Dietary Ulva lactuca and CAZyme supplementation improve serum biochemical profile and hepatic composition of weaned piglets

Ulva lactuca is a seaweed with antinutritional cell wall for monogastrics. Carbohydrate-Active enZymes (CAZymes) supplementation can potentially cause its disruption. This study evaluates four diets: Ctrl—control diet; UL—control + 7% U. lactuca (wild caught, powdered form); ULR—UL + 0.005% Rovabio® Excel AP; ULU—UL + 0.01% ulvan lyase on piglets’ haematologic and serologic profiles, hepatic lipids and minerals. White blood cells and lymphocytes reached the highest values in piglets fed UL compared to control, and to control and ULR; respectively (P < 0.05). IgG levels were boosted by seaweed incorporation compared to control (P = 0.015). The glycaemic homeostasis was assured by the seaweed inclusion. Dietary seaweed decreased serum lipids (P < 0.001), with the exception of ULU, due to HDL-cholesterol increase (P < 0.001). Cortisol was decreased in ULR and ULU (P < 0.001). No systemic inflammation was observed (P > 0.05). While hepatic n-3 PUFA increased in piglets fed with seaweed diets due to increment of beneficial 22:5n-3 and 22:6n-3 fatty acids (P < 0.05), the opposite occurred for n-6 PUFA, PUFA/SFA and n-6/n-3 ratios (P < 0.05). Hepatic pigments were unchanged (P > 0.05). ULR reduced α-tocopherol levels (P = 0.036) and increased serum potassium levels (P < 0.001) compared to control. Seaweed contributed to overcome piglets’ weaning stress, with some benefits of including CAZyme supplementation.

The environmental impact of livestock production has become an important and controversial global issue 1 , whose effects are expected to worsen as both world population and animal product demand increase. High quality and locally produced feedstuffs can be alternatives to conventional ones, such as maize or soybean meal that are imported for instance into Europe from the Americas with high economic and environmental costs. Seaweeds are among such alternatives. They are interesting, albeit heterogeneous, organisms providing an abundant source of biomass 2 . However, they have a recalcitrant cell wall with a complex cross-linked matrix of insoluble polysaccharides that reduce the digestive availability of intracellular nutrients during monogastrics' digestion, thus limiting dietary incorporation levels for highly productive animals 3 . The supplementation with Carbohydrate-Active enZymes (CAZymes) stands out as a possible approach to increase the digestive utilization of such feedstuffs at higher inclusion levels in the feed 4 . Seaweeds are capable of improving quality traits and nutritional value of meat as well as animal health 2 , as previously described for Laminaria digitata 5 and similar to some microalgae 6,7 .
The green seaweed Ulva lactuca is considered a good source of nutrients 2 . Like other edible green seaweeds, U. lactuca's protein concentration, although very variable and depending on culture and environmental conditions, can represent up to 32% of the algae's dry mass 8 . This makes it a putative, partial alternative to conventional protein sources, such as soybean meal. Moreover, it contains a high amount of carbohydrates, such as ulvan with Haematologic and serologic profiles. The effect of dietary U. lactuca, with or without CAZyme supplementation, on blood cells and serum biochemical metabolites from piglets is shown in Table 1. White blood cells count (P = 0.004) was higher in piglets fed U. lactuca diet mostly due to lymphocyte (P = 0.010) increase, reaching 53%. The opposite occurred for the granulocytes percentage (P = 0.009), which reached a minimum of 41.7%. Monocytes, red blood cells and thrombocytes were not affected (P > 0.05) by the diets.
