Background
Why We Chose to Investigate Certain Proteins and Pepsin:
We chose to investigate certain proteins based on their concentration and significance to certain properties found in human milk. The proteins we found to be the most significant are albumin, beta-casein, kappa-casein, lactoferrin, lysozyme C, IgA Fc receptor, and IgG receptor. Albumin, casein, lactoferrin, lysozyme, and IgA are the most abundant proteins found in human milk [1]. IgG was chosen due to its role as the infant’s primary immune defense during the first few months of life [2]. Pepsin is an enzyme produced in preterm and term infants. As early as 16 weeks (about 3 and a half months) of gestation, pepsin can be found in the stomach of infants [3]. We chose to consider pepsin at both pH 1.3 and pH >2 in order to investigate the different cleavage sites between the two pH’s. Pepsin hydrolyzes proteins at acidic pH and denatures at higher pH values. We chose to study pepsin because of its significance in the breakdown of proteins in the digestive system of infants compared to other enzymes.
Anti-microbial Property:
To predict anti-microbial properties in the peptides of interest, we utilized the “Antimicrobial Peptide Scanner vr.2” by Daniel Veltri, Uday Kamath, and Amarda Shehu. We input peptide sequences in FASTA format and were given a probability that the sequence showcased antimicrobial properties. Any probability above 0.5 is considered to possess antimicrobial properties.
To make conclusions based on the antimicrobial property predictions, we looked at net charge and isoelectric point of the sequences of interest. Most antimicrobial peptides have a positive net charge that allows them to interact with negatively charged bacterial membranes [4]. The isoelectric point is the pH at which that sequence has zero net charge. When the pH is equal to the isoelectric point, the peptide tends to lose its biological function [5]. Based on the studies done by Marc Torrent, it was concluded that most antimicrobial peptides have an isoelectric point close to 10 [6]. To test the correlation between our sequences of interest’s probabilities, net charges, and isoelectric points, we first found the predicted net charge and isoelectric point. To find the net charge, we used the University of Nebraska Medical Center’s “Antimicrobial Peptide Calculator and Predictor” application. To find the isoelectric points, we used the “PepDraw” application by the Wimley Lab. After gathering data, we plotted our data on MATLAB to draw conclusions.
The first set of plots shown above are the plotted probability of antimicrobial property compared to that sequence’s isoelectric point, for pH 1.3 and pH greater than 2. The red points signify a probability of antimicrobial property greater than 0.5, meaning a positive probability, and the blue points signify a probability less than or equal to 0.5, or a negative probability. Based on the graphs, mostly all positive probability points are above an isoelectric point of 8, with only a few outliers. From this, we can draw conclusions that sequences with a positive antimicrobial probability are likely to have a higher isoelectric point. The second set of plots are the probability of antimicrobial property compared to its net charge, again for pH 1.3 and pH greater than 2. From these plots, we can see that for both pH environments, almost all positive antimicrobial property sequences have a net charge of zero or higher. Therefore, from these plots, it can be concluded that there is some correlation between a positive antimicrobial property probability and a positive net charge. To further investigate these trends, more data needs to be collected from online biostatistics tools or experiential peptide sequences need to be analyzed.
Anti-inflammatory Property:
To predict anti-inflammatory properties in the peptides of interest, we utilized the “Prediction of Anti-Inflammatory Peptide” Application by Kurata’s Lab. We input the peptides sequences of interest in FASTA format and were given a probability value that corresponded to the level of confidence that the sequence was anti-inflammatory. Scores of greater than or equal to 0.468 are considered high confidence. Scores greater than or equal to 0.388 and less than 0.468 are considered medium confidence. Scores greater than or equal to 0.342 or less than 0.388 are considered low confidence. Scores lower than 0.342 are considered negative confidence.
