Molecular characterization of pig LEG1a protein

SUN Minghao, HUANG Yuqi, DANG Yanna, HE Jin ()

Abstract Liver-enriched gene 1 (LEG1) is a newly identified gene that plays an important role in the liver development of zebrafish and the innate immunity of platypus. However, little is known about LEG1 in eutherians such as mice and humans.In this study,we explored the molecular characteristics of the LEG1 protein in a pig model (pLEG1a).A rabbit polyclonal antibody against pLEG1a was produced and validated. Then, we detected the pLEG1a protein in both the salivary gland and lung. Moreover, pLEG1a could be detected in the saliva, suggesting its secretory nature. Ectopic expression of pLEG1a in HEK293T cells further confirmed that pLEG1a was a secretory protein. In addition, glycosylation tests showed the existence of glycosylated bands which could be enzymatically removed, indicating that the secreted pLEG1a was N-glycosylated. In conclusion,the pLEG1a protein is similar to the LEG1 proteins of other species, especially eutherian LEG1s, in terms of expression profile, glycosylation, and secretion ability, suggesting that pLEG1a is a good model to further study the function of eutherian LEG1s.

Key words pig;LEG1a protein;secretory protein;glycosylation

Theliver-enriched gene 1(LEG1orC6orf58) is a newly identified, evolutionarily conserved gene with rare functional annotation. ZebrafishLEG1s(zleg1s) were first studied for their enrichment in the liver, suggesting their potential roles in liver development[1]. This hypothesis was later confirmed by knocking downzleg1susing a morpholino,resulting in a small liver phenotype under stress conditions[2]. MammalianLEG1was first studied in the platypus, in which the monotreme lactation protein encoding gene (MLP) has anti-Gram-positive bacterial activity[3]. However, the function ofLEG1in eutherian species remains poorly understood. Due to two tandem duplication events in mammals,LEG1shave three paralogs,LEG1a,LEG1b, andLEG1c, of which pigs have all three genes (pLEG1a,pLEG1b,andpLEG1c), while humans and mice only haveLEG1a(hLEG1aandmLeg1a). An evolutionary study indicated that vertebrateLEG1sexperienced purifying selection; however, different paralogs ofLEG1smight retain distinct subfunctions of ancestralLEG1[4]. In a liver cell line, transcriptomic prediction confirms that the platypusMLPhas antibacterial activity, andpLEG1cmay play a dispensable role in the endoplasmic reticulum stress response and protein folding regulation, andpLEG1aandpLEG1bhave little function in the liver[5]. However, furtherin vivostudies are required to elucidate the biological roles of theLEG1genes.

The LEG1 protein is characterized by the presence of the LEG1 domain (or domain of unknown function 781, DUF781), preceded by a signal peptide,indicating that it belongs to a secretory protein, as accumulating evidence has shown its existence in milk, saliva, and seminal plasma[3,6-7].Studies of hLEG1a, zLEG1, and platypus MLP have suggested that glycosylation may be another hallmark of the protein family. As we proposed thatpLEG1sare good models to study the gene family in eutherian species[4], the pLEG1 proteins need to be characterized at the protein level. In the current study, a pLEG1aspecific antibody was first validated and used to demonstrate that pLEG1a is similar to other LEG1 proteins. This information would facilitate the research ofLEG1in eutherian species.

1 Materials and methods

1.1 Ethical statement

This work, involving the use of porcine tissue samples, has been approved by the Animal Welfare Committee of Zhejiang University (approval No.11834).

1.2 Plasmid construction

Two types of expression plasmids forpLEG1awere generated as follows. For pCMV-pleg1a-CFLAG,primers(forward:CTCAGTGGATCCGCCGCC ATGGCTTTCCTTCCTCCTTGG,and reverse:TCAG AAATGTTGGAATGCTGCAA), includingBamHⅠandXhoⅠsites, were used to amplifypLEG1afrom salivary gland cDNA. Then, the amplifiedpLEG1afragment was digested withBamHⅠandXhoⅠand ligated to pCMV-C-FLAG (Beyotime Biotechnology,Shanghai, China) to construct pCMV-pleg1a-CFLAG. A similar strategy was used to construct pCAG-pleg1a-3×FLAG. For the generation of pCAG-pleg1b-3×HA, thepLEG1bsequence was directly synthesized and cloned into the pCAG-3×HA plasmid.

