Characterization of Smoc-1 uncovers two transcript variants showing differential tissue and age specific expression in Bubalus bubalis

Characterization of buffalo Smoc-1

The cloning strategy for isolation of full length Smoc-1 CDS of 3474 bp and 1933 bp is demonstrated in Fig. 1A &1B. Clone I contained an insert of 1263 bp (PSmoc-1) lacking 5'/3' UTRs and signal peptide sequence [26]. Clone II (1414 bp) covered complete coding sequence from nucleotides 119–1603 but with partial 5' & 3' UTRs. The 3' UTR was covered by three fragments represented by clone III, IV & V of which clone III covered nucleotides from 1461–2473; clone IV, 2307–3328; and clone V, 2435–3428. The polyadenylation signals were accessed with 3'RACE followed by sequencing of 30 recombinants. This resulted in the identification of two other clones (Clone VII & VIII). Clone VII represented nucleotides 1407–1915 followed by 17 mer Poly(A) tail. Clone VIII covered 2435–3474 encompassing 18 mer Poly(A) tail. The 5' UTR represented by clone VI (nucleotides 1–649) was generated by 5' RACE. Following this strategy, full length CDS of Smoc-1 [FSmoc-1, GenBank: DQ159955 and EF446167] was deduced from different overlapping fragments (Fig. 1 & Additional file 1). The GC rich 5'-UTR of 239 bp was followed by an initiating ATG codon and the terminating TAA codon fell at nucleotide 1545. Thus, translation of the sequence from nucleotide 240 to 1544 encodes a putative protein of 435 amino acids with a calculated molecular mass of 48332 Da with a predicted oxidoreductase activity. Multiple sequence alignments showed that this protein is 95% and 98% identical to human & cattle Smoc-1, respectively (Table 1).

S.N.	Species	Accession numbers (Ensembl/NCBI)	Transcript length	Full length gene (In Kb)	No. of exons	Amino acid residues	Chromo-somal location	Homology with buffalo Smoc-1
								CDS	Amino acids	Secondary structure
1.	Bubalus bubalis	DQ159955	3274	NA	12	435	11	100%	100%	100%
2.	Homo sapiens	ENSG0000198732/AJ249900	3666	172.94	12	434	14q24.2	91%	95%	93%
3.	Pan traglodytes	ENSPTRT0000011881/XM_510036	3,669	154.09	14	435	14	90%	94%	94%
4.	Bos taurus	XM_612029	3473	NA	NA	434	10	98%	98%	97%
5.	Mus musculus	ENSMUSG0000021136/NM_022316	3472	179.58	13	463	12d3	84%	93%	89%
6.	Rattus norvegicus	ENSRNOG00000005998/NM_0102835	1359	193.42	12	452	6q24	85%	92%	89%
7.	Canis familiaris	ENSCAFT0000026288	1452	152.3	12	459	8	87%	95%	85%
8.	Gallus gallus	ENSGALG00000009415	1404	98.76	12	468	5	78%	84%	83%

Buffalo Smoc-1 shows two transcript variants

Northern blot detected two bands of 3.5 kb and 2.0 kb (Fig. 2) which were confirmed to be two transcript variants of Smoc-1 with RACE and sequencing (Fig. 1 and Additional file 1), variant-01 of 3474 bp [GenBank: DQ159955] and variant-02 of 1933 bp [GenBank: EF446167]. Both the variants encoded for identical proteins but showed differences in their 3'UTR length, polyadenylation signals and Poly(A) tails. In the 3' UTR of variant-01 & -02, five & two copies of mRNA instability motif (ATTTA), respectively, were observed. In addition, there were orthodox polyadenylation signals (AATAAA), 1787 and 334 bp downstream of the translation termination codon, in transcript variant-01 and -02, respectively (Additional file 1). Interestingly, two types of transcripts of Smoc-1 have been reported independently in the literature for human [GenBank: AJ249900 and BC011548] and cattle [GenBank: XM_612029 and NM_01079771]. Database search and multiple nucleotide sequence alignment of both the variants showed their conservation across various species (Additional file 2).