Young infants are susceptible to infections and disease due to the immune system not being fully developed yet. Therefore, the primary line of defense for the first few months of life depends on the maternal IgG antibody being transferred to the infant, along with the adaptive immune system mechanism that is driven by antigen exposure in order to build immunological memory [3]. It has been found that synthesis of methionine (M) and cysteine (C) are important factors in the inflammatory response. Both contribute to being a precursor to glutathione (GSH), which is an antioxidant molecule in cells. GSH helps neutralize reactive oxygen species generated as a by-product of immune cells, thus preserving immune cell function [7]. In addition, GSH has been shown to modulate functions of T cells, (what the IgG immune response mechanism depends on), B cells, and antigen-presenting cells. Lastly, cysteine residues that form disulfide bonds in proteins improve stabilization and functions, some of which are related to immune response by modulating changes in the redox state of cysteine residues [8].
It is not proven yet that if the correlation of C and M percentages in a peptide sequence can indicate a peptide sequence to have a higher chance of becoming a protein with anti-inflammatory properties, but based on previously mentioned findings, we decided to investigate this hypothesis.
From this chart, we can see a generalized trend where most of the peptide sequences that contain Cs and Ms are indicated to have a medium to high confidence probability of having anti-inflammatory properties. To further investigate this trend, more data needs to be collected from online biostatistics tools or experiential peptide sequences need to be analyzed.
Trace Metal Binding Property:
To predict the binding affinity for each metal we used Oxford Academic’s mebipred tool for identifying metal-minding potential in protein sequences. We input sequences in FASTA format to obtain a likelihood of metal binding. The cutoff for likely to bind is 0.5 with a 26% precision rate.
We chose to investigate the trace metals calcium (Ca), copper (Cu), cobalt (Co), iron (Fe), potassium (K), magnesium (Mg), manganese (Mn), sodium (Na), nickel (Ni), and zinc (Zn). We organized these metal ions based on periodic table groups. We compared the four main group elements, Na, K, Mg, and Ca, and the six d-block transition metals including Mn, Fe, Co, Cu, Zn, and Ni. These metal ions are essential for redox reactions and catalyzing a wide range of DNA replication and forms of energy generation. Certain enzymes tend to have a higher binding affinity for certain trace metals. For example, lactoferrin has been found to be the most important amongst the whey proteins when it comes to iron-binding [9]. Lactoferrin is responsible for the bioavailability of iron which plays a large role in catalytic functions in many proteins. Serum albumin is an important protein for binding zinc and copper. It binds 28% of total zinc and 39% of total copper. The affinity for metal binding in serum albumin is highly dependent on pH [10]. The metal ions with the strongest coordination that we consider are copper (II), iron (II), and manganese (II). This has been tied to properties including a small density and small atomic radius [11]. Zinc has been found to be the most abundant metal to bind to proteins. It most highly binds to serum albumin and lactoferrin [12]. Histidine binds strongly to trace metals including Co, Ni, Zn, Cu, and Fe due to it’s imidazole side chain [13].
These graphs show the relationship between histidine percentage and trace metal binding percentage. The cutoff for likely to bind to a trace metal is 0.5. We separated the charts to compare at pH1.3 and pH 2 as well as separating main group metals and transition metals. PH 2 transition metals are the least likely to bind to trace metals regardless of histidine percentage. We see more likelihood of metal binding at lower levels of histidine. Main group metals of pH 1.3 and pH 2 both show similar likelihood of trace metal binding. Both show more likelihood at lower levels of histidine. pH 1.3 transition metals show the most correlation between histidine percentage and trice metal binding however trace metal binding is still found at lower levels of histidine.
References:
[1] Ballard O, Morrow AL. Human Milk Composition. Pediatric Clinics of North America. 2013;60(1):49-74. doi:https://doi.org/10.1016/j.pcl.2012.10.002
[2] Dallas DC. Digestion of Protein in Premature and Term Infants. Journal of Nutritional Disorders & Therapy. 2012;02(03). doi:https://doi.org/10.4172/2161-0509.1000112
[3] Wiertsema, S. P., Jeroen van Bergenhenegouwen, Johan Garssen, and Leon. The Interplay between the Gut Microbiome and the Immune System in the Context of Infectious Diseases throughout Life and the Role of Nutrition in Optimizing Treatment Strategies. Nutrients 13:886–886, 2021.
