Schizochytrium sp. and lactoferrin supplementation alleviates Escherichia coli K99-induced diarrhea in preweaning dairy calves

Calf diarrhea, a common disease mainly induced by Escherichia coli infection, is one of the main reasons for nonpredator losses. Hence, an effective nonantibacterial approach to prevent calf diarrhea has become an emerging requirement. This study evaluated the microalgae Schizochytrium sp. (SZ) and lactoferrin (LF) as a nutrient intervention approach against E. coli O101: K99 -induced preweaning calve diarrhea. Fifty 1-d-old male Holstein calves were randomly divided into 5 groups (n = 10): (1) control, (2) blank (no supplement or challenge), (3) 1 g/d LF, (4) 20 g/d SZ, or (5) 1 g/d LF plus 20 g/d SZ (LFSZ). The experimental period lasted 14 d. On the morning of d 7, calves were challenged with 1 × 10 11 cfu of E. coli O101: K99, and rectum feces were collected on 3, 12, 24, and 168 h postchallenge for the control, LF, SZ, and LFSZ groups. The rectal feces of the blank group were collected on d 14. Data were analyzed using the mixed procedure of SAS (version 9.4; SAS Institute Inc.). The E. coli K99 challenge decreased the average daily gain (ADG) and increased feed-to-gain ratio (F:G) and diarrhea frequency (control vs. blank). Compared with the control group, the LFSZ group had a higher ADG and lower F:G, and the LFSZ and SZ groups had lower diarrhea frequency compared with the control group. In addition, the LFSZ and SZ groups have no differences in diarrhea frequency compared with the blank group. Compared with the control group, the blank group had lower serum nitric oxide (NO), endothelin-1, d-lactic acid (D-LA), and lipopolysaccharide (LPS) concentrations, as well as serum IgG, IL-1β, IL-6, IL-10, and TNF-α levels on d 7 and 14. On d 7, compared with the control group, all treatment groups had lower serum NO level, the SZ group had a lower serum D-LA concentration, and the LF and LFSZ groups had lower serum LPS concentration. On d 14, compared with the control group, the fecal microbiota of the blank group had lower Shannon, Simpson, Chao1, and ACE indexes, the LFSZ group had lower Shannon and Simpson indexes, the SZ and LFSZ groups had a higher Chao1 index, and all treatment groups had a higher ACE index. In fecal microbiota, Bifidobacterium and Actinobacteria were negatively associated with IL-10 and d-lactate, while Akkermansia was negatively associated with endothelin-1 and positively correlated with LPS, fecal scores, and d-lactate levels. Our results indicated that LF and SZ supplements could alleviate E. coli O101: K99 -induced calf diarrhea individually or in combination. Supplementing 1 g/d LF and 20 g/d SZ could be a potential nutrient intervention approach to prevent bacterial diarrhea in calves.


