Vol. 10, No. 1 (2017) pp. 7− 13 https://doi.org/10.13160/ricns.2017.10.1.7
Three Dimensional Structure Prediction of Neuromedin U Receptor 1 Using Homology Modelling
Santhosh Kumar Nagarajan1,2 and Thirumurthy Madhavan2†
Abstract
Neuromedin U receptor 1 is a GPCR protein which binds with the neuropeptide, neuromedin. It is involved in the regulation of feeding and energy homeostasis and related with immune mediated inflammatory diseases like asthma. It plays an important role in maintaining the biological clock and in the regulation of smooth muscle contraction in the gastrointestinal and genitourinary tract. Analysing the structural features of the receptor is crucial in studying the pathophysiology of the diseases related to the receptor important. As the three dimensional structure of the protein is not available, in this study, we have performed the homology modelling of the receptor using 5 different templates. The models were subjected to model validation and two models were selected as optimal. These models could be helpful in analysing the structural features of neuromedin U receptor 1 and their role in disorders related to them.
Keywords: Neuromedin U receptor 1, GPCR, neuromedin, NMUR1, Homology modelling
1. Introduction
Neuromedin U is a neuropeptide, expressed exten- sively in the gastrointestinal, central nervous system and genitourinary[1]. It has potent contractile activity on smooth muscle, which reside within the C-terminal por- tion of the peptide, which is highly conserved between species. There are various other functions for the pep- tide that include: regulation of blood flow and ion trans- port in the intestine, increased blood pressure and regulation of adrenocortical function[2]. In the central nervous system, the functions of Neuromedin U are yet to be clearly understood. But the functions might include: neuroendocrine regulation, control of food intake, modulation of dopamine actions and involve- ment in neuropsychiatric disorders[3].
Two neuromedin U receptors have been identified and shown to bind neuromedin U. They are G protein- coupled receptor subtypes, with differing expression
patterns[4]. The receptors also bind with the neuropep- tide, neuromedin S. They are found throughout the body, and have diverse but specific roles. Neuromedin U receptor 1 is found pre-dominantly in the peripheral nervous system mainly in the gastrointestinal tract, whether neuromedin U receptor 2 is found pre-domi- nantly in the central nervous system[5,6]. NmU is involved in the regulation of feeding and energy homeostasis[7-9]. It is found to be one of the links between stress and cancer. NmU is also related with immune mediated inflammatory diseases like asthma, and in maintaining the biological clock, in the regula- tion of smooth muscle contraction in the gastrointestinal and genitourinary tract, and in the control of blood flow and blood pressure[10-13]. Drug development selective towards these receptors, might help in identifying their pathophysiological roles in the diseases related to them.
Homology modelling is an alternate tool helps in pre- dicting the three-dimensional conformation of a protein, when only the sequence data of the protein is available.
Due to the enormous amount of time required to prepare protein for crystallization using experimental process such as protein expression, purification and crystalliza- tion, the number of protein structures resolved experi- mentally lags behind the sequence data available[14]. Homology modelling can provide as a tool for the
1Department of Bioinformatics, School of Bioengineering, SRM Uni- versity, SRM Nagar, Kattankulathur, Chennai 603203, India.
2Department of Genetic Engineering, School of Bioengineering, SRM University, SRM Nagar, Kattankulathur, Chennai 603203, India.
†Corresponding author : thiru.murthyunom@gmail.com, thirumurthy.m@ktr.srmuniv.ac.in
(Received: February 1, 2017, Revised: March 17, 2017, Accepted: March 25, 2017)
experimental procedures in finding the structure of the protein in a rather short time. In this study, we have developed three-dimensional models of neuromedin U receptor 1 based on homology modelling and validated them. The developed models could provide as a tool for further studies on the structural features and binding features of neuromedin U receptor 1-neuromedin U interaction.