Structure of the buffalo Smoc-1 and its phylogenetic delineation

The buffalo Smoc-1 has 12 exons, varying in length from 48 to 1916 bp in variant-01 and 48 to 402 bp in variant-02 (5^th being the smallest and 12^th, the longest one) (Fig. 1 & Additional file 1). Each domain of Smoc-1 is encoded by one or more exons and the domain border coincide with the conserved splice sites. Buffalo Smoc-1 differed from other species in various aspects (Additional file 3), briefly, exon 8 was found to be more diverse at nucleotide level whereas exon 2 comprising the follistatin domain was most divergent at amino acid level. Exon 6 was the most conserved across the species. The buffalo Smoc-1 cross-hybridized to genomic DNA from 13 different species with almost equal signal intensity, confirming its faithful conservation across the species (Additional file 4A). Phylogenetic analysis demonstrated cattle as the closest species followed by human and chimpanzee while dog was the most distant one (Additional file 4B). In addition, the buffalo Smoc-1 also showed homology with other species such as birds, rodents and bony fishes (Additional file 4C). Details of the Smoc-1 gene(s) from different species along with their accession numbers are given in the Table 1.

Domain Organization of Smoc-1

Homology search for the buffalo Smoc-1 protein demonstrated presence of all the domains characteristic to the BM-40 family (Fig. 3). Accordingly, first 26 amino acids at the N-terminus conform well to the signal peptide consensus ending with a signal peptidase cleavage site [28]. Mature Smoc-1 comprises of 409 amino acids. Like in the human, all the essential features of each domain are conserved in buffalo. Since no transmembrane-spanning hydrophobic domain is present in the sequence, Smoc-1 is presumably secreted out from the cells. Further scrutiny allowed the distinction of five modules, an FS domain (Fig. 3A), a TY domain (Fig. 3B), a Smoc-1 unique domain (Fig. 3C), a second TY domain (Fig. 3D) and an EC domain (Fig. 3E). Residues 42–88 are homologous to the canonical FS domain, composed of two sub-domains with the second being similar to the Kazal domain. Structure-based alignment showed that all the six cysteines and the features of secondary structure are conserved in both the TY domains of buffalo Smoc-1. Further, two TY domains (residues 89–159 and 222–293) are separated by 62 amino acids unique to the Smoc-1. Detailed in-silico analysis unveiled a potential N-glycosylation site at Asn-214 and five O-glycosylation sites at Thr-155, -82, -184, -187 and -345 in buffalo Smoc-1. The C-terminus is homologous to the characteristic amphipathic α-helix and the helix-loop-helix motifs of the ECD of BM-40. However, based on the in-silico analyses, in contrast to the ECD of BM-40, both EF-hand motifs of Smoc-1 were found canonical indicating the calcium binding sites (Fig. 3E). Compared to other species, buffalo Smoc-1 showed some specific alteration like R49K in FS domain and V431 insertion in ECD. Some of these changes were from polar to non-polar amino acids and vice versa, similar to that in mouse/rat. In case of human/chimpanzee, these changes always maintained similar biochemical nature (Fig. 3). Interesting enough, predicted secondary structure showed alterations at the N-terminus involving replacement of 8 alpha-helices by equal number of beta-sheets and insertion of 3 helices in FS domain in comparison to other mammals (Additional file 5).

Single copy of the Smoc-1 gene located on chromosome 11 in buffalo

For copy number status, using SYBR green assay and Real Time PCR, a straight curve was obtained with a slope -3.2 using 10 fold dilution series of buffalo blood/semen genomic DNA and FSmoc-1 plasmid as template. Ct increase of 3.3 per dilution and a single dissociation peak indicated maximum efficiency and high specificity of the primer sets. Extrapolation of this standard curve demonstrated the single copy status of the Smoc-1 per haploid genome in buffalo (Additional file 6). Chromosomal mapping of the same was also performed using Fluorescent in situ hybridization (FISH) which confirmed the localization of Smoc-1 gene on the distal end of the acrocentric chromosome 11 in buffalo (Fig. 4).

Recombinant expression of Smoc-1

Affinity-purified recombinant Smoc-1 expressed in E. coli BL2 1(DE3) revealed a major band at ~70 kDa in 10% SDS-PAGE under reduced conditions. The deduced molecular mass of mature Smoc-1 is 45.4 kDa and remaining ~25 kDa represent GST tag. The Anti-PSmoc 1-pAb recognized native protein of ~70 kDa in the western-blot analysis. Anti-SySmoc-1-pAb generated against commercially synthesized 26 amino acids (69S to 95G) also showed the similar result (Additional file 7), substantiating its high specificity. The pre-immune serum did not detect any protein in the western blot (not shown).