INTRODUCTION
Digestive problems, especially calf neonatal diarrhea, is a common disease during the first 3 wk after birth and caused 15.4% of nonpredator losses to cattle producers in 2015 (Cho and Yoon, 2014;Urie et al., 2018;NAHMS, 2017).A major pathogen associated with newborn calf diarrhea is enterotoxigenic E. coli (ETEC), which expresses K99 fimbriae (Shams et al., 2012).Enterotoxins produced by ETEC disturb gut microbiota, stimulate water and electrolytes loss, and cause diarrhea (Nagy and Fekete, 2005).The use of antimicrobial drugs to control and prevent intestinal diseases such as diarrhea in dairy calves has been limited due to antimicrobial resistance concerns in China, Europe, and the United States (Centner, 2016;Ricci et al., 2017).Therefore, the growing pressure to reduce subtherapeutic antibiotic applications in livestock necessitates the development of nonantibiotic alternatives (Cowles et al., 2006).Recent proposals have included n-3 PUFA and lactoferrin (LF), which benefit intestinal health (Yekta et al., 2011;Kieckens et al., 2015;Parolini, 2019).
Lactoferrin is a bioactive whey protein in colostrum and milk and has multiple beneficial effects.In vitro studies showed that lactoferrin has antimicrobial activities (Tomita et al., 1991;Shin et al., 1998), anti-inflammatory effects (Vega-Bautista et al., 2019), and antioxidant activity (Satué-Gracia et al., 2000;Sandomirsky et al., 2003).Lactoferrin could also enhance the host defense mechanisms in mice (Wen et al., 2017;Khanum et al., 2023), and modulate immunity in human and mice (Mulder et al., 2008;Lutaty et al., 2020).Robblee et al. (2003) reported that 1 g/d of LF supplementation in milk replacer significantly reduced fecal scores and the number of medicated days in Holstein dairy calves.Before weaning, calves fed either 1 or 10 g/d LF had significantly (P < 0.05) higher BW, ADG, and feed efficiency, and a tendency to consume more starter feed, even though their fecal scores were unaffected (Joslin et al., 2002).Other studies reported several beneficial LF effects, including growth promotion and intestinal mucosa maturation (Zhang et al., 2001), inhibiting the proliferation of several Clostridium and Enterobacteriaceae species (Teraguchi et al., 1994), and stimulating glucose absorption in mice (Ogata et al., 1998).Earlier studies indicated that LF effectively inhibited a wide range of infectious agents, such as E. coli (Bullen et al., 1972), Streptococcus mutans (Arnold et al., 1980), and Candida albicans (Valenti et al., 1986).However, inconsistent results have been reported by different studies.Yekta et al. (2011) showed that oral LF administration reduced E. coli O157:H7 fecal shedding, in terms of duration and colony-forming units, from sheep challenged with the E. coli serotype.However, Kieckens et al. (2015) found that oral LF administration did not effectively prevent persistent E. coli O157:H7 colonization in calves, but rectal LF administration cleared the pathogen.
The PUFA were shown to reduce inflammatory cytokine levels, prevent inflammation (Kew et al., 2004), regulate immune function (Calder and Grimble, 2002), and improve antioxidant capacity (Shimazawa et al., 2009;Tsiplakou et al., 2017).However, conflicting results are prevalent due to different models and different n-3 PUFA sources (Ballou and DePeters, 2008;Flaga et al., 2019).Therefore, more comprehensive investigations are required to understand the effects of n-3 PUFA on intestinal health in calves, especially resources rich in n-3 PUFA rather than pure PUFA to ensure cost-effectiveness at farm levels.
We previously indicated that SZ enhanced antioxidant capability in calves (Zhu La et al., 2021); however, its ability to improve newborn calf growth or intestinal health was not evaluated.In addition, LF showed an intestinal benefiting bioactivity, which showed potential as a nonantimicrobial approach to control calve digestive problems.Our hypothesis was that the SZ and LF and could alleviate calves' bacteria-induced diarrhea individually or in combination.Therefore, in this study, our objective was to examine the effects of SZ and LF dietary supplementation, alone or in combination, on the growth, development, and intestinal health of preweaning dairy calves challenged with E. coli K99.

Animals, Treatments, and Management
This study was conducted in August 2019 on a dairy farm in Dongying, Shangdong Province, China.Animals were handled and raised following protocols established by the Institute of Animal Science, Chinese Academy of Agricultural Sciences (Protocol number: IAS 20180115).Fifty newborn male Holstein calves (1 d old and 39.8 ± 4.49 kg [SD] initial BW) were randomly assigned to 1 of 5 treatment groups (n = 10 per group): (1) control (CTRL) group, fed whole milk only; (2) LF group, fed whole milk plus LF (1 g/d; 99% purity, Warrnambool Cheese & Butter Factory Company Holdings Ltd., Australia); (3) SZ group, fed whole milk plus SZ (20 g/d; ALGAMAC-3050, Aquafauna Bio-Marine, CA); (4) combination (LFSZ) group, fed whole milk plus LF (1 g/d) and SZ (20 g/d); and (5) blank group, fed whole milk only, without challenge.The LF dose was chosen based on Robblee et al. (2003), and the SZ dose was chosen based on our preliminary results (Zhu La et al., 2021).Total sample size was calculated by G-power 3.1, The effect size was calculated from our previous study of 20g/d of Schizochytrium powder on the ADG of calves, set the effect size as 0.286, power (1-β err) >80% (Zhu La et al., 2021).
Calves were separated from the dam immediately after birth and placed in individual hutches (Calf-tel; Hampel Corp., Germantown, WI) approximately 1.5 m apart for single-lap feeding.Hutches were placed on a sand base, and calves were offered 4 L of pasteurized pooled maternal colostrum within 1 h after birth, and then another 2 L at 6 h.The IgG concentration of pooled colostrum was estimated by using a PA202X digital refractometer (MISCO Refractor, Solon, OH), and the average Brix value of all colostrum fed to the calves was 35.19%, which was ≥22% higher than what was needed for any colostrum used for first feeding.Calves were then fed commercial pasteurized whole milk (Eurolac Blue, the Netherlands) twice a day at 0800 h and 1500 h (2.5 L/meal from d 2-7, then 4.5 L/meal from d 8-14).Clean, fresh water was provided daily throughout the study.Calves were offered a daily starter feed (Rubeiyou8100, Yuan Xing, China) starting from d 4 after milk feeding in the morning.Starter feed ingredients included steam-flaked corn, soybean meal, limestone, sodium chloride, vitamin A, vitamin D3, dl-α-tocopheryl acetate, copper sulfate, manganese sulfate, and zinc sulfate.Lactoferrin or SZ or both were mixed with whole milk in individual milk pails and fed once a day to calves from d 2 to 14 during the morning feeding.On d 7, calves, apart from the blank group, were challenged with E. coli O101: K99 (1 g freeze-dried powder [10 11 cfu/g] in 10 mL saline; final concentration = 1 × 10 10 cfu/mL; China Veterinary Culture Collection Center) by oral gavage at 0700 h.

