Sequence-based structural analysis and evaluation of polymorphism in buffalo Nod-like receptor-1 gene

Abstract

In this study, we have sequence characterized and analyzed the polymorphism in buffalo NOD1 (nucleotide-binding oligomerization domain 1) gene as well as its expression analysis. Full-length sequence analysis of NOD1 revealed this gene in buffalo being conserved with respect to the domain structures, similar to other species. Alternate splice variants having exon3 skipping also identified for the first time in the gene expressed in buffalo-purified peripheral blood mononuclear cells (PBMCs). Phylogenetically ruminant species were found to be clustering together and buffalo displaying maximum similarity with cattle. Sequencing of NOD1 across 12 Indian buffalo breeds identified 23 polymorphic sites within coding region, among which 16 were synonymous and 7 changes found to be non-synonymous. Four SNPs (single nucleotide polymorphisms) of them were genotyped in 393 animals belonging to 12 riverine, swamp and hybrid (riverine × swamp) buffalo populations of diverse phenotypes and utilities, showing variable allelic frequencies. Principal component analysis revealed, riverine and swamp buffaloes being distinctly placed with the distribution of breeds within the group based on the geographical isolation. Further, quantitative real-time PCR detected NOD1 expression in multiple tissues with PBMCs and lungs showing highest expression among the tissues examined. Structural analysis based on the translated amino acid sequence of buffalo NOD1 identified four protein interaction motifs LxxLL important for ligand binding. Molecular interaction analysis of iE-DAP and NOD1-LRR and their complex stability and binding-free energy studies indicated variable binding energies in buffalo and cattle NOD1. Overall, the study reveals unique structural features in buffalo NOD1, important for species-specific ligand interaction.

Electronic supplementary material

The online version of this article (10.1007/s13205-018-1534-2) contains supplementary material, which is available to authorized users.

Introduction

Among higher eukaryotes, the conserved microbial structures known as microbe-associated molecular patterns (MAMPs) are recognized by germ line-encoded pattern recognition molecules (Koller et al. 2009). Four kinds of germ line-encoded receptors are involved in pathogen control and elimination, which include the Toll-like receptors (TLRs), the Nod-like receptors (NLRs), the retinoic acid-inducible gene-1 (RIG-1)-like receptors (RLRs) and the C-type lectin receptors (CLRs). Individual molecules recognize pathogen-derived molecular components from all major classes of microbes, including bacteria, viruses, yeast and parasites. These microbial components include lipopolysaccharides (LPS), lipopeptide, peptidoglycan (PGN), flagellin, dsRNA, ssRNA and CpG DNA, among others (Takeda and Akira 2005). Identification of these MAMPs by pattern recognition molecules triggers the activation of signaling pathways including the nuclear factor kB (NF-kB), mitogen-activated protein kinases (MAPKs) and type I interferon (IFN) molecules, which promote inflammatory and antimicrobial responses (Kawai and Akira 2009).

Nod-like receptors are implicated in the induction of inflammatory responses through activation of nuclear factor kB (NF-kB), upon stimulation by other cellular danger signals (Ogura et al. 2001). Characteristic of NLR proteins is the presence of a NACHT (NAIP-neuronal apoptosis inhibitory protein), which is flanked by highly variable leucine-rich repeat (LRR) domain, play a critical role in the sensing of Gram-negative peptidoglycan. NOD molecules, a family of intracellular proteins including NOD1 (CARD4) and NOD2 (CARD15), represents a new group of intracytoplasmic pattern recognition molecules, thus allowing detection of intracellular invasive bacteria. NOD1 recognizes g-d-glutamyl-meso-diaminopimelic acid (iE-DAP) which is found to be part of PGN in all gram-negative bacteria as well as some gram-positive bacteria (Chamaillard et al. 2003). In particular, NOD1 expressed in epithelial cells considered as the first line of defense against invasive pathogens has been reported to play an important role in sensing the intracellular Gram-negative bacterial cell wall components (Girardin et al. 2003). A study has shown that Helicobacter pylori introduces iE-DAP (combat with NOD1) into the cytosol using its type IV secretion system (Viala et al. 2004). Inflammatory cytokines and other antimicrobial genes which contribute to host defense, are induced by signaling via NOD1 and NOD2 (reviewed by Mogensen 2009). Mutations in NOD1 gene are strongly associated with ischemic stroke (Tiszlavicz et al. 2009) and with susceptibility to bowel disease (McGovern et al. 2005) in humans. Various splice variants of NOD1 have also been identified in human beings with the truncated protein lacking number of LRR domains across tissues (Hysi et al. 2005).

