Identification of Turtle Shell with Chemical Methods


Student thesis: Doctoral Thesis

View graph of relations


  • Linqiu LI

Related Research Unit(s)


Awarding Institution
  • Hon Yeung CHEUNG (Supervisor)
Award date15 Jan 2016


Turtle shell has been used as a medicine for a thousand years, and it is widely utilized as one of the main ingredients in functional foods. However, many forged functional foods that supposedly contain turtle shell are actually made from other substitutes, which can be attributed to its limited supply, high price, and enormous demand. Even if functional foods are made from turtle shell, the species used may be different. Therefore, it is necessary to develop reliable methods to identify turtle shell. In this work, turtle shell was characterized through various methods, such as spectroscopy, chromatography, and proteomics etc., aiming to distinguish turtle shell from inferior replacement products.

A micellar electrokinetic chromatography (MEKC) method with dansyl chloride (Dns-Cl) as a derivatization reagent was developed for separation and quantification of 20 amino acids, namely hydroxyproline, DL-hydroxylysine, and 18 common amino acids found in turtle shell. In order to discover an optimum experimental condition, the derivatization condition, stability of the derivatives, separation condition, and injection condition were investigated. It was found that the highest yield of derivatives was achieved when the volume ratio of sample, NaHCO3 (0.5 M, pH 8.35), and acetone (containing 0.02 M Dns-Cl in) was 1:1:1 with 40 min of incubation time at 65°C in dark. The stability study revealed that amino acid derivavtives were stable for 1.5 years if they were stored at 0-4°C and no quantitative loss within 2 years at -20°C. Separation condition comprising of 20 mM borax and phosphate containing 6% methanol and 0.1 M SDS (pH 8.74) was found the best if samples were eluted at 25 kV and 25°C while injection parameter was 9 kV and 9 s at 25°C.

Since the use of species and parts (carapace and plastron) of turtle shell may be varied, a comparative study between the ventral and the dorsal shells of seven species of turtle was explored in terms of their inorganic elements and amino acid content. The results from the inductively coupled plasma atomic emission spectrometer (ICP-AES) demonstrated that turtle shell was rich in with Ca and Mg with various essential trace elements, such as Zn, Se, Fe, Mn, Cr, and Cu, in which Se is claimed to be an anti-cancer element. The content of inorganic elements based statistical analysis indicated that no obvious differences existed among these seven types of turtle shells. The amino acid content was analysed by MEKC, and it was found that all species of turtle shell contained 20 AAs, with abundant glycine, arginine, proline, and hydroxyproline. By applying the principle component analysis (PCA), Trachemys scripta (TS) and CG (Cuora galbinifrons) were successfully differentiated from Cuora trifasciata (CT), while the content of amino acid in Chinemys reevesii (CR), Cuora flavomarginata (CF), Cuora aurocapitata (CA), and Hardella thurjii (HT) was found to be similar to that in CT. Both the results of ICP-AES and MEKC revealed that no differences exited between carapace and plastron, no matter what the species was.

With the above-mentioned MEKC method, 15 fresh and 6 preserved plastrons (Chinemys Reevesi) along with11 different foods and tissues were investigated, aiming to identify the authenticity of plastrons. The resutls indicated that 18 common AAs and 2 special AAs— hydroxyproline and one form of DL-hydroxylysine were found in the plastron, with glycine (57.985 mg/g), arginine (31.430 mg/g), proline (31.185 mg/g), and hydroxyproline (24.981 mg/g) most abundant. This method allows people to differentiate plastron from chicken, egg, fish, milk, pork, nail, and hair, as these latter materials lacking hydroxyprolineand hydroxylysine. Through statistical analysis, plastron was successfully distinguished from pigskin, chicken tendon, fish skin, and calf tendon.

However, when the MEKC method was applied to turtle shell analysis, some serious matrix effects (ME) were noted. The sensitivity, resolution, and reproducibility of separation were affected and reduced. Therefore, effort was made to find out the relationship between separation efficiency and current, and to figure out an effective, simple, and economic solution to overcome the negative impact of ME. The study showed that small amounts of NaCl (≤0.005 mg/mL) in the sample had no impact on the separation but enhanced the sensitivity. However, when the concentration of NaCl increased above 0.005 mg/mL, it decreased the separation efficiency, sensitivity, and migration time. Additionally, increasing NaCl concentration resulted in an increasing turning point (TP). The study of the relationship between current and NaCl concentration indicated that when the TP of a sample was higher than 62.36 μA, desalination was necessary. Since the reported desalination methods are either expensive or complicated, we developed a simple and economic method by simply adding 12 times (by volume) chloroform/methanol (2:1, v/v) into the sample. When this method was applied to turtle jelly, the number of theoretical plates (N) of 20 amino acids was enhanced up to 3-fold.

In addition to the desalination method with the organic solvent, a pipette tip SPE method using graphene as sorbent (PTG-SPE) was developed as another alternative for desalination. In this method, 100 μL of derived amino acids were loaded into the PTG-SPE cartridge, and then eluted with 500 μL methanol. The number of aspirating/dispensing cycles was 18, giving about 1 min of extraction time. Compared with other commercial sorbents, graphene provided many advantages, such as stability, high recovery, and compatibility with various organic solvents. After desalination, the matrix effect was greatly reduced, and both theseparation and sensitivity were improved. Comparing to the organic solvent method, this method could help to reduce the use of organic solvents and easily establish therobustness of the method.

The MEKC method was further validated by both high performance liquid chromatography (HPLC) and ultra performance liquid chromatography (UHPLC) methods. In this comparison study, two derivatization methods, two columns, and separation conditions were optimized. It was found that the 6-aminoquinolyl-N-hydroxysuccinimidyl-carbamate (AQC) based method was more sensitive, time saving and producing fewer by-products although it was more expensive than using Dns-Cl as derivatization reagent. The run time of both HPLC and UHPLC methods was 10 min, which was 45 min shorter than that of the MEKC method. Furthermore, these two LC methods gave better repeatability, good efficiency of separation, and an absence of ME, in contrast to the MEKC method. The disadvantages of these two methods, however, were environmental unfriendly and expensive as they required large quantity of organic solvents. Results revealed a good accordance (relative error < 0.2%) amongst these three methods in terms of content of amino acids in turtle shell and turtle jelly.

As turtle shell is abundant in collagen, an attempt was also made to extract and characterize collagen in different species of turtle shell. After a series of treatment (removal of lipid, calcium, and non-collagen), the collagen in turtle shell became extractable after pepsin digestion and could be purified by 4% NaCl (w/v) at pH 8.0. All samples were found to have the common properties of collagen: a high content of Hypro, Gly, Pro, and Arg, as well as some Hylys. My results also suggested that type I collagen is a major component in all seven turtle shells. CT could be distinguished from other proximate materials. It was also shown that collagen of the plastrons could be divided into three groups: (1) TS & CF & CA, (2) CR & HT & CT, and (3) CG.