Observations and Sampling
Maternal colostrum was sampled on d 1, and whole milk and starter feed sampled on d 7 and 14.For composition analysis, a 50-mL aliquot of each was stored at −20°C after adding a preservative (bronopol tablet, D&F Control System, San Ramon, CA); 50 g of starter feed was also collected, vacuumed, sealed in plastic bags, and stored at −20°C.Protein, fat, lactose, milk density, TS content, and solid nonfat were analyzed by mid-infrared procedures using a Milk Oscan Minor machine (MilkoScan Type 78110; Foss Electric, Hillerød, Denmark).The IgG concentrations in maternal colostrum were estimated using a digital refractometer (PA202X; MISCO, Solon, OH).Dry matter intake was recorded for both milk and starter feed.Neutral detergent fiber was determined following van Soest et al. (1991) using α-amylase and sodium sulfide.Acid detergent fiber and CP levels were analyzed according to the official methods of analysis (AOAC, 1990, andAOAC International 2000), respectively.Ether extract content was analyzed as described by Thiex et al. (2003) using a Soxtec 2050 system (Foss Analytical A/S, Hillerød, Denmark).Calcium (Ca) and phosphorus (P) levels were determined by atomic absorption spectroscopy and spectrophotometry (AOAC, 1990), respectively.Fat and fatty acid content were analyzed as described by Sündermann et al. (2016) and Folch et al. (1957).Whole milk and maternal colostrum chemical composition levels are shown (Table 1), and the chemical composition and fatty acid profiles of starter feed and SZ are shown (Table 2).
At the beginning of the study (d 1), d 7, and at study end (d 14) before morning feeding, BW was measured using a weighing scale (XK3190-A12+, measuring range ≤500 kg, Ronghua, Shanghai, China).The ADG was then calculated over d 1 to 14. Feed efficiency was calculated as feed: gain (F:G) based on DMI (both whole milk and starter feed) and ADG.Fecal scores were recorded daily from d 2 to 14 according to a 4-point scale by Magalhães et al. (2008) in which 1 = firm, 2 = soft or of moderate consistency, 3 = runny or mild diarrhea, 4 = watery and profuse diarrhea.Calves with fecal score higher than 2 were considered to suffer from diarrhea.
On d 7 and 14 after afternoon feeding, 20 mL of blood was collected from the jugular vein of all calves using blood-collecting needle and vacuum tubes without anticoagulant (Jiangsu Kangjian medical apparatus, Jiangsu, China).After 30 min storage in room temperature, blood samples were centrifuged at 3,000 × g for 15 min at 4°C, serum collected (Bu et al., 2007), and then analyzed for nitric oxide (NO) levels using a nitric oxide assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China).d-Lactic acid (D-LA) concentrations were detected using an ELISA kit (Enzyme-linked Biotechnology Co. Ltd, Shanghai, China), endothelin-1 (ET-1) concentrations were analyzed using the bovine endothelin-1 ELISA kit (BIM Biosciences Inc., San Francisco, CA), and LPS concentrations were determined using an assay kit (Xiamen Bioendo Technology Co. Ltd, Xiamen, China).Blood serum bovine IgG, IL-1β, IL-6, IL-10, and TNF-α levels were determined using respective ELISA kits (OmnimAbs), the coefficient of variation of all ELISA kits was <10% for interplate and <15% for intraplates.Six calves from each group (except the blank group) provided fecal samples, which were collected from the rectum using sterile gloves at 3, 12, and 24 h postchallenge and the morning of d 14 (7 d postchallenge).Fecal samples from the blank group were collected only on d 14.Samples were snap-frozen in liquid nitrogen and stored at −80°C.

