitive response of acute-phase proteins to protein scarcity (Houdijk et al., 2007) or a high priority for the use of AA for protein gain in young pigs, occurring at the expense of the synthesis of haptoglobin. It is unexpected that this interaction was not observed for WBC counts, which were consistently higher for LP pigs compared with NP pigs (P = 0.02), being unaffected by SC. Possible differences in migration of WBC into the lymphatic system or tissues complicates clear conclusions on this point (Ganusov and Auerbach., 2014; Marelli-Berg et al., 2010) .

Supplementation of AA in diets has been shown to influence the immune system by increased variation of lymphocytes, increased production of antibodies and cytokines, and activation of lymphocytes, natural killer cells, and macrophages (Daly et al., 1990; Li et al., 2007; Negro et al., 2008) . Although increasing the dietary protein concentration reduced monocyte counts, supplementation of Met, Thr, and Trp increased monocyte counts, regardless of SC. Elevated levels of Trp in the AA-S-fed pigs might have stimulated production of monocytes, as Trp is known to play a role in functionality of monocytes and lymphocytes . The administration of 300 mg Trp to rats increased monocyte phagocytosis and the innate immune response (Esteban et al., 2004) . Overall, there was no clear interaction for sanitary conditions and AA profile present on the measured blood parameters.

In summary, poor sanitary conditions imposed in our study, in a practical setting, reduced ADG over the entire BW range of 17 to 110 kg. Low sanitary conditions increased pleuritis scores at slaughter and increased indicators for the innate immune response and serum haptoglobin concentrations. The vaccination strategy for the HSC pigs in early life may have triggered a compensatory performance response (particularly ADFI) during the starter phase. Dietary supplementation of Met, Thr, and Trp improved the G:F, particularly under LSC, illustrating that dietary requirements for these AA are affected by sanitary conditions. Furthermore, this study provides indications that dietary protein concentration and Met, Thr, and Trp supplementation modify immune status.

Our study suggests that dietary supplementation of Met, Thr, and/or Trp, in addition to provision of AA for covering basal requirements for protein deposition, is beneficial for animal performance, particularly under poor sanitary conditions.

Dietary proteins in the regulation of food intake and body weight in humans

Amino acid composition of growing pigs is affected by protein and energy intake

The role of methionine as methyl donor for choline synthesis in the chick

Chemical compositions and nutritional values of feed ingredients

Effects of protein insufficiency on immune responsiveness

Effect of dietary protein and amino acids on immune function

Association of natural (auto-) antibodies in young gilts with osteochondrosis at slaughter

Immune system stimulation reduces the efficiency of tryptophan utilization for body protein deposition in growing pigs

Lean meat prediction with HGP, CGM, CSB-Image-Meater, with prediction accuracy evaluated for different proportions of gilts, boars and castrated boars in the pig population

Effect of orally administered ltryptophan on serotonin, melatonin, and the innate immune response in the rat

Mathematical modelling reveals kinetics of lymphocyte recirculation in the whole organism

The use of free amino acids allows formulating very low crude protein diets for piglets

Immunonutrition: Role of sulfur amino acids, related amino acids, and polyamines

Periodate-sensitive immunological cross-reactivity between keyhole limpet haemocyanin (KLH) and serodiagnostic Schistosoma mansoni egg antigens

The porcine acute phase response to infection with Actinobacillus pleuropneumoniae. Haptoglobin, C-reactive protein, major acute phase protein and serum amyloid A protein are sensitive indicators of infection

Effects of sub-clinical post-weaning colibacillosis and dietary protein on acute phase proteins in weaner pigs

Modulation of nutrient metabolism and homeostasis by the immune system

International Organization for Standardization (ISO). 2005b. ISO 5983-1:2005. Animal feeding stuffs -Determination of nitrogen content and calculation of crude protein content -Part 1: Kjeldahl method

Evaluation through literature data of the amount and amino acid composition of basal endogenous crude protein at the terminal ileum of pigs

The effect of diet composition on tryptophan requirement of young piglets

Dietary amino acid deficiency reduces the utilization of amino acids for growth in growing pigs after a period of poor health

Induced lung inflammation and dietary protein supply affect nitrogen retention and amino acid metabolism in growing pigs

Effect of feeding reduced protein, amino acid-supplemented diets on nitrogen and energy balance in grower pigs

Influence of dietary protein level, amino acid supplementation, and dietary energy levels on growing-finishing pig performance and carcass composition

Whole body protein depositions and plasma amino acid profiles in growing and/or finishing pigs fed increasing levels of sulfur amino acids with and without Escherichia coli lipopolysaccharide challenge

Nutrition and the immune system

Monokines in growth and development

Ratio of total sulfur amino acids to lysine for finishing pigs

Markers in nutrition

Understanding and making use of human memory cells

Performance and utilization of dietary energy and amino acids in piglets fed low protein diets

Towards amino acid recommendations for specific physiological and patho-physiological states in pigs

Importance of sanitary environment for growth performance and plasma nutrient homeostasis during the post-weaning period in piglets

The effect of sanitary status degradation and dietary tryptophan content on growth rate and tryptophan metabolism in weaning pigs

Modifications of protein and amino acid metabolism during inflammation and immune system activation

Dietary tryptophan helps to preserve tryptophan homeostasis in pigs suffering from lung inflammation

Amino acids and immune function

Memory T-cell trafficking: New directions for busy commuters

Chronic lung inflammation affects plasma amino acid concentrations in pigs

Protein metabolism in the growing pig

Current research on acute phase proteins in veterinary diagnosis: An overview

Technical note: A procedure for the preparation and quantitative analysis of samples for titanium dioxide

Branchedchain amino acid supplementation does not enhance athletic performance but affects muscle recovery and the immune system

Effect of protein and lysine levels in the diet on body gain composition and energy utilization in growing pigs

Nutritional requirement of swine. 11th rev

Sanitary housing conditions modify the performance and behavioural response of weaned pigs to feed-and housing-related stressors

Meta-analysis of feed intake and growth responses of growing pigs after a sanitary challenge

Pig-MAP and haptoglobin concentration reference values in swine from commercial farms

Genetic variation of natural antibodies in milk of Dutch Holstein-Friesian cows

Immune system stimulation of growing pigs does not alter apparent ileal amino acid digestibility but reduces the ratio between whole body nitrogen and sulfur retention

Impact of immune system stimulation on the ileal nutrient digestibility and utilisation of methionine plus cysteine intake for whole-body protein deposition in growing pigs

Serum acute phase proteins as biomarkers of pleuritis and cranio-ventral pulmonary consolidation in slaughter-aged pigs

A model for predicting feed intake of growing animals during exposure to pathogens

Determination of titanium dioxide added as an inert marker in chicken digestibility studies

Effect of single or combined climatic and hygienic stress on natural and specific humoral immune competence in four layer lines

Evaluation of the acute phase protein haptoglobin as an indicator of herd health in slaughter pigs

The effect of various levels of ractopamine hydrochloride on the performance and carcass characteristics of finishing swine

Effect of level of chronic immune system activation on the growth and dietary lysine needs of pigs fed from 6 to 112 kg

Amino acids: Metabolism, functions, and nutrition