Effects of cashew nut shell extract supplementation on production, rumen fermentation, metabolism, and inflammatory biomarkers in transition dairy cows

Cashew nut shell extract (CNSE) is a byproduct of the cashew nut industry containing bioactive compounds that alter rumen fermentation patterns. Therefore, study objectives were to evaluate the effects of CNSE (59% anacardic acid and 18% cardol) on production, rumen fermentation variables, metabolism, and inflammation in transition dairy cows. Fifty-one multiparous Holstein cows were used in a randomized design and assigned to treatment based on their previous 305-d mature equivalent milk and parity and assigned to 1 of 2 treatments 21 d before expected calving: 1) CON (control diet; n = 17) or 2) CNSE-5.0 (control diet and 5.0 g/d CNSE granule [containing 50% CNSE]; n = 34). Following parturition, 17 cows (preselected at initial treatment assignment) from the CNSE-5.0 treatment were reallocated into a third treatment group: CNSE-2.5 (control diet and 2.5 g/d CNSE granule; n = 17) resulting in 3 total treatments postpartum: 1) CON, 2) CNSE-2.5, and 3) CNSE-5.0. Prepartum rumen pH was unaltered by treatment; however, postpartum rumen pH was increased (0.31 units) in CNSE cows relative to CON. Prepartum rumen ammonia N concentration tended to be decreased (34%) in CNSE-5.0 cows compared with CON, and there tended to be a quadratic effect on postpartum ammonia N as it was decreased in CNSE-2.5 compared with CON and CNSE-5.0. Prepar-tum dry matter intake (DMI) was unaffected by treat-ment; however, postpartum DMI was increased (8%) in CNSE cows relative to CON. No treatment differences were observed in pre or postpartum digestibility measurements. Milk and protein yield from cows fed CNSE tended to be increased (6 and 7%, respectively) relative to CON. No treatment differences were detected for energy


INTRODUCTION
The weeks surrounding parturition are critical in determining a cow's welfare and future productivity as well as the producer's profitability, and farm's sustainability.Lactation onset is a demanding period in which energy output (maintenance and milk synthesis) exceeds the capacity for energy intake, resulting in a period of negative energy balance, which has been associated with adverse health events (Grummer, 1993;Drackley, 1999).Inadequate feed intake around parturition reduces the availability of diet-derived gluconeogenic precursors such as propionate (Bell, 1995).This metabolic scenario results in the cow enlisting multiple glucose sparing mechanisms characterized by tissue mobilization (Bauman and Currie, 1980;Bell, 1995).Glucose is essential for lactation because it is the precursor for lactose synthesis; the primary osmotic regulator of milk yield (Guinard-Flament et al., 2006) and is also the preferred fuel of activated immune cells (Lang et al., 1993).Lactogenesis and immune system activation almost always occur concurrently restulting in a considerable net glucose deficit.Thus, increasing the gluconeogenic precursor supply has numerous benefits in early lactation and is the foundation for why more energy dense diets (and specifically more nonstructural carbohydrate dense rations) are fed following parturition.
Feed supplements have been widely utilized in transition cow diets to modulate microbial populations and alter rumen fermentation patterns (Nocek and Kautz, 2006;Zaworski et al., 2014).Compounds that can safely increase ruminal propionate synthesis often improve feed efficiency and production while decreasing enteric methane production by providing a reducing equivalent sink (Honan et al., 2021).Cashew nuts (Anacardium occidentale) are produced worldwide with over 50% grown in Asia (Maeda et al., 2020).Cashew nut shell extract (CNSE; interchangeably denoted in the literature as cashew nut shell liquid or CNSL) is a byproduct containing anacardic acid, cardanol, and cardol; molecules that have antibacterial activity against grampositive bacteria (Kubo et al., 1993;Watanabe et al., 2010;Shinkai et al., 2012).Additionally, CNSE appears to alter rumen fermentation by increasing propionate production both in vitro and in vivo (Watanabe et al., 2010;Shinkai et al., 2012;Kang et al., 2018), but the effects of feeding CNSE during the pariparturient period has not been fully examined.Increased propionate production would energetically benefit transition dairy cows that simultaneously require a large amount of glucose for both milk synthesis and to mount an effective immune response (Horst et al., 2021).Therefore, study objectives were to evaluate the effects of CNSE on production, rumen dynamics, post-absorptive energetic metabolism, and inflammatory markers in transition dairy cows.