No change was detected for glucose, or insulin resistance index (HOMA-IR) (P > 0.05). Urea (P < 0.001) and creatinine (P = 0.049) were significantly increased in ULU (11.2 and 0.739 mg/dL, respectively) piglets compared respectively to all treatments and with UL. The same pattern of variation was found for cholesterol (P < 0.001), HDL-cholesterol (P < 0.001), VLDL-cholesterol (P < 0.001), total lipids (P < 0.001) and triacylglycerols (P < 0.001), in which piglets fed the combination of U. lactuca with the recombinant lyase had the highest values, whereas piglets fed the combination of U. lactuca with commercial Rovabio ® showed the lowest values. Still regarding lipemia, LDL-cholesterol (P < 0.001) was reduced by at least 26.9% with the combination of U. lactuca and commercial Rovabio ® by comparison to the other groups. Total protein was decreased by both feed enzymes relative to the other dietary groups (P < 0.001). Concerning hepatic markers, the highest values of alkaline phosphatase (ALP, P < 0.001) were seen in the control group (335 U/L) by comparison to the remaining diets. In turn, aspartate aminotransferase (AST, P = 0.004) had the highest levels in piglets fed the combination of U. lactuca with commercial Rovabio ® (43.8 U/L), while gamma-glutamyltransferase (GGT, P < 0.001) had the lowest levels in this same group (17.8 U/L). Regarding immunoglobulins, IgA was found below the minimum detection levels (< 4.0 mg/dL), IgM was lowest in U. lactuca without enzymes (25.0 mg/dL, P = 0.015), and IgG (P = 0.015) was decreased in the control group (91.9 mg/dL) by comparison to the U. lactuca dietary group (132 mg/dL). Cortisol (P < 0.001) was diminished by feed enzymes in relation to the other dietary treatments. U. Lactuca diets reduced IGF-1 (P < 0.001) concentrations by comparison to the control (191 µg/L). The interleukin-10 was higher in piglets fed U. lactuca combined with the recombinant ulvan lyase (29.4 pg/mL), intermediate in U. lactuca alone or combined with commercial Rovabio ® and lower in the control (P < 0.001). C-reactive protein (< 0.03 mg/ dL), IL-6 (< 1.5 pg/mL) and ApoA1 (< 3.0 mg/dL) were found below the minimum detection levels. In terms of electrolytes, potassium (P < 0.001) and chloride (P = 0.017) were higher in piglets fed U. lactuca combined with commercial Rovabio ® . Concerning redox status, piglets fed on the control diet tended to have higher concentrations of total antioxidant capacity (P = 0.053) when compared to piglets fed U. lactuca. In turn, glutathione peroxidase activity in piglets' serum did not change between dietary treatments (P = 0.378).
Hepatic α-tocopherol and pigments. Contrarily to α-tocopherol, no differences were found for hepatic pigments (Table 3). α-Tocopherol was increased in the control group by 1.4-fold relative to piglets fed on U. lactuca combined with the commercial Rovabio ® (P = 0.036). www.nature.com/scientificreports/ Hepatic mineral profile. Micro-and macromineral profiles are presented in Table 4. No major variations were observed for mineral contents in the liver, except for calcium and sulphur. Calcium was increased in piglets fed the combination of U. lactuca with ulvan lyase (20.8 mg/100 g liver) relative to piglets fed U. lactuca alone (18.0 mg/100 g liver) P = 0.023). In turn, sulphur content was lower in piglets fed the control diet (162 mg/100 g liver) when compared to piglets fed U. lactuca, alone (184 mg/100 g liver) or combined with the commercial Rovabio ® (176 mg/100 g liver, P = 0.001).

Principal component analysis (PCA).
Two principal component analyses, carried out with serum and liver metabolites, are depicted in Fig. 1 (panels A and B, respectively). The plot loadings are presented in supplementary material 2. Regarding serum, there was no clear clustering of any of the experimental groups. Nevertheless, there was a high heterogeneity of UL when compared to the other groups, particularly due to variables such as IgG and creatinine. The PCA carried out for liver metabolites revealed an overlap of all groups, but the controls tended to cluster separately, despite the high heterogeneity. Indeed, fatty acid variables such as 17:0, 15:0 and total n-6 PUFA explain such separation, which is in line with our previously mentioned statistical analysis, in which they were the highest in the control group.