The dairy industry of India largely depends on the 108.7 million buffaloes (Livestock Census, 2012, http://dahd.nic.in/dahd/bahs-2012.aspx), which account for 56.7% of the world’s buffalo population, contributing about 52.5% to total milk production in the country. Besides milk, buffaloes are an important source of meat, leather and draft power. Association of polymorphism in immune response receptor genes with several infectious diseases has been reported in bovines, including mastitis in Indian dairy buffaloes (Dhiman et al. 2017). Hence, it is expected that the investigation of genetic diversity in immune genes, particularly pathogen-binding receptors that play a key role in innate immune defense, would shed focus on the disease susceptibility or resistance in riverine and swamp buffaloes.

Among the microbial-sensing receptor molecules, NOD-like receptors have emerged as potential candidates for investigating the genetic structure and polymorphism analysis. While the work focused on the TLRs’ role in innate resistance of buffalo is sparse (Dubey et al. 2012), very less information is available on riverine and swamp-type buffaloes related to characterization of immune response elements like NOD1. In this study, we report cloning-based sequence characterization of buffalo NOD1 gene as well as tried exploring the polymorphism and generating gene expression profile across tissues. Sequence-based structural analysis of NOD1 data available for bovine species has also been attempted to investigate the possible role of this gene in species-specific immune response.

Materials and methods

Sample collection

Genomic DNA isolated from blood samples of 393 buffaloes was used in the study, which comprised 221 riverine types—Murrah (30), Nili Ravi (34), Toda (47), Marathwadi (46), Jaffarabadi (30) and Chilika (34) and 48 Assamese hybrid buffaloes and 124 pure swamps belonging to Assamese Silchar (27), Assamese Dibrugarh (12), Nagaland (36), Mizoram (15), Manipuri (34). Blood samples collected from 12 different buffalo breeds, including 6 riverine buffaloes and 6 swamps, representing different utilities and geographical locations from India, were utilized for polymorphism detection. Necessary approval from the Institute Animal Ethic Committee was taken for the blood samples’ collection. The genomic DNA was isolated using the standard SDS-proteinase K method (Sambrook and Russell 2001) and the quality/quantity of DNA was assessed using a Nanodrop (ND-1000).

Sequencing and SNP detection in buffalo NOD1 gene

For the characterization of buffalo NOD1 gene, a total of 2865 nucleotides comprising ORF region including 11 exons were amplified, and 15 sets of overlapping primers were designed from gene sequence of cattle sequence available in Ensembl genome browser (ENSBTAT00000061183), for PCR amplification (Table 1-SI in supplementary information). PCR was performed on DNA samples of panel representing 12 pooled samples of 3 animals from each breed above, in a total volume of 20 µl containing approximately 80 ng genomic DNA, 10 × PCR buffer having 15 mM MgCl₂, 0.5 µl of 10 mM dNTPs, 10 pmol of each primer and 1 U of Taq DNA polymerase (Bangalore Genei, India). Following the initial denaturation step (95 °C for 3 min), samples were subjected to 32 cycles of PCR consisting of 94 °C for 30 s, primer-specific annealing temperature for 30 s (Table 1-SI in supplementary information); and 72 °C for 1 min, followed by a final extension for 10 min at 72 °C. Amplified products were checked on 1.5% agarose gel and sequenced from both ends after column purification. Sequencing of purified PCR products was carried out using BigDye^® Terminator v3.1 Cycle Sequencing Kit on ABI PRISM^® 3100 Genetic Analyzer (Applied Biosystems, CA, USA). Data was further analyzed using SeqMan and MegAlign program of the Lasergene software (DNASTAR Inc., USA). Polymorphic sites were documented by sequence alignment and crosschecking the chromatograms of individual animal.

Genotyping of polymorphic nucleotide by PCR-RFLP and tetra-ARMS PCR

Identified SNPs, were further analyzed for their position, structural domain location and nucleotide substitution status. A PCR-RFLP genotyping protocol was developed for the three non-synonymous SNPs—(g.338C>T, g.797A>G and g.1482G>C), using the NEBcutter online software (http://tools.neb.com/NEBcutter2/). Primers were designed and a tetra-primer Amplification Refractory Mutation System (ARMS)-PCR procedure (Table 1-SI in supplementary information), was also used for genotyping of one synonymous SNP-(g.2394A>C) across different buffalo breeds. PCR-RFLP genotyping of these three SNPs-g.338C>T, g.797A>G and g.1482G>C were carried out using AciI, MspI and PvuII restriction enzymes (NEW ENGLAND BioLabs^® Inc.), respectively. Amplified PCR products were digested with the respective restriction enzyme and resolved in 3% agarose gel and genotypes were recorded. Genotypic frequency and genetic distance (DA, Nei’s standard genetic distances) were calculated using GenAlex6.2 program (Peakall and Smouse 2006, http://www.anu.edu.au/BoZo/GenAlEx/). Genetic distance was determined based on allelic frequencies at three loci in buffalo NOD1 gene. Principal component analysis (PCA) was performed using GenAlex6.2 program, to determine population relationships directly based on the variability in allele frequencies of different genotypes taken as the largest possible variance (Manly 1986).