Microbial DNA Extraction, PCR, and Sequencing
To analyze fecal microbiota, 1 g of rectum fecal samples, which were collected at 3, 12, and 24 h, and 7 d postchallenge (d 14) of 6 random calves in treatment and control groups and fecal samples on d 14 for the blank group (a total of 30 calves) were selected and sent to Novogene Bioinformatics Technology (Beijing, China).Briefly, microbial DNA was extracted from fecal samples using the cetyltrimethylammonium bromide method (Griffith et al., 2009).The V3-V4 hypervariable region of the 16S rRNA gene was amplified using purified template DNA (10 ng).The primers were synthesized by Novogene Bioinformatics Technology (Beijing, China) and sequences were 515F (forward; 5′-GTG CCA GCM GCC GCG GTA A-3′) and 806R (reverse; 5′-XXX XXX GGA CTA CHV GGG TWT CTA AT-3′; Caporaso et al., 2011;Klindworth et al., 2013).Each PCR reaction was performed in a 30-µL volume containing 15 µL of 2× Phusion High-Fidelity PCR Master Mix (New England Biolabs) and 0.2 µM forward and reverse primers.Thermal cycling consisted of an initial denaturation step at 98°C for 1 min, followed by 30 cycles of denaturation at 98°C for 10 s, annealing at 50°C for 30 s, and elongation at 72°C for 30 s.A final 72°C extension for 5 min completed reactions.The PCR products were purified using a Qiagen Gel Extraction Kit (Qiagen, Dusseldorf, Germany) and purified amplicons were used for library preparation using the TruSeq DNA PCR-free sample preparation kit (Illumina) following manufacturer recommendations.Index codes were also added.Libraries were quantified using a Qubit@ 2.0 Fluorometer (Thermo Scientific) and quality controlled using a Bioanalyzer 2100 system (Agilent).Libraries were then pooled in an equal molar ratio and sequenced on a NovaSeq platform (Illumina) using 2 × 250 paired-end protocols.

Sequence Analysis and Statistics
Paired-end sequence reads were demultiplexed according to unique barcodes by deMULTIplex.Sequences were further analyzed using Quantitative Insights into Microbial Ecology v1.9.1 (Caporaso et al., 2010).Each sequence read was trimmed at the first base where the average Q score was >20 over a 3-base sliding window.Then, corresponding R1 (forward) and R2 (reverse) reads were merged using FLASH V1.2.7 (Magoč and Salzberg, 2011).Sequences <75% of the expected length were also discarded (Hu et al., 2018).Chimeric sequences were identified and removed using UCHIME.Quality-filtered sequences were de novo clustered into operational taxonomic units (OTU) at 97% identity thresholds using UPARSE (V7.0.1001, http: / / drive5 .com/uparse/ ; Edgar, 2013).One representative sequence was selected from each out, and the Silva Database (http: / / www .arb-silva .de/; Quast et al., 2013) was used, based on Mothur algorithms, to annotate taxonomic information.To examine the phylogenetic relationship of different OTUs, and differences in the dominant kingdom, phylum, class, order, family, genus, and species in different samples, multiple  sequence alignments were conducted using MUSCLE software (version 3.8.31,http: / / www .drive5.com/muscle/ ; Edgar, 2004).Samples were normalized to a common sampling depth using the sequence number corresponding to the sample with the least sequences.Subsequent α-diversity analyses were performed using these normalized data and measurements calculated using QIIME v1.9.1, and included observed OTU numbers, Chao1 richness estimates, and Shannon and Simpson diversity indexes.β-diversity was analyzed in R studio (version 4.2.2) by nonmetric multidimensional scaling (NMDS) using Bray-Curtis dissimilarity, the PERMANOVA analysis was also based on the Bray-Curtis distance.
Statistical analysis was performed using SAS v. 9.4 (SAS Institute, Cary, NC), with data tested for normality.Data were analyzed using PROC MIXED to generate a variable linear model, including challenge (1,0), SZ level (1,0), LF level (1,0), and time, with fixed effects of SZ, LF, LFSZ interactions, time, and 2-and 3-way SZ and LF interactions.Effects of factors are determined by the F-test of the PROC MIXED model.Because blood and fecal samples were collected at different time points (d 7 and 14 for bloods and 3, 12, 24 h, and 7 d postchallenge for feces) and were not equally spaced, the covariance structure for repeated measurements was modeled using the spatial power option.The Kenward-Roger option was used for computing denominator degrees of freedom for hypothesis testing.Spearman correlation analysis was operated on R studio (version 4.2.2) to evaluate the correlation between microbiota taxa and serum indices and the P-value was adjusted by least significant difference.Significance was accepted at P < 0.05, high significance was accepted at P < 0.01, and tendencies were accepted at 0.05 ≤ P ≤ 0.10.The correlation heatmap was generated by ggplot2 on the R studio.