Discussion
To our knowledge, the use of U. lactuca seaweed as a feed ingredient, with 7% of dietary incorporation in weaned piglet's diets has not been previously reported. Indeed, some studies have used Ulva sp. extracts to improve the immune status of weaner 21 and weaned 22 piglets but did not use the whole biomass. We initially hypothesized that feeding the seaweed at ingredient level (> 3%) without enzymatic supplementation might lead to antinutritional effects, causing lower availability (and consequent accumulation) of micronutrients, such as fatty acids. The objective of this study was to use it as a partial protein source for weaned piglets, taking advantage of its several bioactive properties in the process. The enzymatic supplementation aimed at maximizing its potential as a nutritional source. We therefore measured the piglet's response in both blood and liver metabolites, thus focusing on systemic and central metabolism. www.nature.com/scientificreports/ A moderate effect of experimental diets on blood cells was observed. Even if red blood cells and thrombocytes did not vary across dietary groups, lymphocytes increased as a consequence of dietary seaweed. On the contrary, the granulocytes count was reduced, and monocytes were kept unchanged in the experimental groups when compared to controls. A blood cell modulation of dietary seaweed has been previously reported by Shimazu et al. 23 when applying 1% of dietary wakame (Undaria pinnatifida) which increased natural killer cells in pig's blood. Our results seem to indicate that the seaweed promotes a positive modulation of the adaptative immune response, given that it increases lymphocytes by comparison to controls. However, the differences between the www.nature.com/scientificreports/ enzyme-supplemented groups were not significant, which suggested that the positive effect could be reversed. The reason for these variations is unclear. Lipemia was influenced by dietary treatments. Interestingly, a consistent pattern of reduction was observed for total cholesterol, VLDL-cholesterol, total lipids and triacylglycerols levels when U. lactuca was fed to piglets, alone or combined with the commercial Rovabio ® , but not with the recombinant ulvan lyase, by comparison to the control diet. In addition, LDL-cholesterol decreased across all U. lactuca diets. Even if total cholesterol marginally surpassed the reference values (36-54 mg/dL) for pigs 24 , such hypocholesterolaemic effect is known to reduce the risk of cardiovascular morbidity and human mortality 25 . In some clinical studies, U. lactuca has been established as a source of antioxidants capable of decreasing serum total cholesterol, LDL cholesterol, and triacylglycerols levels 11 , all of which are significant risk factors for coronary diseases. The reversion of this effect by the recombinant enzyme could originate the degradation of the U. lactuca cell wall, making the lipid fraction more available for digestion. This would explain, for example, the increased circulating triacylglycerol levels in the ulvan lyase group. In addition, no such increase was recorded in the Rovabio ® supplemented group, putatively demonstrating the inability of the commercial enzyme mix to efficiently degrade the recalcitrant seaweed polysaccharides.
No changes were detected either for glucose or for insulin resistance index (HOMA-IR), suggesting the maintenance of glycaemic homeostasis. This happened despite reduced insulin-like growth factor (IGF-1) levels by U. lactuca inclusion, with control and ulvan lyase groups being similar. Insulin-like growth factor has the ability to decrease blood glucose levels, which does not agree with our data. Insulin is a well-known stimulator of lipogenesis 26 that increases fatty acid synthesis in the liver with formation and storage of triacylglycerols 27 . The reason why glucose levels were maintained might be putatively related to nutrient unavailability of seaweed.
Concerning hepatic function markers, different effects were promoted by feed enzymes. While the recombinant ulvan lyase decreased AST, the commercial Rovabio ® and U. lactuca alone decreased GGT and ALP, respectively, compared to the other diets. Despite the variations observed for aminotransferase activities, it is worth noticing that the levels found are still within the reference figures for pigs, which are 31-58 U/L for ALT, 32-84 U/L for AST, 10-52 U/L for GGT and 118-395 U/L for ALP, respectively 24 . Therefore, these differences do not seem to have biological significance.
