Identifying Proteins Using Size Exclusion Chromatography

Laboratory Write Up:

The aim of this practical was to identify three unknown proteins by using size exclusion chromatography and ion exchange chromatography to separate the proteins and then analyse the unknown proteins using SDS-Page.

Size exclusion chromatography is also known as gel filtration and separates proteins depending on molecular weight (MW) which differs in each protein due to the difference in amino acids (Noronha et al., 2007). Size exclusion chromatography is unique as the molecules do not bind to the resin compared to ion exchange chromatography where the resin and its pH determines whether the protein is eluted first or not (Shabram et al., 2016). Size exclusion chromatography uses a resin which has pores and is made up of bead like particles (Paul Shabram, 2016). This allows the proteins with a smaller MW to pass through the column via diffusion in the pores whereas the proteins with larger MW’s are eluted (Shabram et al., 2016). The medium has a specific MW where if the MW of the protein is above the medium, it will be eluted whereas if it has a MW below it will bind to the column due. Some of the most common resins that are used in size exclusion are Sephadex and Bio-Beads (Fletouris, 2007) and Sephadex G-100 was used in this experiment.

Figure 1: A diagram explaining what happens within the column during size exclusion chromatography (ref)

Ion exchange chromatography separates proteins depending on the charge of the molecule for example whether it is positively or negatively charged. As amino acids are zwitterionic and therefore are positively and negatively charged, whether a protein is positively or negatively is based on the pH of the environment it is in (Cheriyedath, 2018). This type of chromatography uses a resin which can either be strongly acidic or weakly acidic cation exchange resins but can also be either strong basic or weak basic anion exchange resins (Marshall and Marshall, 2017) It is cation exchange for acidic resins because protons are positively charged whereas hydroxide ions are negatively charged and therefore it is anion exchange for bases. Cation exchange is where positively charged ions are attracted to a negative surface whereas anion exchange is where negatively charged ions are attracted to a positive surface which occurs during the stationary phase (Cheriyedath, 2018). The isoelectric point is the pH at which the protein has a charge of zero (Bhatia, Kiran Sharma and Randhir Dahiya, 2015)

.

Figure 2: A diagram explaining what happens within the column during ion exchange chromatography (ref)

SDS-Page is a type of electrophoresis that is used to identify the separation of proteins. SDS-Page uses sodium dodecyl sulfate and polyacrylamide gel to separate the proteins based on MW (Oswald, 2008). This is because the larger MW’s don’t move through the gel as quickly as the proteins with smaller MW’ (Oswald, 2008). As SDS-page is a type of electrophoresis this means that the bands are produced from the moving of proteins through an electric field, where there is an anode and a cathode (Fritsch and Krause, 2003). The proteins with smaller MW will migrate more quickly towards the anode (Oswald, 2008). Once the sodium dodecyl sulfate is applied it disturbs the proteins tertiary structure to unfold the proteins to allow the sodium dodecyl sulfate to “coat” each protein with a negative charge, allowing them to all migrate towards the anode to enable the distance to be measured (Oswald, 2008). From this bands are produced and unknown proteins can be identified from the distance the band has travelled. The process and final product of SDS-page can be seen in figure 3.

Figure 3 A diagram explaining how to perform SDS-page and the results after staining (ref)

Figure 4: A table showing the native MW’s subunit molecular weight, pI and the ratio between 410nm and 280nm absorbance of eight proteins. Reference for unkown proteins. (ref)

As there was a medium that had a molecular weight of 30kDa this means that the proteins in figure 4 that have a molecular weight above 30kDa are paired with the proteins in figure 4 that have a molecular weight below 30kDa. For size exclusion chromatography this is the list for pairwise combinations of proteins that would be separated:

1. Bovine serum albumin and cytochrome c (horse heart)
2. Bovine serum albumin and $\alpha$

   -Lactalbumin (bovine milk)

3. Bovine serum albumin and myoglobin (horse heart)
4. Bovine serum albumin and ribonuclease A (bovine pancreas)
5. Haemoglobin (bovine) and cytochrome c (horse heart)
6. Haemoglobin (bovine) and $\alpha$

   -Lactalbumin (bovine milk)

7. Haemoglobin (bovine) and myoglobin (horse heart)
8. Haemoglobin (bovine) and ribonuclease A (bovine pancreas)
9. Catalase (bovine liver) and cytochrome c (horse heart)
10. Catalase (bovine liver) and $\alpha$

