Immunological detection of denatured proteins as a method for rapid identification of food residues on archaeological pottery

Journal of Archaeological Science 73 (2016) 25e35

Contents lists available at ScienceDirect

Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas

Immunological detection of denatured proteins as a method for rapid identification of food residues on archaeological pottery Jaroslav Pavelka a, Ladislav Smejda b, c, *, Radovan Hynek d, Stepanka Hrdlickova Kuckova d a

Centre of Biology, Geosciences and Environmental Education, University of West Bohemia, Univerzitni 8, 30614, Plzen, Czech Republic Department of Archaeology, University of West Bohemia, Univerzitni 8, 30614, Plzen, Czech Republic c Department of Ecology, Czech University of Life Sciences Prague, Kamycka 961/129, 165 00, Praha 6-Suchdol, Czech Republic d Department of Biochemistry and Microbiology, University of Chemistry and Technology, Prague, Technicka 3, 166 28, Prague 6, Czech Republic b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 June 2015 Received in revised form 18 June 2016 Accepted 11 July 2016

Our understanding of human diet in different periods of history can be enhanced by investigating direct evidence represented by accidentally preserved food remains found on pottery. So far, this task has been accomplished by the application of gas chromatography/mass spectrometry, often in combination with stable isotope analysis. These methods require specialised laboratories and their cost prevents wider penetration into the daily practice of archaeology and related disciplines. We have tested commercially available immunochromatographic kits for this task, which are designed to detect contaminants and allergens in the modern food industry. Unlike the previously published studies on archaeological material, we focus specifically on the identification of damaged and denatured proteins, which correspond better to the state of preservation of proteins in desiccated and carbonised organic residues that have survived from antiquity. We report the first successful qualitative detection of bird eggs, animal meat, milk (and species of origin), and to some extent also the presence of plant food, especially cereals and hazelnuts. The immunoassay is a methodology that is well suited for use in the field and resource-poor environments, so it is ideal for most archaeological excavations and museums. With necessary caution, the results can be used as a proxy for human diet in the past and reconstructions of anthropogenically modified environments. © 2016 Published by Elsevier Ltd.

Keywords: Immunological assays ELISA Food Organic residues Archaeological pottery Past diet patterns

1. Introduction It is occasionally possible to find ancient food residues attached to archaeological pottery, usually in the form of accidentally desiccated or charred organic material. Analysis of such remains is highly desirable, as the results may suggest the final purpose of a particular vessel, before it went out of use and was discarded (Pollard et al., 2007:22e23). The significance of this type of research is twofold. It may reveal a cultural association between certain types of pots and meals that were cooked in them and possible variations in the use of ceramic inventories of households. Secondly, it provides data on human diet in specific cultural and social contexts, if these can be recognised from the archaeological record. Alternatively e and complementarily e approaches that

* Corresponding author. Department of Archaeology, University of West Bohemia, Univerzitni 8, 30614, Plzen, Czech Republic. E-mail address: [email protected] (L. Smejda). http://dx.doi.org/10.1016/j.jas.2016.07.004 0305-4403/© 2016 Published by Elsevier Ltd.

help study subsistence on the basis of archaeological finds include the compositional analysis of animal bones and finds of preserved botanical macrofossils. Well preserved bones and plant remains are not ubiquitous on ancient sites, depending heavily on local taphonomic conditions (Dincauze, 2000; van der Veen, 2007; Wright, 2003). Indirect approaches, which can detect only broad types of diet, include complex dental analysis and the analysis of stable isotopes in human bones (Day, 2013; Forshaw, 2014; Klippel, 2001; Landon, 2005; Reitsema et al., 2010). For a detailed analysis of carbonised food residues usually found as attachments on the surface of cooking vessels, there are currently two dominant approaches available, each with its own pros and cons. The more widespread method is the analysis of lipids through the identification of fatty acids by gas chromatography-mass spectrometry (GC-MS), sometimes coupled with the analysis of carbon stable isotopes, which allows the distinction between ruminant and non-ruminant fat in remains of animal origin (Cramp et al., 2014; Dudd et al., 1999; Regert, 2011; Salvini et al., 2008). A possible advantage of the study of lipids

26

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

over proteins is seen in the fact that lipids should be less susceptible to leaching and diagenetic degradation than proteins (Craig et al., 2005). The second major approach aims to identify source-specific proteins in the preserved organic residues. Today this task can be performed either by mass spectrometry or by more traditional immunological methods. Recently, mass spectrometry made significant advances in this direction and has been applied to analyses of protein components in historic art works (Hynek et al., 2004; Kuckova et al., 2013) and in historic mortars (Kuckova et al., 2009). The approach to food residue analysis by peptide mapping  et al., 2010). MS proteomics currently is very analogous (Kuckova uses several instrumentation variants based on liquid chromatography (LC) combined with matrix-assisted laser desorption ionisation-time of flight mass spectrometry (MALDI-TOF-MS). Contrary to lipid analysis, this technique should be able to detect distinct differences in the amino acid loci of alpha S1 casein characteristic for individual species raised for milk (Buckley et al., 2013; Hong et al., 2012). A relatively high cost of peptide mapping and its high sensitivity to preservation of target protein structures versus contamination (e.g. by human agents or microorganisms) presents barriers to its routine use in archaeology. In this paper we explore current prospects in the field of protein identification in ancient food residues by the immunochromatographic method. Although this technique has been known for a long time, we pay attention to certain recent trends in this field, which are promising for the large-scale investigation of organic residues on pottery vessels. 2. Immunological analysis of food residues Immunological detection of proteins is a well-known and routinely used technique in many disciplines. Pavelka et al. (2011) used this method for taxonomic determination of animal bones from archaeological sites to enhance the rate of successful identification of species by standard osteological approach. Examples of its application to ancient food residues have already been published and verified by an independent method (Craig et al., 2005). However, immunological tests can be prepared in many particular configurations requiring somewhat different laboratory protocols. Their performance and sensitivity mainly depend on the state of protein preservation in largely variable diagenetic conditions of recovery contexts. The methods used for analysis should therefore specifically target types of proteins that are likely to be present in the samples. The same analytical approaches can be applied to either organic remains visibly adhering to the surface of pots or to the detection of organic residues soaked into the porous mass of ancient ceramics. However, proteinaceous crusts of organic material attached to the surface of cooking pots can be analysed more readily, while proteins adsorbed to the ceramic matrix and siliceous minerals are generally more difficult to extract (Craig and Collins, 2000). In our study, we mainly focused on the analysis of carbonised attachments to the pottery surface, opened by recent advances in commercial immunological tests used in the food industry (Dzantiev et al., 2014). This novel approach opens broad possibilities of application in the study of past diet patterns. Although organic materials can be in principle tested for the presence of DNA, the origin of which could subsequently be assessed, this approach seems to be rather unpromising for the investigation of archaeological finds of food remains. In the existing archaeological samples of food remains, which are typically small and degraded by heat and other diagenetic factors, the preservation of ancient DNA is likely to be poor. Proteins are more resistant than DNA and only very small regions of protein molecules, known as