Animals, Experimental Design, and Diets
All procedures were approved by the Iowa State University Institutional Animal Care and Use Committee.Fifty-one multiparous Holstein cows calving between June and September 2019 were assigned to treatment based upon their previous 305-d mature equivalent milk and parity.Cows were assigned to 1 of 2 treatments 21 ± 4 d before expected calving: 1) CON (control diet; n = 17) or 2) CNSE-5.0 (control diet and 5.0 g/d CNSE [containing 50% CNSE]; n = 34).Following parturition, 17 cows (preselected at initial treatment assignment) from the CNSE-5.0treatment were reallocated into a third treatment group: CNSE-2.5 (control diet and 2.5 g/d CNSE; n = 17) resulting in 3 total treatments postpartum: 1) CON, 2) CNSE-2.5 and 3) CNSE-5.0.Treatments were mixed with ground corn to equal 100 g (ground corn and CNSE) and top-dressed on the TMR daily (the controls received 100 g of ground corn).The CNSE doses were selected based upon extrapolated data from previous reports (Kang et al., 2018;Shinkai et al., 2012) and only the CNSE-5 was evaluated prepartum because a higher dose was hypothesized to be more effective in a higher forage-based diet.Cows remained on their respective postpartum dietary treatment until 28 DIM.Immediately postpartum, cows were milked and processed according to standard operating procedures implemented at the Iowa State University Dairy Farm and moved to the lactation freestall barn.One cow in the CNSE-2.5 treatment was removed from the experiment due to health issues (lung abscess) postpartum and her data was not included in the final data set.
Cows were individually fed using the Calan Broadbent feeding system (American Calan, Northwood, NH) a diet formulated to meet or exceed the predicted requirements (NRC, 2001; Table 1) of energy, protein, minerals, and vitamins for close up dry and early lactation cows, respectively.Samples of TMR were obtained weekly and composited into a monthly sample for nutrient analysis (Dairyland Laboratories Inc., Arcadia, WI).Respective rations were delivered once daily, and orts were measured, recorded, and discarded 1 h before feeding.Cows were milked twice daily (1100 and 2200 h), and yield was recorded using the Boumatic Smart Dairy parlor system and reported by PCDart dairy management software at the Iowa State University Dairy Farm.A sample for composition analysis was obtained on d 3, 7, 14, 21, and 28 ± 1 from the 1100 h milking (Boumatic SmartControl metering system).Samples were stored at 4°C with a preservative (bronopol; D & F Control System, San Ramon, CA) until analysis by Dairy Lab Services (Dubuque, IA) using AOAC approved infrared analysis equipment and procedures (AOAC International, 1995).

Body Measurements and Fecal pH
Body weight and body condition score (BCS) were obtained weekly throughout the experiment.Body condition scores were measured by the same individual utilizing Wildman et al. (1982) scoring system but reported in 0.25-unit increments.Fecal samples were col- lected (0530 h) via manual rectal extraction biweekly from wk −3 to 4 relative to calving.To ensure accuracy, 50 g of fecal matter were diluted with 50 g of distilled water and homogenized (Lab-Blender Stomacher 80, Seward Ltd.) before pH was determined as previously described (Branstad et al., 2017).Following sample homogenization, fecal pH was assessed using a portable pH meter (Orion Star A121, Thermo Fisher Scientific) and later averaged into weekly values for analysis.

