Short communication
Non-protein nitrogen determination: A screening tool for nitrogenous compound adulteration of milk powder

https://doi.org/10.1016/j.idairyj.2016.12.003Get rights and content

Abstract

Kjeldahl and Dumas crude protein measurement methodologies do not distinguish between nitrogen native to milk and nitrogen in low molecular mass, nitrogen-rich adulterants. Measuring the non-protein nitrogen (NPN) content is one possible means of closing this loophole. Four methods were considered, with three selected for further research and validation: protein precipitation using trichloroacetic acid, protein precipitation using tannic acid, and molecular mass cut-off filtration. Kjeldahl assay was used on the supernatant or filtrate for all three methods. An NPN reference concentration range was established using 15 milk powder samples. This was followed by spiking experiments using seven low-molecular-weight, nitrogen-rich adulterants. Tannic acid precipitation and molecular mass cut-off filtration, both simple techniques available for routine use in food analysis laboratories, proved to be suitable for detecting the adulteration of milk powders with a variety of nitrogenous contaminants. NPN concentrations of ≥0.34% in milk powders are suspected to result from adulteration.

Introduction

The economically motivated adulteration of milk products with melamine in China in 2008 was a public health tragedy that resulted in more than 300,000 cases of illness among Chinese infants and young children, six reported deaths, significant impairment of international trade, and the temporary collapse of the Chinese dairy industry (Gossner et al., 2009, Ingelfinger, 2008, Moore, 2014). The incident revealed the vulnerability of the standard methods for the determination of the protein content of milk-based ingredients and infant formulas to verify product specifications. Methods based on the Dumas combustion principle and the Kjeldahl digestion principle measure the nitrogen content and calculate the protein content using suitable nitrogen-to-protein conversion factors. These methods lack the specificity to discriminate protein from other nitrogen-rich chemicals. Hence, the addition of melamine or other nitrogen-rich compounds to fraudulently inflate the apparent protein content of milk ingredients cannot be detected by these methods (Moore, DeVries, Lipp, Griffiths, & Abernethy, 2010). New analytical approaches are needed to more specifically characterize the inherent nitrogen composition of milk, milk ingredients, and milk products to detect economically motivated adulteration.

Milk ingredients contain nitrogen in the form of protein and non-protein nitrogen (NPN) components. Measurement of the quantity of the NPN fraction can be utilized to determine whether adulterants have been added. The addition of low-molecular-weight nitrogen-containing chemical(s) (e.g., melamine, urea, etc.), to increase the apparent protein concentration that is measured by these traditional nitrogen test methods, would increase the total NPN content and would result in a failure to meet product acceptance criteria. Likewise, milk powder dilution with a nitrogen-free adulterant (e.g., water or maltodextrin) would lower the NPN content and, if carried out at a significant level, would decrease the measured total NPN.

The NPN fraction of milk and other foods can be isolated using different techniques. Trichloroacetic acid (TCA) was used by Rowland (1938) to separate the protein and NPN fraction of milk. This approach has been standardized in AOAC methods 991.21, 991.22, and 991.23 (AOAC, 2012a, AOAC, 2012b, AOAC, 2012c) to improve the accuracy of milk protein determination compared with crude total protein analysis by AOAC method 991.20 (AOAC, 2012d) or ISO 8968-1|IDF 20-1:2014 (ISO/IDF, 2014). A similar TCA procedure was recently used to calculate an NPN index to detect adulteration in raw fluid milk (Gao, Li, Zan, Yue, & Shi, 2015). Cationic zinc (Zn2+) has been used to precipitate milk proteins followed by centrifugation, a second precipitation, and a filtration step (Scott, 1934). Tannic acids (TAs), which are used to precipitate proteins in beer production, have been used for milk protein isolation and for other applications (Courtney and Brown, 1930, Silanikove et al., 2001, Turley et al., 1990). TA and Zn2+ procedures are of interest for milk screening, given their simplicity and use of nonhazardous reagents. Other milk protein recovery and precipitation techniques that include the use of organic solvents alone (e.g., acetone or ethanol; Beaudoin et al., 2003, Hewedi et al., 1985), sulfosalicylic acid (Du and Bo, 2006, MacWilliam, 1891), phosphotungstic acid (Courtney & Brown, 1930), pH adjustments, salt treatments, or combinations thereof with or without heat (McKenzie, 1967, Ottenhof, 1985, Palmer, 1934, Scott, 1952, Spreer and Mixa, 1998, de Haast et al., 1987) have been reported. All of these additional protein recovery and precipitation approaches have low NPN recoveries or are considered to be too complicated for routine screening analysis. Methods for the precipitation of plant storage proteins from milk powders using borate buffers and turbidimetry (Cattaneo et al., 1994, Scholl et al., 2014) do not separate the NPN from the milk proteins.

In contrast to chemical procedures, molecular mass cut-off filtration is a physical approach that can be used to separate milk protein from the NPN fraction (Atkinson & Begg, 1988). A filtration approach using a 10 kDa or smaller membrane is attractive given its simplicity and the fact that it requires no reagents other than water. The various methods discussed above have not been fully validated using known or potential adulterants found in milk powder.

The objectives of this study were to develop and validate NPN measurement methodologies and to establish reference NPN ranges that would be suitable for screening skim milk powder (SMP) and nonfat dry milk (NFDM) to detect economically motivated adulteration. Although SMP and NFDM are similar, SMP manufacturers are allowed to adjust the protein concentration to 34% by adding protein-rich or protein-reduced milk fractions (ADPI, 2016). Thus, it was deemed to be important to study both commodities. Four candidate approaches for isolating the NPN fraction of SMP and NFDM were investigated: precipitation with TCA, TA, or Zn2+, and molecular mass cut-off filtration. Typical NPN ranges for authentic materials were established, upon completion of which studies were conducted using nitrogen-rich adulterants to assess the performance of each method.