Sequencing and characterization of buffalo NOD1 cDNA

To amplify buffalo NOD1 transcript, PCR primers specific to amplify complete ORF were designed and got synthesized (Table 1-SI in supplementary information). Total RNA was extracted from PBMCs of Murrah buffalo, using TRIzol method (Invitrogen, Bangalore, India), and subjected to further purification and on-column DNase digestion using RNeasy mini kit (Qiagen India). Concentration of isolated RNA was determined spectrophotometrically using the Nano-Drop ND-1000 (Thermo Scientific). Good-quality samples with optical density OD260/OD280 ratios within the range of 1.9–2.1 were used further for reverse transcription-PCR. cDNA was synthesized from 1 µg of total RNA using RevertAid First Strand cDNA Synthesis kit (Fermentas, Thermo Fisher Scientific, India), with the help of Oligo (dT) primers, as prescribed by the manufacturer. PCR was performed using 3 µl of cDNA in a 50 µl reaction volume under the conditions of one cycle of initial denaturation at 95 °C for 3 min followed by 32 cycles of 94 °C for 30 s, 60 °C for 30 s, 72 °C for 75 s and a final extension at 72 °C for 10 min.

PCR product was analyzed in 1.5% agarose gel, the single specific 2.9 kb band was gel purified by agarose gel purification kit (Roche, Germany) and cloned using InsTAclone PCR Cloning Kit (Thermo Scientific), following manufacturer protocol. Both strand sequencing was carried out with M13 forward and reverse as well as gene-specific internal (Table 1-SI in supplementary information) primers (ABI prism 3000). Overlapping sequences of different fragments were converted into single ORF coding coting and analyzed further. Molecular weight and an isoelectric point (pI) of buffalo NOD1 was analyzed using ExPASy software on the Proteomics Server (https://expasy.org/tools/pi_tool.html) and the cellular location was predicted by pSORT (https://worfpsort.org). An online program SMART was used for the analysis of protein domain architecture of buffalo NOD1 gene. Online software PolyPhen-2 (https://genetics.bwh.harvard.edu/pph2/) was used to predict the possible impact of non-synonymous SNPs on the protein structure. PCR amplification of cDNA synthesized from RNA isolated from PBMCs of one animal showed amplification of two fragments of 2.9 kb and 1.2 kb of equal intensity, supposed to be splice variants. To confirm the lower size band to be from a splice variant, cloning and sequencing of both the fragments was performed.

Tissue distribution and expression analysis of buffalo NOD1 gene

Quantitative real-time PCR was performed for comparative analysis of NOD1 gene expression across ten adult (4–6 years) buffalo tissues—small intestine, lung, liver, heart, mammary gland, stomach, kidney, spleen, skeletal muscle, and PBMCs, collected from three animals each representing Murrah breed. RNA was subsequently isolated and checked for concentration and purity as described above using Nanodrop ND-1000 (Thermo Fisher Scientific). Quality of RNA samples was further checked by Experion Bioanalyzer (BioRad Laboratories, CA, USA). RNA samples with RIN (RNA integration number) value of more than 8.0 were used further for gene expression analysis. An amount of 1 µg of total RNA from each tissue sample was reverse transcribed into cDNA using RevertAid First Strand cDNA Synthesis Kit (Fermentas, Thermo Fisher Scientific, India), using Oligo(dT) primers as described previously.

NOD1 gene-specific primers were designed and used for real-time-based transcriptional profiling (Table 1-SI in supplementary information). Primer pair was tested for specificity through amplification of cDNA synthesized from PBMCs through normal PCR amplification. Real-time PCR was performed using SYBRGreen-based detection on a LightCycler480 (Roche) machine, with each 20 µl reaction containing 2 µl of cDNA synthesized from 1 µg of RNA, 0.5 µl (5 pmol) of sense and antisense primers, 2 × SYBRGreen qPCR Master Mix (USB, OH, USA). The mixture was subjected to the following real-time PCR conditions: 95 °C for 10 min; followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Each reaction was performed in duplicate, using two internal control genes ribosomal protein S9 (RPS9) and ubiquitously expressed prefoldin-like chaperone (UXT) for normalization. PCR efficiency was determined from the slope of a standard curve generated using a tenfold dilution series of the cDNA template. The data of target NOD1 gene was normalized against geometric mean of two reference genes. Fold changes in target transcript levels were determined using the delta Ct method (Pfaffl 2001). ΔΔCt values were calculated using kidney tissue, showing lowest expression (highest ΔCt) as control.