epitopes, are needed for binding of specific antibodies (Hofreiter et al., 2012). The immunological detection therefore usually has a higher chance of being successful compared to ancient DNA extraction and identification. This is especially true when denatured or thermostable forms of proteins are targeted, which can be expected to be preserved in food remains attached to cooking vessels. We used relatively simple tests based on the immunological antigen-antibody interaction. The most innovative part of our approach is aimed specifically at those types and forms of proteins, which can be expected to be the most characteristic for archaeological samples in the studied region of temperate Europe. A number of earlier attempts using antibody tests on archaeological material did not yield consistent results and were consequently viewed rather critically. The problem of cross-reactions was mentioned in Child and Pollard (1992) and studied experimentally by Collins et al. (1992). Dongoske et al. (2000) repeated the warning that the degradation of proteins may result in antigen binding that accounts for nonspecific reactions of ELISA tests. Problems linked with immunological techniques applied to old samples were elaborated in a more detailed study by Brandt et al. (2002), who investigated the effects of diagenesis and contamination on the detection success of non-collagenous proteins from human and other mammal samples of modern and ancient origin. It must be understood that types of antibody tests used in past studies were mostly developed for use with modern biological samples. This factor indeed limits their applicability in studies working with archaeological material. However, it is possible to use antibody tests targeted specifically at denatured/degraded proteins, which can make a profound difference in the quality of obtained results. A wide range of detection tests based on immunological reaction are commercially available today, which were developed for the identification of various components in heat-treated foods. This category of commercial test is purposefully designed to work with degraded traces of biological tissues in the food industry and they are undergoing rigorous evaluation of their ability to detect the ingredients in processed food correctly (Bjorklund et al., 2001). Many food components can be identified on the basis of the presence of some typical proteins which have a well-known chemical structure; the producers of commercial kits make use of these properties for their reliable detection. Currently used foodsafety tests are constantly improving to reduce the possibility of cross-reactions e.g. with bacteria (this risk is described in Brandt et al., 2002). The kits we have chosen for testing archaeological material are highly sensitive for the detection of allergens in the food industry. This means that they are expected to identify even trace amounts of specific proteins that may present unwanted contamination in food. Manufacturers of these kits indicate that the detection limit for the target content can be at single-digit ppm, although this may vary depending on sample type and extraction efficiency. We evaluate their performance on archaeologically retrieved food remains, as it can be anticipated that these modern, specifically targeted assays could be more apt for such a complex task than traditional tests applied in archaeological studies in previous decades. All the kits used in this study were based on commercially available antibodies (Table 1). The market of immunochromatographic tests has become very dynamic in the last decade. The production of different kits for food-safety testing is growing globally every year (Dzantiev et al., 2014), and the competition among companies is strong. This implies that the names of producers and their marketed kits may be subject to rapid change. However, alternative tests should be available and improved technologies will lead to more varied products tailored to the specific needs of individual groups of users. Due to differences

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

27

Table 1 Overview of immunological tests used in this study. Manufacturer

Product name

Type of test

Intended use

Measurement

Possible cross reactivity

Gen-Probe Life Sciences Ltd., Deeside, UK

BIOKITS Cooked Species Identification Test Kit

sandwich enzyme immunoassay

detection of beef, pork, poultry and sheep content in meats, meat products and feedstuffs

photometric at 450 nm

beef test may cross-react with bison and weakly with buffalo and red deer; sheep test cross-reacts with goat; the poultry test indicates well chicken, turkey, duck and quail

Gen-Probe Life Sciences Ltd., Deeside, UK Gen-Probe Life Sciences Ltd., Deeside, UK Neogen, Lansing, USA

BIOKITS RAPID 3-D Casein test BIOKITS RAPID 3-D Hazelnut Test BioKits BLG Assay Kit

lateral flow test

detection of casein in cooked and uncooked foods detection of hazelnut protein

visual visual

walnut

indirect competitive enzyme immunoassay

detection of bovine BLG

visual

to a small extent with BLG from sheep and goats

Neogen, Lansing, USA

Reveal 3-D Egg Test

lateral flow test

visual

Neogen, Lansing, USA

Reveal 3-D gluten test

lateral flow test

R-Biopharm AG, Darmstadt, Germany Tepnel Biosystems, Deeside, UK

RIDASCREEN GIS

in vitro enzyme immunoassay sandwich enzyme immunoassay

presence of egg white protein ovomucoid presence of gluten in cooked and uncooked foods (bread wheat, durum wheat, rye and, to a lesser extent barley detection of goat's milk in sheep's milk detection of beef, pork, poultry and sheep content in meats, meat products and feedstuffs

BioKits (Cooked) Species Identification Test Kit

lateral flow test

between the designs of various tests not being equally suited to pre-modern samples and their comparative efficiency and reliability, they need to be validated for particular tasks. 3. Materials and methods 3.1. Samples Finds of ancient pottery used in this study were obtained from archaeological sites located in the Czech Republic. Dating of finds varies from the Eneolithic period (fourth to third millennium BC) to Late Middle Ages (fifteenth century AD); an overview of the analysed samples and obtained results is presented in Table 2. As macroscopically visible organic residues are very rare finds occurring only accidentally in archaeological material, we had to look for suitable specimens mostly in museum collections. For these specimens, soil controls were not available. The investigated ceramic finds (Figs. 1 and 2) were selected from larger assemblages of artefacts by archaeologists and curators responsible for the individual excavation projects and submitted for analysis to the authors. Although residues attached to pottery largely prevail in our research, most of them carbonised to a large extent, in several cases we also worked with material that appeared merely desiccated, adhering either to ceramic (#21 in Table 2) or bronze (#5e6) artefacts. The analysed samples are mixed with regard to age, pottery type, and recovery contexts; whether this heterogeneity causes any patterns in results is considered later (in the discussion section). The random character of sampling therefore reflects opportunities arising in museums and fieldwork projects, resulting in a diverse set of material used for testing and evaluation of the method. 3.2. Description of used tests Milk. The presence of casein can be detected using the Biokits

visual

brown linseed

photometric at 450 nm photometric at 450 nm

Rapid 3-D Casein Test Kit (lateral flow test). This test can prove the presence of dairy products in a sample. Casein is a natural part of milk and represents 80% of milk proteins. The detection is based on the reaction of casein with its specific antibody. When the presence of milk was confirmed in our study, its origin was specified by the following two tests. Bovine milk. Tests for the identification of the origin of dairy products from cattle are performed using the BIOKITS BLG (blactoglobulin) Assay Kit. Results are determined on microtitre plates. It is an ELISA competitive test where negative reactions reveal themselves by a specific colour. Goat milk. The test for the detection of goat milk RIDASCREEN® GIS is an enzyme immunoassay for the detection of goat milk in sheep milk. The test is primarily designed to reveal where sheep milk or cheese has been unlawfully declared. Sheep milk is more expensive than goat milk, and for this reason the production of kits for the detection of sheep milk is not a commercially attractive option; therefore, tests for sheep milk were not available for our study. However, if bovine and goat milk antibodies would give negative results, sheep could be a likely source of milk in our geographical setting. Meat. The identification of domestic animal proteins was performed using (Cooked) Species Identification Kits. This kit allows for proteins specific for beef, pork, poultry and mutton products to be distinguished. This test uses ELISA microwell modules and antibodies for thermostable species-specific muscle proteins. It is a non-competitive sandwich type of assay. Eggs. The Biokits Rapid 3-D Egg Test was used for testing for the presence of bird egg proteins, namely ovomucoid (Gal d 1), which is the main thermostable allergen of egg white (Matsuda et al., 1982). This test detects the presence of egg white protein in raw and cooked food products. Cereals containing gluten. The presence of gluten was indicated by the Biokits Rapid 3-D™ Gluten Test (lateral flow test). Gluten