1.3 Cell culture and plasmid transfection

HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (HyClone) in a 6-well plate until reaching 80%confluency.Then,3 μg of pMax(Lonza Group, Switzerland) and eukaryotic expression plasmids were used to transfect the cells with lipofectamine 3000 (Thermo Fisher Scientific Inc.,U.S.) according to the guidelines. Forty-eight hours later,the cells were harvested and subjected to protein extraction. The cell culture medium was concentrated by trichloroacetic acid (TCA) method[8]. Briefly, the cell culture medium was collected and centrifuged at 1 000gfor 5 min at 4 ℃. Then, the TCA solution(500 g TCA in 350 mL H2O) was added to the sample at a volume ratio of 1∶4. After incubation for 10 min on ice,the mixture was centrifuged again at 1.3×104gfor 5 min. Then, the protein pellet formed after carefully removing the supernatant. Then, 200 μL of cold acetone was added to wash the pellet three times.Finally, the dried pellet could be directly dissolved in the protein loading buffer for sodium dodecyl sulfatepolyacrylamide gel electrophoresis(SDS-PAGE).

1.4 Antibody preparation, protein extraction,and Western blotting analysis

The pCMV-pleg1a-C-FLAG plasmid was digested withBamHⅠ andXhoⅠ. ThepLEG1acontaining fragment was then purified and ligated to pET-32a to obtain the prokaryotic expression plasmid pET-32a-pleg1a. The subsequent antibody production was accomplished by Hangzhou HuaAn McAb Biotechnology Co. Ltd., which finally provided us with rabbit anti-pLEG1a polyclonal antibody and antiserum(No.1869).

Whole protein extracts from the tissues or HEK293T cells were prepared using cell lysis buffer for Western and IP (Beyotime Biotechnology,Shanghai, China) supplemented with 1 mmol/L phenylmethanesulfonyl fluoride (Beyotime Biotechnology,Shanghai, China) as a protease inhibitor. For tissue protein extraction, 200 μL of lysis buffer was added to 20 mg tissues,which were then homogenized using an automatic homogenizer (Shanghai Jingxin Company,China). After the adherent HEK293T cells were washed with phosphate buffered saline (PBS) twice,200 μL of cold lysis buffer was added to the 6-well plate with gentle agitation for 5 min before collecting the lysed solution. After centrifugation of the lysates from tissues or cells, the supernatant was evaluated using the bicinchoninic acid protein assay kit(BeyotimeBiotechnology,Shanghai,China).Approximately 30 μg of protein extracts was used for Western blotting analysis.

Pig saliva was collected from three pigs and subjected to Western blotting analysis to detect the presence of pLEG1a. Before direct loading for Western blotting analysis, approximately 5 mL of saliva was concentrated using the TCA precipitation protocol. Western blotting analysis was performed using a previously established protocol[1]. The antibodies used were rabbit anti-human LEG1(antibody No. 897, a gift from Dr. PENG Jinrong),rabbit anti-pLEG1 (No. 1869), anti-FLAG (No. M2,Sigma Aldrich, Germany), and anti-HA (No. HA7,Sigma Aldrich, Germany) at a dilution of 1∶1 000.Ponceau staining (Beyotime Biotechnology, Shanghai,China) was employed to quantify the loaded protein according to the manufacturer’s guidelines.

1.5 Glycosylation analysis

The bioinformatics tools NetNGlyc 1.0 and NetOGlyc 4.0[9]were applied for the prediction of potential glycosylated sites in pLEG1 proteins(GenBank accession Nos. XP_003121259.1, XP_020930551.1,XP_020940144.1)with default parameters.

For experimental validation ofN- orOglycosylation of pLEG1, PNGase F and endo-α-Nacetylgalactosaminase plus neuraminidase (NEB),which cleaveN-linked andO-linked glycans,respectively, were used for treating concentrated saliva and tissue protein extracts. The treated and untreated samples were analyzed by Western blotting for band shifts.