Growth Performance and Diarrhea Frequency
No differences (P > 0.050) were observed in initial BW, final BW, prechallenging diarrhea frequency or starter DMI indexes in the 5 groups, whereas milk DMI in the blank group were lower when compared with the other groups (P < 0.01 to all groups, Table 3), and the postchallenge diarrhea frequency of blank group was also lower than control group (P < 0.01).
The postchallenge daily diarrhea frequency of control group was also higher than the SZ (P < 0.01) and LFSZ groups (P < 0.01).The F:G in the control group was higher when compared with blank, SZ, and LFSZ groups (P = 0.02, 0.036, and 0.021 for blank, SZ, and Different lowercase letters refer to significant differences within one group at different time points (P < 0.05).LFSZ).The K99 challenge exerted effects on milk DMI (P < 0.01) and postchallenge diarrhea frequency (P < 0.01).The SZ group also showed effects on the postchallenge diarrhea frequency (P < 0.01).
On d 14, TNF-α concentrations were lower in calves fed SZ compared with control animals (162.50 pg/mL for SZ and 160.70 pg/mL for LFSZ vs. 166.88pg/mL for control, P < 0.05, Table 5), whereas SZ and LFSZ treatments showed a trend to elevate IgG and IL-10 (P = 0.056 and P = 0.066, respectively).
We identified 18 phyla in fecal samples from calves challenged with K99, with predominant Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, and Fusobacteria levels regardless of treatment (Figure 2a).Importantly, Firmicutes dominated all treatment groups.Supplementation of LF and SZ and resultant Different lowercase letters refer to significant differences within one group at different time points (P < 0.05).
Using Spearman correlation analyses, we identified correlations between the top fecal bacterial taxa, serum indices, and diarrhea frequency at 7 d postchallenge (Figure 3).At phylum levels (Figure 3a), Verruco- Different lowercase letters indicate significant differences between groups within the same time point (P < 0.05).

A,B
Different uppercase letters indicate significant differences in the same group between time points (P < 0.05).