As long described, algae inclusion in feeds has immune-modulatory properties. This is what is expected to occur. For instance, their anti-inflammatory activity has been associated with a minor secretion of pro-inflammatory cytokines 28,29 . This fact is corroborated by the absence of effects on IL-6, and also by enhanced levels of IL-10, across U. lactuca diets. IL-10 is a pleiotropic cytokine known for its potent anti-inflammatory and immunosuppressive effects 30 that maintain normal tissue homeostasis 31 , while IL-6 is a multifunctional cytokine that plays a central role in host defense due to its wide range of immune and hematopoietic activities and its potent ability to induce the acute phase response 32 . Polysaccharides in U. lactuca have also been demonstrated to have significant analgesic and anti-inflammatory characteristics and may well be related with such differences. These are particularly relevant for the weaned piglet, given that they endure severe enteral inflammation associated with increased exposure to pathogens and undernutrition 16,33 .
Accordingly, we found that U. lactuca by itself increased IgG immunoglobulins, which are known to be the first line of defence of the organism against infections 34 alongside with IgM. This might have contributed for increased serum protein content. IgG and IgM antibodies act in a coordinated way in short-and long-run protection against infections 35 . Upon infection, the IgM level will rise for a short period of time and then will begin to drop as the IgG levels increase, protecting the organism in the long-run 35 . In accordance, white blood cells, as a whole and lymphocytes, in particular were also increased in piglets fed U. lactuca-based diets. Bussy et al. 21 reported that a sulphated polysaccharide extract from U. armoricana in sow diets increased serum IgG of nursing sows and their piglets. The partial degradation of the cell wall by targeting ulvan might explain why the effects of enzyme supplementation of our study reversed IgG increases in the non-supplemented diets. Indeed, since ulvan is a major cell wall polysaccharide and has immunomodulating properties 36 , its degradation will theoretically prevent its effects. Altogether, these changes suggest an improved immune response during the critical post-weaning period of piglets promoted by the seaweed, which is at least attenuated by the enzymatic supplementation. Some n-3 fatty acids, which are highly present in marine algae, have immunomodulatory properties 37,38 . Moreover, antioxidants, β-carotene and vitamin B12 are also available in seaweeds, and can also modulate the immune system, even in different animal models 39 . In this study, IL-6, C-reactive protein and ApoA1 were found below detection limits. IL-6, a pro-inflammatory cytokine, is inhibited by long-chain n-3 PUFA, through the signalling of NF k B transcription factor 25 . ApoA1 is a protein involved in lipid metabolism, which is upregulated by n-3 PUFA signalling of PPARα in the liver 25 . The de novo lipogenesis that is putatively occurring in the liver because of dietary seaweed (discussed below) might provide fatty acids, such as DHA to signal these pathways. The absence of changes in these biomarkers might indicate that feeding U. lactuca does not trigger these inflammatory pathways. This is additionally supported by the fact that the acute phase C-reactive protein was unaffected by dietary treatments. All things considered, these variations reflect a boost on cellular and humoral immune response stimulated by U. lactuca that contribute to guarantee piglets' survival at the critical post-weaning period.
Cortisol is a steroid hormone that regulates a wide range of fundamental processes, including metabolism and immune response. It also has a very important role in stress response 40,41 . Stress is a phenomenon with multifactorial causes, which produces an organic response that generates adverse effects in pigs' health and welfare, and general productive performance. Herein, both exogenous feed enzymes displayed a beneficial effect in piglets by promoting a reduction of this stress hormone 42 . This could be related with decreased metabolic stress caused by higher nutrients availability, but more data is necessary to confirm this hypothesis.
Renal failure disorders are usually associated with common electrolytes, such as sodium and potassium 43  www.nature.com/scientificreports/ reverse was found with the recombinant ulvan lyase. Additionally, the kidney plays a fundamental role in keeping chloride balance in the organism. Renal chloride transport is coupled with sodium transport 44 , in accordance with the values observed for both parameters in this study. Although statistically significant, the variations found for potassium in serum are believed to be devoid of clinical physiological relevance, as the mean values obtained are very similar among dietary treatments together with a minimal standard error. The tendency for reduced antioxidant capacity in piglets fed U. lactuca combined with commercial Rovabio ® compared to controls was supported partially by a reduction in hepatic α-tocopherol, the most potent fat-soluble antioxidant available in nature 45 . No further variations in pigments with antioxidant function, like chlorophylls and carotenoids, were found in the liver, suggesting that they might be metabolized in another, hitherto unknown, fashion.