28

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

Table 2 List of analysed samples and results (key: þþ high positive, þ low positive, ? dubious positive, - negative, empty cell ¼ not tested). #

Site

Age

Casein Cattle Goat Meat Meat type bLG IgG proteins

1 2 3 4 5 6

Otmí c hill Radkovice Radkovice  Pilsen-Hradiste dka Zahra dka Zahra

Eneolithic Eneolithic Eneolithic Late Bronze Age Early Iron Age Early Iron Age

þþ ? þþ þ þ

e e þ þ

e þþ þ þ

7 8 9 10 11 12 13 14 15

 Rovna  Rovna  Rovna  Rovna  Rovna  Rovna tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube

Early Early Early Early Early Early Early Early Early

Age Age Age Age Age Age Age Age Age

þ e e þ ? e e e ?

e

þþ

e e e

þþ þþ þþ

tín 16 Prague-Hloube

Early Iron Age

?

e

tín 17 Prague-Hloube

Early Iron Age

þ

e

tín 18 Prague-Hloube tín 19 Prague-Hloube tín 20 Prague-Hloube

Roman Period e Roman/Migration Period þ Early Middle Ages þ

Iron Iron Iron Iron Iron Iron Iron Iron Iron

e

e

þþ e

tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube tín Prague-Hloube  Boleslav-Karmel Mlada

Early Middle Ages Early Middle Ages Early Middle Ages Early Middle Ages Early Middle Ages Early Middle Ages Early Middle Ages High Middle Ages

þþ e e þ e e þ þþ

þþ

 Boleslav-Karmel 29 Mlada  Boleslav-Karmel 30 Mlada  Boleslav-Karmel 31 Mlada

High Middle Ages High Middle Ages High Middle Ages

þþ þþ þþ

e þþ þþ

21 22 23 24 25 26 27 28

e

e

e e

þ

?

32 33 34 35 36 37 38 39

 Boleslav-Karmel Mlada Horní Bríza Horní Bríza Aldasín Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre

High Middle Ages High Middle Ages High Middle Ages High/Late Middle Ages High/Late Middle Ages High/Late Middle Ages High/Late Middle Ages High/Late Middle Ages

þ þþ þ þ e e e e

þþ þ e e e e e e

e e e e

40 41 42 43

Pilsen-historic Pilsen-historic Pilsen-historic Pilsen-historic

High/Late High/Late High/Late High/Late

Ages Ages Ages Ages

e e e e

e e e e

e e e e

44 Pilsen-historic centre

High/Late Middle Ages

e

e

e

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

High/Late Middle High/Late Middle High/Late Middle High/Late Middle High/Late Middle High/Late Middle High/Late Middle High/Late Middle High/Late Middle Late Middle Ages Late Middle Ages Late Middle Ages Late Middle Ages Late Middle Ages Late Middle Ages Late Middle Ages

? ? ? ? ? ? þþ e e þ e e e þ e þ

þ e þ þþ þþ e e

þ þþ

e e e e e e e

þþ e e e þ e þ

centre centre centre centre

Prague-Republic Square Prague-Republic Square Prague-Republic Square Prague-Republic Square Prague-Republic Square Rabí Rabí Rabí Rabí Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre Pilsen-historic centre

Middle Middle Middle Middle

Ages Ages Ages Ages Ages Ages Ages Ages Ages

e e e e e e

þ þ e þ þ þ ? e þ þ þþ þþ þ þ þ e e þþ þ ? e ? e ? e þþ þþ þþ e e e

Ovomucoid Gluten Hazelnut MS verification

þ þ ? þ þ

þþ þ þþ Pork poultry ? Pork poultry e Pork poultry e beef e ? e pork poultry mutton pork ? beef mutton e beef e e mutton ? beef þ mutton ? þ poultry þþ þ mutton ? þ þþ beef pork poultry e e ?

poultry pork

e

e

þþ þþ

e

þ e þ e þ þ e þ e þ þ þ þ e e e e e e e e e e e e e e þ e

mutton poultry pork poultry

pork poultry pork poultry

e

Bovine serum albumin (5 matchesa) a lkov (S a et al., 2015)

e e e e e e e e e e e e e

e e e e þþ e e e e þ e

e e

þ þ

e þ þþ þþ þ þþ e e e

þ e þ e

poultry

poultry

e þ þ þþ þþ þþ

e e e e e

e e þ e ? e þ

MS sample Q39001: Ovalbumin (1 matchb)

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

29

Table 2 (continued ) #

Site

Age

61 62 63 64 65

Pilsen-historic centre Prague-Republic Square Prague-Republic Square Prague-Republic Square Prague-Republic Square

Late Late Late Late Late

Middle Middle Middle Middle Middle

Ages Ages Ages Ages Ages

e þþ þþ þþ þþ

e e e e e

e þ þ þþ þþ

e e e e e

66 67 68 69 70 71 72 73 74 75

Prague-Republic Square Prague-Republic Square Prague-Republic Square Prague-Republic Square Kyzy Kyzy Hony Hony Hory Mlad a Boleslav-Minorite monastery Mlad a Boleslav-Minorite monastery Mlad a Boleslav-Minorite monastery Control bovine milk on pot replica, buried for 6 months Control goat milk on pot replica, buried for 6 months Control beef on pot replica, buried for 6 months Control pork on pot replica, buried for 6 months

Late Late Late Late Late Late Late Late Late Late

Middle Middle Middle Middle Middle Middle Middle Middle Middle Middle

Ages Ages Ages Ages Ages Ages Ages Ages Ages Ages

þþ þþ ? ? e e ? þ þ þþ

e e þ þ

þþ þ

? e e e e e

mutton

þ e þ e

e þ þþ

þþ þ beef mutton

e

Late Middle Ages

þ

e

?