2. Material and Methods
2.1. Template Selection
The amino acid sequence of the human neuromedin
U receptor 1 (accession No: Q9HB89) was retrieved from the Uniprot database. Protein BLAST[15] search was performed against PDB[16] to find suitable templates for modelling the receptor. Based on sequence identity, query coverage and E-value, 5 different templates were selected. The selected templates were – 4BWB, 4BVO, 3ZEV, 4XEE and 4XES. If the level of sequence iden- tity is above 30%, then up to 90% of the polypeptide conformation tends to be modelled well[17-19]. All the templates were having sequence identity ≥ 30%. As the identities of the templates were above 30% (Table 1), we have performed single template based homology modelling. Query coverage for the templates was Table 1. The query coverage and identity values of the templates
PDB ID Max Score Total score Query Coverage % E Value Identity %
4BWB 156 156 73% 8e-44 33%
4BVO 155 155 73% 2e-43 33%
3ZEV 151 151 72% 8e-42 32%
4XEE 155 155 79% 1e-41 31%
4XES 152 152 79% 1e-40 315%
Fig. 1. Alignment between the target (neuromedin U receptor 1) and template (4BWB).
greater than 70%. Also, all of the templates retained the seven transmembrane helix regions, which is the char- acteristic feature of the GPCR proteins.
2.2. Homology Modelling
Using the modelling platforms, EasyModeller 4.0[20]
and IntFold[21], the three dimensional structures of human neuromedin U receptor 1 were developed. Easy- Modeller 4.0 uses MODELLER 9.12[22] and Python 2.7.1 in the backend. IntFold is an integrated server used for modelling protein three dimensional structures from their amino acid sequences[23]. Initially, the predicted
Table 2. Model validation results - RMSD and Ramachandran plot values
Model No
Templates Used
Homology Modeling Validation RMSD
Ramachandran Plot Number of residues in
favored region (%)
Number of residues in allowed region (%)
Number of residues in outlier region (%)
1 4BWB 0.284 92.0 4.7 3.3
2 0.239 93.2 4.0 2.8
3 0.235 89.9 5.4 4.7
4 0.223 91.0 5.9 3.1
5 0.390 91.7 4.2 4.0
6 4BVO 0.555 91.3 5.0 3.8
7 0.199 95.0 3.5 1.4
8 0.255 92.7 5.2 2.1
9 0.284 91.3 5.2 3.5
10 0.169 93.2 5.2 1.7
11 3ZEV 0.237 92.0 5.0 3.1
12 0.316 93.9 4.0 2.1
13 0.268 92.7 3.8 3.5
14 0.302 91.0 6.8 2.1
15 0.279 92.9 4.5 2.6
16 4XEE 0.950 91.0 6.4 2.6
17 1.009 94.6 3.1 2.4
18 0.877 92.2 4.5 3.3
19 1.963 88.9 6.6 4.5
20 0.909 91.3 6.6 2.1
21 4XES 0.668 93.2 4.4 2.4
22 0.449 90.6 5.9 3.5
23 0.525 91.5 5.7 2.8
24 0.679 89.9 6.6 3.5
25 0.474 92.2 4.2 3.5
26 2KS9 1.96 90.6 6.1 3.3
3D4S 2.61
27 2KS9 1.83 88.9 6.8 4.2
3D4S 3.05
4N6H 3.16
28 2KS9 0.75 90.8 6.1 3.1
29 2KS9 1.64 92.5 5.4 2.1
2VT4 1.71
30 2KS9 1.22 85.4 9.9 4.7
2KSB 1.22
models were validated using the RMSD values. Then, using RAMPAGE web server, Ramachandran plots for the models were plotted[24]. Ramachandran plot pro- vides a way to visualize backbone dihedral angles ψ against ϕ of amino acid residues in protein structure, which identifies the sterically allowed regions for these angles. Later, validation by Verify3D and ERRAT plots were carried out. Verify3D determines the compatibility of the predicted model with its own amino acid sequence by assigning a structural class based on its location and environment (alpha, beta, loop, polar, non- polar etc.) and comparing the results to good struc-
tures[25]. ERRAT plots are plotted as a function of the position of a sliding 9-residue window. The error func- tion is based on the statistics of non-bonded atom-atom interactions present in the structure[26]. ProSA-web, an interactive web server is used to identify the errors in three-dimensional structure of the protein[27].