DISCUSSION
Diarrhea is a common intestinal disease in newborn calves (Heinrichs and Heinrichs, 2011;Soberon et al., 2012) and reduces growth production performances (Kim et al., 2021).Most calf diarrhea cases are reported between 0 and 2 wk old (Yimer et al., 2015).As the prophylactic use of antibiotics is currently limited or banned in Europe and the United States (FDA, 2013;Ricci et al., 2017), the need for nonantibiotic alternatives to control ETEC infections and reduce calf diarrhea is essential.This prompted us to evaluate the potential effects of LF and SZ dietary supplementation to protect newborn calves from diarrhea when challenged with ETEC, in which serotype O6, O27, O148, O159, O149, and O101 (Duchet-Suchaux, 1988;Konishi et al., 2011) are commonly associated with calf diarrhea (Liu et al., 2010).Therefore, we sought to establish an E. coli O101-induced diarrhea calf model.
A transient E. coli challenge induced long-term effects on growth performance and fecal microbiota (Brubaker et al., 2021).Fecal scores reflected fecal traits; a lower fecal score indicated more solid-like feces and calves with fecal scores of 3 to 4 were considered to have moderate or profuse diarrhea.As expected, the CTRL group had increased diarrhea frequency after the K99 challenge when compared with the blank group.This was attributed to E. coli adhesion to the intestinal tract, which induced dysfunction in intestinal epithelial cytoskeletal proteins, disrupted tight junction complexes in epithelial cells, decreased ion and solute absorption rates, increased intestinal secretions, and promoted body fluid loss (Gareau and Barrett, 2013).Hence, the significantly higher milk DMI levels in the E.coli K99-challenged groups may compensate for the body fluid loss in diarrhea-induced dehydration.Lactoferrin and SZ dietary supplementation significantly reduced fecal scores and delayed peak diarrhea times in calves.
Moreover, fecal scores in calves who received LFSZ remained relatively stable before and after the K99 challenge, suggesting reduced ETEC-mediated diarrhea risks, or at least risks caused by K99.However, the time after the challenge showed a decreasing E. coli abundance, while the LFSZ group showed a trend of increasing E. coli abundance with time.In a previous study, LF inhibited at least 2 pathogenic E. coli strains in mice (Teraguchi et al., 1994) and prevented rotavirus infection in enterocyte-like HT-29 cells (Superti et al., 1997).LF binds to the lipid A portion of LPS on the surface of Gram-negative bacteria, thereby destroying outer bacterial membranes (Ochoa and Sizonenko, 2017).In our studies, we found that the LF group have a lower serum LPS concentrations on d 7 but have no difference on d 14, which may because inflammation was limited by the LF supplementation.
Schizochytrium sp. is rich in several n-3 PUFA, including DHA.Lapillonne et al. (2014) observed that DHA added to infant formula decreased infant diarrhea.We evaluated LF and SZ supplementation effects on several inflammation markers, including TNF-α, IL-1β, IL-10, and IL-6.Supplementation with SZ decreased TNF-α serum concentrations in challenged calves but increased IL-10 and IgG serum concentrations, which indicated some anti-inflammatory and pro-immune effects mediated by SZ, and this may contribute to alleviating E. coli K99-induced diarrhea by inhibiting inflammation and promoting pathogen elimination.Flaga et al. (2019) reported that TNF-α and IL-1β mRNA expression levels in bovine lymphocytes decreased linearly with increasing supplemental DHA-rich algae doses.
Another DHA-rich feed source, fish oil, decreased serum TNF-α concentrations in swine and TNF-α mRNA expression levels in peripheral mononuclear cells in calves (Carroll et al., 2003;Karcher et al., 2014).Decreased TNF-α levels were attributed to n-3 PUFA, especially EPA and DHA, by decreasing inflammatory cytokine and adhesion molecule expression (Calder, 2006(Calder, , 2015)).Previous studies indicated that n-3 PUFA reduced pro-inflammatory cytokine levels, regulated immune functions, protected vital organs, and enhanced clinical outcomes (Lei et al., 2015;Mavrommatis et al., 2021a,b).Teama and El-Tarabany (2016) also found that n-3 acid (mainly EPA and DHA) supplementation increased serum IgG concentrations in goats.Immunoglobulin is the main component of antibodies, thus increased IgG levels indicate good immune status in animals (Teama and El-Tarabany, 2016).Similarly, in our study, SZ improved immune functions and reduced inflammation in challenged calves.However, on the d 14 (7 d postchallenge), IgG concentrations appeared to be positively correlated with several pathogens and clustered into the same group as fecal scores and other inflammation indicators, which may have elevated intestinal permeability via bacterial infection-induced inflammatory and immune responses.
The intestinal microbiota of young calves develops rapidly, with many factors, like starter feed components and antimicrobial application, affecting diversity and composition (Malmuthuge and Guan, 2017).We observed that LF increased fecal microbiota species richness, consistent with previous studies showing that recombinant human LF and bovine LF increased the abundance of a wide range of microbiota members in piglets (Hu et al., 2012b;Berding et al., 2016).However, Simpson and Shannon diversity indices in the LFSZ group were lower compared with other groups.One reason may be the synergistic effects of SZ and LF; however, such action mechanisms require further study.At phylum levels, fecal bacterial communities in all groups were dominated by Firmicutes, Bacteroidetes, Proteobacteria, and Actinobacteria, irrespective of challenge.This finding accorded with a previous study which identified Firmicutes, Actinobacteria, Proteobacteria, and Bacteroidetes predominance in preweaning calves challenged with E. coli K99 (Bi et al., 2017).Higher Actinobacteria abundance also reflects a healthy intestine condition in calves (Gomez et al., 2017).Hildebrandt et al. (2009) reported that Actinobacteria levels were elevated in mice on a high-fat diet, further suggesting that Actinobacteria levels were influenced by dietary fat types (Liu et al., 2012).Bifidobacterium, which belongs to the Actinobacteria phylum, is an intestinal bacteria probiotic in dairy calves (Yu et al., 2013).In our study, Bifidobacterium had the highest overall relative abundance in SZ-fed calves and may have contributed to gut health by producing organic acids and soluble factors (Fujiwara et al., 1997;Liévin et al., 2000;Ray et al., 2014), modulating gastrointestinal and immune system responses (De Simone et al., 1992;Yasui et al., 1995).Correlation analyses in our study also indicated that Bifidobacterium have negative correlations with D-LA, which might be related to the above-mentioned immune-promotion, and microbialinhibiting metabolites improved intestinal barrier and alleviated inflammation (Laursen et al., 2021).The dietary intake of n-3 PUFA or higher levels from fish oil in rat models increased the relative abundance of Bifidobacterium and Lactobacillus (Caesar et al., 2015;Robertson et al., 2017).Gomez et al. (2017) reported that Bifidobacterium were increased in the feces of healthy dairy calves when compared with diarrheic dairy calves, similar to our