Rumen pH, Ammonia N, and VFA Analysis
Rumen fluid was collected before feeding (0530 h) from a subset of rumen cannulated cows (n = 12; 4/ treatment).Fluid was obtained from the cranial, caudal, dorsal, and ventral quadrants of the rumen, strained through a paint strainer (Reaves & Co., Inc.), and pH was immediately determined using a portable pH meter (Orion Star A121, Thermo Fisher Scientific) on d −14 ± 3, −7 ± 3, −3 ± 1, 3, 7, 14, 21, and 28 relative to calving.A subsample was stored at −80°C and later submitted to the University of Georgia for ammonia N and volatile fatty acid (VFA) analysis on d −14 ± 3, −7 ± 3, 7, 14, and 21 relative to calving.Ammonia N concentrations were assayed using a modified colorimetric method (Chaney and Marbach, 1962;Russell and Jeraci, 1984).Briefly, thawed rumen fluid was centrifuged to remove bacteria and undigested feed (10,000 × g, 15 min, 5°C), and the cell-free supernatant was stored at ‴15°C.A 25 μL sample or standard (0 to 11.99 mM ammonium sulfate) was mixed with 3.0 mL phenol reagent followed by 3.0 mL hypochlorite solution and was incubated at 39°C for 20 min before absorbance was read at 630 nm (Genesys 30 Visible Spectrophotometer; Thermo Fisher Scientific).Analysis of VFA was performed according to the procedure described previously (Lourenco et al., 2020).Rumen fluid samples were centrifuged at 10,000 × g for 10 min, and 1 mL of the supernatant was mixed with 0.2 mL of metaphosphoric acid solution (25% wt/vol).Samples were vortexed for 30 s and stored at −20°C overnight.The following morning, samples were thawed and centrifuged at 10,000 × g for 10 min.The supernatant was removed and transferred into polypropylene tubes combined with ethyl acetate in a 2:1 ethyl acetate to supernatant ratio.Tubes were vortexed for 10 s to thoroughly mix them and allowed to settle for 5 min for optimum separation.Then, 600 μL of the top layer was transferred into screw-thread vials.Volatile fatty acid analysis was performed using gas chromatograph (Shimadzu GC-2010 Plus; Shimadzu Corporation) with a flame ionization detector and a capillary column (Zebron ZB-FFAP; 30 m x 0.32 mm x 0.25 μm; Phenomenex Inc.).Sample injection volume was set to 1.0 μL, and helium was used as a carrier gas.Column temperature started at 110°C and increased to 200°C for 6 min.The injector temperature was set to 250°C, and the detector temperature was set to 350°C.

Digestibility
A subset of cows (n = 24; 8/treatment) were utilized for digestibility analysis on d −10 ± 4 and 21 relative to calving.Cows were orally dosed twice daily (0530 and 1730 h) with a gelatin capsule containing 10 g of titanium dioxide (TiO 2 ) for 12 d.On d 11 and 12, fecal samples were obtained from the rectum (0000, 0800, 1600 h on d 11 and 0400, 1200, and 2000 h on d 12) and frozen at −20°C until analysis.Sampling times were staggered such that the entire 24-h day was represented to capture diurnal variation.A similar volume of feces (~500 g) on a wet basis was collected and composited for each cow.Fecal samples were analyzed for dry matter (DM), organic matter (OM), neutral detergent fiber (NDF), acid detergent fiber (ADF) using methods previously described (Russell et al., 2016).Fecal samples were sent to a commercial laboratory (Dairyland Laboratories Inc., Arcadia, WI) for analysis of starch content.Samples were prepared using methods outlined by Myers et al. (2004), and TiO 2 content was determined colorimetrically (Eon Microplate Spectrophotometer; Biotek).Individual feed intake was calculated by averaging intakes for the final 4 d of TiO 2 supplementation period for each cow before fecal collection.Fecal output was calculated by dividing individual TiO 2 intake (g) by the fecal TiO 2 concentration (g TiO 2 / g dry feces).Fecal output (g, DM basis) and fecal nutrient concentrations (DM, OM, NDF, ADF and starch) were multiplied to calculate amounts excreted for each cow (g).Similarly, individual feed intake and analyzed diet nutrient concentrations were multiplied to determine nutrient intake (g).Apparent digestibility (%) for each nutrient was calculated as [1 -(output/input)] × 100.A 10% coefficient of variation threshold was maintained for fecal output calculations.

Blood Metabolites, Hormones, and Acute Phase Proteins
Blood samples from a subset of cows (n = 36; 12/ treatment) were collected before feeding (0530 h) via coccygeal venipuncture on d −14 ± 3, −7 ± 3, 3, 7, 14, 21, and 28 relative to calving.Samples were collected into a tube containing K 2 EDTA (Becton, Dickinson and Co.; for plasma collection) and a tube containing a silicone clot activator (Becton, Dickinson and Co.; for serum collection).Serum samples were allowed to clot at room temperature for ~1 h before centrifugation.Plasma and serum samples were harvested following centrifugation at 1,500 × g for 15 min at 4°C and were subsequently frozen at −20°C until analysis.

