Evaluation of in-field efficacy of dietary ferric tyrosine on performance, 1 intestinal health and meat quality of broiler chickens exposed to natural 2 Campylobacter jejuni challenge

Campylobacter is an important pathogen commonly found in chickens that can cause severe acute 25 gastroenteritis in humans. Despite intensive efforts to inhibit food-borne transmission of 26 Campylobacter no effective strategy exists to reduce Campylobacter loads in farmed broilers. This 27 study examined the capacity of a novel feed additive to lower Campylobacter jejuni populations 28 and to improve growth efficiency of broiler chickens. A total of 384 male one-day-old broiler 29 chicks were used in a 42-day trial. Birds were randomly allocated into four treatments with six 30 replicates of sixteen chicks per pen. Three groups were fed the basal diets further supplemented 31 with TYPLEX™ chelate (ferric tyrosine) at various concentrations (0.02, 0.05 and 0.20 g/kg, 32 groups T 2 -T 4 , respectively). Control group (T 1 ) was fed basal diets in mash form that did not 33 contain added ferric tyrosine. Feed and water were provided ad libitum . At 20 days of age, broilers 34 were exposed to natural C. jejuni challenge by introducing contaminated litter from a commercial 35 farm. At day 25, pen litter samples analysed positive for C. jejuni , and the infection intensity was 36 homogeneous among pens. At the end of the study C. jejuni counts in bird caeca were significantly 37 reduced, by 2 log 10 in the T 4 group, compared to the T 1 Control and T 3 groups ( p = 0.004). During 38 this study, a natural infection with Eimeria tenella occurred at days 26-29. For animal welfare 39 reasons all birds were treated with an anti-coccidial drug as recommended, for two consecutive 40 days. At day 42, diarrhoea was observed on the litter in only 1 of 6 pens in the T 4 group, but in 5 41 of 6 pens in the T 1 Control group. In addition, autopsies showed that the T 4 group had the highest 42 percentage of birds with normal intestinal tracts. The T 1 group had the lowest percentage of birds 43 with infection-free tracts, and higher incidence of coccidiosis and bloody diarrhoea. At 42 days of investigation. Collectively, our data suggest that addition of ferric tyrosine at 0.20 g/kg exerts a protective effect against C. jejuni and coccidiosis.


INTRODUCTION
Antibiotics have been used extensively in diets of livestock to prevent disease and/or Control group (T1) was fed basal diets in mash form; (starter feed, 1-21 days; grower feed, 134 22-42 days), without added iron. The basal diets of the other groups were supplemented with ferric 135 tyrosine at 0.02 g/kg feed (T2), 0.05 g/kg (T3) or 0.20 g/kg feed (T4). The ferric tyrosine, brand 136 name TYPLEX TM chelate (Akeso Biomedical Inc., Waltham, USA) is an iron chelate (III) with L-137 tyrosine (4-hydroxyphenylalanine). All diets were formulated to meet or exceed NRC (1994) 138 recommendations and then analysed (AOAC, 2007) for crude protein, ether extract, dry matter, 139 iron and ash (Suppl. Table 1). Coloured tracers (Micro-Tracers Inc, San Francisco) were initially 140 added to the ferric tyrosine at 10% w/w, to enable visual confirmation of ferric tyrosine content 141 and uniform mixing in feed samples. Proximate analyses of feed samples acted as a double check 142 on feed homogeneity and confirmed that feed nutrients were within the expected ranges (Table 1). 143 Diets did not contain any added iron compounds, coccidiostats or antibiotic growth promoters. infection spreads very rapidly throughout the entire shed in just few days. Therefore, at 20 days 152 of age broilers were exposed to natural C. jejuni challenge by means of contaminated litter, from 153 commercial broilers, sourced from a local farm that tested positive for C. jejuni at 44 days of age. 154 A previous study in Scotland used litter artificially contaminated with C. jejuni (Khattak et al., 155 2018) but in this study our main intention was to use a completely natural mode of infection from 156 a different geographical region in order to evaluate campylobacter replication behaviour in commercial units. The infecting inoculum was prepared by mixing thoroughly 6 kg of 158 contaminated litter to ensure an even distribution throughout and using 200 g to contaminate each 159 pen. C. jejuni is highly infectious and it has been shown before that even a single bird infected with   176 Faecal swabs, caecal and litter samples were taken on day 25 and 42, respectively, for PCR 177 amplification (Suppl. Table 2) to confirm the presence of C. jejuni (Suppl. Fig. 1). In addition, litter 178 samples from days 25 and 42 and caecal samples from day 42 and were collected and analysed for 179 C. jejuni analysis by conventional culture (Suppl. Fig. 2, 3 and 4). The caeca of two birds per pen  Coccidial OPGs were also determined in excreta samples taken from each subgroup daily 197 for the first and second day that blood presence was noticed in faeces. Sampling was carried out 198 by collecting randomly 50 g samples of excreta, two times per day from each cage for 2 consecutive 199 days. OPGs were also determined in excreta samples from each subgroup at the end of the trial at 200 the birds that had bloody diarrhoea. Samples collected from each subgroup were placed in separate 201 airtight plastic bags, homogenized thoroughly by a domestic mixer, and kept refrigerated until 202 assessed for total oocyst counts. Homogenized samples were ten-fold diluted with water to be 203 further diluted with saturated NaCl solution at a ratio of 1:10. OPGs were determined using 2.6. DNA extraction 207 In a PCR tube (300 μl; Starlab PCR Product), 5-10 random colonies were dissolved into 208 100 μl TE 10:1. The DNA was denatured by boiling for 10 min. The tube was centrifuged at 20,000   Table 2). In addition, a 16S primer set has been included as quality assurance  288 Performance parameters measured from 0 to 42 days on trial indicated that the dietary 289 supplementation of ferric tyrosine did not have any significant effects on body weight, weight gain, 290 feed intake and feed efficiency (Table 2). General health was good with low mortality until the end 291 of the trial (Suppl.