þ þ e

? þþ e e e e e þþ þ e

e

e

Late Middle Ages

þþ

þþ

þ þ

beef mutton

e

e

76 77 78

79 80 81 a b

Casein Cattle Goat Meat Meat type bLG IgG proteins

Recent

Recent

Ovomucoid Gluten Hazelnut MS verification

e þ þ þ

þ e e þ þþ

MS sample M31056: Ovalbumin (2 matches), Collagen alpha-1(I) chain (8 matches), Collagen alpha-2(I) chain (2 matches) b

þ

þ

Alpha-S2-casein (1 match), Beta-casein (1 match)a

Recent

þ

beef

Recent

þ

pork

  Bude jovice, Czech Republic. Proteomic Centre, Faculty of Science, University of South Bohemia, Cesk e Department of Biochemistry and Microbiology, University of Chemistry and Technology, Prague, Czech Republic.

reacts in the testing device with its tightly held monoclonal antibodies specific for omega gliadin. A positive test is shown by coloured stripes visible in the device reading window. In this way, we could look for the presence of some of the following grains: wheat (Triticum spp.), rye (Secale cereale L.), to a less extent barley (Hordeum vulgare L.) but not oat (Avena sativa), millet (Panicum miliaceum) and other non-gluten crops. Hazelnut. This plant ingredient was tested only exceptionally in our study. The Biokits Rapid 3-D Hazelnut Test was used for this purpose. As in the case of other Rapid 3-D lateral flow tests, the result is read from coloured stripes in the device reading window. 3.3. Preparation of samples Organic residues from archaeological contexts represent precious historic evidence and usually they are rare finds, available only in small amounts of varying states of preservation. This fact calls for careful planning of analytical procedures, because they inevitably consume a certain amount of sample material. Sometimes available volume of extracted analyte suffices only for a limited number of tests. As each immunological kit on the contemporary market targets a single type of protein, researchers are forced to consider which tests will be prioritised in particular projects, depending on their environmental and cultural contexts. The preparation of samples in our study for Biokits Rapid 3-D tests (lateral flow type of test) differed in some aspects from the manufacturer's instructions. We developed a special procedure for utilisation of a limited amount of archaeological material. A sample of small size (1/10 to 1/1000 g) was ground to pellets in an

eppendorf tube and mixed with buffer(s), which are part of the test kit, to get a final volume of 200e300 ml, according to the ratios and temporal order of application stated in the guidelines provided by the producer. It is possible to use a very small amount of sample material e in the lower range indicated above e but larger amounts, if they are available, may improve the performance of the test. The sample mixed with buffer was incubated for 10e30 min at room temperature, during which time the sample was intensively shaken using a vortex at least three times (for approx. 1 min). Approximately 100 þ 100 ml of the mixture of sample and buffer was pipetted into the analyte cavity of the test device so that it was saturated with the liquid. In the case of the microtitration plate ELISA test BioKits e cooked species identification test kit, we used 2e5 mg of grinded/ crushed carbonised sample and 400e500 ml of saline distilled water (concentration based on the manufacturer's guidelines) and the sample was incubated for 10e30 min at room temperature, during which period the sample was intensively shaken using a vortex at least three times (for approx. 1 min). For the ELISA cooked species identification test, 100 ml of the mixture was always used in each testing well on the microtitration plate following the manufacturer's specifications. The larger total volume of sample solution for this test is given by the need to divide it amongst more microwells dedicated to testing for the presence of individual species. For the BioKits bLG (b-lactoglobulin) competitive ELISA test a similar approach was adopted, but overall the dilution of the sample powder with 150 ml of extraction buffer (0.05 M carbonate/ bicarbonate buffer pH 9.6) was sufficient for one analysis. BLG biotin was added to the extract solution as the secondary antibody.

30

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

Fig. 1. Analysed pottery fragment with an attached food residue, from the site of m e stí Republiky. Prague-Na

Similarly, during the detection of goat's milk using the RIDASCREEN® GIS kit, 100 ml distilled H2O usually sufficed for one sample. General note: regardless of the type of kit, it is sometimes necessary to use up to 200 ml distilled H2O to dissolve some of the crushed or grounded residues, as the dry sample powder soaks up a part of the liquid. A smaller extract volume is always left for the subsequent analysis compared to the amount of liquids used for the extraction.

3.4. Laboratory procedure ELISA: Specific antibodies are bound to the bottom of microwells in a microwells module. The antibodies are able to bind to the specific proteins in a sample. These reactions were made visible using biotinylated species-specific antibodies and a sequence of washing and colouring steps. Positive reactions were detected by a yellow colour similar to the simultaneously generated positive control, which differs from the negative control. We used spectrophotometric detection of the results on Elisa reader VERSAmax™ (Molecular Devices) at 450 nm. Rapid 3-D: These lateral flow tests are ready-to-use devices delivered in a plastic holder that has a small opening with the analyte sample pad at one end and a reading window. Analyte (proteins dissolved in buffer) is delivered on the sample pad and the solution migrates into the interior of device, where it first

Fig. 2. Analysed pottery fragment with an attached food residue, from the site of Aldasín.

reaches the conjugate pad. Here the fluid coming from the sample pad is mixed with the particulate conjugate, thus preparing the analyte for the next step. Capillary forces make the solution travel through a porous membrane (reaction matrix) toward detection strips where it meets immobilised molecules (antibodies or antigens, depending on the design of the particular assay) which interact with the flowing analyte. Excess reagents continue to migrate through the matrix past the detection strips and finally are entrapped in the wick, which thus enables constant movement of the analyte through the device (O'Farrell, 2013; Posthuma-Trumpie et al., 2009; Posthuma-Trumpie and Van Amerongen, 2012). Within 5e15 min, distinct lines become visible in the reading window. When working with archaeological material, where the concentration of recognisable protein may be very low, it is advisable to allow for more time e we checked the results after at least one hour. The type of test used in our study has three detection lines (Fig. 3). One (marked C ¼ control) indicates that the method was successfully applied (analyte correctly reached the end of test strip) and the next two lines may show either none, low or high degree of detection according to their appearance (visible T line ¼ positive test, invisible O line ¼ high presence of/overload with target antigen). 3.5. Verification Three types of verification experiments were conducted to test the reliability of the kits selected for our study.

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

Fig. 3. Rapid 3D lateral flow test (positive and negative results indicated by detection lines).

