**3. Discussion**

While supplementation with LAM was previously shown to be effective in improving pig performance from weaning until day 14 post-weaning [5], the results of this study sugges<sup>t</sup> that, under good sanitary conditions, there is no additional benefit on performance to continued supplementation with either LAM or FUC supplementation up to day 35 post-weaning. It is important to note, however, that pigs fed the LAM supplemented diet had increased colonic butyrate production and positive alterations in the colonic microbiome characterized by reduced *Campylobacter* and increases in the genus *Roseburia* and *Faecalbacterium*. Such effects could contribute to a better host immune response to any potential environmental/pathogen challenge during this time period.

There were substantial differences in the microbiome profile between the three experimental groups. There was a contrasting effect between the two SWE supplemented treatments, particularly in relation to changes in the phylum Proteobacteria. The pigs fed the LAM supplemented diet had a decrease in this phylum, while pigs fed the FUC supplemented diet had an increase in the abundance of this phyla compared to pigs fed the basal diet. These differences were supported at both family and genus taxonomic levels with differences in the family *Campylobacteraceae* and the genus *Campylobacter.* The magnitude of the differences was interesting as pigs fed LAM had 23 times lower *Campylobacter* compared to pigs fed the basal diet, while pigs fed the FUC supplemented diet had 4 times higher *Campylobacter* compared to pigs fed the control diet. *Campylobacter* is a major foodborne pathogen often solely linked to infections from broilers through *C. jejuni*, however, *C. coli* is implicated with infections originating from swine [9]. A potential anti-bacterial effect following the addition of LAM is potentially of major benefit to the pig industry [10]. While *Campylobacter* is often considered to form part of the commensal flora of pigs and cause no negative impact on animal performance, there is some debate as high counts of *Campylobacter* are often observed in conjunction with other bacteria during incidences of diarrhea [11]. From either the aspect of reducing a zoonotic pathogen or improving intestinal health, the observed effects on *Campylobacter* by the addition of LAM can be considered a positive finding while the large increase in *Campylobacter* following the addition of FUC is likely to have negative connotations.

Coinciding with the lowest abundance of Proteobacteria in this study, pigs fed the LAM supplemented diet had the greatest abundance of the phylum Firmicutes. This increase was predominantly related to increases in the *Ruminococcaceae* and the *Lachnospiraceae* families. At genus level this coincided with differences in *Faecalbacterium* and *Roseburia.* Within the genus *Faecalbacterium* only one species has been identified in the literature, which is *Faecalbacterium prausnitzi*, implying that this change at genus level relates to a change in abundance of this species. This genus has been associated with increased intestinal health based on a low incidence of post-weaning diarrhea in pigs [12]. In humans, *Faecalbacterium prausnitzi* is used as an indicator of intestinal health and is also considered to be a very promising probiotic for human use [13,14]. The end products of carbohydrate fermentation by *Faecalbacterium prausnitzi* are formate, lactate and substantial quantities of butyrate [15]. Further corroborating the potential improved gu<sup>t</sup> health in pigs fed LAM are an increase in the genus *Roseburia*. *Roseburia* is a member of the Firmicutes phylum and is associated with the production of butyrate [8]. Therefore, the increase in *Faecalbacterium and Roseburia* could be associated with the increase in butyric acid identified in pigs from this study fed the LAM supplemented diet. Butyrate is produced from bacterial fermentation to provide energy to colonocytes. It also has functions as a cellular mediator, regulating multiple functions of gu<sup>t</sup> cells and beyond, including gene expression, cell differentiation, gu<sup>t</sup> tissue development, immune modulation, oxidative stress reduction and diarrhea control [16]. These changes sugges<sup>t</sup> an improvement in gu<sup>t</sup> health in pigs fed the LAM supplemented diet.