294
On day 25 (5 days after introduction of contaminated litter) C. jejuni was isolated from pen litter 295 samples (Suppl. Fig. 2). All pens were infected with C. jejuni and the counts were evenly 296 distributed among the pens. However, C. jejuni counts were significantly lower in T4 birds 297 compared to those in T1, T2 and T3 (p = 0.007). At the end of the study (day 42), C. jejuni counts 298 in the litter did not differ significantly between groups, but lower contamination was observed in 299 the T4 group (Suppl. Fig. 3). C. jejuni counts in bird caeca on day 42 were significantly lower in 300 the T4 group, compared to the T1 Control, T2 and T3 groups (p = 0.004, Fig. 1 and Suppl. Fig. 4).

301
None of the pens were negative either at day 25 or 42 and the counts were evenly distributed among 302 the pens while at day 42 all birds were infected (Suppl. Fig. 2, 3, and 4) and no negative counts 303 were observed, suggesting that all birds were exposed initially to a similar level of infection at day

326
Diarrheal scores of all pens were checked from day 24 to the end of the study (day 42). From 327 Day 26 to 36, the incidence of diarrhoea was 100% (6/6) in all pens from all treatment groups.

331
On day 42, the intestines of 37 birds per treatment group were examined post mortem.

332
Significantly more birds fed the T4 diet had normal intestinal tracts compared to the control birds 333 (p = 0.007, Suppl. Table 5). 336 In the breast meat samples, the T4 group tended to have a higher protein content compared to the 337 T3 group (p = 0.087, Suppl. Table 6). In the thigh meat samples, the T3 group had significantly 338 lower fat (p = 0.006) compared to groups T1 and T4, and significantly higher moisture (p = 0.001) 339 compared to groups T1, T2 and T4. The T4 group had significantly lower (p = 0.002) protein content 340 compared to groups T1, T2 and T3.  343 A sensory panel of 14 members recorded their degree of liking of cooked breast and thigh 344 meat. Regarding breast meat, the T3 group had significantly better scores in tenderness (p = 0.002) 345 and juiciness (p = 0.008) compared to T1 and T4 (Fig. 3). T2 and T3 groups had significantly better 346 scores in "like overall" than T4 (Fig. 3). No significant differences (p > 0.05) in sensory parameters 347 were noted for cooked thigh meat.

351
Considerable global efforts are being made to prevent human campylobacteriosis by non- For this reason, even a partial reduction of contamination of the environment and the carcasses is 374 very important when considering the risks of human campylobacteriosis (Hermans et al., 2011). 375 For example, it has been reported that the incidence of disease in humans could be reduced by 48%, 85% and 96%, if carcass contamination by Campylobacter can be reduced by 1, 2 or 3 log10 CFU,       Effect of dietary addition of TYPLEX TM on C. jejuni infection. The CFU counts (log 10 ) from caecal samples taken at study end (42 days on trial) (mean ± SEM). Replicates; 2 birds per pen , 6 pens per treatment i.e. 12 samples in total, and 3 plate replicates for each sample i.e. a final total of 2 x 6x 3 = 36 replicate samples. Values in the same treatment with no common abc superscript differ significantly (p ≤ 0.05 and ns = no significance; One way ANOVA). Effect of dietary addition of TYPLEX TM on the incidence of coccidian, diarrhea and bloody diarrhea at the end of the trial (42 days).

Figure 3.
Effect of dietary addition of ferric tyrosine on sensory panel scores of cooked breast meat on appearance, tenderness, juiciness and overall. Nº replicates: 48 (2 carcasses per pen/treatment) scored by 14 panelists. Results range: From 1 (negative perception) to 9 (positive perception). p < 0.05.