**5. Detection methods for surveillance of prion diseases**

As international trade continues to increase, in order to successfully enforce food and feed safety regulations and avoid economic loss, many nonclinical methods for the detection of the contaminated products have been developed. Overall, these detection methods can be classified into two types: non-immunochemical detection (i.e., chromatography, spectroscopy, and polymerase chain reaction (PCR)) and immunochemical detection methods.

Among the non-immunochemical detection methods, chromatography is typically used to separate molecules based on differential absorption and elution, which involves the flow of a fluid carrier over an immobile absorbing phase. Using chromatography, brain-specific fatty acids such as lignoceric acid and the *cis*/*trans* isomers of nervonic acid[122, 123] had been used as the analyte for the detection of the presence of CNS tissue in meat products through their characteristic component patterns. Spectroscopic techniques are based on the unique absorb‐ ance profiles of the sample components at specific wavelengths of the electromagnetic spectrum.

Near-infrared spectroscopy[124] and attenuated total reflectance Fourier transform infrared spectroscopy[125] have both been used to analyze bovine spinal cord in ground beef. Some CNS fatty acids have also been used as markers in gas chromatography–mass spectrometry (GC– MS).[126, 127] However, these methods require expensive instruments and reagents and highly trained staff; their sophisticated nature and the laborious sample preparation involved have limited the utility of these instrumental methods for routine analysis. Alternatively, its ability to detect GFAP mRNA makes PCR a sensitive technique for the detection of bovine CNS tissue in meat products.[128-130] However, although PCR can achieve a very low detection limit, it suffers from serious drawbacks, once again requiring expensive instrumentation and re‐ agents and expert technicians. PCR methods are also prone to contamination.[130-132]

Immunochemical detection methods based on the specific immunoreactions between an antibody and its target antigen have been extensively employed for the detection of CNS contamination in meat (Table 1), generally in the form of an enzyme-linked immunosorbent assay (ELISA),[76, 128, 133-143] immunoblotting[137-140, 144-149] or immunohistochemistry. [145, 146, 150-152] There are a number of advantages associated with using an immunoassay. For example, no serious instrumentation is required; it is easy to operate and it employs minimal reagents. In the case of ELISA, it not only has a large-scale screening and field test capability, but is also a rapid, specific, and sensitive technique. Several markers have been explored to detect bovine CNS tissue in meat products with immunoassays, either using monoclonal (mAb) or polyclonal (pAb) antibodies.

#### **5.1. CNS markers for non-immunochemical detection methods**

#### *5.1.1. Cholesterol*

Cholesterol (MW: 386.7 g/mol) is a sterol component of cell membranes, hormones, and bile acids. It has been reported that 85 g of brain tissue may contain as much as 2,640 mg of cholesterol, while the same amount of a meat sample will only contain up to 85 mg cholesterol. [153] Although cholesterol is not specific to CNS, it can serve as a useful marker for a screening test due to the low-cost and rapid procedures available for measuring it. Lucker *et al*.[146] pioneered the use of cholesterol to analyze CNS tissue in meat products, using it to detect the cholesterol content of 402 heat-treated meat products from different food markets in Germany. Sixteen field samples (4%) were identified as being possibly contaminated with CNS tissue using this procedure, but the presence of CNS tissue was confirmed by immunoblotting for both neuron-specific enolase (NSE) and GFAP in just 7 of these 16 meat products. This suggests that cholesterol is not a reliable marker for CNS tissue due to its non-tissue specificity. It can, however, be used to roughly screen for CNS residue in meat products even though the accuracy is much lower than that of markers such as GFAP and NSE.

