Initial results from the analyses of geological and anthropological samples for amino acids were difficult to accept because of the high enantiomeric purities of the analytes (i.e., predominantly L-amino acids). Consequently, sources of contamination had to be considered. All sources were eliminated except for direct atmospheric contamination. Essentially invisible, microscopic, aerosol/dust was found to rapidly contaminate the surface of samples and sample containers even after brief exposure times in clean laboratories. Contamination increased with exposure time. The aerosol/dust amino acids were contained predominantly in a proteinaceous material. Aerosol/dust from different locations can contain different percentages of proteinoid/ amino acid material. However, the relative concentrations of the amino acids were similar for both laboratory and residential samples. The enantiomeric purity of the
L-amino acids studied in aerosol/dust appear to be 99% or greater for the samples examined. Thus, even slight contamination of any sample with microscopic dust or aerosol particles can skew the results of trace amino acid analyses and amino acid e.e. determinations.
One of the most common and important types of analysis for biological samples involves amino acids.1–3 The amino
acids can be either free entities or, more commonly, part of larger compounds (e.g., peptides, proteins, etc.). Recently, the enantiomeric composition of many amino acids was found to provide additional information as to disease states, nutrition, age, biological role, environmental conditions, and purity.3–10 As with all types of analyses, background contamination must be avoided in order to obtain accurate and reproducible results. Contamination is less of a problem when dealing with samples that have high concentrations of the analytes of interest, compared to the surrounding environment. This is generally the case when measuring amino acids in biological samples which contain significant quantities of amino acids, peptides, proteins, and related conjugates. However, when analyzing very small samples or samples that contain very little of the component of interest, contamination can be a very real
problem. Examples of this, in the realm of amino acid analysis, include the analysis of geological samples,11 extraterrestrial samples (meteorites, moon rock, etc.),12,13, or even biological samples in which only a minuscule amount of material is available.8,9,14 The analysis of such samples usually requires specialized sample pretreatments and handling, as well as ultraclean experimental environments.
Recently, we became involved in a project that involved the analysis of prehistoric cave paintings. One of the questions to be answered concerned the composition of the binder in the paint used by prehistoric artists. It was hoped that amino acid analysis (after hydrolysis) would provide information as to the nature of the binder (i.e., whether or not it was proteinaceous and perhaps the type of protein material used). Only small amounts of the samples were available and they consisted mainly of inorganic material. Analyses of the paint, as well as the underlying limestone and other samples of virgin limestone, indicated the presence of amino acids with high e.e.’s (of L-amino acids). We found these results difficult to believe, even when repeated analyses with extensive cleaning, washing, and careful handling of the samples gave similar results. Contamination was suspected but its source was difficult to confirm. After checking all reagents and equipment, empty sample vials were taken and subjected to the entire cleaning, hydrolysis, and handling process. Analysis of the “empty” vials (blanks) revealed the presence of amino acids with high e.e.’s. These amino acids could only be eliminated when the entire process was carried out in a sealed, clean, glove.
the box under a positive pressure of filtered nitrogen. This indicated that the source of the contamination was atmospheric in origin. This appeared to be true even though the sample vials contained no visible material (even under 20× magnification) and may have been exposed to the atmosphere for only a few seconds. Hence, we were forced to examine the role of microscopic (and macroscopic) aerosols/dust as the main source of amino acid contamination.
Aerosols are known to be ubiquitous in most environments and a common “transport vehicle” for a variety of substances and pollutants.15 However, we were unable to find published information on the content and chirality of amino acids in aerosols or dust from indoor environments.
In this work, we examine dust from several laboratories and residential enclosures and compare the results to those found in the previous blank and limestone experiments.
Amino acid standards and hydrochloric acid (constant boiling) were purchased from Sigma (St. Louis, MO). Microhydrolysis tubes (1 ml) were purchased from Kontes (Vineland, NJ). Two HPLC column-switching systems
were used. System 1 was a Shimadzu (Kyoto, Japan) system consisting of two LC-6A pumps, an SCL-6A system
controller, a Rheodyne 7125 injection valve, a C-18 column, an SPD-6A UV spectrophotometric detector, a C-R6A Chromatopac. This was connected with a Rheodyne 7000 switching valve to a second HPLC system consisting of an LC-6A pump, a Rheodyne 7125 injection valve, a b-cyclodextrin column (Cyclobond I-2000; Astec, Whippany, NJ), an RF-535 fluorescence detector, and a C-R3A Chromatopac. System 2 consisted of a BAS (West Lafayette, IN)
PM-80 pump, a Rheodyne 7125 injection valve, a C-18 column, and a BAS UV-116A UV-VIS detector. This was connected with a Rheodyne 3092 switching valve to a second HPLC system consisting of a Shimadzu LC-6A pump, a
Rheodyne 7125 injection valve, a Cyclobond I-SN column (Astec), and a Shimadzu RF-535 fluorescence detector.
