High, but variable prevalence of Sarcocystis cruzi infections in farm-raised American bison (Bison bison) beef destined for human consumption

Abstract

Background

Bison (Bison bison) and cattle (Bos taurus) are closely related (can interbreed) and they also share many parasites. Cattle are commonly infected with one or more of the eight named Sarcocystis species: Sarcocystis hirsuta, S. cruzi, S. hominis, S. bovifelis, S. heydorni, S. bovini, S. sigmoideus and S. rommeli. Among these, the full life-cycle is known only for S. cruzi. Sarcocystis cruzi (transmitted via canids) is recognized as the most pathogenic Sarcocystis species, causing abortion, low milk yield and poor body growth. It has been experimentally cross-transmitted from cattle to bison and vice versa.

Methods

We tested 200 bison tongues from three commercial sources (farms) (Nebraska #141; South Dakota #36; New Jersey and Pennsylvania #23). Frozen tongues were purchased and examined for Sarcocystis infection using light microscopy, histology and quantitative PCR (qPCR) targeting 18S ribosomal DNA (18S rRNA) of S. cruzi. Lesions associated with degenerating sarcocysts were studied. The intensity of Sarcocystis infection in histological sections was quantitated.

Results

Sarcocystis cruzi-like infections were detected in 129 of 141 (91.5%) tongues from Nebraska, 36 of 36 (100%) tongues from South Dakota and two of 23 (8.6%) tongues from New Jersey and Pennsylvania. Sarcocysts were detected in histological sections stained with hematoxylin and eosin in 167 of 200 samples. Light microscopy examination revealed that the sarcocysts had thin walls (< 1 µm thick) and appeared to be S. cruzi. However, in two samples, sarcocysts had thicker walls measuring up to 2.3 µm wide and 154 µm long and the sarcocyst wall was not striated; these two samples could not be characterized further. In three tongues, degenerating sarcocysts were recognized; two of these were associated with thick-walled sarcocysts. Molecularly, S. cruzi from bison was identical to that in cattle.

Conclusions

In the present study of bison tongues, S. cruzi was the only species identified in bison using both molecular and morphological methods. An unidentified species of Sarcocystis found in two bison samples needs further study.

Graphical abstract

Background

American bison (Bison bison) is the biggest mammal in North America, and it was named the first national mammal of the USA in 2016 through the National Bison Legacy Act [1]. Historically, the American bison sustained the economy and societal well-being of indigenous peoples (Native Americans) in their native range. During the 1800s, millions of bison were slaughtered by European and American settlers, decimating the primary source of food and clothes as well as the culture of indigenous peoples via “The Great Slaughter” [2]. Bison hunting became a popular activity, and no protective measures were put in place; consequently, the animal was driven to near extinction [3].

According to the IUCN (International Union for Conservation of Nature and Natural Resources) Red List Assessment of 2017 [4], the American bison is listed as Near Threatened (NT). This listing is due to the small size of the wild bison population that is managed exclusively for conservation purposes, with most animals being managed for commercial purposes. The Red List shows the likelihood of a species becoming extinct in the near future; as such, the NT classification warns that American bison will soon likely qualify for the threatened category [4].

According to the National Agricultural Statistics Service (NASS) Census of Agriculture of 2022 published by the United States Department of Agriculture (USDA), the production and consumption of bison meat has increased within the past 5 years in the USA. The 2017 census shows that 52,937 bison on 1049 farms were managed for commercial purposes in 2017, increasing to 60,804 animals on 1201 farms in 2022, an increase of approximately 14.5%.