    -Lactalbumin (bovine milk)

11. Catalase (bovine liver) and myoglobin (horse heart)
12. Catalase (bovine liver) and ribonuclease A (bovine pancreas)
13. Ferritin (horse spleen) and cytochrome c (horse heart)
14. Ferritin (horse spleen) and $\alpha$

    -Lactalbumin (bovine milk)

15. Ferritin (horse spleen) and myoglobin (horse heart)
16. Ferritin (horse spleen) and ribonuclease A (bovine pancreas)

As the resin used had a pH of 7.5 this means that the proteins in figure 4 that have a pI above 7.5 are paired with the proteins in figure 4 that have a pI below 7.5. For ion exchange chromatography this is the list for pairwise combinations of proteins that would be separated:

1. Bovine serum albumin and cytochrome c (horse heart)
2. Bovine serum albumin and myoglobin (horse heart)
3. Bovine serum albumin and ribonuclease A (bovine pancreas)
4. $\alpha$

   -Lactalbumin (bovine milk)and cytochrome c (horse heart)

5. $\alpha$

   -Lactalbumin (bovine milk) and myoglobin (horse heart)

6. $\alpha$

   -Lactalbumin (bovine milk) and ribonuclease A (bovine pancreas)

7. Haemoglobin (bovine) and cytochrome c (horse heart)
8. Haemoglobin (bovine) and myoglobin (horse heart)
9. Haemoglobin (bovine) and ribonuclease A (bovine pancreas)
10. Catalase (bovine liver) and cytochrome c (horse heart)
11. Catalase (bovine liver) and myoglobin (horse heart)
12. Catalase (bovine liver) and ribonuclease A (bovine pancreas)
13. Ferritin (horse spleen) and cytochrome c (horse heart)
14. Ferritin (horse spleen) and myoglobin (horse heart)
15. Ferritin (horse spleen) and ribonuclease A (bovine pancreas)

Materials and methods:

Size exclusion chromatography:

The materials that were used in this experiment were:

• Size exclusion column
• Size exclusion buffer
• 1.0% (w/v) Blue Dextran solution
• UV-Vis spectrophotometers
• UV-Vis microtitre plate
• Eppendorf tubes and rack
• Pasteur pipettes and teats
• Micro-pipettes and tips

The experiment was carried out by following a set of instructions in the BIO2090 practical schedule (pages 5-6) however some modifications were made. The void volume that was collected in a measuring cylinder was 14.7ml. The volume of buffer that was added was not numerically measured, it was added to allow enough buffer so that the column didn’t run dry. 40 1.5ml Eppendorf tubes were used to collect 40 fractions. The number of fractions remained at 40 as this was the point at which all of the proteins had been eluted. From the elution profiles drawn, fractions 7,8 and 9 were put aside for the first peak and then fractions 21, 22, and 23 were put aside for the second peak.

Ion exchange chromatography:

• DEAE cellulose ion exchange column
• Ion exchange elution buffer A (‘low salt’, 50mM Tris-HCl, pH 7.5)
• Ion exchange elution buffer B (‘high salt’, 50mM Tris-HCl + 0.5M NaCl pH 7.5)
• UV-Vis spectrophotometers
• UV-Vis microtitre plate
• Eppendorf tubes and rack
• Pasteur pipettes and teats
• Micro-pipettes and tips

This experiment was carried out by following a set of instructions in the BIO2090 practical schedule (pages 6-7) however some modifications were made. Instead of adding 30ml of buffer A approximately 42ml of buffer A was added as 28 fractions were collected, each of 1.5ml in Eppendorf tubes. Furthermore, 45ml of buffer B was added as 30 fractions were collected, again each of 1.5ml in Eppendorf tubes. From the elution profiles drawn, fraction 18 was put aside from buffer A for SDS-Page and fraction 52 was put aside from buffer B for SDS-Page.