The hepatic fatty acid profile was significantly improved by the experimental diets. The most striking evidence is that U. lactuca based diets significantly increased n-3 PUFA in the liver, which explains the decreased n-6/n-3 ratio found in this tissue mostly due to 18:2n-6 (LA), 20:2n-6, and 20:4n-6 (AA) variations. Under the same rationale, 22:5n-3 (DPA) and 22:6n-3 (DHA) fatty acids were increased by the seaweed inclusion. Major n-3 PUFA, such as 18:3n-3 (ALA) and 18:4n-3, were present in high levels in U. lactuca diets, with the latter fatty acid not being detected in the control. Therefore, dietary availability is most likely the major driver for such changes. Overall, the aforementioned variations contribute to improved nutritional composition of this edible tissue, in addition to its health promoting properties for the weaned piglet. Indeed, n-3 PUFA accumulation could contribute to a downregulation of PUFA oxidation pathways and improved hepatic oxidative status 46 , counterbalancing the lower hepatic α-tocopherol levels found in seaweed diets. Similarly to what has been reported by Coelho et al. 20 using 5% of dietary Chlorella vulgaris in growing-finishing pigs, increased hepatic n-3 PUFA contents could be at least partially explained due to increased intake of the major precursor of PUFA elongation and desaturation, the 18:3n-3 (ALA) fatty acid. This is additionally supported by the fact that we did not detect these long-chain PUFA in the diets but found them in the liver. Nevertheless, lowering the n-6/n-3 ratio ultimately improves animal health and nutritional quality of this edible tissue. Indeed, lowering it has been associated with reduced hepatic inflammation and increased fatty acid oxidation in the liver, contributing for reduced hepatic lipid accumulation 47 . It also contributes for an increased intake of n-3 PUFA by the consumer with known cardiovascular-health promoting effects 48 . Unexpectedly, 18:1c9 monounsaturated fatty acid was similar across dietary groups when a twofold increase was observed with the recombinant ulvan lyase in vitro 12 .
The reasons for such a lack of differences are not clear.
In general, seaweeds have high mineral content, depending on several factors including species, environmental conditions and harvest season 2,3,49 . In this study, minor variations promoted by U. lactuca seaweed were found for mineral content in the liver. Indeed, microminerals, in particular copper, iron, manganese and zinc, presented no changes across dietary treatments. Copper and manganese are particularly important cofactors for antioxidant enzymes, such as superoxide dismutase 50,51 which contributes to overall antioxidant capacity. The lack of differences in these microminerals could be associated with similar antioxidant enzyme activity in the four groups. Indeed, the only variations observed were for a few macrominerals, namely calcium and sulphur. Calcium levels were higher in piglets fed U. lactuca combined with the recombinant ulvan lyase by comparison with piglets fed on U. lactuca alone, possibly reflecting increased intracellular calcium availability. In turn, sulphur was increased by U. lactuca diets reflecting its composition in seaweed diets.
Finally, our PCAs performed with serum and liver metabolites demonstrate that there is no clear clustering of the experimental groups. However, U. lactuca alone and control tended to be separated from the other groups in serum and liver, respectively, despite being more heterogeneous. Overall, this could demonstrate that feeding piglets with this high level of dietary U. lactuca has no major detrimental effect in systemic and central metabolism of piglets, regardless of feed enzymes supplementation.

Conclusions
Herein, we reported for the first time the effect of 7% of dietary U. lactuca in combination with feed enzymes in the haematological profile, serum biochemical metabolites, and hepatic lipid compounds, pigments, and mineral composition in recently weaned piglets. U. lactuca inclusion promoted a boost of the immune system, through the increase of lymphocytes and IgG immunoglobulins. This fact is corroborated by the non-variation of IL-6, and by the enhanced levels of IL-10, a pleiotropic cytokine known for its potent anti-inflammatory and immunosuppressive effects in the three U. lactuca fed groups. In addition, a hypolipidemic effect affecting total cholesterol, VLDL-cholesterol, total lipids and triacylglycerols was observed for U. lactuca dietary inclusion, alone or in combination with commercial Rovabio ® , but not for the seaweed combined with the recombinant ulvan lyase. In addition, U. lactuca decreased the hepatic n-6/n-3 ratio, which is a consequence of high deposition of beneficial n-3 fatty acids (as the case of DPA and DHA). Hepatic pigments remained unchanged across dietary groups, but α-tocopherol was slightly reduced by U. lactuca reaching statistical significance when combined with Rovabio ® .