Bison meat has more protein and less fat than beef (USDA Agricultural Research Service: FoodData Central Food Data; 2018 [https://fdc.nal.usda.gov/fdc-app.html#/food-details/168608/nutrients]), and many consumers prefer the taste. American bison (Bison bison) and cattle (Bos taurus) are closely related (can interbreed) and share many common parasites. Cattle are commonly infected with one or more of the eight named Sarcocystis species: Sarcocystis hirsuta, S. cruzi, S. hominis, S. bovifelis, S. heydorni, S. bovini, S. sigmoideus, S. rommeli. Two Sarcocystis species, S. hominis and S. heydorni, are zoonotic [5, 6]. Of these, S. cruzi and S. heydorni form thin-walled (< 1 μm thick) sarcocysts whereas the remaining species have thick-walled (> 3 μm thick) sarcocysts. Among these, the full life-cycle is known for only S. cruzi (transmitted via canids); it is recognized as the most pathogenic species, causing abortion, low milk yield and poor body growth in cattle [7]. Thick-walled sarcocysts have been associated with abattoir condemnation due to infected meat [8].

Nothing is known of the prevalence of Sarcocystis spp. in the American bison raised for human consumption. Four decades ago, thin-walled sarcocysts were reported in the muscles of two of 15 (13%) feral bison from the National Bison Range, near Missoula, Montana [9]. Thin-walled sarcocysts were also found in sections of muscles of a bison that was found dead near Livingston, Montana; a laboratory-raised coyote excreted S. cruzi sporocysts in its feces after feeding on the muscles of this bison [10]. In another experiment, thin-walled sarcocysts were detected in sections of the esophagus and heart of a hunter-killed bison from Montana; 11 days after feeding on the infected parts, a laboratory-raised coyote (#13) excreted S. cruzi-like sporocysts [11]. In a subsequent study, S. cruzi from cattle (Bos taurus) was transmitted successfully to laboratory-raised bison and S. cruzi from bison was transmitted to cattle [12]. In this study, two 1-week-old bison were fed S. cruzi sporocysts from the feces of a laboratory-raised coyote that had been fed S. cruzi-infected meat of a cow (Bos taurus) from Montana [12]. One bison was fed 10 million sporocysts and developed acute sarcocystosis; the second bison was fed 100,000 sporocysts and developed muscle sarcocysts. Additionally, three cattle calves inoculated with 100,000 or 700,000 sporocysts from coyote #13 developed clinical Sarcocystis infections [13]. Two calves had diarrhea on 19–25 days post-inoculation (dpi); one of the two calves fed 700, 000 sporocysts became moribund on 28 dpi. Although limited, data from this study indicated that the bison strain of S. cruzi was more pathogenic than the cattle strain of S. cruzi [13] and that clinical signs might vary with host species. The diarrhea observed in one bison and three experimentally infected calves was unusual because it was normally not observed in cattle fed the bovine isolate of S. cruzi [13].

Here, we report the first survey of Sarcocystis infections in commercially raised bison in the USA.

Methods

Naturally infected bison

Two hundred frozen bison tongues were purchased from three farms located in different states (Nebraska #141, South Dakota #36, New Jersey and Pennsylvania #23) between October 2023 and March 2024 (Table 1). Bison from Nebraska and South Dakota were raised on open pasture and were 100% grass-fed; bison from New Jersey and Pennsylvania (1 owner) were raised on semi-open range and fed primarily on grass, enriched with a mix of minerals and grains as a supplement.

Table 1 Positivity of bison tongue samples for Sarcocystis spp. separated by source and methodology

Testing for Sarcocystis infection

All samples were examined microscopically using several procedures.

Microscopic examination of muscle by the compression method

Small pieces (approx. 2 × 2 cm) of each tongue were compressed between a glass slide and coverslip and examined microscopically for sarcocysts. A sample was considered to be negative for Sarcocystis infection when no sarcocysts was detected in any of 30 preparations.

Histological examination

Four or more pieces (enough to fill a paraffin 4 × 2-cm cassette) of each tongue were fixed in 10% buffered formalin and processed for histology using paraffin embedding. Three 5-μm-thick sections were stained with hematoxylin and eosin (HE) and examined microscopically.

Tongues stored in freezer

Approximately 50 g of each tongue was stored at - 80 °C for future study.