Sequence-based homology modeling

To find suitable templates for the modeling of consensus, river, swamp and cow NOD1 protein sequence, BLASTp search was performed against the protein databank (PDB) (https://www.rcsb.org/pdb/). Best scoring template 5IRL resulting from fold recognition servers were considered for model construction using MODELLER 9.12. Both single-template and multiple-templates approaches were adopted for building NOD1 protein of each species. The 3D models were also constructed using the automated modeling programs, SWISS-MODEL, BioSerf, and Easy Pred 3D for comparison. The MODELLER-derived 3D models of buffalo NOD1-LRR were evaluated based on discrete optimized protein energy (DOPE) scores. Five models with lowest DOPE scores were chosen for further refinement using AMBER force field.

Comparative modeling, validation and annotation of NOD1

The complete 3D structure of 954 amino acids sequence of river, swamp and cow NOD1 protein, (NCBI: gb| AJC01625.1, AJC01625.1 and NP_001243492 respectively) was predicted. The predicted structure was further visualized in PyMol program. The stereochemical quality of the final energy-optimized model was verified using SAVE (http://nihserver.mbi.ucla.edu/SAVE/), bond length and bond angle analyses of the model which was performed in MolProbity. Furthermore, the Z score of hydrogen bond (H-bond) energy, packing defects, bump score, radius of gyration and deviation of Y angles of the refined model was tested in VADAR, GeNMR and PROSESS web servers. Furthermore, the PROCHECK analysis, used for the validation of the modeled structure, yielded the corresponding Ramachandran plot with 95% residues being noticed in the most favored regions.

Molecular docking of NOD1-LRR with iE-DAP

The 2D structure of iE-DAP (CID_45480617) was obtained from the PubChem database (http://pubchem.ncbi.nlm.nih.gov). The Automated Topology Builder (ATB) server was used to generate the 3D coordinates of iE-DAP employing energy minimization with the Gromos 96-53a6 force field. Molecular docking was carried out using AutoDock 4.2 to identify key ligand-binding residues. The NOD1 model was assigned with Kollman charges and Geister partial charges, that were applied for iE-DAP using Autodock Tools 1.5.

Results and discussion

The innate immune system represents one of the first lines of defense against microorganisms, through the specific detection of highly conserved structures. NOD-like receptor1 is one of the best known cytosolic innate immune receptors of the NLR family in vertebrates (Franchi et al. 2009). NLR1 in human, piscine and pig is very well-characterized (Inohara et al. 1999; Tohno et al. 2008; Sha et al. 2009; Xie et al. 2013), but the polymorphism detection and its association with disease incidences in buffalo NOD1 gene have not been reported so far.

Detection of polymorphism in buffalo NOD1

Sequencing of 15 overlapping PCR fragments ranging from 400 to 700 bp revealed 2.9 kb nucleotide sequence of buffalo NOD1 gene. The ORF of buffalo NOD1 gene consisting of 2865-nucleotides long region comprising 11 exons, was found to be highly conserved across livestock species. The intron–exon boundaries were assessed by comparing the genomic sequence data obtained with buffalo NOD1 mRNA sequence (Accession no. KP280085) as well as predicted NOD1 gene sequence of Mediterranean buffalo from whole genome shotgun data (accession no. NW_005784709), available in GenBank. By comparative analysis, buffalo NOD1 nucleotide and amino acid sequence, respectively, showed the highest identity with that of cow, followed by goat, sheep, pig, horse, human and mice with being the most distant (Table 1). To examine the nucleotide genetic diversity for single nucleotide polymorphisms (SNPs) across breeds of buffalo NOD1, sequencing of different fragments generated from various buffalo breeds were sequenced. Comparative analysis of 2865 nucleotides long buffalo NOD1 sequences covering the entire coding region, acquired for a panel of 12 animals of 6 swamp and riverine buffaloes, revealed the presence of 23 polymorphic sites, including 16 transitions and 7 transversions, among which 16 SNPs were synonymous, and 7 were non-synonymous (Table 2; Fig. 1). Shinkai et al. (2015) tried to assess the polymorphism in pigs NOD1 and identified 21 SNPs deciphered 9 non-synonymous SNPs localized throughout coding region and also in sequences encoding domains important for ligand recognition. Recently in buffaloes, 42 SNPs in the NOD2 transcript (homolog of CARD4/NOD1) have been reported by Dhiman et al. (2017).

Table 1.

Comparative analysis of buffalo NOD1 structural features and homology with different species

Species	Accession no.	CARD	NACHT	No. of LRR	No. of AA	Length of ORF	Percent homology nucleotide	Percent homology amino acid
Buffalo	KP280085	85	173	8	954	2865	–	–
Cow	NM_001256563	85	173	8	954	2865	98.2	99.2
Goat	XM_005679253	85	173	8	954	2865	97.4	97.6
Sheep	XM_012112248	85	173	8	954	2865	97.4	97.3
Pig	AB187219	85	173	7	953	2862	89.2	88.3
Horse	XM_001499566	85	173	7	953	2862	87.7	86.1
Mice	AY160222	86	173	7	953	2862	79.4	77.8
Human	AF113925	85	173	7	953	2862	86.2	83.1

Table 2.