study.Moreover, functional prediction results suggested that Megamonas was related to fatty acid metabolism, sphingolipid synthesis, and tryptophan and vitamin metabolism (Suchodolski et al., 2015).The SZ-treated calves contained more Bifidobacterium and Megamonas levels and suggested diarrhea levels were lower when compared with other groups.Our correlation analyses were consistent with these results; Bifidobacterium was negatively related to IL-10, IL-1, and D-LA levels.On the species level, Streptococcus gallolyticus, an opportunistic pathogen (Boleij and Tjalsma, 2013), was positively correlated with IL-1β, IL-10, D-LA, LPS, IL-6, and TNF-α, which suggests that Streptococcus gallolyticus may be correlated with higher intestinal inflammation, which is consistent with the findings of Slanzon et al. (2022).Redding et al. (2021) suggested that the R. gnavus abundance was increased in calves infected with Clostridioides difficile.In our study, R. gnavus abundance in the control group was higher than LFSZ and blank group on d 7, and was positively correlated with NO, D-LA, and TNF-α and negatively correlated with ET-1 in this study.Our result may suggest that fecal R. gnavus abundance may be related to intestinal infection.
Intestinal barrier integrity is essential for normal epithelial cell functions and prevents pathogen entry, which limits inflammation (Peng et al., 2019).Disrupted intestinal barriers increase epithelial permeability (Hu et al., 2012a).Serum D-LA is a fermentation product of some bacteria inherent to the gastrointestinal tract (Peng et al., 2019).An intestinal microbiota imbalance can cause local pathogen proliferation and excessive D-LA production, which passes into the bloodstream via damaged intestinal barriers (Zhao et al., 2011).As a consequence, increased blood D-LA levels may indicate increased intestinal permeability due to damaged intestinal barriers.In our study, SZ dietary addition reduced serum D-LA levels in dairy calves, indicating a potential protective mechanism toward barrier function in the intestinal tract.This could be due to n-3 PUFA in SZ working together to inhibit intestinal inflammation, thus protecting the intestinal barrier from E. coli K99induced diarrhea (Parolini, 2019).Our inference was also supported by the findings of Whiting et al. (2005), which stated that n-3 PUFA could reduce chronic inflammatory bowel disease severity via regulating immune cell activation or enhancing the epithelial barrier.
Lipopolysaccharide is a major gram-negative bacterial outer membrane component and causes severe intestinal inflammation and local and systemic complications such as abdominal pain, diarrhea, and weight loss (Burrin et al., 2005).We suspect that lower LPS serum levels from the LF group were caused by LF-mediated antibacterial effects (Ellison et al., 1988), and also competition with LPS in binding to cell surface molecules (Elass-Rochard et al., 1998).Neonatal diarrhea causes intestinal inflammation in dairy calves and increases the serum concentrations of specific inflammatory mediators (e.g., NO; Kojouri et al., 2012).Lactoferrin decreased NO serum concentrations compared with the CTRL group after the challenge.When administered as a therapeutic postinduced endotoxic shock, LF reduced TNF-α and NO serum concentrations (Kruzel et al., 2010).It was also hypothesized that LF may modulate intestinal microbial composition, reduce opportunistic pathogen proportions (Hu et al., 2012b;Berding et al., 2016), promote intestinal cell maturation, differentiation (Vega-Bautista et al., 2019), and help maintain intestinal wellness (Comstock et al., 2014).Decreased NO, D-LA, and LPS serum concentrations in response to LF and SZ supplementation in challenged calves possibly indicated that LF helped maintain intestinal barrier functions and alleviated diarrhea in these animals.Suchodolski et al. (2015) reported that chronic hemorrhagic diarrhea was associated with Lachnospiraceae, Coprobacillaceae, Helicobacter, and Ruminococcus genera.Overall, we observed similar results where R. gnavus relative abundance was significantly increased in CTRL diarrheal calves.Several specificinterest genera were identified from classified genera in our study.Supplementation with LFSZ increased Fusobacterium relative abundance; Fusobacterium is a potential pathogen that translocates to other organs, causes host abscesses or dermatitis digitalis, or spreads into the environment via feces (Puniya et al., 2015).However, the LFSZ group did not show a greater diarrhea occurrence when compared with other groups, which may be partly due to Lactobacillus, which had an increased relative abundance in this group.Previously, it was reported that the co-aggregation of certain Lactobacillus strain with K88 fimbriae positive ETEC and decreased ETEC colonization (Spencer and Chesson, 1994), so this mechanism in LFSZ calves may have alleviated diarrhea in these animals.The LFSZ group also showed an increased relative abundance of Akkermansia (Supplemental Table S2), which regulates intestinal mucus thickness and permeability of the intestinal barrier and maintains intestinal integrity (Li et al., 2012).Liu et al. (2015) revealed that Akkermansia was more abundant in healthy piglets when compared with piglets with epidemic diarrhea.Fish-oil-based diets were associated with Lactobacillus and Akkermansia muciniphila blooming, and reduced gut inflammation (David et al., 2014;Derrien et al., 2017).Interestingly, we showed that Akkermansia was related to higher inflammation indicators and lower ET-1 concentrations.As a mucolytic bacteria, A. muciniphila may use mucin as a metabolic substrate, and this mucolytic stimulation could promote mucin secretion and maintain physiological intestinal permeability (Chelakkot et al., 2018).However, A. muciniphila could also exacerbate inflammation during Salmonella infections (Ganesh et al., 2013), and mucin consumption in the developing intestines of preweaning calves may weaken intestinal barriers (Sheng et al., 2012).Bovine LF is also considered a probiotic analog due to its ability to stimulate the growth of certain beneficial bacteria, including Lactobacillus and Bifidobacterium (Roberts et al., 1992).Therefore, the probiotic effects Bifidobacterium likely protect calves against Fusobacterium and E. coli O101: K99 pathogens.
Even though we set up a separate blank group for the baseline, the potential individual difference among calves, especially initial serum IgG levels, may still be a limitation.The other limitation of this study is that the serum inflammation indicators, despite being closely correlated with the intestinal barrier (Bianchi et al., 2012), do not provide direct evidence for the changes.
In summary, our study showed that the supplement of 1 g/d LF, 20 g/d SZ powder, and their combination could alleviate the ETEC-induced calf diarrhea and growth performance losses.The combination of lactoferrin and Schizochytrium sp.powder have a better effect on inhibiting ETEC-induced intestinal inflammation.And the LF and SZ supplement may not influence the circular level of inflammation related cytokines.Our study showed that the combination of lactoferrin and Schizochytrium sp.(LFSZ) could be a potential nutrient intervention approach to prevent diarrhea-induced growth performance losses.