[4] Angela Di Somma, Moretta A, Canè C, Cirillo A, Duilio A. Antimicrobial and Antibiofilm Peptides. Biomolecules. 2020;10(4):652-652. doi:https://doi.org/10.3390/biom10040652
[5] Osorio D, Rondón-Villarreal P, Torres R. Peptides: A Package for Data Mining of Antimicrobial Peptides. The R Journal. 2015;7(1):4. doi:https://doi.org/10.32614/rj-2015-001
[6] Torrent M, Andreu D, Nogués VM, Boix E. Connecting Peptide Physicochemical and Antimicrobial Properties by a Rational Prediction Model. PLOS ONE. 2011;6(2):e16968-e16968. doi:https://doi.org/10.1371/journal.pone.0016968
[7] Lan, W., Y. Ren, Z. Wang, J. Liu, and H. Liu. Metabolic Profile Reveals the Immunosuppressive Mechanisms of Methionyl-Methionine in Lipopolysaccharide-Induced Inflammation in Bovine Mammary Epithelial Cell. Animals 11:833–833, 2021.
[8] Yang, H., P. Lundbäck, Lars Ottosson, H. Erlandsson-Harris, E. Venereau, M. E. Bianchi, Yousef Al-Abed, U. Andersson, and K. J. Tracey. Redox modifications of cysteine residues regulate the cytokine activity of HMGB1. Molecular Medicine 27:, 2021.
[9] Author links open overlay panelCarla Mariane Costa Pozzi a, a, b, Highlights•In total nine spots of proteins were identified as human milk secretory IgA (sIgA) components.•One spot of protein was identified as the transmembrane secretory component.•Eight spots of proteins were identified as the alpha-1 heavy chain const, A. presence of calcium, O. Ballard, S. Beranova-Giorgianni, C. A. Blindauer, G. Fanali, B. Lönnerdal, P. M. Moraes, D. S. Newburg, G. Picariello, R. T. Radulescu, W. Shi, I. L. Alcantara, M. A. Z. Arruda, J. Bates, P. Cayot, S. P. Dallas, L. A. Davidson, B. T. Doumas, J. W. Froehlich, X. Gao, and H. Haraguchi. Metal ions bound to the human milk immunoglobulin A: Metalloproteomic approach. , 2014.at <https://www.sciencedirect.com/science/article/abs/pii/S0308814614009261#:~:text=In%20milk%2C%20two%20examples%20of%20proteins%20that%20utilize,al.%2C%201982%2C%20Fanali%20et%20al.%2C%202012%29%2C%20and%20lactoferrin.>
[10] LS;, L. B. B. Zinc and copper binding proteins in human milkat <https://pubmed.ncbi.nlm.nih.gov/7148736/>
[11] Rodzik, A., P. Pomastowski, G. N. Sagandykova, and B. Buszewski. Interactions of whey proteins with metal ions. , 2020.at <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7139725/>
[12] Vegarud, G. E., T. Langsrud, and C. Svenning. Mineral-binding milk proteins and peptides; occurrence, biochemical and technological characteristics: British Journal of Nutrition. , 2007.at https://www.cambridge.org/core/journals/british-journal-of-nutrition/article/mineralbinding-milk-proteins-and-peptides-occurrence-biochemical-and-technological-characteristics/3E327A15B9783E925D129272221994FC
[13] 3.Moro, J., D. Tomé, P. Schmidely, T.-C. Demersay, and D. Azzout-Marniche. Histidine: A systematic review on metabolism and physiological effects in human and different animal species. , 2020.at <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7284872/>