1) The immunochromatographic testing kits were verified on present-day defined food samples (n ¼ 4: bovine and goat milk, pork and beef), which were deliberately carbonised on a shard of a modern unglazed ceramic. The pottery with adhering residues was buried for six months (February to July 2010) in the open landscape characterised by the brown soil, in the depth of 0.4 m (Location: Úlice near Pilsen, Czech Republic). 2) In the second experiment, we used real archaeological samples  dka (S  et al., 2015), dated from a recent excavation at Zahra alkova to the Iron Age (n ¼ 2). These samples (#5 and 6) were analysed independently by two methods: 1) by antibody testing, using the methodology described in this paper; and 2) by liquid chromatography-electrospray ionisation-quadruple-time of flight tandem mass spectrometry (LC-ESI-Q-TOF tandem MS). 3) We re-analysed some residues after several years since the first analysis, which had originally yielded positive results by immunoassays. These consisted of the residues stored in different conditions. Two samples (#10 and 11) were re-analysed by the same immunological methodology after they had been kept for four years in a laboratory cold storage (at 4  C) while the other two samples (#58 and 59) were re-analysed after whole pots with attached residues had been kept for seven years in a normal office with a temperature of about 20  C. Four other archaeological samples from the same office were also reanalysed by MS (#32, 64, 65 and 67), plus two controls (bovine and goat milk residues #78 and 79), buried in the ground five years previously (experiment 1 above). This last experiment was intended to compare the survival of biomolecules in various types of collections of archaeological material. Laboratory procedure of the proteomic analysis. Prior to mass spectrometry analysis the samples were submitted to specific proteolytic cleavage. Approximately 10 mg of each sample were placed into 20 mL of solution of 10 mg/mL of sequencing grade trypsin (Promega) in 50 mM NH4HCO3 and incubated at room temperature for two hours. The solution containing released peptides was desalted using ZipTips packed with reversed phase (C18) resin and evaporated to dryness. Subsequently samples were dissolved in the mixture of water: acetonitrile: formic acid (97:3:0.1%), and then loaded on trap column Acclaim PepMap 100 C18 (100 mm  2 cm, particle size 5 mm, Dionex, Germany) with mobile phase flow rate of A (0.1% formic acid in water) 5 mL/min for 5 min. The peptides were eluted from trap column to analytical column Acclaim PepMap RSLC C18 (75 mm  250 mm, particle size 2 mm) by mobile phase B (0.1% formic acid in acetonitrile) using the following gradient: 0 min 3% B, 5 min 3% B, 85 min 50% B, 86 min 90% B, 95 min 90% B, 96 min 3% B, 110 min 3% B. The flow rate during gradient separation was set to 0.3 mL/min. Peptides were eluted directly to the ESI source e Captive spray (Bruker Daltonics, Germany). Measurements were carried out in positive ion mode with precursor ion selection in the range of 400e2200 Da; up to ten

31

precursor ions were selected for fragmentation from each MS spectrum. Measurements were carried out using UHPLC Dionex Ultimate3000 RSLC nano (Dionex, Germany) connected with mass spectrometer ESI-Q-TOF Maxis Impact (Bruker, Germany). Peak lists were extracted from raw data by Data Analysis (Bruker Daltonics, Germany). Proteins were identified using Mascot version 2.2.04 (Matrix Science, UK) by searching protein database Uniprot version 20110e12, specifically against Bos taurus and Gallus gallus databases with a set modification - oxidized proline. Parameters for the database search were set as follows: oxidation of methionine and hydroxylation of proline as variable modifications, tolerance 50 ppm in MS mode and 0.05 Da in MS/MS mode. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD004337 and 10.6019/PXD004337 (Vizcaíno et al., 2016). 4. Results 4.1. Samples dated to prehistory Although we are only dealing with 13 samples from several periods, it appears that remains attached to the sampled cooking vessels represent a relatively rich range of ingredients (Table 2). In particular, we worked with material from the Eneolithic period (fourth to third millennium BC) (John and Pavelka, 2010), detecting either casein (of indeterminate origin) or, with a weaker signal, ovomucoid. A single tested Late Bronze Age sample (tenth to seventh century BC) proved positive for goat casein and gluten, the presence of egg white protein was questionable. The material from the Early Iron Age Hallstatt culture (eighth to sixth centuries BC) seems to be divided into two groups. One showed a strong reaction against gluten antibodies plus a weaker signal for both cattle and goat milk, as well as for egg white and hazelnuts, the last mentioned being tested only in these two samples. The second group of potsherds lacked gluten, but they were positive for animal meat proteins, specifically for pork, mutton, and e with a weaker signal e beef and poultry. Reaction to casein and ovomucoid was weak or dubious in this group. In the remaining two prehistoric samples dated to the Roman/Migration period (first to seventh century AD); one was negative for all types of the tested antibodies, the other indicated the presence of cattle milk. 4.2. Samples dated to the Middle Ages Mediaeval pottery analysed in our study can be classified according to morphological traces and ceramic fabric into several chronological stages: Early Mediaeval (our samples dated specifically to the fifth to sixth century), High (twelfth to thirteenth century), High/Late (thirteenth to fourteenth century), and Late (fourteenth to fifteenth century) Mediaeval periods (Table 2). We analysed eight samples dated to the Early Middle Ages (Pavelka and Vareka, 2008). The most frequently detected protein was ovomucoid, although the positive reactions were mostly only weak. Similarly, casein produced only one strong response, attributable to cattle milk, and three weaker detections. Animal protein was also found; we detected one standard reaction against mutton antibodies and several weak ones (mutton, beef, less likely poultry). The group of High Mediaeval pottery consisted of seven samples coming from farmsteads, i.e. non-urban environments. The strongest identified component in food remains was casein; in Mlad a Boleslav it originated from cattle milk, in Horní Bríza from goat milk respectively. In one case, the goat milk was accompanied by glutencontaining cereals. The High/Late Mediaeval category of finds comes from urban

32

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

contexts. Interestingly, positive reactions were less frequent in this group and weaker in comparison with the previous groups. Samples from Prague provided some indications for cattle milk, from Rabí there were minor traces of goat milk, and from Pilsen weak detections of animal meat (poultry and/or pork). The last group of samples belongs chronologically to the Late Mediaeval period. We analysed 24 pieces of pottery with mixed results. The urban environment of Prague repeatedly provided a combination of goat milk (less often cattle milk) and eggs, sometimes together with gluten cereals and in one case possibly with mutton. Finds from Pilsen (town) and Kyzy (village) were very poor with regard to positive detections, with faint traces of milk or poultry. Samples from the village of Hony showed the presence of casein and gluten. Lastly, out of three finds recovered in Mlad a Boleslav monastery, one was negative, with an inconclusive trace of milk, while the other two provided strong signals for a combination of cattle milk and animal meat protein (beef and mutton). 4.3. Results of method validation experiments 4.3.1. Experimental burial of charred residues Four different samples of charred food residues containing known ingredients (bovine milk, goat milk, pork and beef) were excavated after spending six months in the ground. All four samples were positively tested for the presence of the respective proteins by the immunological tests used in this study (samples #78e81 in Table 2). 4.3.2. Mass spectrometry of archaeological samples For two samples it was possible to carry out mass spectrometry analysis alongside the immunochromatography from the freshly excavated material. The first analysis (#5) yielded no protein match, while the second (#6) identified bovine serum albumin that may correspond to the immunologically detected bovine milk. Only one or two peptide matches were found. This may seem as not very significant confirmation, but we must take into account that the two methods work on a different basis, which will be considered in the discussion. 4.3.3. Sample storage effect To get a better understanding of how denatured proteins survive in various storage conditions, we repeated some tests on material kept for several years 1) in cold storage and 2) at room temperature. On samples from laboratory cold storage (#10e11) the immunological tests were successfully repeated, while residues on the pottery located since the first analyses in a normal office (#64 and 67) proved to be negative on the second attempt, contrary to the original tests. Concurrently conducted proteomic analysis (LC-MS/MS) of the latter material (i.e. kept in an office at room temperature) resulted mostly in a high number of matches corresponding to human keratins. Besides the keratins, no positive results identified as food residues were obtained on archaeological samples (#64 and 67).  Boleslav-Karmel, however, yielded a One sample (#32) from Mlada weak signal for ovalbumin, thus suggesting the presence of eggs, although egg white protein ovomucoid tested in the same residue by immunology was negative. Slightly better correspondence appeared in one sample from Prague-Republic Square (#65) with ovalbumin, collagen alpha-1(I) and collagen alpha-2(I) chain identified by MS, where egg had previously been detected by immunology as well. The recent control sample known to have originated from bovine milk (#78) produced no match 5 years after the original immunological analysis, while in the goat milk control (#79) alpha-S2-casein and beta-casein (one peptide each) were detected.