Polymorphic nucleotides detected in buffalo NOD1 with their position and amino acid changes

Location	SNP	Nucleotide position as per accession no. KP280085	Transition/transversions	Syn/non-syn	Amino acid change	Domain region
Exon-1	A/G	113	Transition	Syn	Gln38Gln
	G/A	196	Transition	Non-syn	Asp66Asn	CARD
Exon-2	G/C	312	Transversion	Syn	Leu104Leu
	C/T	338	Transition	Non-syn	Pro113Leu	Unknown
Exon-3	A/G	556	Transition	Non-syn	Thr186Ala	Unknown
	C/A	606	Transversion	Syn	Gly202Gly
	C/T	795	Transition	Syn	Pro265Pro
	G/A	797	Transition	Non-syn	Gly266Glu	NBD
	C/T	990	Transition	Syn	Leu330Leu
	G/C	1137	Transversion	Non-syn	Leu379Asp	Unknown
	C/T	1275	Transition	Syn	Phe425Phe
	C/T	1362	Transition	Syn	Arg454Arg
	C/T	1365	Transition	Syn	Ala455Ala
	C/T	1428	Transition	Syn	Phe476Phe
	G/C	1482	Transversion	Non-syn	Gln494His	Unknown
	G/A	2016	Transition	Syn	Ala672Ala
	C/T	2058	Transition	Syn	Cys686Cys
Exon-5	C/T	2316	Transition	Syn	Val772Val
Exon-6	A/C	2394	Transversion	Syn	Thr798Thr	LRR
Exon-7	C/T	2475	Transition	Syn	Ile825Ile
Exon-9	C/T	2649	Transition	Syn	Asp883Asp
	G/C	2661	Transversion	Non-syn	Glu887Asp	LRR
Exon-11	A/C	2832	Transversion	Syn	Val944Val

Fig. 1.

Schematic representation showing structural placement of non-synonymous SNPs identified in the buffalo NOD1

Among the seven non-synonymous SNPs, three were predicted to lie in functionally important domains: aspartic acid to asparagine at position 66 (Asp66Asn) was in CARD domain, glycine to glutamic acid at position 266 (Gly266Glu) in the NACHT and glutamic acid to aspartic acid at position 887 (Glu887Asp) was found in LRR7 domain, while four SNPs proline to leucine at position 113 (Pro113Leu), threonine to alanine at position 186 (Thr186Ala), leucine to aspartic acid at position 379 (Leu379Asp) and glutamine to histidine at position 494 (Gln494His) were detected in the region of unknown function. PolyPhen-2 analysis was performed on translated amino acids to comprehend the probable impact of nsSNPs on the function and structure of buffalo NOD1 protein (Fig. 2). PolyPhen-2 analysis suggested that two amino acid changes, Asp66Asn and Glu887Asp, possibly damaged the protein structure and function of functionally important CARD and LRR domains, respectively. These changes may impact the overall structure and biological function of the protein, which needs to be explored further. Earlier, polymorphism within or adjacent to the LRR muramyl dipeptide recognition domain of NOD2 has been found to associated with altered NF-kB activation and this increased the susceptibility to Crohn’s disease (Hugot et al. 2001).

Fig. 2.

Non-synonymous SNPs in buffalo NOD1 analyzed by Polyphen-2 for their sensitivity and specificity score and the possible/probable damaging effect on protein structure and function

Genotyping and allele frequency distribution

Genotyping of three polymorphic sites, SNPs—g.338C>T (AciI) transition (Pro113Leu) in NOD1 exon-2, g.797A>G (MspI) transition (Gly266Glu) and g.1482G>C (PvuII) transversion (Gln494His) in exon 3 was carried out using restriction fragment length polymorphism (RFLP), while one synonymous SNP—g.2394A>C found in functionally important leucine-rich repeat-4 (LRR4) domain analyzed by tetra-ARMS PCR in a total of 393 buffaloes was represented by different breeds/populations. These SNPs were selected based on the type of amino acid changes (mostly hydrophobic to hydrophilic or charged) resulting in the non-synonymous nature of SNPs and the possibility of designing simple RFLP or tetra-ARMS PCR genotyping protocols. Variable allelic frequencies were observed across riverine and swamp buffaloes at all the four loci. SNP variation g.1482G>C was observed only in swamp buffaloes with an overall frequency of 0.65 for G and 0.35 for C allele, whereas in riverine and hybrid animals, this polymorphic site was fixed for G allele (Table 3). Though premature to say, these variable allelic frequencies could be responsible for the variation in immune response to diseases as reported earlier (Mingala et al. 2009). Individual frequencies of the genotypes were in the Hardy–Weinberg equilibrium for most of the SNPs, as determined by Chi square test. The calculated Chi square values ranged from 0.02 to 34.00 (Table 2-SI in supplementary information). NOD1 polymorphism has been demonstrated to modulate pro-inflammatory cytokine responses induced through Toll-like receptor or Nod-like receptor ligands in rheumatoid arthritis-affected human beings (Plantinga et al. 2013). Principal component analysis based on genetic distance among haplotypes of four SNPs examined also showed classification of different riverines, swamps and their hybrid buffaloes into distinct quadrants. Within riverine and swamp types, classification was in correspondence with the geographical distribution of breeds (Fig. 3). The results further supported the distinct allelic distribution among riverines, swamps and their hybrids, at these four loci, which could be responsible for the variable immune response to pathogens. Similar findings have been reported with other innate immune response genes also (Dubey et al. 2013a).

