Long-term persistence of robust antibody and cytotoxic T cell responses in recovered patients infected with SARS coronavirus

Abstract

Most of the individuals infected with SARS coronavirus (SARS-CoV) spontaneously recovered without clinical intervention. However, the immunological correlates associated with patients' recovery are currently unknown. In this report, we have sequentially monitored 30 recovered patients over a two-year period to characterize temporal changes in SARS-CoV-specific antibody responses as well as cytotoxic T cell (CTL) responses. We have found persistence of robust antibody and CTL responses in all of the study subjects throughout the study period, with a moderate decline one year after the onset of symptoms. We have also identified two potential major CTL epitopes in N proteins based on ELISPOT analysis of pooled peptides. However, despite the potent immune responses and clinical recovery, peripheral lymphocyte counts in the recovered patients have not yet been restored to normal levels. In summary, our study has, for the first time, characterized the temporal and dynamic changes of humoral and CTL responses in the natural history of SARS-recovered individuals, and strongly supports the notion that high and sustainable levels of immune responses correlate strongly with the disease outcome. Our findings have direct implications for future design and development of effective therapeutic agents and vaccines against SARS-CoV infection.

Figure 1. Sequential analysis of peripheral changes in total lymphocytes (LY), CD3, CD4, CD8, B cell, and NK cell counts in recovered patients over 2-year period after onset of symptom.

Grey stars indicate the significant differences between the recovered patients at 24 months after onset of symptom and normal healthy controls. Their respective p values are presented.

Figure 2. Sequential analysis of antibody responses against the whole viral lysates (A), N protein (B), and pseudotyped virus with S protein (C) of SARS-CoV in 30 recovered patients over 2-year period.

Serum samples were diluted 100- and 900-fold prior to ELISA and neutralization studies. Both dilutions were used to carried out experiments presented in (B) and (C), where the only 100-fold diluted serum were used for experiments presented in (A). The top and bottom of each rectangular box denote the 75^th and 25^th percentiles, respectively, with the median shown inside the box. Horizontal bars extending from each box represent the 90^th and 10^th percentiles. Significant differences between samples are indicated by grey stars with their respective p values. Otherwise no significance was found.

Figure 3. Sequential analysis of CTL responses against a pool of peptides derived from N protein in 30 recovered patients over 2-year period.

(A) The average number of spot forming cells (SFC) per million of PBMC from recovered patients at month 3, 12, and 18 post onset of symptom. No significant differences were found among these samples. The top and bottom of each rectangular box denote the 75^th and 25^th percentiles, respectively, with the median shown inside the box. Horizontal bars extending from each box represent the 90^th and 10^th percentiles. (B) The average number of SFC per million of PBMC against various pools of N protein peptides (open rectangle) and percent of PBMC samples recognizing various pools of N protein peptides (closed rectangle). An arbitrary line was draw at 3000 SFC per million PBMC to identify peptide pools that are preferentially recognized by the recovered patients in this cohort.

Figure 4. The actual composition and cross-broad layout of each peptide pools derived from the N protein (upper panel).

The 57 peptides (one is not listed) were pooled into 15 groups, 7 of which (NX1-7) are listed vertically and 8 are (NY1-8) listed horizontally. The actual amino acid residue sequences preferentially recognized by the recovered patients are highlighted (lower panel), which correspond to residue sequences between position number 211 to 235 and 330 to 354.