SDS-Page:

This experiment was carried out by following a set of instructions in the BIO2090 practical schedule (page 7) however there were specific steps into making the 20ul mixture for each lane in SDS-Page. Firstly, to produce the sample for the first peak from size exclusion, 10ul of each fraction set aside for the first peak were pipetted into one Eppendorf tube. 20ul of this mixture was then pipetted into another Eppendorf tube that was labelled. For the second tube, the same method was applied as the first peak but using fractions. This method was again applied for ion exchange using fractions for the first peak and using fractions for the second peak. In the method in the BIO2090 hand out it expresses to heat the mixtures for 10-20 minutes, however ours were heated for 5 minutes. Finally, 10ul of the sample was loaded into a lane, 10ul of the mixture for the first peak for size exclusion was loaded into the lane and 10ul of the mixture for the second peak of size exclusion was loaded into the second lane. 5ul of the mixture for the first peak for ion exchange was loaded into the next lane and the same for the second peak for ion exchange.

Results:

Figure 5: A table of results for size exclusion chromatography showing the absorbance at 410nm and 280nm as well as the ratio between the two. The ratio increases up to fraction 25, the absorbance readings for 280nm increase slightly to produce a peak at fraction 9 and then increases again to produce a peak at 22. Whereas the absorbance readings for 410nm increase once to show one peak at fraction 23.

Figure 6: A graph produced from the absorbance readings at 280nmand 410nm from ion exchange chromatography. The graph shows two absorption peaks for 280nm, one broad between fractions 7 and 11 and one stronger peak at fraction 22. There is also one peak for 410nm at fraction 23.

Figure 7: A table of results for ion exchange chromatography for buffer A showing the absorbance at 410nm and 280nm as well as the ratio between the two. The ratio increases up to fraction 19, and the absorbance readings at both 280nm and 410nm increase showing a peak both a fraction 18.

Figure 8: A table of results for ion exchange chromatography for buffer B showing the absorbance at 410nm and 280nm as well as the ratio between the two. The ratio between A₄₁₀ and A₂₈₀ decreases, whereas the absorbance readings at 280nm increases to show one peak at fraction 23 and the absorbance readings at 410nm doesn’t vary a huge amount showing no peak at 410nm.

Figure 9: A graph produced from the absorbance readings at 280nmand 410nm from ion exchange chromatography. The graph shows two absorption peaks, one at 410nm at fraction 18  and one at 280nm at fraction 18 for buffer A. There is one absoprtion peak for buffer B 280nm at fraction 51 .

Figure 10: An SDS-Page gel with the markers labelled as well as which lane corresponded to which chromatography. This shows a band for size exclusion corresponding with unknown protein 1 and a second band showing the mixture of the two smallest proteins. Ion exchange shows the separation of these two proteins as two bands have been produced.

Figure 11: A table showing the distance travelled for each band in millimetres, the ratio between the base line and the distance travelled and the log of the molecular weight of the band.

Figure 12: A line graph plotting logMW against R(BPB) with a trendline that shows the R² equation on the line to be 0.83745, meaning there is a strong correlation between logMW and R(BPB).

Figure 13 An extrapolated line graph plotting logMW against R(BPB) with a trendline that shows the R² equation on the line to be 0.83745, meaning there is a strong correlation between logMW and R(BPB).

.

Figure 14: A table showing the distance travelled for each unknown protein, the ratio between the base line and the distance travelled, the log of the molecular weight which was calculated from the figure 13 using the ratios. Mass of the unknown proteins were then calculated from logMW.

Results:

For size exclusion chromatography, by looking at the absorbance’s at 280nm and 410nm it can be seen that two peaks are produced for 280nm whereas there is only one peak for 410nm. There is an increase in absorbance for 280nm between fractions 6 and 12, as the absorbance increases from fraction 6 to 9 and then decreases to 12 with a peak being at fraction 7, where the ratio is 0.273. In addition to this there is an increase between fractions 14 to 22 and then decreases to fraction 40 with a peak being at fraction 22, where the ratio is 2.322. Whereas for 410nm there is only one peak produced due to only one main increase and decrease between absorbance’s. The increase can be seen from fractions 13 to 22 and then decreases to fraction 34 where it then plateaus, with the peak having a ratio of 2.092.

For ion exchange chromatography, by looking at the absorbance’s at 280nm and 410nm it can be seen that two peaks are produced for 280nm whereas there is only one peak for 410nm. There is an increase in absorbance for 280nms between fractions 12 and 27, as the absorbance increase from fraction 12 to 18 and then decreases to 27, where the peak has a ratio of 2.182. There is also another increase between fractions 20 and 30 where it increases from fraction 20 to 23 and then decreases to fraction 30 with the peak having a ratio of 0.086. However, for 410nm only one peak can be seen as there is an increase between fractions 14 and 28, with an increase between fractions 14 to 18 and a decrease to fraction 28, where the peak has a ratio of 2.182.