5. Discussion 5.1. Methodological issues The sample collection used for this study covers a very long period of human history spanning some five millennia. It contains a relatively small number of specimens for individual sub-periods. Therefore, the results (Table 2) mostly provide evidence for the applicability of the method on samples of diverse ages and recovery contexts. On a preliminary basis, and with an obvious sampling bias in mind that could only be overcome through an extended largescale analytical program, the data can contribute to a partial reconstruction of past cooking practices and diets. Our study, conducted solely on archaeological material from the Czech Republic, is lacking any samples from the Neolithic period of the first farmers in the Central European sense (which pre-dates the Neolithic as it is defined in NW Europe). It starts with the Eneolithic, but the number of analyses is quite low for the entire prehistoric period to draw a sketch of the dietary practices that could be convincingly relied upon. The important lesson learned from our study is that the application of immunochromatographic assays is technically possible to samples of considerable date (one to six millennia old). On the other hand, the biological nature of the residues makes them susceptible to contamination and degradation, which means that their analysis will hardly ever be an easy and routine task. It is highly advisable to continuously educate archaeologists and museum curators about the unique information potential of organic residues found at excavations and that this precious material requires special handling and storage. It should be kept separately from other finds, ideally in a cold and sterile environment, and should be submitted to a specialised laboratory as soon as possible after its recovery. We could see in the samples included in our study that residues may become seriously deteriorated from chemically aggressive diagenetic conditions, and this poor state of preservation may only be worsened if such finds are cleaned and processed during the post-excavation phase the same way as regular finds without organic attachments. Even the long-term storage should happen in laboratory facilities with controlled conditions, not in common offices or storage rooms, where inconvenient temperature, humidity and the presence of microorganisms may cause irreparable damage to the remaining preserved biomolecular structures. Exposure to such an environment is also an open door for potential contamination from various sources, which further complicate the analysis (we concluded that e.g. human keratins easily enter the porous target material, as dead skin cells represent a significant part of dust circulating in the buildings). While immunological methods by principle target only specific biomolecules of interest, in MS-based proteomics signals of lowabundance ancient proteins may easily be suppressed by signals from high-abundance proteins that may well be such unwanted admixtures, impurities and recent contaminations. 5.2. Contribution to the study of past human diet The results show and confirm what could be anticipated from other sources: the use of prehistoric pots for cooking cow and goat milk (sheep milk could not be tested in our study), frequently together with glutenous cereals and eggs, perhaps in some form of porridge. It is notable that there seems to also be another separate group of pots, which served for cooking meat (pork, mutton, and less markedly beef and wild poultry) where any traces of the formerly mentioned ingredients are missing or dubious. Very similar results were found for the beginning of the Early Middle Ages in our study region, where the subsistence strategies were in

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

most aspects just a continuation of prehistoric practices. More analyses of this kind are certainly desirable and expansion of our database will bring more light to our knowledge of this distant time period. For the mediaeval period, we have found a significant difference between the food residue patterns identified in urban centres and small settlement units located in the countryside. The farmstead excavated at Mlad a Boleslav-Karmel (13the14th century) showed regular use of cow and goat milk, in one case together with meat from farmhouse animals (beef, pork, poultry), eggs and glutenous cereals. A similar situation seems to be witnessed in other countryside village settlements, but so far very few samples have been analysed (Pavelka and Vareka, 2008; Trnka and Zelenka, 2012). Slightly different types of residues, combining cow milk and animal meat (beef and mutton together), come from the late mediaeval  Boleslav. monastery in Mlada The organic remains on ceramics from middens and cesspits from high/late mediaeval town of Pilsen were also tested (17 samples altogether), representing refuse accumulated from the thirteenth and fourteenth centuries to the end of the fifteenth century. The ceramics were extracted from cesspit fills consisting of dark, damp, and malodourous material. Here, the results of the tests were significantly weaker compared to the previously mentioned cases from a relatively dry environment (Pavelka and Orna, 2011). Although organic attachments to pottery finds were abundant, we could not identify any foodstuffs in many of the samples with the methodology described in this paper. Nevertheless, we identified weak signals of cereals (gluten) and animal proteins from the contexts dated to the earlier phase (13the14th century), which were mostly detected separately, i.e. not combined in one sample. In several cases, we detected pork and poultry together. It was only the later phase samples dated to the fourteenth to fifteenth centuries in which weak traces of milk were confirmed, and notably this was always goat milk. It is interesting that wherever we found evidence of animal meat protein there was no trace of milk or cereals. We can therefore consider the possibility of the specialised use of vessels for heat-treatment of either meat or cereal-based meal. Because the samples used for the analysis were excavated 20e50 years ago, it was not possible to analyse the character of the infills in which the shards remained since the Middle Ages. It is likely that the cesspit character of the sediments could have helped with the preservation of some organic finds, like wood or seeds, but proteins important for detection by our assays were probably washed away or badly deteriorated due to a chemically aggressive environment. It was not possible to find out whether detergents were used for cleaning of the pottery in a museum, but it is probable as it was routine museum practice at the time of their discovery; this factor plus less suitable storage conditions might also have caused the comparatively weak reactions against antibodies in our assays. Our data for the high and late Middle Ages in Prague also suggest that the analysed pots were used for cooking simpler meals when compared to the sampled farmhouse and monastery. The samples we could use dated from the second half of the fourteenth century and the first half of the fifteenth century indicate the prevalence of goat milk over cattle milk in urban contexts as well as the near absence of meat proteins. The signals obtained for milk were stronger when compared to Pilsen, which may be the result of different chemical characteristics of recovery contexts (the samples from Prague come from a drier environment without concentrated biological waste). The type of meal prepared in a pot could often be determined from identified ingredients. A sample obtained from a jar excavated in the town hospital in Prague can suggest the serving of heated