3. Results and Discussion
3.1. Model Generation
Using EasyModeller, 5 models are modelled for each of the five templates 4BWB, 4BVO, 3ZEV, 4XEE and
Table 3. Model validation results – ProSA, ERRAT and Verify3D values
Model No ProSA Z-Score ERRAT
Overall quality factor
Verify3D
(% of the residues had an averaged 3D-1D score >= 0.2)
1 -2.49 55.610 52.82
2 -2.39 56.566 58.92
3 -2.40 52.206 60.80
4 -2.58 54.386 59.62
5 -2.83 45.844 64.32
6 -0.72 50.000 43.43
7 -3.75 63.446 57.98
8 -3.05 62.005 57.04
9 -2.49 58.090 51.64
10 -3.82 64.675 53.99
11 -2.05 51.175 58.45
12 -2.09 51.571 50.94
13 -2.20 42.021 55.63
14 -2.62 50.262 57.98
15 -2.19 47.733 56.81
16 -2.94 50.242 53.05
17 -4.09 56.522 60.80
18 -2.64 53.589 58.22
19 -2.48 45.455 57.28
20 -3.11 49.761 51.64
21 -2.48 69.930 75.82
22 -2.15 70.890 69.48
23 -2.39 68.836 63.85
24 -1.59 72.165 61.27
25 -2.09 69.580 65.73
26 -1.23 53.269 56.34
27 -2.18 50.365 53.76
28 -1.50 46.747 44.37
29 -1.78 47.573 41.78
30 0.02 43.541 55.40
4XES. Therefore totally 25 models were developed using EasyModeller. Five best models developed from IntFold server were selected. The alignment of the tem- plate 4BWB with the modelled receptor was repre- sented in Fig. 1.
3.2. Model Validation
The predicted models were validated using various validation techniques. Root mean square deviation (RMSD) of all the predicted models with their respec- tive template was calculated. Ramachandran plot was generated for each model and the number of residues in favourable, allowed and disallowed region was identi- fied. The statistics of both RMS deviation and Ramach- andran plots are represented in the Table 2. Only models scoring acceptable results are displayed and are numbered. Verify3D and ERRAT plots were developed for the models. Using ProSA web server Z-scores were calculated. The results from ProSA, Verify3D and ERRAT plots are represented in Table 3. Based on the statistics, from the models 7 and 10, developed using Easymodeller were found to be the best models. Espe- cially, model 7 scored well in all the validation and is found to be the most reliable among the developed
models. Also, all the developed models have similar structure. The best models – 7 and 10 are represented in Fig. 2. RC plot and ERRAT plots of the selected models were represented in Fig. 3 and Fig. 4 respec- tively.
Fig. 3. RC plot for selected models 7 (a) and 10 (b).
Fig. 2. Best models (Model 07 and Model 10) selected after validation.
4. Conclusion
Three dimensional models for neuromedin U receptor 1 were generated using single template based approach.
Model numbers 7 and 10 were selected as best, based on their RMS deviation, Ramachandran plot, ERRAT plot and Verify3D values. The selected models showed similar structures. Based on the results after model val- idation, it is found that all the generated models are sim- ilar and the structures are reliable. These predicted models would be useful in the studying the interaction of neuromedin with the receptor neuromedin U receptor 1. Also, these models may serve as a reliable tool for analysing the important structural features and function of neuromedin U receptor 1.
References
[1] R. Raddatz, A. E. Wilson, R. Artymyshyn, J. A.