The SDS-Page shows six bands on the marker lane showing the bands of 105, 78,55,45,34 and 6kDa, as well as markers 16kDa and 17kDa which had merged. In one lane, there is a large band that can be seen for size exclusion which shows the heaviest protein and the second lane for size exclusion shows two bands showing the two lighter proteins that were then further separated by ion exchange chromatography. In the first lane for ion exchange chromatography where one band can be seen for the protein that has a pI of above 7.5 and in the second lane for ion exchange chromatography one band can be seen for the protein that has a pI of below 7.5.

Discussion:

In figure 6, there are two peaks, one weak and strong peak for 280nm and one strong peak for 410nm. The first peak shows the separation of a protein with a larger MW than 30kDa as proteins with a larger MW than the resin was eluted first due to the protein not being stuck in the pores. The second peak is larger because it is a mixture of two proteins with a lower MW than 30kDa. This is the reason ion exchange was then used to separate the mixture of the two smaller proteins. In figure 4 there are two peaks, the first peak identifying a protein with a pI of above than 7.5 and the second peak identifying a protein with a pI of less 7.5 as the pH of the resin is 7.5.

The separation was successful as two peaks were observed in both figure 3 and figure 4 for both size exclusion and ion exchange meaning that three proteins have been separated. However, the smaller peak at for 280nm in figure 3 could have been more distinct showing that there may have been some errors encountered in the practical.

One error is the disruption of beads in the columns when adding the buffer or protein sample. Precautions were taken to eliminate this error however it is unlikely that the beads were undisturbed throughout the experiment and therefore this could have affected the separation of the proteins. By not having 1ml of sample for size exclusion and 1.5ml of sample in Eppendorf tubes for fractions may have affected the absorbance readings and therefore the peaks produced. This in turn would have affected the fraction samples used to produce a sample for the SDS-page. The use of SDS-page in itself produces many areas for error to occur. One being the loading of the samples, if the full sample was not loaded correctly into the gel this in turn would affected the bands produced on the gel as the sample would not be able to be separated correctly. Another error is the measurement of the distance travelled by the proteins as the gel produces thick bands and therefore measurements need to be taken from the middle of the band for accuracy which  is difficult to identify and therefore this would have affected the measured distance travelled and in turn the R(BPB) and therefore the graph used to identify the proteins.

Finally, in terms of analysis and calculation of the MW of unknown proteins, there is error due to the extrapolation of the graph. This is because extrapolation is based on unknown y values and therefore isn’t as accurate as a normal trendline.

Proteins absorbance’s are measured at 280nm because amino acids strongly absorb UV light at this wavelength mainly due to the amino acids tryptophan, cystine and tyrosine (Schmid, 2001) as well as proteins that contain aromatic rings. Protein absorbance’s are also measured at 410nm because heme groups absorb at around 400nm and iron is an element that is commonly found in amino acids.

Figure 6 shows that unknown protein 1 eluted therefore it has the largest MW, above 30kDa and it only produced one absorption peak, at 280nm. This means that this protein contains tryptophan, cystine and tyrosine. However, there is not a peak at 410nm, therefore there are no haem groups present in this protein. The mass of the protein calculated from SDS-Page is 79.4kDa meaning that the protein could be either bovine serum albumin (BSA) or catalase (bovine liver). BSA has a subunit MW of 66.5kDa (Sigma-Aldrich, 2018) and catalase has a subunit MW of 60kDa (Sigma-Aldrich, 2018) therefore accounting for error, the unknown protein could be either one of these proteins.  Catalase contains 4 haem groups (Kirkman and Gaetani, 1984) as well as tyrosine and tryptophan (Jakopitsch et al., 2003) accounting for two peaks produced in size exclusion, one at 280nm and one at 410nm showing that this protein must be BSA. Furthermore, the ratio for this peak is 0.273, which is close to the ratio of 0 for BSA (figure 4) accounting for error whereas catalase has a ratio of 1 (figure 4).