Highest expression of Smoc-1 transcript variants in liver

Northern blot analysis showed abundant Smoc-1 transcripts in liver and faint signals in testis and ovary. After prolonged exposure, the spleen, lung, kidney and heart also showed negligible to faint signals (Fig. 2). Similarly, RT-PCR followed by Southern hybridization detected reduced signals in spleen (Fig. 5A &5B) and negligible ones in lung, kidney and heart after prolonged exposure (not shown). Using quantitative expressional analysis, β-actin as an internal control and lung cDNA as calibrator, the highest level of expression of transcript variant-01 (165–364 folds) and -02 (360–697 folds) was observed in liver (Fig. 5C). This was substantiated further by expression data from the five additional animals. In the same assay, sperm cDNA showed similar level as that of testis. However, relatively higher amount (1.2–3.5 folds) of variant-02 was observed in all the tissues examined as compared to that of variant-01 (Fig. 5D). Based on these observations, the variant-02 may be addressed as "major transcript" and -01 as "minor" one.

Since all the mRNA transcripts may not translate into protein, the relative quantitation mentioned above was substantiated by the Western blot analysis using total tissue proteins and anti-SySmoc 1-pAb which detected the ~45 kDa bands (mature protein). Interestingly, the strongest signal was detected in liver and faint ones in testis, ovary and spleen (Fig. 5E) whereas lung, kidney and heart were found to be devoid of Smoc-1 protein corroborating the RT-PCR and relative expressional analysis. As a control, anti-β-actin-mAb showed almost equal signal intensity in each tissue (Fig. 5F).

Age specific progression of expression of Smoc-1 in water buffalo

We report for the first time, expression profile of Smoc-1 in water buffaloes of varying ages starting from 20 days to 15 years. Lowest expression of Smoc-1 was detected in the blood lymphocytes of animals aged 20 days with gradual increase in the expression (1.5–2 times) from 1 month-10 months. However, a sharp enhancement in the expression (25–30 times) was detected at the age of 1–1.25 years and this remained consistent up to the age of 15 years and beyond (Fig. 5G). The dramatic increase in expression of Smoc-1 at around 1 year of age was further confirmed by western blotting using anti-SySmoc 1-pAb and total protein isolated from blood samples of the same animals (not shown).

The comparative expression analysis of the two transcript variants revealed a gradual increase of the variant-02 compared to that of variant-01 with the progression of age (Fig. 5H). The expression of variant-01 was higher in animal aged up to ~6 months, after which the variant-02 expression starts increasing gradually during the ages 6 to 15 months and remained consistent thereafter.

Association of Smoc-1 with basement membranes

Smoc-1 was present abundantly in the basement membrane zone of discontinuous endothelial cell layer or the tunica media around the central veins in liver (Fig. 6A). In addition, Smoc-1 was found to be ubiquitously distributed in the connective tissues surrounding each lobule and extracellular matrices of space of Disse. In testis, Smoc-1 was abundant in the basement membrane zone surrounding coiled seminiferous tubules below the columnar epididymis and scarcely in the interstitial tissues (Fig. 6B). Smoc-1 was localized in the zona pellucida of ovary in buffalo (not shown) similar to that in mice. In other tissues also, it remained localized within the basement membrane zones.

Sample collection and isolation of genomic DNA, total RNA and cDNA synthesis

Blood and tissue samples of both the sexes of water buffalo were collected from local slaughterhouse, Delhi following strictly the guidelines of Institute's Ethical and Biosafety Committee. Buffalo semen samples were collected from the local dairy farm. Details of the genomic DNA from different species used in this study for cross-hybridization have been given earlier [26, 42]. Total RNA was isolated from all the tissues, semen and blood samples from different age group of animals using standard protocols [42]. The cDNA synthesis was conducted using a commercially available kit (ABI, California, USA) and confirmed by PCR amplification using a set of bubaline derived β-actin (forward 5'CAGATCATGTTCGAGACCTTCAA3' and reverse 5'GATGATCTTGATC TTCATTGTGCTG 3') primers.