#### *5.1.2. Nervonic acid*

Nervonic acid (15-tetracosenoic acid, MW: 366.6 g/mol), a monounsaturated long chain fatty acid (C 24:1), is enriched in nervous tissue and is mainly present in the sphingolipids of the brain. As it is seldom found in non-CNS tissue, nervonic acid has been used as a marker indicating the presence of CNS for the detection of SRM in meat products by chromatographic techniques. Based on the different ratios of *cis*/*trans* isomers of nervonic acid in different animal species, nervonic acid has also been used to differentiate CNS tissue from various animal species in meat products using an on-line liquid chromatography–gas chromatography (LC– GC) method developed by Barcarolo *et al*.[123] The authors reported that the ratio of the *cis*/ *trans* isomers can be used to provide a rough estimate of the age of the animal but may not be an accurate method for the detection of CNS adulterant. Biedermann *et al*.[154] analyzed the concentrations of fatty acids typical in bovine CNS, including docosahexaenoic acid (C22:6), lignoceric acid (C24), nervonic acid, and cerebronic acid (C24OH), using GC–MS to determine the CNS content The detection limit for CNS using this GC–MS method was 0.01%. In 2003, Agazzi *et al*.[122] reported an enhanced GC method for the detection of isomers of nervonic acid in meat products, reporting that the ratio of the *cis*/*trans* isomers of nervonic acid not only exhibit species specificity but also show a significant difference between cow and calf, and between pig and piglet.

Biedermann *et al*.[127] went on to study the structural characterization of nervonic acid as a marker for SRM using GC–MS and found that the ratio of *cis*/*trans* isomers used previously should be replaced by the ratio of positional isomers, i.e., the ratio of ω9/ω7-nervonic acid (15c-C24:1/17c-C24:1). Overall, although nervonic acid may be a specific and stable marker, the GC– MS-based CNS method is costly, requires a lengthy analysis time, and the data interpretation is highly complex. In addition, the lack of species and age specificity of nervonic acid for the detection of CNS disqualifies the GC–MS-based method as an effective method.

#### *5.1.3. Glial Fibrillary Acidic Protein (GFAP) mRNA*

In the CNS, GFAP (MW: about 50 kDa) is the major component of the filaments found in the astrocytes that support the functions of nerve cells.[155] In the peripheral nervous system (PNS), GFAP is mainly expressed by Schwann cells. The PCR method has been used in several studies to detect GFAP mRNA as a potential marker of CNS tissue contamination in meat products.

Seyboldt *et al*.[130] developed a reverse transcription-PCR (RT-PCR) assay coupled with restriction fragment length polymorphism for the detection of GFAP mRNA from bovine CNS tissue. Although this approach suffered from cross-reactions with unheated heart and skeletal muscle tissues, heat treatment (70 °C for 20 min) prior to RNA extraction was found to reduce the cross-reaction with a detection limit of 0.5% (g/g) heated bovine brain in bovine minced meat. In another study[143], a similar method was able to detect 0.25% (g/g) bovine CNS tissue in pork liver sausage (after a heat treatment of 80 °C for 80 min) that had been stored up to 28 days at 4 °C, although native peripheral nerves showed positive results.

Abdulmawjood and co-workers[129, 131, 156] developed a quantitative real-time RT-PCR to detect GFAP mRNA. Their system was capable of detecting down to 0.1% of CNS tissue in strongly heated (120 °C for 15 min) or medium-heated (80 °C for 90 min) meat samples (50% pork and 50% beef), and 0.2% of CNS tissue in pork liver sausages (80 °C for 1 h). However, low levels of GFAP could still be detected in peripheral nerves (sciatic and axillary nerves) and non-neuronal tissues (parotid gland, pancreas and adrenal gland).

#### **5.2. CNS Markers for immunochemical detection**

#### *5.2.1. GFAP*

The immunochemical differences between CNS-GFAP and PNS-GFAP are well known [157], and GFAP was first exploited to indicate the presence of brain or spinal cord tissues in meat as early as 1999 [134]. It has since been utilized as a marker for the detection of CNS tissue in meat products in a number of immunoassay systems [74, 76, 133, 141, 158]. Schmidt *et al*. were the first to develop a colorimetric sandwich ELISA in 1999 [134] and went on to develop an improved fluorescent sandwich ELISA in 2001 [135]. Although this fluorescent method was capable of detecting the presence of as little as 0.05% (g/g) bovine brain and spinal cord in beef, once again this assay suffers from a cross-reaction with PNS-GFAP, which is present in sciatic nerve and cervical ganglion tissue and is not considered an SRM [135]. These results suggest that the anti-GFAP pAb used in their assay was not actually specific to the CNS tissue.