Taken together, our findings indicate that U. lactuca at high incorporation levels is safe for weaned piglets, as no kidney or liver toxicity was recorded, and promotes a strong activation of their immune system, mediated by the increase of lymphocytes and IgG immunoglobulins. Also, U. lactuca does not affect glycaemic homeostasis and induces a hypolipidemic effect, involving both cholesterol and triacylglycerol compounds. In addition, both exogenous enzymes seem to display a beneficial effect on piglets' stress by promoting a reduction in cortisol concentrations. Further research using the swine model is needed to clarify the mechanisms associated with the immunostimulatory, anti-inflammatory and hypolipidemic effects promoted by U. lactuca, combined or not with feed enzymes. Moreover, studying possible negative impacts of heavy metals accumulated by the seaweed on animal metabolism would be important to fully ascertain its feed safety. An integrated proteomics and metabolomics approach could putatively be of high relevance for the elucidation of such mechanisms. The seaweed powder had 28.2% and 2.9% of crude protein and crude fat on a DM basis, respectively. The main ingredients of feeding treatments, fatty acid composition, pigments and minerals are presented in Table 5. Each piglet was individually housed in metabolic cages with free access to water. After 5 days of adaptation period to minimize stress and stabilize all metabolic conditions, the experimental animal trial began and lasted two weeks. During the trial, piglets were fed on a pair-feeding basis (50 g/kg of live weight) each day and were weighed at the beginning and end of each week. Faecal consistency was scored daily (0-normal faeces, 1-soft faeces, 2-diarrhoea, 3-severe diarrhoea). By the end of the experiment, all piglets were slaughtered according to standard commercial practices. Blood was collected into Sarstedt Z serum tubes (ref.: 32,329, Sarstedt, Nümbrecht, Germany) followed by centrifugation at 1500 g for 10 min at room temperature to separate serum which was frozen at − 80 °C, until further analyses. Hepatic samples were harvested from within the tissue, minced and frozen at − 80 °C.
Hepatic lipids determination. Freeze-dried hepatic samples were used for total lipid extraction, using the Folch et al. 57 procedure with methanol and dichloromethane (1:2 v/v) 58 . Fatty acids were transesterified with NaOH in anhydrous methanol (0.  59 . The nonadecanoic acid (19:0) was the internal standard, converting peak areas into weight percentages. The identification of fatty acids was done according to their retention times and expressed as g/100 g of total fatty acids.
Hepatic pigments determinatio n. Chlorophylls a and b, and total carotenoids were determined using the Tolpeznikaite et al. 60 protocol with minor modifications. In total, 2.5 g of fresh liver was weighted and 5 mL of 100% acetone added. Then, samples were homogenised with a T25 UltraTurrax homogenizer (IKA, Königswinter, Germany) and then centrifuged at 3000 rpm for 5 min at 4 ºC. The supernatant was separated and immediately analysed. These compounds were measured using UV-Vis spectrophotometry (ThermoScientific-Genesys 150, Waltham, MA, USA). The pigments content was determined using equations described by Dere et al. 61 .
Hepatic minerals determination. Hepatic minerals were determined as described by Ribeiro et al. 62 .
Briefly, 300 mg of freeze-dried liver was weighed into a digestion tube. HCl and HNO 3 solutions were added to each tube, followed by an overnight incubation. Before digestion, H 2 O 2 was added to each tube followed by 1 h of gradual increase to 95 ºC, and another hour at constant 95 °C. Afterwards, the subsequent solution was filtered and processed with inductively coupled plasma-optical emission spectrometry (ICP-OES).  www.nature.com/scientificreports/