Photographs

Cysts were photographed using a digital DP73 camera (Olympus Optical Ltd., Tokyo, Japan) attached to an Olympus AX 70 microscope (Olympus Optical Ltd.).

DNA isolation and amplification

Each tongue muscle (50 g) was homogenized with 250 ml of 0.85% aqueous NaCl (saline) in a blender and passed through a cheesecloth; part of the filtrate was collected in a 50-m; tube and centrifuged for 10 min at 400 g [5]. The supernatant was discarded and the pellet resuspended with 2.5 ml of 0.85% NaCl (saline); a portion of each suspended pellet was processed for DNA extraction. For this, part of the homogenization product was transferred to a 2-ml microtube and centrifuged; the supernatant was discarded and the pellet was stored at - 80 °C. Part of the homogenization product was transferred to a 2-ml microtube and centrifuged again in a microtube centrifuge for 10 min at 100 × g. The supernatant was discarded, and the pellet was extracted using the QIAGEN DNeasy® Blood and Tissue Kit (QIAGEN, Hilden, Germany) utilizing a modification of the manufacturer’s protocol (the lysis step was carried out at 56 °C overnight in a dry bath incubator on an Eppendorf® ThermoMixer F1.5 [Eppendorf, Hamburg, Germany] and DNA samples were eluted in 100 μl of QIAGEN® Nuclease Free Water [QIAGEN] during last step). Extracted DNA samples were quantified and assessed using the NanoDrop® 1000 full-spectrum spectrophotometer (V3.8; Thermo Fisher Scientific, Waltham, MA, USA) and kept at - 20 °C until further analysis.

Quantitative analysis

We performed quantitative real-time PCR (qPCR) targeting the 18S ribosomal RNA (18S rRNA) subunit of S. cruzi in the first four samples, employing four dilutions (10, 20, 40 and 60 ng) compared to non-diluted samples. This validated the use of stocks diluted to 10 ng. Those few samples estimated by spectrophotometry (NanoDrop® 1000 system) to contain < 10 ng/μl were concentrated using a SpeedVac Vacuum Concentrator (Thermo Fisher Scientific).

The qPCR tests targeted the 18S rRNA gene marker of S. cruzi using qPCR probes. Forward and reverse primers and a probe were designed using Primer3 [14] by incorporating the S. cruzi reference sequence (GenBank accession number: KT901173). The primers and probe designed were forward primer q_S. cruzi (5’-ATA GTC ATA TCA GAT GAA AAT CTA C-3’), reverse primer q_S. cruzi (5’–CAG CCA TAT AAA ATG ACC ATA-3’) and probe q_S. cruzi (5’–ATC TGT TAA CAG CAG GTG GTG TAA AAA AGG T/3BHQ–3’). The PCR reactions were carried out using the Applied Biosystems™ QuantStudio 7 Flex Real-Time PCR System 7, and 96-well plates were used in addition to the IDT PrimeTime Gene Expression Master Mix Protocol (Integrated DNA Technologies, Inc., Coralville, IA, USA). The PCR tests were carried out in a total reaction volume of 20 μl, containing 10 μl PrimeTime MasterMix 2×, 0.5 μl forward primer, 0.5 μl reverse primer, 0.3 μl probe, 6.7 μl molecular biology grade water and 2 μl DNA) (Additional file: Table S1).

Negative samples according to the qPCR results were submitted to a second run using 60 ng (6 μl DNA) and the same qPCR conditions, with changes only in the quantity of DNA and molecular biology grade water. Accordingly, PCR tests were carried out in a total reaction volume of 20 μl, containing 10 μl PrimeTime MasterMix 2×, 0.5 μl forward primer, 0.5 μl reverse primer, 0.3 μl probe, 2.7 μl molecular biology grade water and 6 μl DNA.