GFAP has also been used as a CNS marker in a number of immunoblotting techniques [137, 145-147]. In a study by Lucker *et al.* [146], the immunoresponse of heat-treated samples was reduced significantly, however, and the sensitivity of the assay was inadequate when inten‐ sively heat-treated meat products were analyzed. Several other studies have also shown that the ability to detect GFAP is diminished when meat samples are subjected to severe heat treatment [133, 135].

#### *5.2.2. Neuron-Specific Enolase (NSE)*

NSE (MW: 48 kDa) is an enzyme that is found in both neurons and neuroendocrine cells. The immunological CNS specificity of γγ-enolase has been used in clinical research for a long time [159], and in 1999 Lucker *et al*. were the first to adapt this clinical method to permit the use of NSE as a marker for the detection of CNS tissue in meat products using immunoblotting [148]. The detection limit of this assay was 1% (g/g) spiked brain in cooked (100 °C for 1 h) sausages of the emulsion type. They suggested that the sensitivity of the assay could be dramatically improved by removing the sample's fat content (30 to 40%) using Soxhlet extraction, although this is exceptionally time-consuming, requiring 8 h to extract the fat.

Several commercial immunochemical test kits have been developed [160]. Hughson *et al*. [140] compared the performance of two commercial kits, the RIDASCREEN GFAP kit and the Brainostic NSE kit (Table 3), for the detection of CNS tissue in meat products; these kits were based on the methods proposed by Schmidt *et al*. [134] and Lucker *et al*. [148], respectively. Hughson *et al*. [140] reported that levels down to 0.1% (g/g) of ovine spinal cord tissues could be detected in raw lamb meat using the GFAP kit, while 1.0% (g/g) was detected by the NSE kit. The detection limits of the GFAP and NSE methods for cooked samples (80 °C for 40 min) were 0.2% (g/g) bovine brain in pork and 0.1% (g/g) bovine spinal tissue in pork, respectively. They noted that the GFAP assay was easier to use and to interpret the data than the NSE assay. In a similar study by Agazzi *et al*. [138], the sensitivity of the NSE test kit was reported as being 0.5% (g/g) CNS tissue in raw and medium-heated (80 °C for 20 min) meat samples, rising to 2.0% (g/g) in strongly heated (120 °C for 20 min) meat materials. Another study [146] also found that the NSE immunoreactivity in samples of pure brain and brain muscle dropped signifi‐


cantly with time at 100 °C, almost disappearing after a 2 h heat treatment. The properties of these commercial kits are summarized in Table 3. In the CNS, myelin proteolipid protein (MPP, MW: 30 kDa) makes up approximately 50% of the myelin proteins [161]. This protein can be readily extracted from brain white matter using organic solvents [162]. Although MPP is the most abundant myelin protein in the CNS, mRNA for PLP is also expressed in the PNS and small amounts of MPP are synthesized [162]. This protein has also been used as a marker for the detection of CNS tissue in meat products [163]. The detection limit for CNS in raw meat can be as low as 0.025% using Western blot [164], and 0.01% bovine brain in raw minced beef using dot blot [165]. However, the antibody used in these studies was not species-specific and was

**Table 3.** Commercial immunoassay kits for the detection of prohibited CNS and ruminant tissue contaminants in meat and feed products

**5.2.3.** 

**Myelin proteolipid protein**

#### *5.2.3. Myelin proteolipid protein*

In the CNS, myelin proteolipid protein (MPP, MW: 30 kDa) makes up approximately 50% of the myelin proteins [161]. This protein can be readily extracted from brain white matter using organic solvents [162]. Although MPP is the most abundant myelin protein in the CNS, mRNA for PLP is also expressed in the PNS and small amounts of MPP are synthesized [162]. This protein has also been used as a marker for the detection of CNS tissue in meat products [163]. The detection limit for CNS in raw meat can be as low as 0.025% using Western blot [164], and 0.01% bovine brain in raw minced beef using dot blot [165]. However, the antibody used in these studies was not species-specific and was unable to differentiate between mammalian species. Another problem is that MPP is not thermostable; the reaction signal decreased significantly when the heating time was increased to 95 °C. MPP was not detectable after 3 h of heat treatment at this temperature using dot blot [164].

