Antigenicity Analysis of Different Regions of the Severe Acute Respiratory Syndrome Coronavirus Nucleocapsid Protein

sequence analysis and primer design

The sequence of the gene encoding N protein from the SARS-CoV (strain BJ01) was retrieved from GenBank (accession no. {"type":"entrez-nucleotide","attrs":{"text":"AY278488","term_id":"30275666","term_text":"AY278488"}}AY278488) and analyzed for putative epitopes with the DNASTAR software package (DNASTAR, Inc.; http://www.dnastar.com/). On the basis of the predicted epitope locations, 33 primers were designed and synthesized to amplify the fragments spanning these epitopes (Table 1 ). The designed primers were synthesized by SUNBIO Tech Company.

cloning, production, and purification of full-length n protein and fragments

The SARS-CoV (BJ01) was cultured in Vero E6 cells at 37 °C for 48 h; the medium from the infected cells was then collected for RNA extraction by the QIAamp viral RNA Mini Kit (QIAGEN, Inc.). From the RNA template, single-stranded cDNA was synthesized with use of the primer NO-R (Table 1 ) and the SuperScript cDNA system (Invitrogen). The full-length N gene was amplified from the cDNA templates with the primer pair NO-F and NO-R. The PCR products were purified with Montage PCR reagent sets (Millipore Corporation), digested with BamHI and SalI (Promega Corporation), and ligated into similarly digested plasmid pET-32a (Novagen). The ligation products were transformed in BL21 (DE3; Novagen Inc.), and the recombinant clones were identified by PCR and confirmed by sequencing. Various fragments spanning different putative epitopes (Table 2 ) were amplified from the recombinant clone with the full-length N protein gene by different primer combinations and cloned into pET-32a as described above. To produce the target proteins, the recombinant clones were cultured at 37 °C to A₆₀₀ = 0.6 and then induced with isopropyl β-d-thiogalactoside (Sigma). The proteins were then purified by use of Ni-NTA agarose (Qiagen) according to the manufacturer’s instructions.

immunoblotting analysis of the proteins

For immunoblotting analysis, the proteins were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membranes were blocked in Tris-buffered saline-Triton (150 mmol/L NaCl; 50 mmol/L bis-Tris, pH 7.5; 0.5 mL/L Triton X-100) containing nonfat milk (10 mL/L) for 2 h at room temperature and then incubated with pooled patient sera (1:200 dilution) for 2 h. After three washes in Tris-buffered saline-Triton, the membranes were incubated with horseradish peroxidase-conjugated goat anti-human IgG (Santa Cruz Biotechnology) for 2 h and then washed with Tris-buffered saline-Triton three more times. The membranes were developed with diaminobenzidine substrate for ∼5 min, and the reaction was stopped by the addition of water.

protein microarray analysis

The purified proteins were dissolved at different concentrations (400, 100, 25.5, and 0.5 mg/L) in phosphate-buffered saline (PBS; pH 7.5) containing 400 mL/L glycerol and printed in triplicate on silylated glass slides (CEL Associates, Inc.) by use of the GSI Flexys arrayer (Genomic Solutions). Goat anti-human IgG and human IgG were printed as positive controls, and cell lysates of E. coli BL21 (transformed with pET-32a) and PBS (containing 400 mL/L glycerol) were printed as negative controls (see Fig. S1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol50/issue6/). Printed slides were stored at 4 °C and used within 2 weeks after being printed. To eliminate antibodies against E. coli in the sera, all of the sera were incubated with lysates of E. coli BL21 carrying pET-32a before microarray analysis. For microarray probing, the slides were first blocked with 30 mL/L nonfat milk in PBS for 1 h and then incubated with diluted sera (1:200) for another 1 h at room temperature. The incubated slides were washed three times with PBS containing 1 mL/L Tween-20 and once with PBS and then incubated for another 2 h with the fluorescently labeled secondary antibody (goat anti-human IgG) generated by use of a Cy5 antibody labeling reagent set (Amersham Biosciences). The resulting slides were washed with PBS-Tween twice and PBS once, followed by a 2-min ethanol rinse. After being dried with hot air, the slides were scanned with the GenePix Personal 4100A Scanner (Axon Instruments). Each analysis was repeated at least once.