Calculations and Statistical Analysis
Milk composition samples obtained on d 3 and 7 were averaged and assigned to wk 1, whereas samples obtained on d 14, 21, and 28 were assigned to wk 2 to 4, respectively.Energy, fat, and solids corrected milk (ECM, FCM, and SCM, respectively) were calculated as described by Tyrrell and Reid (1965) and Van-Baale et al. (2005)  Milk yield and DMI were analyzed as daily values, but all other daily measurements were converted to weekly averages for analysis, and any week with less than 4 daily measures per animal was removed from the data set and was considered a missing data point.The sample size calculation was determined using PROC POWER and was based on the anticipated effect of treatment on milk yield of 3.0 ± 2.8 kg/d using 86% power.Data were analyzed using the MIXED procedure of SAS version 9.4 (SAS Inst.Inc.; Cary, NC).Each animals' respective parameter was analyzed using repeated measures with an autoregressive (equally spaced measures) or spatial power (unequal spacing between measures) covariance structure and time (i.e., day or week) as the repeated factor.Cows were blocked by their month of calving, and pre and postpartum variables were analyzed separately.The model included treatment, time (day or week), treatment by time interaction, and block as fixed effects, with cow used as the random effect.Each animals' previous 305 mature equivalent milk served as a covariate.Preplanned con- trasts were used to determine the effects of CNSE (i.e., CON vs. CNSE-2.5 and CNSE-5.0),whereas orthogonal contrasts were used to determine linear and quadratic effects of increasing CNSE level.Data are reported as least squares means and considered significant if P ≤ 0.05 and a tendency if 0.05 < P ≤ 0.10.

Rumen pH, Ammonia N, and VFA
Overall, prepartum rumen pH was unaffected by time or diet (P ≥ 0.19; Table 2), but postpartum rumen pH increased in CNSE cows relative to CON (0.31 units; P < 0.01).Rumen ammonia N concentration tended to decrease in CNSE-5.0 cows compared with CON prepartum (34%; P = 0.08) and tended to be quadratically decreased postpartum, such that in CNSE-2.5 fed cows it was decreased relative to CON and CNSE-5.0(P = 0.07; Table 2).There were no overall treatment differences in prepartum rumen molar proportions of acetate, butyrate, and total VFA (P ≥ 0.25; Table 2); however, CNSE supplementation increased prepartum molar proportions of propionate compared with CON cows (7%; P = 0.05).Postpartum molar proportions of VFA and total VFA were similar between treatments (P ≥ 0.30).

Milk Yield and Composition
Milk yield was progressively increased in all treatments over time (P < 0.01) and tended to be increased in CNSE cows relative to CON (6%; P = 0.06; Figure 1B).Overall, no treatment differences were observed for concentrations of milk fat, protein, lactose, total solids, MUN, or SCC (P ≥ 0.17; Table 4).Milk fat content steadily decreased in all treatments (P < 0.01; temporal pattern not shown); however, on d 3 the decrease was greater in CNSE-fed cows relative to CON cows (11%; P = 0.05) whereas on d 21, milk fat was decreased in CNSE-5.0 cows relative to CON (13%; P = 0.03).Milk protein yield tended to be increased in CNSE cows relative to CON (7%; P = 0.08; Table 4).No overall treatment differences were observed for ECM, FCM, or SCM (P ≥ 0.40); however, as week of lactation progressed ECM, FCM, and SCM increased  Values within a row with differing superscripts denote differences (P < 0.05) between treatments.
Circulating SAA, LBP, Hp, and cortisol were similar between treatments pre and postpartum (P ≥ 0.12; Table 5).Serum amyloid A, LBP, and Hp concentrations were highest on d 3 and then gradually decreased as lactation progressed (P < 0.01; Table 5; temporal pattern not shown).