Polymerase chain reaction

Based on the cycle threshold (CT) values and qPCR results and comparing these with the light microscopy (LM) results, we selected nine bison tongue samples (sample numbers: 8, 94, 108, 148, 165, 181, 191, 193 and 199) for conventional PCR testing using specific primers targeting the 18S, 28S (28S RNA) and cytochrome c oxidase subunit 1 (cox1) regions of Sarcocystis species.

The 25-µl PCR mix reaction volume consisted of 2 μl of DNA template, 12.5 μl of Platinum Hot Start PCR Master Mix (Invitrogen, Thermo Fisher Scientific), 1 μl of 10 pmol/μl of each primer (Integrated DNA Technologies, Inc.) (Table 2) and 8.5 μl of molecular grade water. The PCR cycling parameters consisted of an initial denaturation at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 30 s and elongation at 68 °C for 20 min, with a final elongation of the incubated products at 68 °C for 5 min. In the case of negative results or low amplifications, this first PCR round was followed by a second round of PCR using the amplified products. All PCR products were analyzed by electrophoresis in a 2% agarose gel, and size was estimated by comparison with the 100-bp Plus DNA Ladder followed by DNA cleaning [15]. The final purified PCR products were sent for sequencing to the DNA sequencing Unit of Psomagen (Rockville, MD, USA), a commercial sequencing company for direct sequencing on an ABI 3500xl Genetic Analyzer (Applied Biosystems™, Thermo Fisher Scientific) using the primer sets specially designed for this study to obtain both strand reads. The sequences were deposited in the GenBank database, and the accession numbers were obtained (Table 2).

Table 2 Accession numbers and PCR primers used for the amplification and sequencing of different gene markers for Sarcocystis cruzi

Phylogenetic reconstructions

Amplifications were attempted for all nine bison samples using 18S rRNA and mitochondrial cox1 genes. We also succeeded in amplifying 28S rRNA from two bison samples (#94 and #148). The resulting sequences were trimmed and aligned by MAFT alignment (using the L-INS-i algorithm and 1.53 gap open penalty) in Geneious Prime, followed by ClustalW multiple alignment in MEGAX [16]. Any ambiguous bases were clarified by checking the respective chromatograms. The consensus sequences were then analyzed using a standard online Basic Local Alignment Search Tool (BLAST) (https://blast.ncbi.nlm.nih.gov/Blast.cgi) [17] against the genetic dataset available at the National Center for Biotechnology Information (NCBI).

The web server GUIDANCE2 [18] was used to align and remove ambiguously aligned positions. Specifically, the sequences were aligned with the MAFFT algorithm under the options Max-Iterate: 1000 and Pairwise Alignment Method: –local pair. Positions with a score < 0.93 were removed. Phylogenetic relationships were reconstructed under the maximum likelihood (ML) criterion. ML analyses were performed with the program IQ-TREE version 1.6.12 [19]. The analyses were run with the options –m MFP –b 1000. All codon positions were used. The models selected based on the Bayesian information criterion (BIC) criterion were JC + G + I and K2 + G + I for 18S and cox1, respectively.

Results

Microscopy

Microscopic examination of compression preparations detected the presence of sarcocysts in 157 of 200 (78.5%) samples (Table 1); following staining with HE, sarcocysts were detected in 167 of 200 the histological sections (Table 1). The sarcocysts had thin walls (< 1 um thick) and appeared to be S. cruzi (Fig. 1a, c). However, in two bison samples (sample #94 and #108), sarcocysts had thicker walls measuring up to 2.3 μm wide and 154 μm long (Fig. 1b, d).

Fig. 1

Comparison of thin-walled and thick-walled sarcocysts detected in histological sections (hematoxylin and eosin staining) of the tongue of bison #94 from Nebraska. Arrowheads point to wall thickness. A Thin-walled Sarcocystis cruzi sarcocyst at low resolution. B Unidentified sarcocyst with relatively thicker sarcocyst wall at low resolution. C Sarcocystis cruzi sarcocyst at higher resolution; note thin septa (se) and robust bradyzoites (br); D Unidentified sarcocyst with relatively thicker cyst wall at higher resolution; note robust bradyzoites (br)

The number of sarcocysts per histologically stained slide ranged from 0 to 260 (Table 3). No sarcocysts were seen in 33 tongues, and 50 tongues had fewer than six cysts.