Section snippets

Chemicals

ACS reagent grade TCA was purchased from Alfa Aesar (Ward Hill, MA, USA). ACS reagent grade TA was purchased from Sigma–Aldrich (St. Louis, MO, USA). Sodium chloride, zinc sulfate, and sodium hydroxide solutions were all prepared from ACS reagent grade or equivalent chemicals. Melamine, urea, ammonium phosphate dibasic, dicyandiamide, and l-arginine were purchased from Sigma–Aldrich. Aminotriazole was purchased from Alfa Aesar. Isobutylidenediurea, in an agricultural grade of unknown purity,

Results and discussion

The key success factors for the application of an NPN technique to detect adulteration include: (1) accurate and precise measurement of the NPN; (2) satisfactory determination of the range of the NPN inherent in the SMP and NFDM; (3) satisfactory recovery of added adulterants through the measurement of the NPN fraction. Initial attempts using Dumas methodology for NPN determination were unsuccessful because of poor precision that resulted from the very small sample size used in the combustion

Conclusions

Methodologies based on TA precipitation or molecular mass cut-off filtration, followed by Kjeldahl analysis for NPN determination, were effective in detecting nitrogen-containing adulterants in milk powder. Milk powders with NPN concentrations ≥0.34% are suspected to be adulterated. The US Pharmacopeial Convention (USP) has adopted methods based on the TA and molecular mass cut-off principles as standard procedures in the Food Chemicals Codex of the USP.

Further validation work is currently

Acknowledgments

Andrew Mackey at Mondelez International for carrying out Raman analysis on the TCA-melamine precipitates; Tony Qin at USDA for carrying out Raman hyperspectral analysis on the TCA-melamine precipitates. From Medallion Laboratories, Tim Peters for conducting differential scanning calorimetry analyses, and Phil Laudolff and Daniel Thurston for conducting Kjeldahl assays.

References (33)

  • T.M.P. Cattaneo et al.

    Sample preparation for selective and sensitive detection of soya proteins in dairy products with chromatographic and electrophoretic techniques

    Netherlands Milk and Dairy Journal

    (1994)
  • A.M. Courtney et al.

    The protein and non-protein fractions of some samples of woman's milk

    Archives of Disease in Childhood

    (1930)
  • Z. Du et al.

    Composite protein precipitator

    (2006)
  • P. Gao et al.

    A non-protein nitrogen index for discriminating raw milk protein adulteration via the Kjeldahl method

    Analytical Methods

    (2015)
  • C.M. Gossner et al.

    The melamine incident: Implications for international food and feed safety

    Environmental Health Perspectives

    (2009)
  • J. de Haast et al.

    Protein co-precipitates from milk and blood plasma: Preparation and investigation of some functional properties

    Journal of the Science of Food and Agriculture

    (1987)
  • Cited by (18)

    • A comparison of analytical test methods in dairy processing

      2021, Food Control
      Citation Excerpt :

      Plant's B and C use similar methods, a crude protein calculation based on the Kjeldahl principle and a block digestion method respectively. The difference between Kjeldahl and Dumas methods have been widely researched (DeVries et al. 2017; Thompson et al. 2002). The Dumas method determines total nitrogen including inorganic matter while Kjeldahl determines only organic nitrogen (Müller, 2017).

    • A novel approach to detect milk adulteration based on the determination of protein content by smartphone-based digital image colorimetry

      2020, Food Control
      Citation Excerpt :

      It is also possible to estimate the protein content in milk from the absorption of UV radiation (measurements at 205 nm and 280 nm) (Simonian & Smith, 2006); however, selectivity is a critical aspect. Other approaches to evaluate the authenticity of milk samples involved determination of non-proteic nitrogen in whey after protein precipitation (DeVries et al., 2017) or association of videos acquired with a smartphone to pattern recognition tools (Song, Jiang, Wang, & Vincent, 2020). Currently, the standard procedure adopted in the dairy industry to determine the physicochemical composition of milk and its somatic cell count is near infrared (NIR) spectroscopy.

    • Rapid analysis of food raw materials adulteration using laser direct infrared spectroscopy and imaging

      2020, Food Control
      Citation Excerpt :

      Nitrogen-rich compounds are normally used in legal commercial applications by industries (e.g. as fertilizers, herbicides, non-protein nitrogen source in animal feed and various commercial products). They are therefore easily accessible and can be used by fraudsters to increase the apparent protein concentration of food which is normally measured by nitrogen test methods (De Vries et al., 2017). Several analytical methods have been developed for the detection of nitrogenous compounds in foods.

    • Cricket powder (Gryllus assimilis) as a new alternative protein source for gluten-free breads

      2019, Innovative Food Science and Emerging Technologies
      Citation Excerpt :

      Protein content (% N x 6.25) was quantified by Kjeldahl method (Method 46-13.01). In order to avoid an overestimation of the protein content, non-protein nitrogen was quantified using the TCA (Trichloroacetic acid) precipitation method described by DeVries et al. (2017), and then deducted from the total nitrogen found in the protein analysis. Total lipids content was determined using a Soxhlet extractor and petroleum ether as a solvent (Method 30-25.01).

    View all citing articles on Scopus
    1

    Since retired.

    View full text