#### *5.2.4. Myelin Basic Protein (MBP)*

MBP is located in CNS on the cytoplasmic face of myelin membranes in the white matter corresponding to the major dense line. MBP makes up about 30% of the total protein in the mammalian CNS myelin. It is the only structural protein found so far that is essential for the formation of CNS myelin [166]. As a water-soluble protein, MBP can be extracted with either acid or salt solutions [162]. There are four isoforms of MBP, each with a different molecular weight: 21 kDa, 18.5 kDa, 17 kDa and 14 kDa [162]. Different animal species have different isoforms. In particular, 18.5-kDa MBP, the major isoform protein in the CNS, is highly unfolded with essentially no tertiary structure in solution and the amino acid sequence of this major basic protein is similar in a number of animal species [162]. The sequence identity of the 18.5 kDa MBP in bovines and pigs is 93% homologous [167]. It has been reported that bovine 18.5 kDa MBP is not only very stable in solution at neutral pH (pH 7.2), but also highly thermostable [168]. Its immunoreactivity has been shown to remain clearly detectable even after severe heat treatment at 133 °C for 30 min [168].

MBP has been used as a marker for the detection of CNS contamination using immunoassays in several studies.[136, 143, 145, 147, 151] Levels as low as 1% (g/g) of bovine brain have been detected in a heated luncheon meat-type product (115 °C for 1 h) using a pAb in immunohis‐ tochemistry[151] and a mAb in immunoblotting.[147] In another study,[136] a detection limit of 10% was recorded for CNS tissue using a mAb in indirect competitive ELISA. Using an improved MBP extraction method,[143] compared with the detection limit reported using the same antibody and the same ELISA method,[136] the detection limit for brain tissue could be enhanced from 10 to 0.05% (~200 times). This improved extraction method will provide a useful analytical basis for the subsequent development of a specific immunoassay for the detection of bovine CNT in processed meat and feed products.

#### **5.3. Detection methods for the control of prion diseases in feed products**

To safeguard livestock, a number of non-immunochemical and immunochemical detection methods have been developed to detect the presence of banned animal materials in feeds. A high-performance liquid chromatography method has been developed that utilizes dipeptide carnosine and related dipeptides as the markers,[169] and several other studies have used nearinfrared reflectance spectroscopy (NIRS) to predict the amino acid content in animal feeds. [170-172] PCR has also been used for the detection of animal tissue in feeds. For example, Gao *et al.*[173] amplified a highly conserved eukaryotic DNA region of the 18S ribosomal gene using multiplex PCR (MPCR) and were able to detect levels as low as 0.02% bovine MBM in feedstuffs. Bellagamba *et al.*[174] detected 0.25% ruminant or pig adulterants in fish meal by MPCR.

A number of immunoassays have been developed to detect prohibited ruminant animal materials in feed. In our laboratory, a panel of species-specific mAbs have been produced for the detection of animal tissues in both food and feed products. [175-177] A skeletal muscle protein, troponin I (sTnI, 24 kDa), was first identified as a thermal-stable species marker protein; sTnI maintains its solubility and antigenicity even after undergoing severe heat treatment (126 °C for 120 min).[177] Given its even distribution in skeletal muscles, sTnI appears to be an ideal marker, allowing antibodies to be developed for the detection of animal tissues in severely heat-treated samples such as MBM. sTnI-specific mAbs have demonstrated their usefulness in an indirect non-competitive ELISA for the detection of rendered muscle tissues in animal feedstuffs, with detection limits of the mammalian and ruminant assays being between 0.3 and 2%.[175] In 2004, a sandwich ELISA was reported for the detection of ruminant proteins in feedstuffs.[176] This assay used a capture mAb specific to bovine and ovine sTnI and a biotin-conjugated detection mAb that reacts to all animal sTnI. The optimized assay achieved detection limits for bovine and ovine sTnI as low as 5.0 and 4.0 ng/ml, respectively. [176] Based on the detection of sTnI, several commercial immunoassay kits are now available for the enforcement of ruminant feed ban regulations (Table 3).