In contrast to the clear beneficial effects of supplementation with LAM on members of the Firmicutes and Proteobacteria phyla, the increase in abundance of the Spirochaetes phylum are less clear. Within this phylum associated changes were identified with increases in the family *Spirochaetaceae* and the genus *Treponema* were also identified, with pigs fed the LAM supplemented diet having greater abundance than both the two other experimental groups. The genus contains both pathogenic as well as commensal bacteria [17]. Pathogenic bacteria include *Treponema pallidum*, and others such as *T. denticola* (Tde), *T. putidum* (Tpu) *T. pedis* (Tpe), *T. brennaborense* (Tbr), and *T. paraluiscuniculi* (Tpar) which are the causative agents of various disease in both humans and animals [18]. Porcine intestinal Spirochaetes form a diverse group of organisms including the beta-hemolytic *Serpulina*/*Treponema hyodysenteriae* which is the main etiological agen<sup>t</sup> of swine dysentery [19]. While the classification of OTUs at species level is unreliable, in this study, species classified as part of the genus Treponema include *berlinense*, *bryantii*, *succinifaciens*, *suis* and *porcinum*. While data related to these species is limited, *Treponema porcinum*, *Treponema succinifaciens*, *Treponema berlinense* and *Treponema bryantii* are associated with improved feed efficiency in pigs [20,21]. Further analysis such as shotgun metagenomics will need to be conducted to better understand the impact of changes in this genus have on the function of the microbiome. However, based on the other positive effects on gu<sup>t</sup> health in pigs fed the LAM supplemented diet and the lack of negative impacts on gu<sup>t</sup> health, it seems the changes in the Spirochaetes phylum are not negatively impacting the pigs.

Pigs fed both the LAM and the FUC supplemented diets had reduced bacterial diversity based on the Observed, Chao1, ACE and Fisher measures of alpha diversity. While there are conflicting opinions in the literature, lower bacterial diversity is generally considered to have negative connotations as greater microbial diversity is directly associated with ecosystem stability [22]. For example, in humans, reductions in diversity is associated with a range of health issues and diseases [23]. Therefore, while a reduced diversity in the pigs fed the FUC supplemented agrees with the changes in the microbiome and performance for this group, the reduction in diversity in pigs fed the LAM supplemented diet is surprising. The changes in the microbiome in this group particularly in the reduction in *Campylobacter* sugges<sup>t</sup> improved gu<sup>t</sup> health and therefore, it is unlikely that a reduction in diversity is having a negative influence in this group of pigs.

In relation to the gene expression data in this study, there are contrasting effects between the diets supplemented with the seaweed extracts. In relation to supplementation with LAM only minor changes in gene expression were identified relative to the control while more substantial changes were identified in pigs fed the FUC supplemented diet relative to pigs fed the control diet. However, there was region specific differences. In the duodenum, pigs fed the FUC supplemented diet had increased expression of SLC6A19 and CNDP1 compared to pigs fed the basal and also the pigs fed the diet supplemented with LAM. The two SWE supplemented groups had reduced expression of SLC16A10 compared to pigs fed the basal non-supplemented diet. These three transporters are all involved in the transport of amino acids [24]. This suggests there is increased availability and transport of amino acids in the duodenum of pigs fed the diets containing seaweed extracts compared to pigs fed the control. However, nutrient transporters are regulated by dietary substrate levels meaning the increased expression of protein transporters is a response to an increase in nutrient availability in the duodenum which can be impacted by a number of factors such as feed intake, but also may be a response to changes in intestinal architecture such as reductions in villus height [25]. A further influence of the SWE on nutrient transport was identified in the jejunum with pigs fed the FUC supplemented diets having increased expression of MCT1/SLC16A. SLC16A1 is responsible for the transport of monocarboxylates such as L-lactate and ketone bodies and also with the transport of butyrate [26,27]. These data, while speculative sugges<sup>t</sup> an increased bacterial fermentation in the jejunum with the addition of FUC leading to increased production of butyrate and subsequent absorption with MCT1. While the influence of bacterial fermentation in small intestine is not as pronounced as the large intestine, the production and utilisation of butyrate in the small intestine and in particular the jejunum, as is the case in this study is important in the modulation of balance between apoptosis and proliferation [28]. In contrast to the upregulation of expression of genes related to nutrient digestion and absorption in the duodenum and jejunum, in the ileum, FUC supplementation led to a reduction in the expression of SLC2A2, GCG, FABP2, SGLT1 and SI in the ileum. Similar, results were identified where the feeding of a similar level of FUC in the post-weaning period had a similar influence on nutrient transporters suggesting that the addition of FUC has an impact in nutrient availability [6]. In the study of Rattigan et al. [6] the authors attributed the reduction in gene expression to the presence of alginates and fucoidan in the extract. Both FUC and alginates have been attributed with increasing the viscosity of digestive contents, reducing the flow of digesta and reducing the mixing of digestive contents leading to lower rates of nutrient breakdown [29,30]. The impact of these contrasting effects in different regions of the small intestine are difficult to quantify particularly due to the fact ileal digestibility's

was not measured. Unfortunately, the material required to conduct an analysis of nutrient digestibility was unavailable due to the volume of digesta in the digestive tract at the time of slaughter. Further research will be required to further understand the impacts of these extracts and nutrient breakdown and absorption.