Both detectors on System 2 were interfaced with a BAS DA-5 ChromGraph Interface to a Gateway 2000 computer
(486, 66 MHz).
All HPLC columns were purchased from Advanced Separation Technologies (Astec). The AccQFluor Reagent Kit (Waters, Milford, MA) was used to derivatize the amino acids with 6-aminoquinolyl Nhydroxysuccinimidyl carbamate (AQC).
two residential properties. They were placed in glass vials previously cleaned and capped in a glove box and only
opened at the site of collection. Subsequently, all samples were handled prior to hydrolysis in a glove box (containing
a filtered N2 atmosphere at a positive pressure) to avoid contamination from airborne sources. Approximately 3 mg
of dust was weighed out and put in a hydrolysis tube with approximately 500 µl of constant boiling HCl. The tube was then closed and subjected to a vacuum followed by a nitrogen purge repeated three times with a final vacuum step. The hydrolysis tubes were then placed in a 100°C oven for 96 h to hydrolyze any proteins present. The samples were then transferred to a 1.5 ml Eppendorf tube and put in a vacuum centrifuge to remove the HCl. Samples were then dissolved in 100 µl 20 mM HCl and derivatized with AQC (AccQFluor Reagent Kit; Waters). Originally we hydrolyzed blanks in the laboratory and from samples of limestone rock and found trace levels of amino acids, necessitating the use of the glove box to avoid contamination from airborne sources. Additionally, extracts of samples washed with constant boiling HCl solution were run (without incubation at 100°C for 96 h) to analyze for free amino acids; none were found (i.e., below our limit of detection), suggesting that the main source of amino acids was proteinaceous. The time dependency for dust/aerosol accumulation in an open hydrolysis tube (4 mm opening) was examined by leaving the tubes open for specific periods of time (0.1–24 h) to the ambient air in the laboratory. Constant boiling HCl was then added as before, as well as all other aforementioned steps in the analysis of dust samples. For derivatization, 300 µl buffer was added to 100-µl samples and then 100 µl AQC (3 mM in acetonitrile) was added. The samples then sat for 1 min at room temperature and then 15–20 min at 55°C. Samples were then ready for direct injection into the HPLC system. UV detection was performed at 254 nm and fluorescence detection at an excitation of 250 nm with emission at 395 nm. Figure 1 shows the HPLC separation of the individual amino acids. Figure 2 shows the enantiomeric separation of D, L-leucine that was switched from the achiral reversed-phase column. The relative standard deviation of this method was determined by taking a macroscopic dust sample and dividing it into four approximately equal portions (of about 4 mg each). Each of the four samples was independently subjected to the entire wash, hydrolysis, derivatization, and separation procedure described in this section. Leucine
was taken as a representative amino acid and its concentration was measured in all four samples. The relative standard deviation of this approach was found to be 10%.
RESULTS AND DISCUSSION
Aerosol/dust is ubiquitous in most enclosed work and living environments. While most individuals are familiar
with macroscopic “dust aggregates,” microscopic particles are even more prevalent, although usually unnoticed. Unless precautions are taken, this can be a problem when doing trace analyses on samples for components that are
also present in aerosols/colloidal dust. Dust found in enclosed work and living environments is a complex mixture of many inorganic and organic components. The inorganic components can consist of a variety of minerals (e.g., quartz, clay, etc.) and salts.16,17 The organic components can include cellulosic, proteinaceous, and microbial matter as well as synthetic compounds.18–20 In our work, we were concerned with trace amino acid analysis. It
appeared that the proteinaceous material (from the skin, hair, fibers, microorganisms, etc.) of aerosols/dust severely affected our results when this material was not rigorously excluded. Figure 3 shows that amino acids from aerosol/
dust proteins can enter a clean sample vial if that vial is unsealed, even briefly, in a supposedly “clean” laboratory
environment. Furthermore, the amount of contamination increases rapidly for the first 3 h and more slowly after
several hours (Fig. 3). The rapid onset of the contamination may possibly be due to electrostatic or other surface interactions between the colloidal dust and the clean inner wall of the container (which in this case was glass).
The free amino acid content of laboratory and residential dust samples was found to be nearly negligible compared
to the amino acids inbound or complexed form (e.g., proteins). The proteinaceous content of laboratory dust (7.9 ±
0.5%) was consistently lower than that of residential dust (10.4–14.4%). The protein/amino acid content of dust in
certain agricultural environments can be even higher.20 Table 1 gives the concentration of 14 amino acids found
in hydrolyzed dust samples from three different locations. Since alanine usually was the least prevalent amino acid
(on a weight basis), all of the concentrations can be normalized to it. Figure 4 shows the normalized amino acids.