DISCUSSION
There continues to be growing interest in identifying target molecules that can enhance feed efficiency, livestock performance and system sustainability.Some natural compounds (i.e., plant extracts, essential oils, phenols) have specific biological properties that could benefit multiple aspects of animal agriculture.Due to its unique composition, CNSE displays a wide array of antimicrobial (Kubo et al., 1993), antioxidative (Kubo et al., 2006) and antitumor (Itokawa et al., 1987) activities.Importantly, CNSE's proposed impact in shifting VFA production toward propionate makes it a promising feed additive for use in transition cows.Glucose demands during early lactation are markedly increased due to both the onset of milk synthesis and immune system activation.Thus, the inflamed early postpartum cow has a substantial glucose requirement (Horst et al., 2021) and identifying feed supplements that can safely and sustainably increase propionate production would ostensibly benefit both animal welfare and productivity.
Regardless of dose, feeding CNSE increased rumen pH postpartum which corroborates previous reports (Osmari et al., 2017;Díaz et al., 2018;Maeda et al., 2020).Conversely, other in vivo and in vitro experiments evaluating dietary CNSE in ruminants observed no change (Shinkai et al., 2012;El-Zaiat et al., 2014;Kang et al., 2018;Konda et al., 2019) or even a decrease in rumen pH (Watanabe et al., 2010).Reasons for the inconsistencies are not clear but could include species differences, physiological state, host microbiome, raw versus technical CNSE (extraction method), or in vitro versus in vivo models.Cashew nut shell extract has surfactant action that is thought to physically disrupt  the cell membrane of gram-positive bacteria (Kubo et al., 2003;Oh et al., 2017).Thus, increased rumen pH observed herein is likely a result of the antimicrobial activity of CNSE, more specifically anacardic acid, which inhibits some gram-positive bacteria (e.g., lactic acid producing) which indirectly promotes the growth of lactate utilizing bacteria (Kubo et al., 1993;Watanabe et al., 2010;Osmari et al., 2017).Regardless of the mechanism(s), the CNSE-induced increase in rumen pH could theoretically ameliorate a variety of transition cow maladies related to reduced ruminal pH (Watanabe et al., 2010;Oh et al., 2017).Conversely, while CNSE did not affect postpartum fecal pH, it did slightly decrease prepartum pH (0.12 units), but an adequate explanation for and biological consequence of which are not clear.Concurrent with rumen pH changes, CNSE has been reported (in vitro and in vivo) to alter rumen fermentation end products (Watanabe et al., 2010;Shinkai et al., 2012;Díaz et al., 2018).In the current study, ruminal ammonia N tended to be decreased prepartum and tended to be quadratically decreased postpartum in CNSE-supplemented cows which agrees with results observed in semicontinuous culture conditions (Watanabe et al., 2010).In contrast, Shinkai et al. (2012) reported either a decrease or no change in rumen am- monia N when supplementing CNSE to dry cows with 2 different pellet ingredients.Reduced rumen ammonia N with CNSE supplementation may reflect increased ammonia uptake or the inhibition of proteolytic bacteria (Watanabe et al., 2010).Furthermore, the decrease in ammonia N and BUN could be an indication that ammonia is being sequestered by microbes for proliferation and thus less ammonia leaving the rumen and being converted to urea (Agle et al., 2010).Along with changes in ammonia N, there was also an increase in prepartum rumen propionate with CNSE supplementation.The prepartum increase in the molar proportion of propionate agrees with previous CNSE studies (Watanabe et al., 2010;Shinkai et al., 2012;Oh et al., 2017), but differs from others (Danielsson et al., 2014).Reasons for the inconsistencies are unclear but could be from a lack of a sufficient adaptation period to CNSE, in vitro versus in vivo models, or differences in CNSE dosage.These changes likely reflect a shift in microbial populations (Su et al., 2021), resulting in a redistribution of metabolic hydrogen toward propionate production (Ungerfeld, 2015).Despite the increase in propionate prepartum, no differences were detected postpartum, and this is likely attributed to differences in diet, passage rate, absorption rates and an inadequate number of animals per treatment.The lack of postpartum treatment differences in rumen propionate concentrations are particularly confusing considering the production and metabolic responses.Regardless, VFA concentrations are a snapshot in time and reflect both the production and removal/absorption and thus a more detailed VFA "turnover" evaluation is needed to better understand how CNSE alters fermentation patterns.Regardless, CNSE supplementation appears to alter rumen metabolism, evidenced by decreased ammonia N and increased propionate concentrations.
Prepartum DMI did not differ due to treatment; however, the expected decline in DMI as parturition approached was observed (Hayirli et al., 2002), with most of the decline occurring in the last week of gestation.In contrast to previous reports indicating no change (Coutinho et al., 2014;El-Zaiat et al., 2014;Branco et al., 2015;Osmari et al., 2017;Konda et al., 2019), we observed increased postpartum DMI with CNSE supplementation.The mechanism by which CNSE improved DMI is not clear but may stem from a healthier rumen environment (as indicated by elevated pH and decreased ammonia N), allowing for more effective feed digestion.In addition, milk yield also tended to increase with CNSE, and the improved feed intake was likely due to increased milk synthesis (i.e., increased productivity and energy expenditure controls appetite; Baumgard et al., 2017).This remains to be further investigated as others have reported no changes in milk yield following CNSE supplementation (Coutinho et al., 2014;Branco et al., 2015).
The transition period is characterized by unique homeorhetic metabolic alterations ensuring the mammary gland is supplied with the necessary nutrients for lactation (Bauman and Currie, 1980).In brief, the concurrent   tions allows most extramammary tissues and cell types to utilize fatty acids as fuel and this endocrine and metabolic adjustment spares glucose for the mammary gland, a key reason healthy high-producing cows have increased NEFA (Harrison et al., 1990;Horst et al., 2021).
Essentially all (even seemingly healthy ones) periparturient cows are inflamed, it is only the magnitude of the inflammatory response that differs (Bertoni et al., 2008;Trevisi et al., 2012).Although the temporal pattern (peaking immediately following calving) of acute phase proteins measured were as expected, they were not influenced by CNSE.This is slightly surprising as previous work has demonstrated reduced acute phase protein concentrations with increased glucose supply (McCarthy et al., 2020).Reasons why CNSE cows did not have improved metrics of immune activation are not clear but maybe the increase in propionate and improved rumen pH were not large enough to have a meaningful impact on the inflammatory variables measured.