Significant effects were identified between experimental groups in both the expression of targets related to barrier function and the expression of cytokines. Pigs fed the FUC supplemented diet had reductions in the expression of *OCLN* in both the jejunum and ileum compared to pigs fed the basal diet, while a tendency towards reduced expression was also identified in the duodenum. In agreemen<sup>t</sup> with these data is a reduction in the expression of *CLDN3* and *CLDN5* in the ileum compared to pigs fed the control diet. Proteins in the claudin and occludin families are a main component of tight junctions and form a seal that modulates paracellular transport in the intestinal epithelium [31]. Reduced expression of claudins such as 3 and 5 is associated with intestinal inflammatory disorders [32]. This is in agreemen<sup>t</sup> with effect identified on the expression of IL8/CXCL8 in the jejunum and the increased expression of *TGFB1* and *TNF* in the ileum of pigs fed the diet supplemented with FUC compared to pigs fed the basal unsupplemented diet. For the pigs fed the FUC supplemented diet the changes in gene expression are likely indicators of detrimental effects of the extract on gu<sup>t</sup> health with increased markers of inflammation and reduced gene expression of targets associated with barrier function. In conjunction with the changes in the microbiome in the large intestine these data are possible explanations for the reduction in performance of pigs fed FUC. The pigs fed LAM had reductions in the expression of *MUC1* compared to pigs fed the basal and also pigs fed the diet supplanted with LAM. MUC1 provides a barrier against potential pathogens while also modulating the expression of inflammatory cytokines [33]. Therefore, the increase in expression of this gene is surprising as pigs fed the diet supplemented with LAM had no increases in inflammation, a healthier microbial profile and improved performance. Further research will need to be conducted to establish the reasons for this increase in expression such as a quantification of protein abundance as changes in expression are not always a true reflection of the production of the associated protein.

The aim of this study was to determine if the continued supplementation of two SWEs past the initial post-weaning period continues to be beneficial to performance. It is interesting to note that while SWE supplementation past the immediate post-weaning period up to d 35 in this study did not result in improved performance in healthy pigs under good sanitary experimental conditions, there was evidence to sugges<sup>t</sup> that LAM supplementation had positive effects on gu<sup>t</sup> health. This is significant as previous studies have identified that SWE supplementation can provide substantial performance and health benefits when the pigs become challenged. Heim at al [34] identified that SWE supplemented pigs had lower faecal scores and better performance than basal fed pigs following an ETEC challenge. Rattigan et al. [6] identified that LAM supplementation improves performance when pigs are housed in unsanitary conditions. While the mode of action is not completely understood, the evidence would sugges<sup>t</sup> that bioactives in the SWE can influence both the host immune cells and gu<sup>t</sup> microbiome cells. For instance, bioactives such as laminarin, can be internalized by intestinal epithelial cells and gu<sup>t</sup> associated lymphoid cells [35]. This internalization upregulates the expression of pattern recognition receptors, increasing protective cytokine expression and induces protection against infectious challenge [36]. Hence, this effect would not be evident until the pigs undergo some type of environmental or pathogenic challenge.

#### **4. Materials and Methods**

All experimental procedures described in this work were approved under the University College Dublin Animal Research Ethics Committee (AREC-17-19-O'Doherty, 11 May 2017) and were conducted in accordance with Irish legislation (SI no. 543/2012) and the EU directive 2010/63/EU for animal experimentation.

#### *4.1. Experimental Design and Animal Management*

72 weaned pigs (progeny of Landrace boars × (Large White × Landrace) sows) were sourced from sows from the same farrowing house. Blocking of piglets was done on the basis of initial live weight (8.4 kg, sd 1.05 kg), sex, litter of birth and pigs were assigned to one of three groups: 1) basal diet; 2) basal diet + 250 ppm fucoidan rich extract (FUC); 3) basal diet + 300 ppm of a laminarin (LAM) rich extract with a total of 8 replicates per treatment.

The pigs were penned in groups of three and housed on fully slatted floors (1.68 × 1.22 m) with the initial temperature set at 30 ◦C and subsequently reduced each week by 2 ◦C with the humidity maintained at 65%. Pig weights were recorded on day 0 (weaning) and subsequently on days 7, 14, 21, 28 and 35. Feed and water were available ad libitum from four-space feeders and nipple drinkers, respectively, throughout the experiment.