2 Results

2.1 Establishment of a polyclonal antibody for pLEG1a

In a previous report, we demonstrated thatpLEG1awas specifically expressed in salivary glands,while no expression signals were found in any tissues forpLEG1bandpLEG1cby reverse transcription-polymerase chain reaction (RT-PCR)[4].However, due to the lack of suitable antibodies recognizing the newly identified protein, the expression of pLEG1a protein has yet to be determined. To solve this problem and characterize pLEG1a, we produced a rabbit anti-pLEG1a polyclonal antibody (No. 1869) based on the pLEG1a sequence. According to the RT-PCR results, we tested the specificity of the antibody using proteins extracted from the heart, salivary glands, and HEK293T cells overexpressingpLEG1a. The results of Western blotting analysis showed that the antipLEG1a purified antibody and antiserum both readily recognized a band of approximately 38 kDa(corresponding to the molecular weight of pLEG1a)in the proteins from the salivary glands and the cells rather than the heart, in accordance with the result obtained using anti-hLEG1a serum (No. 897) (Fig.1A). This experiment indicated that the 38 kDa band is pLEG1a because it can be detected using different anti-pLEG1a antibodies. We also noted that the cells transfected with the pCMV-pleg1a-C-FLAG plasmid had a low level of pLEG1a expression,resulting in the nonspecific bands in Fig. 1A. Thus,we reconstructed the expression plasmids using the CAG promoter and 3×FLAG or 3×HA tags in our subsequent experiments. The newly constructed plasmids forpLEG1aandpLEG1bwere used to transfect HEK293T cells. Then, the extracted cells were blotted with anti-pLEG1a (No. 1869) antibody,which could detect both pLEG1a and pLEG1b proteins (Fig. 1B). To confirm the result of the antipLEG1a antibody, we employed the anti-FLAG antibody to detect the blot again, in which a protein band with a similar size was detected in thepLEG1atransfected cells (Fig. 1B). This result confirmed that the raised antibody could be used to identify the pLEG1a and pLEG1b proteins.

Fig.1 Western blotting analysis of the pLEG1 protein

2.2 Expression profile of the pLEG1a protein

To determine the expression profile of pLEG1a,we used anti-pLEG1a to detect various tissue extracts.The results showed that pLEG1a was detectable in the salivary glands (Fig. 2A), consistent with the previous RT-PCR results[4]. Unexpectedly, another bright band was found in the lung extract,wherepLEG1genes were not expressed according to RT-PCR. Both bands (≈38 kDa) detected corresponded to the predicted molecular weight of pLEG1a.These results indicated that pLEG1a existed in both the salivary gland and lung but was absent in the heart,liver,spleen,kidney,brain,intestine,and skeletal muscle. As LEG1s from other species are regarded as secretory proteins, we wanted to determine whether pLEG1a is also present in pig saliva. Western blotting was employed using concentrated saliva from three pigs, in which two bands in each saliva sample were detected (Fig. 2B). The lower band displayed a similar size to those identified in the lungs and salivary glands.In contrast,the upper band(between 45 kDa and 55 kDa) with the same size as those in Fig. 1B was distinct from the nonspecific 55 kDa band found in other tissues (Fig. 2B). This result indicated that pLEG1a could be secreted into the saliva in two forms.

Fig.2 Expression of the pLEG1a protein

2.3 N-glycosylated salivary pLEG1a

As two bands could be recovered in the pig saliva,we speculate that the upper band might be due to a glycosylation event,which is a common phenomenon in other species[3,6-7,10-11].Thus, bioinformatics tools were first utilized to predict the potential glycosylation sites of pLEG1 proteins, and the results showed that pLEG1s exhibit several potentialN-glycosylation sites andO-glycosylation sites. For example,pLEG1a was predicted to show noO-glycosylation but contain several potentialN-glycosylation sites. In contrast,pLEG1b and pLEG1c were both predicted to beN- andO-glycosylated (Fig. 3A). Subsequently,the proteins from the heart and salivary gland were treated with either PNGase F or endo-α-N-acetylgalactosaminidase plus neuraminidase (O-glycosidase/neuraminidase). However, no band shift was observed(Fig.3B),demonstrating that pLEG1a is not glycosylated in tissues. Next, the concentrated saliva was treated with PNGase F orO-glycosidase/neuraminidase. The results showed that a band shift occurred following treatment with PNGase F instead ofO-glycosidase/neuraminidase, which eliminated the upper band. Moreover, the shifted band showed a similar molecular weight to those in the salivary gland and saliva (Fig. 3C). Thus, salivary pLEG1a is aN-glycosylated protein exhibiting noO-glycosylation.In addition, the presence of the 38 kDa band in the saliva suggested that pLEG1a is not completely glycosylated,resulting in a low level of nonglycosylated protein.