CONCLUSION
Cashew nut shell extract has a wide range of biological activity that could potentially benefit transition dairy cows.Supplementing CNSE prepartum favorably altered rumen fermentation with increased propionate and decreased rumen ammonia N production.Furthermore, postpartum CNSE supplementation increased DMI, milk production, and rumen pH.The metabolic adjustments indicated that CNSE cows had reduced insulin, decreased glucose and increased NEFA; metabolic adjustments that allowed for increased milk yield.CNSE appears to be an effective dietary strategy to help improve rumen health while increasing productivity in transition dairy cows.
Goetz et al.: CASHEW NUT SHELL EXTRACT AND THE TRANSITION PERIOD Goetz et al.: CASHEW NUT SHELL EXTRACT AND THE TRANSITION PERIOD Goetz et al.: CASHEW NUT SHELL EXTRACT AND THE TRANSITION PERIOD Figure 1.Postpartum (A) DMI and (B) milk yield in transition dairy cows fed either control (CON), 2.5 g/d of cashew nut shell extract (CNSE-2.5),or 5.0 g/d of cashew nut shell extract (CNSE-5.0).Data were analyzed using PROC MIXED and included fixed effects of treatment, day, and their interaction.Only statistically significant comparisons are provided in the figure.
Goetz et al.: CASHEW NUT SHELL EXTRACT AND THE TRANSITION PERIOD Figure 2. Circulating (A) glucose, (B) insulin, (C) BHB, (D) NEFA, (E) BUN, and (F) glucagon in transition dairy cows fed either control (CON), 2.5 g/d of cashew nut shell extract (CNSE-2.5),or 5.0 g/d of cashew nut shell extract (CNSE-5.0).Cows were fed either CON or CNSE-5.0 for 21 ± 12 d prepartum.Following parturition, cows from the CNSE-5.0treatment were divided equally into CNSE-2.5 and CNSE-5.0groups which were fed until 28 DIM.Data were analyzed using PROC MIXED and included fixed effects of treatment, day, and their interaction.Pre and postpartum variables were analyzed seperately.Only statistically significant comparisons are provided in the figure.

Table 2 .
Effects of cashew nut shell extract (CNSE) on rumen pH and volatile fatty acids (VFA) in transition dairy cows

Table 3 .
Effects of cashew nut shell extract (CNSE) on DMI, BW, BCS, EBAL, and fecal pH in transition dairy cows

Table 4 .
Effects of cashew nut shell extract (CNSE) on production parameters in transition dairy cows