Microscopic aerosol/dust can be a significant contamination problem when doing trace amino acid analysis and
enantiomeric purity determination for amino acids. Microscopic and macroscopic laboratory and residential dust
have similar, relative amino acid levels and high enantiomeric purities. These trends and values may be useful in
identifying contaminated samples. Contamination increases with exposure time but levels off after several
hours. Limestone appears to contain small amounts of indigenous D-amino acids.
1. Barrett GC, Elmore DT. Amino acids and peptides. Cambridge, UK: Cambridge University Press; 1998.
2. Bajusz S, Hudecz F (eds.). Proceedings of the Twenty-Fifth European Peptide Symposium. Akademiai Diade, Budapest; 1999.
3. Ekborg-Ott KH, Armstrong DW. Stereochemical analysis of food components in: chiral separations: applications and technology. Ahuja S, editor. Washington, DC: American Chemical Society; 1997. p 201–270.
4. Armstrong DW, Duncan JD, Lee SH. Evaluation of D-amino acids levels in human urine and in commercial L-amino acid samples.
The protein standard consisted of a-chymotrypsin (see Experimental section for details). b
These values came from exposing a clean, dry glass hydrolysis tube (4 mm opening) to ambient laboratory air for 30 min (see Experimental as well as Results and Discussion sections). No visible contamination was in the device.
c A 3 mg sample of freshly exposed limestone (i.e., from the center of a limestone rock ∼10 cm diameter that was broken in half) was taken and washed with 2N NaOH solution and placed in a hydrolysis tube (see Experimental). The closed sample was acidified with constant boiling HCl and digested at 100°C for 96 h. The resulting solution was analyzed for amino acid content and enantiomeric purity. AMINO ACIDS IN AEROSOL/DUST 157
5. Child AM. Amino acid racemization and the effects of microbial diagenesis. Washington DC: Archeological Chem. ACS Symposium Series, Vol. 625; 1996. p 336–377.
6. Engel MH, Goodfriend GA, Qian Y, Maco SA. Indigenesty of organic matter in fossils. Proc Natl Acad Sci USA 1994;91:10475–10478.
7. Fu S-J, Fan C-C, Song H-W, Wei F-Q. Age estimation using a modified HPLC determination of the ratio of aspartic acid in dentin. Forensic Sci Int 1995;73:35–40.
8. Armstrong DW, Gasper MP, Lee SH, Ercal N, Zukowski J. Factors controlling the level and accurate determination of D-amino acids in the urine and plasma of laboratory rodents. Amino Acids 1993;5:299–315.
9. Armstrong DW, Zukowski J, Ercal N, Gasper MP. Stereochemistry of pipecolic acid found in the urine and plasma of subjects with peroxisomal deficiencies. J Pharm Biomed Anal 1993;11:881–886.
10. Berthod A, Liu Y, Bagwill C, Armstrong DW. Facile liquid chromatographic enantioresolution of native amino acids and peptides using a teicoplanin chiral stationary phase. J Chromatogr A. 1996;731:123–137.
11. Hare PE. Geochemistry of proteins, peptides, and amino acids. In: Eglinton G, Murphy MTJ, editors. Organic geochemistry. New York: Springer-Verlag; 1969. p 438–463.
12. Bada JL. Amino acid cosmochemistry. Philos Trans R Soc Lond B 1991;333:349–358.
13. Engel MH, Maco SA. Isotopic evidence for extraterrestrial nonracemic amino acids in the Murchison meteorite. Nature 1997;389: 265–268.
14. Poinar HN, Hoss M, Bada JL, Paabo S. Amino acid racemization and the preservation of ancient DNA. Science 1996;272:864–866.
15. Spurny KR (ed.). Aerosol chemical processes in the environment. Boca Raton, FL: CRC Press; 2000.
16. Antal L, Hlavay J, Karpati J. Analysis of respirable and sedimented dust samples. Period Polytech Chem Eng 1986;30:209–212.
17. Silva PJ, Carlin RA, Prather KA. Single particle analysis of suspended soil dust from Southern California. Atmos Environ 2000;34:1811–1920.
18. Dashek WV, Olenchock SA, Mayfield JE, Wirtz GH, Wolz DE, Young CA. Carbohydrate and protein contents of grain dust in relation to dust morphology. Environ Health Perspect 1986;66:135–143.
19. Witteman AM, Voorneman R, Van den Oudenrijn S, VanLeeuwen J. Silverfish protein in house dust in relation to mite and total arthropod level. Clinical and experimental allergy. J Br Soc Allergy Clin Immun
20. Raszyk J. Fyzikalni, chemicke a biologicke vysetreni prachu z velkovykrmen prasat. Veterinarni Medicina 1986;31:233–244.
21. Lehninger AL. Biochemistry. New York: Worth Publishers; 1970. p 92–93.