Table 3 Intensity of sarcocysts in 2 × 4-cm sections of bison tongues on hematoxylin and eosin-stained slides

Mononuclear cell infiltrations were detected in 22 tongues (Fig. 2). Degenerating sarcocysts were recognizable in three tongues, of which two (sample #94, #108) were associated with thick-walled sarcocysts. In some of the degenerating sarcocysts, the cyst wall was apparent, whereas in others no cyst wall was recognizable.

Fig. 2

Myositis in the tongue of bison #94 from Nebraska. Hematoxylin and eosin staining. A A partly degenerating sarcocyst with mononuclear cell infiltration; note thick sarcocyst wall (arrow). B A focus of mononuclear infiltration around and inside a degenerating sarcocyst; the sarcocyst wall (arrow) is still visible

Molecular analysis

Quantitative PCR using 18S rRNA for S. cruzi revealed 95.5% positivity in all bison tongue samples (140/141 samples from Nebraska, 36/36 from South Dakota and 15/23 from New Jersey and Pennsylvania) (Table 1).

Based on the results from all methods, we determined that the prevalence of Sarcocystis spp. in bison tongue samples was 95.5% (191/200) (Table 1). Of the 141 tongue samples from the Nebraska bison, Sarcocystis infections were detected in 140. Sample #8 from Nebraska, assessed to be positive for Sarcocystis spp. based on histology showing a thin-walled sarcocyst, tested negative for S. cruzi by qPCR, indicating the possible presence of a distinct species of Sarcocystis. Samples #94 and #108 were positive for S. cruzi and, both thin-walled and thick-walled cysts were detected, suggesting a co-infection with a different Sarcocystis spp. whose identity could be determined from the methods employed in the study.

Of the nine samples studied using PCR, sample #181 did not amplify after many rounds of amplifications. Samples from South Dakota were all positive for S. cruzi based on the results of each of the three diagnostic methods, while infections were less prevalent in samples from New Jersey and Pennsylvania. Additionally, positive samples from New Jersey and Pennsylvania showed a higher CT value, indicating a low infectivity, which correlates with the results (Additional file 1: Table S1).

Sanger sequencing of the 18S and cox1 genes corresponded well with the qPCR results and showed 99% identity to the published S. cruzi sequences. Analyses of the 18S and cox1 genes of Sarcocystis also revealed only a few differences between the obtained sequences and those isolated from wood bison (Bison bison athabascae) [20] and European bison (Bison bonasus) [21] (Additional file 2: Figure S1, Additional file 3: Figure S2).

For constructing consensus trees, we first blasted the edited sequences against the NCBI nucleotide database (https://www.ncbi.nlm.nih.gov/nucleotide/) and then downloaded the most closely related Sarcocystis species. All of the haplotypes used for the tree showed > 95% homogeneity with the sequence we obtained from bison samples #94 and #108. The analysis was performed against the existing partial sequences previously obtained from different isolates of S. cruzi and closely related Sarcocystis spp. Isolates of S. bovifelis were used as outgroups. The final dataset included 16 taxa and 1465 positions for 18S and 16 taxa and 992 positions for cox1.

During the study, apart from interspecific variations, which ranged from 95% to 99% between different isolates of S. cruzi obtained from the different locations and hosts, intraspecific variations (> 99%) were also found between the sequences of different bison tongue samples isolated during the study.