Bonini, B. Borowsky, L. W. Boteju, S. Zhou, E. V.
Kouranova, R. Nagorny, M. S. Guevarra, M. Dai, G. S. Lerman, P. J. Vaysse, T. A. Branchek, C. Ger- ald, C. Forray, and N. Adham. “Identification and characterization of two neuromedin U receptors dif- ferentially expressed in peripheral tissues and the central nervous system”, J. Biol. Chem., Vol, 275, pp. 32452-32459, 2000.
[2] A. D. Howard, R. Wang, S. S. Pong, T. N. Mellin, A. Strack, X. M. Guan, Z. Zeng, D. L. Williams Jr, S. D. Feighner, C. N. Nunes, B. Murphy, J. N. Stair, H. Yu, Q. Jiang, M. K. Clements, C. P. Tan, K. K.
McKee, D. L. Hreniuk, T. P. McDonald, K. R.
Lynch, J. F. Evans, C. P. Austin, C. T. Caskey, L.
H. Van der Ploeg, and Q. Liu, “Identification of receptors for neuromedin U and its role in feeding”, Nature, Vol. 406, pp. 70-74, 2000.
[3] S. Budhiraja and A. Chugh, “Neuromedin U: phys- iology, pharmacology and therapeutic potential”, Fund. Clin. Pharmacol., Vol. 23, pp. 149-157, 2009.
[4] P. J. Brighton, P. G. Szekeres, and G. B. Willars, Fig. 4. ERRAT plot developed for the selected models 7(a) and 10(b)
*on the error axis, two lines are drawn to indicate the confidence with which it is possible to reject regions that exceed that error value
“Neuromedin U and its receptors: structure, func- tion, and physiological roles”, Pharmacol. Rev., Vol.
56, pp. 231-248. 2004.
[5] J. Ballesta, F. Carlei, A. E. Bishop, J. H. Steel, S.
J. Gibson, M. Fahey, R. Hennessey, J. Domin, S. R.
Bloom, and J. M. Polak, “Occurrence and develop- mental pattern of neuromedin U-immunoreactive nerves in the gastrointestinal-tract and brain of the rat”, Neuroscience, Vol. 25, pp. 797-816, 1988.
[6] C. Austin, M. Oka, K. A. Nandha, S. Legon, N.
Khandannia, G. Lo, and S. R. Bloom, “Distribution and developmental pattern of neuromedin-U expres- sion in the rat gastrointestinal-tract”, J. mol. Endo- crinol., Vol. 12, pp. 257-263, 1994.
[7] A. Inui, “Feeding and body-weight regulation by hypothalamic neuropeptides-mediation of the actions of leptin”, Trends neurosci., Vol. 22, pp. 62- 67, 1999.
[8] T. R. Ivanov, C. B. Lawrence, P. J. Stanley, and S.
M. Luckman. “Evaluation of neuromedin U actions in energy homeostasis and pituitary function”, Endocrinology, Vol. 143, pp. 3813-3821, 2002.
[9] M. Nakazato, R. Hanada, N. Murakami, Y. Date, M.
S. Mondal, M. Kojima, H. Yoshimatsu, K. Kan- gawa, and S. Matsukura, “Central effects of neuro- medin U in the regulation of energy homeostasis”, Biochem. Bioph. Res. Co., Vol. 277, pp. 191-194.
2000.
[10] M. Moriyama, T. Sato, H. Inoue, S. Fukuyama, H.
Teranishi, K. Kangawa, T. Kano, A. Yoshimura, and M. Kojima, “The neuropeptide neuromedin U pro- motes inflammation by direct activation of mast cells”, J. Exp. Med., Vol. 202, pp. 217-224, 2005.
[11] N. Minamino, K. Kangawa, and H. Matsuo, “Neu- romedin-U-8 and neuromedin-U-25 - novel uterus stimulating and hypertensive peptides identified in porcine spinal-cord”, Biochem. Bioph. Res. Co., Vol. 130, pp. 1078-1085, 1985.