Table 3.

Allelic frequencies of the genotyped SNPs (g.338C>T, g.797A>G, g.1482G>C, g.2394A>C) in the coding region of NOD1 gene of Indian riverine and swamp buffaloes

Breeda	N (animals)	g.338C>T		g.797A>G		g.1482G>C		g.2394A>C
		C	T	A	G	G	C	A	C
Swamp
DB	12	0.96	0.04	0.63	0.38	0.96	0.04	0.96	0.04
SL	27	0.94	0.06	0.82	0.19	0.88	0.11	0.87	0.13
NGL	36	0.97	0.03	0.83	0.17	0.74	0.26	0.86	0.14
MN	34	1.00	0.00	0.96	0.04	0.38	0.62	0.99	0.02
MZ	15	1.00	0.00	0.83	0.17	0.30	0.70	1.00	0.00
Swamp overall	124	0.97	0.03	0.81	0.19	0.65	0.35	0.93	0.07
Hybrids
ASW	48	0.83	0.17	0.47	0.53	1.00	0.00	0.27	0.73
Riverine
MU	30	0.78	0.22	0.53	0.47	1.00	0.00	0.43	0.57
JAFF	30	0.65	0.35	0.48	0.52	1.00	0.00	0.13	0.87
MTW	46	0.72	0.28	0.30	0.69	1.00	0.00	0.54	0.46
CH	34	0.82	0.18	0.43	0.57	1.00	0.00	0.46	0.54
TD	47	0.85	0.15	0.20	0.79	1.00	0.00	0.59	0.40
NR	34	0.85	0.15	0.60	0.39	1.00	0.00	0.44	0.56
Riverine overall	221	0.77	0.23	0.42	0.58	1.00	0.00	0.43	0.57

^a ASW Assam (hybrids), DB Dibrugarh (Assam-swamp), SL Silchar (Assam swamp), NGL Nagaland (swamp), MN Manipur (swamp), MZ Mizoram (swamp). Riverine buffaloes include—MU Murrah, JAFF Jaffarabadi, MTW Marathwadi, CH Chilika, TD Toda, NR Nili Ravi

Fig. 3.

Genetic distance-based principal component analysis on the basis of haplotypes of four SNPs (g.338C>T, g.2394A>C, g.797A>G, g.1482G>C) genotyped and examined between buffalo breeds [ASW—Assam (riverine and hybrid), DB—Dibrugarh (Assam swamp), SL—Silchar (Assam swamp), NGL—Nagaland (swamp), MN—Manipur (swamp), MZ—Mizoram (swamp). Riverine buffaloes include—MU—Murrah, JAFF—Jaffarabadi, MTW—Marathwadi, CH—Chilika, TD—Toda, NR—Nili Ravi]

Sequencing and characterization of buffalo NOD1 transcript

Sequencing of full-length cDNA revealed buffalo NOD1 gene comprising of an open reading frame (ORF) of 2865 nucleotides, encoding a polypeptide of 954 amino acids found to similar in size to that of other ruminants. The buffalo NOD1 protein had a theoretical molecular weight of 107 kDa and an isoelectric point (pI) of pH 6.86, and the cellular location was predicted to be cytoplasmic as reported in other species (Kufer et al. 2008). Structurally, buffalo NOD1 comprised of one caspase recruitment domain (CARD, residues 20–104) at NH₂ terminal, involved in interaction with RIP2, leading to the downstream signals for the activation of NF-κB and MAPKs pathways (Viala et al. 2004; Maharana et al. 2017). Centrally located NACHT domain (residues 196–368) and eight leucine-rich repeat (LRR) motifs at C-terminal were conserved across all ruminant species including buffalo. Further, within NACHT domain, two functionally important motifs WalkerA (GKS) at position 207–209 and WalkerB (DGLDE) 284–288 were found to be conserved across species (Fig. 4). Mutations within these regions have been found to be associated with several genetic diseases in human (Zurek et al. 2012).

Fig. 4.