Minisatellite Associated Sequence Amplification (MASA) with consensus of 33.15 repeat loci

MASA was conducted using 16 nucleotide long oligo (5' CACCTCTCCACCTGCC 3') primer derived from the consensus sequence of 33.15 repeat loci and cDNA from different tissues. The experimental details of MASA have been described earlier [26]. MASA uncovered several bands and one was found to represent partial sequence of Smoc-1 gene (PSmoc 1: GenBank accession no. AY947405), lacking 5'/3'UTR and signal peptide.

Amplification of full length Smoc-1 CDS using end point PCR and RACE

Full length Smoc-1 CDS was isolated using four sets of primers designed from of 5' and 3' regions of the human and cattle Smoc-1 sequences (GenBank Accession nos. AJ249900 and XM_612029) respectively using primer3 output [43]. Details of the primer sequences, Tm and corresponding sizes of amplicons are given in the additional file 8. End point PCR was conducted to amplify 3'UTR using cattle derived primers (JS275-JS997). The 5' UTR and polyadenylation signal at 3'UTR were identified using 5' & 3'RACE kits (Invitrogen, USA) and Smoc 1-specific primer J5UTR & JS977/JS997, respectively. All the PCR reactions were conducted using Vent Polymerase (NEB, USA) following standard protocols. The amplicons were analyzed by gel electrophoresis followed by cloning and sequencing. The sequences so obtained were assembled into full length Smoc-1 (FSmoc-1). To ascertain the possible insertion of intron 12^th in 3'UTR, the primers JS990-SA991 & JS996-997 were used on buffalo genomic DNA as template to conduct end point PCR.

Cloning, sequencing, Secondary structure prediction and phylogenetic delineation

The PCR products were tagged with dTTP and cloned into pGEM T-Easy vector (Promega, USA). The sequences of all the fragments were analyzed using Blast Search [44], ClustalW [45] and Gene Runner software. Finally, derived full length transcripts were submitted in the GenBank (Accession nos. DQ159955 and EF446167). The secondary structure of predicted protein was ascertained using Phyre software [46]. The calcium binding affinity, and N-glycosylation and O-glycosylation sites were predicted using different bioinformatics tools. Based on the homology, the phylogenetic tree [45] was constructed using Smoc-1 transcript(s) from buffalo and other species (Table 1). Only those species showing significant homologies (85–100%) were taken into consideration for phylogenetic analysis.

Cross hybridization of buffalo Smoc-1 gene across the species

Approximately, 200 ng of heat denatured genomic DNA from different species was briefly run on 0.8% agarose gel and transferred onto the nylon membrane. Hybridization was conducted with labeled PSmoc-1and FSmoc-1 pr obes following standard procedure [42].

Metaphase chromosome preparation and Fluorescence in situ hybridization

Approximately, 400 μl of whole blood from normal buffaloes was cultured for chromosome preparation following standard protocols [47]. Cattle derived BAC clone (Ctg9.CH240-54I18) representing full length Smoc-1 gene was used as probe. The clone was labeled with Fluorescein-12-dUTP using Nick Translation Kit from Vysis, (IL, USA). Fluorescein was detected with biotinylated anti-fluorescein antibody and FITC-avidin DCS (Vector Labs) following prescribed methods [47]. Map position of Smoc-1 gene on the chromosome was carried out following the International System for Chromosome Nomenclature of Domestic Bovids (ISCNDB 2000).

Northern blot, RT-PCR and Southern Blotting

For Northern blot analyses, 5–10 μg of total RNA resolved on the 1% agarose gel was transferred onto the nylon membrane (Amersham Biosciences, USA). Hybridizations were performed under the high stringent conditions using standard procedure [26, 42]. FSmoc 1 probe was labeled with [³²P] α-dCTP using rediprime™ II kit (Amersham Pharmacia biotech, USA). The Northern blot results were confirmed by RT-PCR with the internal primers designed from FSmoc-1 (F 5'-GGTTTCTCATAAGTGACCGTGACC-3', R 5'-TGAGATGACCTTGTCC TTGTTCAG-3') and cDNA of different tissues as template using thermal profile, 95°C-1 min, 59°C-1 min, 72°C-1 min. The products were transferred onto the nylon membrane followed by hybridization with [³²P] α-dCTP labeled FSmoc 1 using standard procedures. Bubaline derived β-actin gene probe and bacterial genomic DNA were used as positive and negative controls, respectively.