[12] N. Minamino, T. Sudoh, K. Kangawa, and H. Mat- suo, “Neuromedins: novel smooth-muscle stimulat- ing peptides identified in porcine spinal-cord”, Peptides, Vol. 6, pp. 245-248, 1985.
[13] M. S. Mondal, Y. Date, N. Murakami, K. Toshinai, T. Shimbara, K. Kangawa, and M. Nakazato, “Neu- romedin U acts in the central nervous system to inhibit gastric acid secretion via CRH system”, Am.
J. Physiol.-Gastr. L., Vol. 284, pp. G963-G969.
2003.
[14] G. C. Baker, J. J. Smith, and D. A. Cowan, “Review and re-analysis of domain-specific 16S primers”, J.
Microbiol. Meth., Vol. 55, pp. 541-555, 2003.
[15] S. F. Altschul, W. Gish, W. Miller, E. W. Myers, and D. J. Lipman, “Basic local alignment search tool”, J. Mol. Biol., Vol. 215, pp. 403-410, 1990.
[16] H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H. Weissig, I. N. Shindyalov, and P. E.
Bourne, “The protein data bank”, Nucleic Acids Res., Vol. 28, pp. 235-242, 2000.
[17] M. Shalini and T. Madhavan, “Homology Modeling of CCR 4: novel therapeutic target and preferential maker for Th2 cells”, J. Chosun Natural Sci., Vol.
7, pp. 234-240, 2014.
[18] B Sathya and T Madhavan, “Homology modeling of cysteinyl leukotriene1 receptor”, J. Chosun Nat- ural Sci., Vol. 8, pp. 13-18, 2015.
[19] S. K. Nagarajan and T Madhavan, “3D structure prediction of thromboxane A2 receptor by homology modeling”, J. Chosun Natural Sci., Vol.
8, pp. 75-79, 2015.
[20] B. K. Kuntal, P. Aparoy, and P. Reddanna, “Easy- Modeller: A graphical interface to MODELLER”, BMC Research Notes, Vol. 3, pp. 226, 2010.
[21] A. Roy, A. Kucukural, and Y. Zhang, “I-TASSER:
a unified platform for automated protein structure and function prediction”, Nature Protoc., Vol. 5, pp.
725-738, 2010.
[22] N. Eswar, M. A. Marti-Renom, B. Webb, M. S.
Madhusudhan, D. Eramian, M. Shen, U. Pieper, and A. Sali, “A comparative protein structure modeling with MODELLER”. Current Protocols in Bioinfor- matics, Vol. 5, pp. 1-5, 2006.
[23] L. J. McGuffin, J. Atkins, B. R. Salehe, A. N.
Shuid, and D. B. Roche, “IntFOLD: an integrated server for modelling protein structures and func- tions from amino acid sequences”, Nucleic Acids Res., Vol. 43, pp. W169-W173, 2015.
[24] S. C. Lovell, I. W. Davis, W. B. Arendall III, P. I.
W. de Bakker, J. M. Word, M. G. Prisant, J. S. Rich- ardson, and D. C. Richardson, “Structure validation by Cα geometry: Φ, Ψ and Cβ deviation”, Proteins., Vol. 50, pp. 437-450, 2002.
[25] J. U. Bowie, R. Lüthy, and D. Eisenberg, “A method to identify protein sequences that fold into a known three-dimensional structure”, Science, Vol.
253, pp. 164-170, 1991.
[26] C. Colovos and T. O. Yeates, “Verification of pro- tein structures: patterns of nonbonded atomic inter- actions”, Protein Sci. Vol. 2, pp. 1511-1519, 1993.
[27] M. Wiederstein and M. J. Sippl, “ProSA-web: inter- active web service for the recognition of errors in three-dimensional structures of proteins”, Nucleic Acids Res., Vol. 35, pp. W407-W410, 2007.