The scanned images were processed, and the generated data were further analyzed by the GenePix Pro 4.0 software (Axon Instruments) and Microsoft Excel software. The fluorescence signal of each spot was calculated as the median fluorescence intensity (FI) minus the median local background intensity. Spots with bad signals were rejected, and those with signal values <1 were set as 1 to reflect a spot signal close to 0 (see Table S1 in the online Data Supplement). The mean human IgG signal from all of the slides was used to calibrate the signals of other spots. The spot signals for each protein fragment in the replicated hybridizations were then averaged and ready for further analysis. The normalized datasets were logarithmically transformed (base 2) and clustered by CLUSTER (16) and viewed by TREEVIEW (http://genome-www4.stanford.edu/MicroArrayy/SMDyrestech.html). To determine the positive serum number of each protein fragment, the cutoff value of each fragment was calculated as the mean value plus 3 SD of the FI from the negative controls.

For further analysis of the epitope locations and antigenicities of different regions, pooled sera (10 of the 52 tested sera) were mixed with excessive amount of specific fragments to eliminate the corresponding antibodies and then probed with protein microarrays.

immunoblotting analysis of the proteins

To locate the linear epitopes in N protein and its fragments, the proteins produced were analyzed with pooled sera (15 of the 52). The results (data not shown) indicated that most of the proteins could react with pooled sera. Fragments that did not react with pooled sera were N114, N214, N215, N315, N416, N822, N922, N1022, F14, F15, N12, and N5, indicating that there were no linear epitopes on these fragments. When equal quantities of proteins were analyzed by immunoblotting, fragments N4, N8, N824, N11, and N13 appeared to have higher reactivities, and fragments N417, N517, N9, and N10 appeared to have lower reactivities, as indicated by the amount of gray that developed on the immunoblot.

correlation between captured IgG and fi

To explore the correlation between the FI and the IgG quantities captured by the printed antigens, human IgG in increasing concentrations was printed on slides to act as an internal standard for calibration. As the results showed, there was a good correlation (r² = 0.9987) between FI and IgG concentration (Fig. 2A ), indicating that the FI can reflect the amount of IgG captured by the printed antigens. Because each pin was estimated to transfer a 40-pL sample, the calibration curve of FI to the amount of IgG could be calculated (Fig. 2B ). This curve could be used to estimate the amounts of antibodies captured by antigens on the microarray. In general, fragments with higher reactivities captured more antibodies. For example, fragment N4 captured ∼9.65 pg of IgG, three times the amount captured by F17 (see Fig. S2 in the online Data Supplement). The correlation between the FI and printed antigen concentration was lower than that of the printed human IgG (data not shown). However, the correlation was good for certain fragments that could react with most (>80%) of the tested sera (e.g., r² = 0.9823 for N4).

Open in a separate window

Figure 2.

Correlation between FI and captured IgG.

(A), correlation between FI and concentration of printed IgG. Human IgG was printed in increasing concentrations and probed with Cy5-labeled goat anti-human IgG. The FI was plotted as a function of the concentration of printed human IgG to generate a dose–response curve, and a good correlation coefficient (close to 1) was obtained. (B), correlation between FI and amount of printed human IgG.

validation of protein microarray analysis

Fragments with different reactivities in microarray assays were selected and analyzed by fragment-specific ELISAs. The absorbance values in ELISA were used for regression analysis against the FI in microarray analysis. The results showed that there was good correlation between IgG absorbance and FI, particularly for fragments that could react with most (>80%) of the sera (data not shown). Fragments with poor reactivity in microarray analysis had a lower correlation coefficient (e.g., r² = 0.8583 for N1), whereas for those that could react with 100% (52 of 52) of the sera, the correlation was high (e.g., r² = 0.9993 for N4).

profile of the antigenicity of n protein fragments

The normalized fluorescence values were logarithmically transformed (base 2) and clustered by CLUSTER and graphed by TREEVIEW (Fig. 3 ). Most of the fragments were clustered into four main groups, among which the fragments in group A had the highest reactivity (>80%), fragments in group B reacted with 40–60% of the tested sera, fragments in group C had lower reactivity (0–40%), and group D included only F17, which also reacted with most (86%) of the sera. N13 (amino acids 221–422) and N2 (1–218) were approximately the same length but reacted with 100% and 86% of the sera, respectively (Fig. 1 ), inferring that there were important antigen determinants at both termini of the N protein but that the antigenicity at the COOH terminus was higher than that at the NH₂ terminus. In addition, as predicted by DNASTAR software, there were more putative epitopes located at the COOH terminus of the N protein (17).