Alignment and architecture of trans-species conservation of deduced amino acid sequences of NOD1. Amino acid sequences of buffalo, cow, goat, sheep, and human NOD1 were aligned. Shaded sequences indicate LxxLL motifs, while numbers indicate the amino acid position. Similar residues are marked with dots. Gaps were introduced to optimize alignment. The domain architecture is highlighted under boxes with relevant stretch of sequence as follows: CARD (green strip), NACHT (blue strip), LRRs (boxed). Within NACHT domain, WalkerA (GKS) at position 207–209 and WalkerB (DGLDE) 284–288 sequences are shown in boxes

Buffalo NOD1 cDNA, amplification from buffalo PBMCs revealed, apart from amplification of expected 2.9 kb full-length transcript, the generation of a smaller product of 1.2 kb in length, with equal intensity in agarose gel (Fig. 5a). Cloning and sequencing of both products divulged a smaller product being generated due to synthesis of a splice variant transcript, having exon3 (1084 nucleotides of amplified cDNA, 1037 nucleotides considering only the ORF) spliced out (Fig. 5b). Expression of this novel splice variant of NOD1 has been identified in ruminants for the first time. This variant with a deletion of exon3 region lacks the functionally important NACHT as well as downstream LRR domains and had an early stop codon resulting in the expression of a truncated protein (Fig. 5c). In human beings, a total of eight NOD1 splice variants has been reported. Cryptic splice sites contribute to resulting in altered protein translation, stability, and expression of NOD1 isoforms, which might be the link between several NOD1-associated inflammatory diseases (Fan et al. 2016). Functions specific to tissues have also been attributed to alternative splicing mechanisms in various genes (Yeo et al. 2004; Zhang et al. 2013), confirmed recently by protein–protein interaction studies of several alternate spliced genes in human (Yang et al. 2016). Keeping in view the multiple functions assigned to NOD1, this functional attribute cannot be ruled out in buffalo also.

Fig. 5.

Splice variant identification in buffalo NOD1 gene: a gel picture showing amplification of buffalo NOD1 cDNA revealing amplification of two fragments of 1086 bp and 2914 bp of equal intensity. b Buffalo NOD1 cDNA sequence analysis giving evidence of alternate splicing. c Domain structure prediction of the splice variant, showing spliced out region

Apart from documentation of splice variants, sequencing also helped in identification of indels to be present in the exon 3 of buffalo NOD1. Alignment of buffalo full-length transcript revealed, one insertion of three nucleotides (1533–1535) in ruminants, as compared to human (Fig. 1). McGovern et al. (2005) found that indels polymorphism in functional NOD1 is associated with inflammatory bowel disease (IBD). However, the role of splice variants and indels (insertion and deletion) observed in buffalo needs to be explored further to ascertain their role if there is any in immune response.

The phylogenetic analysis, based on nucleotides sequence of NOD1 of different species, revealed buffalo having the closest phylogenetic relationship with other ruminant species, cow, goat and sheep (Fig. 6). All the ruminant species clearly branched off in a separate cluster from non-ruminant species. However, among non-ruminants, pig was comparatively more close to ruminant species. Mice NOD1 sequence was found to most diverged among all the mammalian species. Our findings on sequence and phylogenetic analysis of NOD1 gene are similar to other reports for different species and other PRR genes (Dhiman et al. 2017; Dubey et al. 2013b; Goyal et al. 2014), which indicates co-evolution and conserved nature of this receptor protein across closely related ruminant species. Conserved residues found within functional domains across NOD-like receptors are speculated to contribute to the intramolecular interaction of the LRR regions to regulate NLR activation (Proell et al. 2008).

Fig. 6.

Phylogenetic analysis based on NOD1 nucleotide sequence alignment, showing close clustering of ruminant species, buffalo, cow, sheep and goat. Nucleotide sequences were aligned using the ClustalW program and the phylogenetic tree was constructed using MEGA 6 software

Tissue distribution by quantitative real-time assay of buffalo NOD1

The mRNA expression of buffalo NOD1 quantified in ten different buffalo tissues was obtained by real-time PCR assay. The NOD1 transcript was found to be expressed in all the ten buffalo tissues, however, the expression varied between tissues. Our results suggest that buffalo NOD1 transcript was most abundantly expressed in PBMCs. Substantial expression was also found in lung, small intestine and liver tissues, whereas, the spleen, stomach, and mammary gland showed moderate expression. Although, a low expression level was seen in heart, skeletal muscle and kidney tissues (Fig. 7). As kidney tissues showed the lowest expression (highest Ct value), it was included as a calibrator (Table 3-SI in supplementary information). Overall expression analysis across tissues indicates functional significance of NOD1 in these organs, showing variability in transcript levels of this important multifunctional protein. This is the first report on the expression of the NOD1 gene in buffalo tissues, and it is important given the fact that it performs variable functions across tissues (Zhang et al. 2013).

Fig. 7.