Figure 9 shows that unknown protein 2 was eluted first in ion exchange chromatography as it has a pI above 7 and it produced 2 peaks, one for 280nm and one for 410nm. The MW of this protein that was calculated from SDS-Page was 15.85kDa meaning that the protein could be either cytochrome c, myoglobin or ribonuclease A. Cytochrome C has a subunit MW of 12kDa (Sigma-Aldrich, 2018) and a pI of 10.6 (figure 4), myoglobin has a subunit MW of 17kDa (Zaia, Annan and Biemann, 1992) and a pI of 8.2 (figure 4). Finally, ribonuclease A has a subunit MW of 13.7kDa (Sigma-Aldrich, 2018) and a pI of 9.2 (figure 4), therefore accounting for error, the unknown could be either of these proteins. Ribonuclease A contains six tyrosine residues (Noronha et al., 2007) but it does not have a heme group therefore this cannot be the unknown protein as it would produce only one peak. In addition to this the ratio for this peak is 2.182, cytochrome c and myoglobin havea ratio of 4 (figure 4), which is closer to the calculated ratio than the ratio of ribonuclease A which is 0 (figure 4). Cytochrome c not only contains a heme group (Fisher, Taniuchi and Anfinsen, 1972) but also the amino acid tyrosine (Margoliash et al., 1962) and myoglobin also contains a heme residue (Ordway and Garry, 2004) as well as tyrosine and tryptophan (Truong et al., 1967), meaning they are hard to distinguish. However, on the SDS-Page, the band is closer to the 16/17kDa band and therefore is most likely myoglobin.

Figure 9 shows that protein 3 was eluted last in ion exchange as it has a pI below 7.5 and it produced only one peak, at 280nm. The MW of this protein that was calculated from SDS-page was 10.96kDa meaning that the protein could be either $\alpha$

-Lactalbumin (bovine milk) or haemoglobin (bovine). $\alpha$

-Lactalbumin (bovine milk) has a subunit MW of 14.2kDa (Sigma-Aldrich, 2018) and a pI of 4.8 (figure 4) and haemoglobin (bovine) has a subunit MW of 16kDa (Sigma-Aldrich, 2018) and a pI of 6.5 (figure 4), therefore accounting for error, the unknown could be either of these proteins. Haemoglobin (bovine) has four polypeptide chains where each contains one heme group (Alberts et al., 2015), which would’ve produced a peak at 410nm showing that this protein is most likely to be $\alpha$

–Lactalbumin. In addition to this the ratio for this peak is 1.17, $\alpha$

-Lactalbumin has a ratio of 0 (figure 4) which is closer to the calculated ratio that the ratio of haemoglobin which is 4 (figure 4).

As unknown protein 3 was determined to be $\alpha$

–Lactalbumin, this confirms that unknown protein 2 is myoglobin, as based on the theory of SDS-Page, unknown protein 2 should have a higher MW than unknown protein 3.

Conclusion

Through the use of size exclusion chromatography and ion exchange chromatography for protein separation and the use of SDS-Page for analysis, the three unknown proteins were identified. They were identified to be BSA, myoglobin and $\alpha$

–Lactalbumin with allowances between known data and this experimental data due to error within the experiment and analysis.