33

goat milk to the sick. The samples retrieved from urban plots offer interesting insight into the usual diet and/or culinary practices of specific groups of urban dwellers, where animal proteins are completely lacking on the studied pottery refuse. In the recovered pots showing visible food residues, people most often prepared cereal (flour?)-based meals, eggs, and very commonly e goat milk. We can consider some kind of porridge with beaten eggs cooked in goat milk. The presence of goat milk and eggs together can suggest milky soup (Pavelka and Vareka, 2008). The absence of meat proteins probably does not prove the scarcity of meat on most townspeople's tables. We have plenty of literary evidence as well as abundant finds of animal bones in domestic refuse forming archaeological layers showing that meat was a regular part of the diet in mediaeval towns. Therefore, we may conclude that meat was probably not cooked in the pottery vessels that we have tested, but in others, where the process of cooking did not leave characteristic charred residues. We know that meat was also prepared in numerous other ways e for instance by smoking or roasting. Last but not least, it is well attested that townspeople consumed fish , 2006, 2007), which and other sources of animal protein (S uvova could not be detected by the assays we used in this study. The wider knowledge of historic context is therefore always necessary for the interpretation of results. 6. Conclusions The study of food consumption practices in the more distant past has often been based solely on analyses of osteological material and botanical macro-remains from archaeological excavations, and sometimes on written accounts related to historical periods. The use of mass spectrometry certainly represents an important milestone in the study of past diet and culinary practices, but it has been used only in a small number of research centres due to its technical complexity. Immunological tests are therefore a useful tool which can help to reveal the composition of past diets and they will deserve wider application in historically oriented research, ideally together with a range of complementary approaches. Although the GC-MS, or more recently LC-ESI-Q-TOF or LC-MALDITOF MS is currently the leading method in the research of ancient food residues, we think that many questions this method aims to answer could in many cases be studied just as efficiently and perhaps even more easily, by the antibody food tests, targeting specific proteins of individual food components. Our study tested whether the commercially available immunoassays designed for the contemporary food industry can be used for analyses of archaeological finds of food residues. We used ready-to-use kits specifically designed for heat-treated foodstuffs, which target resistant proteins and their denatured/damaged forms. It is notable that some immunological assays developed as food-safety control tools give good results when applied to desiccated or charred remains that are several centuries or even millennia old. Similar tests can be used at almost any archaeological establishment and do not require expensive laboratories. The method described in this paper can bring a better understanding of dietary practices in particular cultural and social contexts, especially when combined with other available data, provided by natural science, literary and iconographic evidence. Our results show that their successful application in archaeology and related disciplines is possible and, with the recent advances in the field of immunochromatography, it would be advisable to develop tests optimised for use on ancient material. Such new designs of immunological assays should take into consideration knowledge of environmental and diagenetic factors in different regions, and make use of a combination of strip and chip technologies to prepare tests that can rapidly analyse more biomarkers of interest in a

34

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35

single test. One factor however seems critical to successful analysis of food residues: potential samples on pottery recovered from archaeological contexts must be treated as biological material that is susceptible to irreparable damage and contamination; if it is treated sub-optimally during on-site processing, cleaning and later museum storage, the bioarchaeological information potential may be irreversibly lost. The development of a relatively cheap and fast methodology that would be applicable already in field and museum conditions (such as immunochromatography) could indicate the type of parent material of these organic residues and suggest where and how they can be best studied in detail. Because every method has its strengths and weaknesses, immunology also can e in its specifically targeted approach e detect low concentrations of denatured biomolecules, which may prove difficult to effectively reveal by methods that are sensitive to a wider spectrum of targets, such as the case of LC-MS. Conflict of interest The authors declare that they have no conflict of interest. Acknowledgements This work was supported by the project of the Faculty of Environmental Sciences, Czech University of Life Sciences Prague: Environmental aspects of sustainable development (IGA FES 4290013123166) and by the Specific University research project (MSMT No 20/2016) carried out at the University of Chemistry and Technology in Prague. The authors wish to thank Peter Koník (Proteomic centre, Faculty of Science, University of South Bohemia,   Bude jovice, Czech Republic) who helped with mass specCesk e trometry measurements used for the verification of the method described in this paper and to Frantisek Jakubec (Keyence International) for digital microscopy of the studied samples. Peter Koník, Luk as Kucera and two anonymous reviewers provided valuable comments and suggestions on an earlier draft of this paper. Appendix A. Supplementary data Supplementary data related to this article (MS proteomic spectra of the sample M31056, corresponding to the sample #65 in Table 2) can be found at http://dx.doi.org/10.1016/j.jas.2016.07.004. References Bjorklund, E., Pallaroni, L., von Holst, C., Unglaub, W., 2001. Method of determination of appropriate heat treatment of animal meal by immunoassay developed for detection of cooked beef: interlaboratory study. J. AOAC Int. 84, 1839e1845. Brandt, E., Wiechmann, I., Grupe, G., 2002. How reliable are immunological tools for the detection of ancient proteins in fossil bones? Int. J. Osteoarchaeol 12, 307e316. Buckley, M., Melton, N.D., Montgomery, J., 2013. Proteomics analysis of ancient food vessel stitching reveals >4000-year-old milk protein. Rapid Commun. Mass Spectrom. 27, 531e538. Child, A.M., Pollard, A.M., 1992. A review of the applications of immunochemistry to archaeological bone. J. Archaeol. Sci. 19, 39e47. Collins, M.J., Westbroek, P., Muyzer, G., de Leeuw, J.W., 1992. Experimental evidence for condensation reactions between sugars and proteins in carbonate skeletons. Geochim. Cosmochim. Acta 56, 1539e1544. Craig, O.E., Collins, M.J., 2000. An improved method for the immunological detection of mineral bound protein using hydrofluoric acid and direct capture. J. Immunol. Methods 236, 89e97. Craig, O.E., Taylor, G., Mulville, J., Collins, M.J., Parker Pearson, M., 2005. The identification of prehistoric dairying activities in the Western Isles of Scotland: an integrated biomolecular approach. J. Archaeol. Sci. 32, 91e103. Cramp, L.J.E., Evershed, R.P., Lavento, M., Halinen, P., Mannermaa, K., Oinonen, M., Kettunen, J., Perola, M., Onkamo, P., Heyd, V., 2014. Neolithic dairy farming at the extreme of agriculture in northern Europe. P Roy. Soc. B-Biol Sci. 281, 17910e17915.