Copy number calculation and quantitative expression using Real Time PCR

Copy number of Smoc-1 gene was calculated based on absolute quantitation using SYBR green assay and Sequence Detection System-7000 (SDS-7000, ABI, USA). Two primer pairs (Additional file 8) specific to 11^th exon of Smoc-1 were designed using Primer Express Software V2.0 (ABI, USA). The relative transcription of Smoc-1 across the tissues was assessed using the same primer set and equal quantity of total cDNA from different tissues and semen samples of buffalo. The relative expression of both the transcript variants was studied using three primer sets, two for both the variants (JSR1015-1018) and one specific for variant-01 (JSR1033-1034). Age specific expression was also carried out using cDNA from blood lymphocytes with the primer sets (JSR1015-1018) which picked up both the variants and another one JSR1033-1034 specific for variant-01. The total cDNA amounts from different tissues were optimized using a set of buffalo β-actin primers (forward 5'TCACGGAGCGTGGCTAC AG3' and reverse 5'TTGATGTCACGGACGATTTCC 3'). Presence of genomic DNA in the cDNA template was ruled out by using mRNA as template in several independent Real Time PCR and end point PCR reactions. Each reaction was repeated three times in triplicates. The details of the copy number calculation of the Smoc-1 mRNA transcripts have been described earlier [26, 48].

Protein Expression and Production of Anti-Smoc-1 Antiserum

Using PSmoc-1 as template, the Smoc-1 was re-amplified to accommodate a Bam HI site at the 5' end (5'-CGGGATCCCACCTCTCCACCTGCCCCAGG-3') and Xho I site at the 3' end (5'-CCCTCGAGTTAGACGAGGCGTCCTACTTC-3'). The resulting amplicon was cloned in pGEX-4T1 vector (Novagen, USA) at BamH I/Xho I sites. Expression of the recombinant GST-tag-Smoc 1 in BL21 (DE3) was induced with 1 mM IPTG at 37°C for 4 h and the recombinant Smoc-1 protein was purified using GST-tag purification resin (Clontech, USA). A rabbit was immunized with purified recombinant pGEX-4T1-PSmoc 1 using alum as an adjuvant to obtain the Anti-PSmoc-1-pAb. To ensure the specificity, primary antiserum (Anti-SySmoc 1-pAb) was obtained for a commercially synthesized 26 amino acid (69S to 95G) long peptide, conjugated to Keyhole limpet hemocyanin (KLH), specific to Smoc-1 domain.

Isolation of total protein from different tissues and Western Blotting

Denaturing 10% polyacrylamide gels were used under reducing conditions for analyzing culture medium and E. coli expressed proteins. Following electrophoresis, proteins were transferred onto nitrocellulose membranes. After blocking with 5% non-fat milk, 1%BSA in PBS for 45 min at room temperature, the membrane was probed with primary antibodies (anti-PSmoc 1-pAb raised against pGEX-4T1-PSmoc 1 anti-SySmoc 1-pAb against synthesized peptide of Smoc-1). Secondary detection was carried out with goat anti-rabbit IgG conjugated with HRP (Bio-rad, USA) following standard protocol [49].

Total protein was isolated from different tissues using Tri-X reagent (MRC) followed by acetone precipitation. Protein quantity was normalized on 12% SDS PAGE followed by its transfer, hybridization and detection using standard procedure [49].

Immunohistochemistry on Buffalo Tissue Sections

The distribution of Smoc-1 protein in different tissues was studied on paraffin sections by indirect Immunohistochemistry using Anti-SySmoc1-pAb (generated against synthesized polypeptide for Smoc-1 as stated earlier). Freshly prepared buffalo tissues were fixed for 1 hr in 4% paraformaldehyde/PBS and after dehydration, were embedded in paraffin. Sections were sliced and processed according to standard procedures [49]. After blocking with 1% BSA/TBS, the sections were incubated with the anti-PSmoc 1-pAb followed by HRP-labeled goat anti-rabbit IgG (Bio-rad, USA) and detected using Di-amino Benzene (DAB) as substrate [49]. The sections were observed under BX-51 microscope (Olympus, JAPAN).