Discussion

Sarcocystis cruzi is one of the most prevalent microbial infections in cattle worldwide. In some surveys, nearly 100% of cattle have been found to be infected, as previously summarized [5, 8]. Several factors explain the highly efficient transmission of this infection: (1) several canids (dogs, coyote, foxes, wolves) can excrete millions of Sarcocystis sporocysts in feces; (2) sporocysts are sporulated and excreted in a fully infective stage; (3) sporocysts are environmentally resistant; and (4) canids can excrete sporocysts for months after a single infective meal [5]. The number of sarcocysts in cattle muscle varies, but to our knowledge, no published record exists. Although S. cruzi can be transplanted transplacentally in cattle, this transmission route is rare, and infected fetuses probably die in utero [5].

From a transmission perspective, the ingestion of only a few sarcocysts by a canid can lead to the excretion of many sporocysts. It should be noted that a bison can weigh up to 1000 kg, and that our estimates of sarcocyst numbers were derived from small sections of muscle on a histology slide (approx. 10 mg of tissue). In this study, we found up to 264 sarcocysts on one histology slide (Table 3).

The very high prevalence of Sarcocystis infections in the South Dakota and Nebraska samples compared with a lower prevalence in the New Jersey and Pennsylvania samples is probably related to management. The semi-open range system, its enriched diet and/or reduced contact with carnivores in the East of the USA may have protected the Eastern herds sampled here. Smaller populations of coyotes and the presence of Eastern Coyotes (Canis latrans var.), which are a canine hybrid of coyotes and wolves, might interfere with the passage of the parasite in the Eastern USA. Recently, wolves in Minnesota were found to be the definitive host for S. cruzi [22]. Wolves are rare in the Eastern USA.

In the phylogenetic tree we constructed based on partial 18S rRNA and cox1 genes (Fig. 3), the sequences obtained during the study formed a strongly supported cluster with sequences attributed to S. cruzi (for 18S: accession no. KP640133, LC171827; for cox1: acccession no. MT796944, MW490605-6, KC209599). In Fig. 3, The “S. cruzi cluster” has been highlighted with brackets in the tree. Sarcocystis gjerdei also formed the extended cluster with S. cruzi. The other Sarcocystis species forming the basal clade are S. morae, S. alces, S. tenella and S. hircicanis.

Fig. 3

Phylogenetic tree reconstructed by maximum likelihood showing the phylogenetic position of Sarcocystis cruzi identified in Bison bison tongue samples using the gene markers 18S ribosomal RNA gene (18S rRNA) (A) and cytochrome c oxidase subunit 1 (cox1) gene (B). The “S. cruzi cluster” is highlighted with orange brackets in both the trees. Species names shown in bold are the those detected during the present study

The tree agrees with previous studies on S. cruzi wherein the molecular analysis and phylogenetic position, prevalence and occurrence have been discussed in the cattle and bison (European and wood bison) [20, 21, 23, 24].

Interestingly, there was a slight deviation in the clustering of these Sarcocystis species infecting closely related hosts, which is evident from the different range of consensus and branch length, even in the closely related species. Despite these variations, the relationships between species within each clade remained relatively consistent, irrespective of the branching order of the major clades involving different isolates of S. cruzi, showing the close evolutionary relationship between the species.

Unidentified sarcocysts

Among the eight species of Sarcocystis in cattle, S. cruzi and S. heydorni have sarcocysts with thin walls (< 1 µm) whereas the cyst walls of the other six species (S. hominis, S. bovifelis, S. bovini, S. sigmoideus, S. hirsuta, and S. rommeli) are thicker than 3 µm and also appear striated under the light microscope [8]. We did not detect any thick-walled sarcocysts in the present study. Rather, in two bison samples (#94 and #108) we detected cysts with walls around 2 µm thick that were not striated (Fig. 1b, d), which we were unable to identify. These latter cysts were present together with thin-walled S. cruzi sarcocysts in these samples.

Conclusions

More research is needed to understand whether species other than S. cruzi, present in cattle, also occur in bison. The high prevalence of S. cruzi in the American bison tongue samples (95.5%) studied indicates the presence of contaminated fecal material from carnivores in the open range of cattle within the same area. The results of the present study show that range management may reduce livestock to exposure.