Day, J., 2013. Botany meets archaeology: people and plants in the past. J. Exp. Bot. 64, 5805e5816. Dincauze, D.F., 2000. Environmental Archaeology: Principles and Practice. Cambridge University Press. Dongoske, K.E., Martin, D.L., Ferguson, T.J., 2000. Critique of the claim of Cannibalism at cowboy wash. Am. Antiq. 65, 179e190. Dudd, S.N., Evershed, R.P., Gibson, A.M., 1999. Evidence for varying patterns of exploitation of animal products in different prehistoric pottery traditions based on lipids preserved in surface and absorbed residues. J. Archaeol. Sci. 26, 1473e1482. Dzantiev, B.B., Byzova, N.A., Urusov, A.E., Zherdev, A.V., 2014. Immunochromatographic methods in food analysis. Trends Anal. Chem. 55, 81e93. Forshaw, R., 2014. Dental indicators of ancient dietary patterns: dental analysis in archaeology. Br. Dent. J. 216, 529e535. Hofreiter, M., Collins, M., Stewart, J.R., 2012. Ancient biomolecules in Quaternary palaeoecology. Quat. Sci. Rev. 33, 1e13. Hong, C., Jiang, H., Lu, E., Wu, Y., Guo, L., Xie, Y., Wang, C., Yang, Y., 2012. Identification of milk component in ancient food residue by proteomics. PLoS One 7, e37053. Hynek, R., Kuckova, S., Hradilova, J., Kodicek, M., 2004. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry as a tool for fast identification of protein binders in color layers of paintings. Rapid Commun. Mass Spectrom. 18, 1896e1900.  analýza rezidua na vnitrní strane  eneoliticke  John, J., Pavelka, J., 2010. Imunologicka dobky z lokality Otmí  hora (okr. Beroun). In: Kristuf, P., na ce eOtmí cska Vareka, P. (Eds.), Opomíjen a archeologie 2007e2008 (Neglected Archaeology  Za pado  univerzity v 2007e2008), Katedra archeologie Fakulty filozoficke ceske Plzni, pp. 156e157. Klippel, W.E., 2001. Sugar monoculture, Bovid skeletal Part Frequencies, and stable carbon isotopes: interpreting enslaved African diet at Brimstone hill, St Kitts, West Indies. J. Archaeol. Sci. 28, 1191e1198. Kuckova, S., Hynek, R., Kodicek, M., 2009. Application of peptide mass mapping on proteins in historical mortars. J. Cult. Herit. 10, 244e247. Kuckova, S., Sandu, I.C.A., Crhova, M., Hynek, R., Fogas, I., Schafer, S., 2013. Protein identification and localization using mass spectrometry and staining tests in cross-sections of polychrome samples. J. Cult. Herit. 14, 31e37.  Va , S., vra, J., Hynek, R., Kodí Ku ckova cek, M., 2010. Analýza obsahu keramických a ných na dob metodou peptidove ho mapova ní. In: Beljak, J., Brezinova , G., sklene gia Barbarov 2009. Hospoda rstvo Germ Varsik, V. (Eds.), Archeolo anov, Nitra, pp. 639e644. Landon, D.B., 2005. Zooarchaeology and historical archaeology: progress and prospects. J. Archaeol. Method Th 12, 1e36. Matsuda, T., Watanabe, K., Nakamura, R., 1982. Immunochemical studies on thermal denaturation of ovomucoid. Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 707, 121e128. O'Farrell, B., 2013. Lateral flow immunoassay systems: evolution from the current state of the art to the next generation of highly sensitive, quantitative rapid assays. In: Wild, D.G. (Ed.), The Immunoassay Handbook, fourth ed. Elsevier, Oxford, pp. 89e107.  , L., Smejda, Pavelka, J., Kova cikova L., 2011. The determination of domesticated animal species from a Neolithic sample using the ELISA test. C R. Pale 10, 61e70.  Pavelka, J., Orna, J., 2011. Výsledky analýzy potravinových zbytk u na pozdne ke  keramice v Plzni. Acta Fak. filoz. Z  univerzity V. Plzni 3, stredove apado ceske 85e98. vek k pozna ní stravy ve vrcholne m a pozdním Pavelka, J., Vareka, P., 2008. Príspe ku: první výsledky analýzy potravinových zbytk stredove u na keramice z j 10, 98e109. archeologických výzkum u. Kude Pollard, M., Batt, C., Stern, B., Young, S.M.M., 2007. Analytical Chemistry in Archaeology. Cambridge University Press, Cambridge e New York. Posthuma-Trumpie, G.A., Korf, J., van Amerongen, A., 2009. Lateral flow (immuno) assay: its strengths, weaknesses, opportunities and threats. A literature survey. Anal. Bioanal. Chem. 393, 569e582. Posthuma-Trumpie, G.A., Van Amerongen, A., 2012. Lateral flow assays. In: Meulenberg, E.P. (Ed.), Antibodies e Applications and New Developments. Bentham Science Publishers, pp. 175e183. Regert, M., 2011. Analytical strategies for discriminating archeological fatty substances from animal origin. Mass Spectrom. Rev. 30, 177e220. Reitsema, L.J., Crews, D.E., Polcyn, M., 2010. Preliminary evidence for medieval Polish diet from carbon and nitrogen stable isotopes. J. Archaeol. Sci. 37, 1413e1423.  lkov k, A., Brezinova , H., Farkasova , K., Houfkova , P., Chvojka, O., John, J., Sa a, T., Bezde  Nova , L., Micha lek, J., Msallamov k, J., Kmosek, J., Koník, P., Kova cikova a, S.,  kov Pavelka, J., Sula a, H., Besta, T., Myskov a, E., Weiter, L., Zronek, P., 2015. Bioarchaeological reconstruction of the funeral rite e case study based on organic material from the Hallstatt Period tumulus at the site Zahr adka (South Bohemia, Czech Republic). Pam atky Archeol. 106, 95e135. Salvini, L., Pecci, A., Giorgi, G., 2008. Cooking activities during the Middle ages: organic residues in ceramic vessels from the Sant'Antimo church (PiombinoCentral Italy). J. Mass Spectrom. 43, 108e115. lu ze trí pozdne  stredove kých S uvov a, Z., 2006. Archeozoologick a analýza materia studen v Plzni. In: Hasek, V., Nekuda, R., Ruttkay, M. (Eds.), Ve slu zb ach dn mia vied, archeologie VII, Muzejní a vlastive a spole cnost-Slovensk a akade Brno-Nitra, pp. 255e260.  na lezy z pozdne  stredove ke  jímky v Perlove  ulici S uvov a, Z., 2007. Archeozoologicke  zastavení). Sborník Za pado ho muzea V. Plzni 18, v Plzni (metodicke ceske

J. Pavelka et al. / Journal of Archaeological Science 73 (2016) 25e35 148e153.  stredove ke  osídlení na katastru Horní Brízy, Trnka, R., Zelenka, A., 2012. Vrcholne   sever. Archeol. za padních Cech okr. Plzen 4, 53e87. van der Veen, M., 2007. Formation processes of desiccated and carbonized plant remains e the identification of routine practice. J. Archaeol. Sci. 34, 968e990. Vizcaíno, J.A., Csordas, A., del-Toro, N., Dianes, J.A., Griss, J., Lavidas, I., Mayer, G.,

35

Perez-Riverol, Y., Reisinger, F., Ternent, T., Xu, Q.W., Wang, R., Hermjakob, H., 2016. 2016 update of the PRIDE database and related tools. Nucleic Acids Res. 44 (D1), D447eD456. Wright, P., 2003. Preservation or destruction of plant remains by carbonization? J. Archaeol. Sci. 30, 577e583.