Figure 1 .Figure 2 .
Figure 1.Measurements of α-and β-diversity of fecal microbiota in calves fed lactoferrin and Schizochytrium sp., individually or in combination.(a-d) Shannon, Simpson, Chao1, and ACE indices for α-diversity; (e) nonmetric multidimensional scaling (NMDS) analysis using Bray-Curtis dissimilarity for β-diversity.Blank, Ctrl, LF, SZ, and LFSZ represent blank, control, lactoferrin, Schizochytrium, and combinations thereof.3 h, 12 h, 24 h, and 7 d represent time post-challenge.In panels a-d, the midline of the boxes indicates the mean.The lower and upper edges of the boxes indicate the interquartile range (IQR: Q1 = 25th percentile; Q3 = 75th percentile).Whiskers indicate the minimum (Q1 − 1.5 × IQR) and maximum range (Q3 + 1.5 × IQR), and dots are outliers from this range.Boxes connected by a line with an asterisk at the top have significant differences (P < 0.05).

Figure 3 .
Figure 3. Correlation heatmaps showing top fecal bacteria taxa and other indices at 7 d (168 h) post-challenge.The correlation heat maps of (a) phyla, (b) genera, and (c) species are shown separately.NO = nitric oxide, ET-1 = endothelin-1, D-LA = d-lactate.An * in a box refers to a significant correlation (P < 0.05), and ** refers to a very significant correlation (P < 0.01).
Ma et al.: SUPPLEMENTATION ALLEVIATES CALF DIARRHEA

Table 1 .
Chemical composition of whole milk and mixed maternal colostrum fed to calves

Table 2 .
Ma et al.: SUPPLEMENTATION ALLEVIATES CALF DIARRHEA Chemical composition and fatty acid content in Schizochytrium sp. and starter feed 1 ND = not determined.

Table 3 .
Ma et al.:SUPPLEMENTATION ALLEVIATES CALF DIARRHEA Effects of feeding lactoferrin and Schizochytrium sp., individually or in combination, on ADG, starter and milk DMI, feed efficiency, and fecal scores in dairy calves from

Table 4 .
Ma et al.:SUPPLEMENTATION ALLEVIATES CALF DIARRHEA Effects of feeding lactoferrin and Schizochytrium sp., individually or in combination, on serum concentrations of nitric oxide, endothelin-1, d-lactic acid, and LPS in dairy

Table 5 .
Ma et al.:SUPPLEMENTATION ALLEVIATES CALF DIARRHEA Effects of feeding lactoferrin and Schizochytrium sp., individually or in combination, on serum concentrations of IgG, interleukins, and TNF-α in dairy calves 1 Ma et al.: SUPPLEMENTATION ALLEVIATES CALF DIARRHEA