References

• Alberts, B., Johnson, A., Lewis, J., Morgan, D., Raff, M., Roberts, K. and Walter, P. (2015). Molecular biology of the cell. 6th ed. New York, NY: Garland Science, p.123.
• Bhatia, S., Kiran Sharma and Randhir Dahiya (2015). Modern Applications of Plant Biotechnology in Pharmaceutical Sciences. Elsevier Science, pp.121-156.
• Cheriyedath, M. (n.d.). How Does Ion Exchange Chromatography Work?. [online] News-Medical.net. Available at: https://www.news-medical.net/life-sciences/How-Does-Ion-Exchange-Chromatography-Work.aspx [Accessed 11 Nov. 2018].
• Fisher, W., Taniuchi, H. and Anfinsen, C. (1972). On the Role of Heme in the Formation of the Structure of Cytochrome c. The Journal of Biological Chemistry, 248(9), p.3188.
• Fritsch, R. and Krause, I. (2003). Encyclopedia of Food Sciences and Nutrition. 2nd ed. Munich, Germany: Elsevier Science Ltd, pp.2055-2062.
• Jakopitsch, C., Kolarich, D., Petutschnig, G., Furtmüller, P. and Obinger, C. (2003). Distal side tryptophan, tyrosine and methionine in catalase-peroxidases are covalently linked in solution. FEBS Letters, 552(2-3), pp.135-140.
• Kirkman, H. and Gaetani, G. (1984). Catalase: a tetrameric enzyme with four tightly bound molecules of NADPH. Proceedings of the National Academy of Sciences, 81(14), pp.4343-4347.
• Margoliash, E., Kimmel, J., Hill, R. and Schmidt, W. (1962). Amino Acid Composition of Horse Heart Cytochrome c. The Journal of Biological Chemistry, 237(7), p.2148.
• Marshall, K. and Marshall, K. (2017). What Is Ion Exchange Resin and How Does It Work?. [online] Samco Tech. Available at: https://www.samcotech.com/ion-exchange-resin-work-process/ [Accessed 11 Nov. 2018].
• Martin, P. (n.d.). How does the ultraviolet absorption of proteins impact analysis?. [online] Microspectra.com. Available at: http://www.microspectra.com/support/learn/protein-absorption [Accessed 10 Nov. 2018].
• Newworldencyclopedia.org. (n.d.). Hemoglobin – New World Encyclopedia. [online] Available at: http://www.newworldencyclopedia.org/entry/Hemoglobin [Accessed 10 Nov. 2018].
• Ordway, G. and Garry, D. (2004). Myoglobin: an essential hemoprotein in striated muscle. Journal of Experimental Biology, 207(20), pp.3441-3446.
• Oswald, D. (2008). How SDS-PAGE Works – Bitesize Bio. [online] Bitesize Bio. Available at: https://bitesizebio.com/580/how-sds-page-works/ [Accessed 5 Nov. 2018].
• Product Information Heme Assay Kit. (2016). [ebook] USA: Sigma-Aldrich, p.1. Available at: https://www.sigmaaldrich.com/content/dam/sigma-aldrich/docs/Sigma/Bulletin/2/mak316bul.pdf [Accessed 10 Nov. 2018].
• Schmid, F. (2001). [ebook] Germany: Macmillan Publishers, pp.1-4. Available at: http://www.life.illinois.edu/biochem/455/Lab%20exercises/2Photometry/spectrophotometry.pdf [Accessed 10 Nov. 2018].
• Sigma-Aldrich. (2018). Bovine Serum Albumin A8531. [online] Available at: https://www.sigmaaldrich.com/catalog/product/sigma/a8531?lang=en&region=GB [Accessed 10 Nov. 2018].
• Sigma-Aldrich. (2018). Catalase from bovine liver E009322. [online] Available at: https://www.sigmaaldrich.com/catalog/product/sigma/e009322?lang=en&region=GB [Accessed 9 Nov. 2018].
• Sigma-Aldrich. (2018). Cytochrome c from equine heart C7752. [online] Available at: https://www.sigmaaldrich.com/catalog/product/sigma/c7752?lang=en&region=GB [Accessed 10 Nov. 2018].
• Sigma-Aldrich. (2018). Ribonuclease A from bovine pancreas R6513. [online] Available at: https://www.sigmaaldrich.com/catalog/product/sigma/r6513?lang=en&region=GB [Accessed 10 Nov. 2018].
• Sigma-Aldrich. (2018). α-Lactalbumin from bovine milk L6385. [online] Available at: https://www.sigmaaldrich.com/catalog/product/sigma/l6385?lang=en&region=GB [Accessed 10 Nov. 2018].
• Truong, T., Bersohn, R., Brumer, P., Luk, C. and Tao, T. (1967). Effect of pH on the Phosphorescence of Tryptophan, Tyrosine, and Proteins. The Journal of Biological Chemistry, 242(12), p.2979.
• Zaia, J., Annan, R. and Biemann, K. (1992). The correct molecular weight of myoglobin, a common calibrant for mass spectrometry. Rapid Communications in Mass Spectrometry, 6(1), pp.32-36.
• Noronha, M., Lima, J., Paci, E., Santos, H. and Maçanita, A. (2007). Tracking Local Conformational Changes of Ribonuclease A Using Picosecond Time-Resolved Fluorescence of the Six Tyrosine Residues. Biophysical Journal, 92(12), pp.4401-4414.
• Fletouris, D. (2007). Food Toxicants Analysis. Elsevier B.V, pp.299-348.
• Shabram, P., Vellekamp, G., Ruan, Q. and Scandella, C. (2016). Adenoviral Vectors for Gene Therapy. 2nd ed. Elsevier Science, pp.197-230.
• Nagy, K. and Vekey, K. (2008). Medical Applications of Mass Spectrometry 2008. Elsevier Science, pp.61-92.