2.4 Secretory protein pLEG1a

To further confirm that pLEG1a is a secretory protein, we used pCAG-pleg1a-3×FLAG and pCAG-pleg1b-3×HA totransientlytransfect HEK293T cells.Extended exposure showed that anti-FLAG and anti-HA antibodies also recognized two bands both in cell lysates and cell culture media,including a lower band with similar size to those detected in the lungs and salivary glands and an upper band corresponding to the size observed in the saliva(Fig. 3D). These results suggested that pLEG1a is a secretory protein.

Fig.3 Glycosylation test of pLEG1a

3 Discussion

In our previous study,pLEG1awas proven to be similar to human and mouse homologs evolutionarily and molecularly[4]. Additionally, at the RNA level,hLEG1a,mLeg1a, andpLEG1aall exhibit salivary gland-specific expression patterns in contrast to those in zebrafish and platypus[1,3-4,12]. These results indicate that human, mouse, and pig orthologs might preserve similar functions due to subfunctionalization, andpLEG1amight be a good model to study eutherianLEG1afunctions. In the current work, we aimed to characterize pLEG1a proteins and determine whether pLEG1a is similar to other LEG1s at the protein level. First, to determine the expression profile of the pLEG1a protein, we prepared an anti-pLEG1a (No.1869) antibody based on the pLEG1a sequence.Then, we identified the antibody validity by detecting pLEG1a in tissues and cells. Using the antibody, we found that pLEG1a was observed in both the salivary gland and lung tissues. However, there was nopLEG1transcription in the lung tissues according to our previous work[4]. As several proteomic studies have shown that LEG1 proteins are secretory proteins that are detected in milk, saliva, and seminal plasma[3,6,10-11,13], we assumed that pLEG1a might be enriched in the respiratory mucus and show a low level of expression in the mucosal epithelial cells.Therefore, future studies using more tissues, such as milk,mucus,and seminal plasma,will be needed.

Glycosylation is an important biological process that plays a critical role in protein stability,trafficking, cell growth, and host-pathogen interactions[14]. It has been reported that glycosylation of zLEG1 is critical in the liver development of zebrafish and that glycosylated MLP can confer antibacterial activity to pups[2-3]. In addition,Nglycosylated hLEG1a was detected in human saliva[7,11]. Thus, glycosylation might be another hallmark of LEG1 proteins. In our experiment,Nglycosylated pLEG1a was observed in concentrated saliva but not in tissue-derived proteins. Furthermore,bothN-glycosylated and nonglycosylated bands could be detected in the saliva of pigs, with glycosylated pLEG1 as the predominant form.NandO-glycosylated sites were bioinformatically predicted in pLEG1 proteins; however, onlyNglycosylation events could be experimentally validated, in accordance with previous experiments conducted on human, mouse, zebrafish, and platypus LEG1s[2-3,7,11-12]. Therefore, the glycosylation pattern might indicate important biological functions of the pLEG1a protein.

Ectopic overexpression of thepLEG1aandpLEG1bgenes in HEK293T cells also confirmed that pLEG1a and pLEG1b are both secretory proteins.Interestingly, two bands with molecular weights correspondingtotheglycosylatedand nonglycosylated pLEG1s were detected in the experiment, although the possible nonglycosylated bands had lower intensities, consistent with the results from saliva. Nonglycosylated secretory proteins, such as transthyretin and growth hormone[15-18], have been reported before, indicating an alternative protein secretion pathway.Additionally,a study that mutated all potentialN-glycosylated sites of carnosinase suggested that a trace amount of the nonglycosylated protein was still secreted.A previous report on zLEG1 protein also observed that when mutating N70A in zLEG1, the nonglycosylated protein could be secreted.These results indicated that either an alternative secretory pathway exists or that these proteins have altered enzymatic activity[19].Based on these studies, the pLEG1a protein is similar at the protein level to other LEG1 proteins, especially eutherian LEG1s, in terms of the expression pattern,glycosylation, and secretion ability. Therefore,pLEG1a might be a good model to study the function of the newly identified gene.