Open in a separate window

Figure 3.

Reactivities of all of fragments printed at the concentration of 400 mg/L.

The fluorescence values were normalized and logarithmically transformed (base 2); the cutoff value 8 was subtracted from each result. The values were then clustered by CLUSTER software and viewed with TREEVIEW software. Black squares represent positive values, white squares indicate negative values, and gray squares indicate intermediate degrees of FI.

regions spanning amino acids 1–50 and 218–311 are not important for the exposure of epitopes

Fragments N7 (amino acids 51–311) and N3 (1–311) reacted with the same percentage of serum percentile (88%), indicating that the 50 amino acids at the NH₂ terminus of N3 were not important for the exposure of epitopes. The results for N6 (amino acids 51–218) and N8 (51–422), which were formed by the deletion of these 50 amino acids from N2 (amino acids 1–218) and N4 (1–422), respectively (Fig. 1 ), also supported this conclusion. Despite a longer peptide, N3 reacted with the same number of sera as N2, suggesting that the region spanning amino acids 218–311 was not important for antigenicity.

regions spanning amino acids 50–70 and 138–208 have important epitopes

Fragment F17 (amino acids 1–208) had positive reactions with 86% (45 of 52) of the tested sera, but fragment N1 (1–134) reacted with only 8% (4 of 52) of the tested sera, indicating that the region spanning amino acids 135–208 contains important antigenic determinants. Fragment N417 (amino acids 72–208) reacted with 62% (32 if 52) of the sera, which was lower than the reactivity of fragment F17, demonstrating that the first 70 amino acids might include epitopes that increase the reactivity of F17. Because the first 50 amino acids (amino acids 1–50) were not important for antigenicity, the epitopes were likely located at the region of amino acids 50–70.

regions spanning amino acids 249–273 and 349–422 contain important epitopes

Fragments N822 (amino acids 249–349), N823 (249–382), and N824 (249–395) reacted with 0 (0%), 47 (90%), and 50 (96%) of the 52 sera, inferring that there are antigen determinants in the region of amino acids 349–395. Fragment N923 (amino acids 273–382), which was formed by deletion of 25 amino acids (249–273) from the NH₂ terminus of N823, did not react with any of the tested sera, suggesting that this 25-amino acid region is very important. N14 (amino acids 317–422), which was formed by deletion of 26 amino acids at the NH₂ terminus and an elongation of 28 amino acids (396–422) at the COOH terminus of N1024, reacted with 25 (48%) of the sera, which further confirmed the important role of this 28-amino acid region in the antigenicity of N14.

four important regions identified and their relationships

From the analysis above, four regions containing important epitopes were identified in the N protein in this study: EP1 (amino acids 51–71), EP2 (134–208), EP3 (249–273), and EP4 (349–422). Fragments that contained only EP1 had lower or no reactivities; for example, F14 and N1 reacted with only 0% and 8% of the tested sera, respectively (Fig. 1 ). Fragments that contained only EP2 (i.e., N9 and N517) reacted with ∼40% of the sera. Fragment N2, which spanned both EP1 and EP2, could react with most of the sera (46 of 52; 88%), which indicated a cooperative relationship between EP1 and EP2. Fragment N9 (amino acids 133–218) was reactive in both the immunoblotting and microarray analyses, indicating that EP2 contains both conformational and linear epitopes.

Fragments that contained only EP4 reacted with approximately one-half of the tested sera, and those that contained only EP3 had no reactivity. Fragments that spanned both of them, however, reacted with 100% (52 of 52) of the tested sera, suggesting that there might be some relationship between EP3 and EP4 for the formation of conformational epitopes. Fragments N922, N923, and N924, which contained only EP4, had no reactivity in protein chip analysis; however, N923 and N924 could react with pooled sera in immunoblotting analysis, indicating that EP4 contains only linear epitopes. Fragments N822, N12, and N922 had no reactivity in both analyses, implying that EP3 could not form epitopes alone.

We thank H. Zhu for helpful discussions and critical review of the manuscript and Q. Guo for proofreading the manuscript. This work was supported by grants from the National 863 Program (NO.2003AA208211) and 973 Program (NO.2003CB514106).