Basal expression of NOD1 gene across buffalo organs by quantitative real-time PCR, taking kidney showing the least expression, as the normalizer tissue. Relative expression of NOD1 in other tissues were represented as fold change from the normalizer (kidney)

Trans-species comparison of key residues in NOD1

NOD1 amino acid residues of buffalo, cow, goat, sheep and human were aligned and evolutionary tracing was used to examine the amino acid conservation (Fig. 4). Comparison among Bovidae NOD1 amino acid residues was found to be highly similar in CARD, NACHT and LRR domains. Disulphide bond formation in buffalo NOD1 through Cys39 residue was found to be conserved in ruminants as well as humans as previously studied by Mayle et al. (2014). In spite of that, buffalo NOD1 shows varying degreesw of conservation of the protein interaction motif LxxLL found in the nuclear receptors. Buffalo NOD1 has four LxxLL motifs L41 (CARD), L250, L314 (NACHT) and L593. Of these, two motifs LQDLL (41) and LCGLL (593) were found conserved across examined species. Residues involved in dimeric interface, implicated in RICK binding and signal transduction of NOD1-CARD were conserved across the examined species with exceptions of Leu41Val, Ala94Ser, and Arg101Lys. Residues that contributed to ubiquitination were found conserved across the ruminants with exception of Ala94Ser in sheep, while His88, Val89, Ile90 were substituted for Tyr88 and Lys89, 90 in humans (Fig. 8).

Fig. 8.

Amino acid conservation in CARD domain of NOD1 across species (buffalo, cow, sheep, goat and human). Conservation patterns observed are around residues implicated in the ubiquitination. Residues are colored according to hydrophobicity (green—hydrophobic, blue—hydrophilic)

Homology modeling and NOD1-iE-DAP interaction

The 3D model of riverine and swamp type buffaloes and cattle NOD1 was generated to have an insight into the structural basis of its function using rabbit NOD2 (5IRL) as the template that shared 35% sequence identity with 85% query coverage. The best models were selected and validated by SAVE server on the basis of core region in Ramachandran plot and Z score value. Riverine, swamp and cow had almost similar DOPE scores, − 98,428.617, − 97,959.078 and − 98,886.242, respectively, altogether having 13 amino acid variations between them. The amino acid variations and ligand interaction of NOD1 of two buffalo types and cattle are shown in Fig. 1-SI (in supplementary information). Homology models of buffalo NOD1-LRRs exhibited a horseshoe-like shape as reported in other species (Boyle et al. 2013) (Fig. 2-SI in supplementary information). Based on homology model, ten LRRs were identified in buffalo NOD1 based on the sequence used in this study. Similarly, Brahma et al. (2015) have also reported, ten LRRs in buffalo NOD1. Buffalo NOD1 showed a consensus in LRRs region LxxLxLxxNxL (Fig. 9). Previous studies through mutagenesis in human NOD1 LRRs have implicated residues His788, Lys790, Gly792, Glu816, Gly818, Trp820 and Trp874 that are important for ligand interaction (Boyle et al. 2013). However, molecular docking in this study with iE-DAP indicates residue interacting sites in riverine buffalo (His789, Leu792, Trp821, Asn845 and Ser847), swamp buffalo (Ser737, Tyr765, Lys791 and Trp821) and cattle (Tyr765, His789, Lys791, Trp821 and Asn845) (Fig. 10). These variable ligand interacting sites could possibly have evolved due to different pathogens to which different species are exposed to. Riverine, swamp and cattle having docking scores of 5.76, 5.32 and 7.10, while their docking energies were − 5.9, − 4.9 and − 6.2, respectively, indicated a stronger interaction of ligand with swamp buffalo NOD1, which is evident from the higher disease resistance ability of this particular species (Mingala et al. 2009).

Fig. 9.

Conservation patterns of LRR domains in buffalo NOD1. Residues highlighted in yellow are involved with iE-DAP ligand interaction (where L = Ile, Arg, leu, Phe, Val, Met; N = Asn; x = any amino acid; signature residues underlined) (green—hydrophobic, blue—hydrophilic)

Fig. 10.

NOD1 receptor ligand interaction. a NOD1-LRR-iE-DAP cattle complex, b NOD1-LRR-iE-DAP riverine buffalo complex, c NOD1-LRR-iE-DAP swamp buffalo complex

Our work thus highlights complete sequence characterization of buffalo NOD1 gene and comparative transcript analysis suggests trans-species conservation across the potential functional domains. Identification of splice variants due to alternative splicing in buffalo NOD1 gene for the first time as well as tissue distribution indicate its diverse functions. A better understanding of distinct variability in allele frequencies sharing among riverine, swamp and hybrid buffaloes with different utilities could have an impact on disease susceptibility. Homology modeling and interaction with iE-DAP of NOD1 revealed potential ligand-binding residues in LRR region among bovines. Overall, this study helps to provide information about expression as well as genetic variability and their impact on buffalo NOD1 function thus supporting a better understanding of innate as well as adapted immunity.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Conflict of interest

On behalf of all the authors, the corresponding author states that there is no conflict of interest for publishing this manuscript.