#### *4.2. Preperation of Experimental Diets and Characterisation of Extracts*

Pigs in this experiment were offered one of three diets: (1) basal diet; (2) basal diet + 250 ppm FUC; (3) basal diet + 300 ppm LAM. The basal diet contained 15.3 MJ/kg digestible energy, 190 g/kg crude protein and 13.5 g/kg total lysine. The basal diet was comprised of wheat (34%), extruded full fat soya (17%), flaked wheat (13%) soya bean meal (48% crude protein) (10.5%), flaked maize (7%) whey powder (5%), provisoy (soya protein concentrate) (6.5%), soya oil (3%) with the remainder compromised of mineral and vitamin supplements (Table 7). Lysine was used a reference for calculation of the other amino acid requirements (NRC, 2012). The extraction procedures and methodology were described in previous publications but are further outlined below [5,6,37].

A laminarin-rich extract (BioAtlantis Ltd., Clash Industrial Estate, Tralee, Ireland) was obtained from *Laminaria Digitata* using a hydrothermal-assisted extraction and preoptimized conditions for maximum yield of laminarin as described previously [38,39]. Briefly, dried and milled seaweed was suspended of 0.1N HCl maintaining a solid to liquid ratio of 1:21 (g/mL). The mixture was thoroughly agitated to ensure uniformity and then subjected to a temperature of 100 ◦C for 30 min. The subsequent crude extract was partially purified to increase the relative polysaccharide content and to remove or reduce other constituents; proteins, polyphenols, mannitol and alginate. This was achieved through mixing the crude extract with pure ethanol (to remove polyphenols) followed by water (to remove protein) and calcium chloride (to remove alginates).

A fucoidan-rich extract (BioAtlantis Ltd., Clash Industrial Estate, Tralee, Ireland) was obtained from *Ascophyllum nodosum* using a hydrothermal-assisted extraction and preoptimized conditions for maximum yield of fucoidan [39]. The crude extract was partially purified to increase the relative polysaccharide content and to remove or reduce other constituents such as proteins, polyphenols, mannitol and alginate. This was achieved through mixing the crude extract with pure ethanol (to remove polyphenols) followed by water (to remove protein) and calcium chloride (to remove alginates). Fractions were separated based on molecular weight by employing a molecular weight cut-off centrifugal concentrator (100 KDa cut-off).

#### *4.3. Post Slaughter Sample Collection*

On d35 following slaughter, tissue sections of 1cm<sup>2</sup> were cut from the duodenum, jejunum and ileum and washed using sterile PBS (Oxoid). The overlying smooth muscle was removed before storage in 5 mL RNAlater solution (Applied Biosystems, Dublin, Ireland) overnight at 4 ◦C which was subsequently removed before long term storage at −80 ◦C. Colonic digesta was removed and stored in sterile containers (Sarstedt, Wexford, Ireland) and frozen ( −20 ◦C) for subsequent 16s rRNA sequencing and VFA analyses.


**Table 7.** Ingredient and chemical composition of basal diet \*.

\* Treatments were as follows: (1) basal diet; (2) basal diet; basal diet + 250 parts per million (ppm) fucoidan; (3) basal diet + 300 ppm of a laminarin rich extract basal diet; † Values calculated based on tabulated nutritional composition [40]. The basal diet was formulated to provide: Cu, 100; Fe, 140; Mn, 47; Zn, 120; I, 0.6; Se, 0.3; retinol, 1.8; cholecalciferol, 0.025; α-tocopherol, 67; phytylmenaquinone, 4; cyanocobalamin, 0.01; riboflavin, 2; nicotinic acid, 12; pantothenic acid, 10; choline chloride, 250; thiamine, 2; pyridoxine, 0.015 (mg/kg diet). Celite was also included at 300 mg/kg.

#### *4.4. Volatile Fatty Acid Analysis*

The VFA concentrations in the digesta were determined using gas liquid chromatography (GLC) as reported by Clarke et al. [41]. 1 g of digesta was mixed with distilled water (2.5 × sample weight) and centrifuged at 1400 g for 10 min (Sorvall GLC-2B laboratory centrifuge, DuPont, Wilmington, DE, USA). One mL of the subsequent supernatant and 1 mL of internal standard (0.05% 3-methyl-n-valeric acid in 0.15 M oxalic acid dihydrate) were mixed with 3 mL of distilled water. The reaction mixture was centrifuged at 500 g for 10 min, and the supernatant was filtered through 0.45 PTFE syringe filter into a chromatographic sample vial. An injection volume of 1 μL was injected into a Varian 3800 GC equipped with a EC ™ 1000 Grace column (15 m × 0.53 mm I.D) with 1.20 μm film thickness. The temperature was set at 75–95 ◦C increasing by 3 ◦C/min, 95–200 increasing by 20 ◦C per minute, which was held for 0.50 min. The detector and injector temperature were 280 ◦C and 240 ◦C, respectively, while the total analysis time was 12.42 min.
