In the history of humans, the Coronavirus disease (COVID-19) caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has made an indelible mark characterized by a contagious respiratory pandemic ^[1][2][1,2]. Consequently, 660 million cases and approximately 6.6 million deaths have been reported as of December 2022 [3]. Since its emergence, studies have been focusing on developing preventive and therapeutic measures against this disease, and several vaccines have been licensed ^[4][5][4,5]. However, the constant mutation of the SARS-CoV-2 virus, which results in different variants, narrows the effectiveness of the licensed vaccines against SARS-CoV-2 infections, thus suggesting a needed continuous effort in developing a SARS-CoV-2 universal vaccine [6]. Besides the pandemic caused by SARS-CoV-2, other human coronaviruses (huCoVs) including human CoV (HCoV)-229E (1962), HCoV-OC43 (1967), SARS-CoV (2002), HCoV-NL63 (2004), HCoV-HKUI (2005), Middle East respiratory syndrome coronavirus (MERS)-CoV (2012) and SARS-CoV-2 (2019) have been implicated in different outbreaks since the start of the 21st century ^{[7][8][9][10][11]}[7,8,9,10,11]. Although the trend of the emergence of coronaviruses remains unclear, adequate preparation must be made for a possible emerging or re-emerging strain. In this case, developing an effective universal vaccine against CoVs should take advantage of the conserved portions of the virus, especially the spike protein (SP).

Levels of similarities have been observed among the emerged CoVs ^[12][13][14][12,13,14]. For instance, the SARS-CoV-2 viral genome sequence analysis revealed its phylogenetic similarity with SARS-CoV (79%) and MERS-CoV (50%) ^[15][16][17][15,16,17]. Like other betacoronaviruses, the SARS-CoV-2 genome consists of 27 proteins encoded by 14 open reading frames (ORF), including non-structural proteins (nsp) that are encoded by the ORF 1 and 2 present at the 5′-terminal regions ^[18][19][18,19]. More importantly, the SARS-CoV-2 structural proteins, notably the SP, envelope protein (E), membrane protein (M), nucleocapsid (N) and eight accessory proteins encoded by the 3′-terminal region of the genomes, are in like manner with other betacoronaviruses [19].

Sequence analysis of the SARS-CoV-2 SP revealed a close association with SARS-CoV regarding amino acid composition as well as comparable binding affinity to human angiotensin-converting enzyme 2 (hACE2) ^[20][21][20,21]. The SP, which is a highly N-glycosylated class I transmembrane fusion protein, plays a critical role during coronavirus infection. The SP mediates the attachment of the virus into the cell receptor and facilitates the viral-host membrane fusion ^[22][23][22,23]. The SP also assembles into trimers on the virus surface and cleaves into two subunits, S1 and S2, during the cell infection ^[24][25][24,25]. The large protein S1 domain contains the RBD that is responsible for binding to the host cell receptor and varies extensively in all isolates of the CoVs ^[20][24][25][20,24,25]. Meanwhile, the S2 domain, which facilitates the virus-cell fusion, is made up of the fusion peptide (FP) and heptad repeat regions (HR1 and HR2), which are conserved among the isolates of the CoVs ^[26][27][26,27], the conserved membrane-proximal external region (MPER) and the transmembrane domain (TM) ^{[28][29][30][31][32][33][34]}[28,29,30,31,32,33,34]. After the binding of the S1 RBD domain to the ACE2, the S2 subunit then inserts its FP into the cell membrane leading to the assemblage of the HR1 and HR2 into a six-helix structure to drive the cellular and viral membrane closely together for viral entry (Figure 1) [35]. Notably, the major determinant of cell tropism in most coronaviruses is typically linked to the structure of the SP ^{[21][25][36][37][38][39][40]}[21,25,36,37,38,39,40]. Phylogenetic, bioinformatic and homology structural modelling analyses showed that the RBD of SARS-CoV-2 only has 64% shared identity with the SARS-CoV, while the NTD has 51% similarity [41]. However, the study revealed that within the S2, the fusion protein (FP) is 93% identical and the HR1 is 88% identical, while the HR2 and the TM are respectively 100% and 93% similar.

SARS-CoV-2 and other SARS-related coronaviruses (e.g., SARS-CoV) can utilize distinct domains within the S1 subunit to identify different attachment and entry receptors in the host cell surface ^[20][40][20,40]. For instance, the differences in the SPs of the known huCoVs contribute to their difference in pathogenesis and the site of infection (lower or upper respiratory) [36]. Taking a closer look at SARS-CoV-2 as an example, Laporte et al. described that the higher transmissibility experienced with the SARS-CoV-2 compared to other human coronaviruses is due to its abundant replication in the upper respiratory tracts [43]. They mentioned that the SARS-CoV-2 SP has an intrinsic temperature preference of 33 °C, like the temperature of human respiratory tracts, instead of the 37 °C required by other human CoVs. In addition to this, it was revealed that the SARS-CoV-2 has multiple cell entry activators, including TMPRSS2 and TMPRSS13 protease, broadening its tropism. As mentioned by Korber et al., a D614G mutation observed in the SARS-CoV-2 variant S1 also resulted in the wider spread of the virus at different geographical regions with higher viral loads in the respiratory tracts [44]. These differences observed in the SP of the CoVs contribute to the difficulties in developing a sustainable vaccine against the emerging variants of SARS-CoV-2. It is therefore a necessity to develop a vaccine that can prevent the spread of all present or emerging variants of SARS-CoV-2.

Developing a vaccine for viruses with multiple strains or variants tends to take advantage of the conserved epitopes present in the virus. Conserved epitopes are epitopes that are relatively the same among different strains of a pathogen. For example, due to the multiple strains of influenza over the years, the means of developing a universal vaccine has been the use of the conserved epitopes on the hemagglutinin (HA stalk) or the matrix ectodomain (M2e) ^{[45][46][47][48][49][50][51]}[98,99,100,101,102,103,104]. The strategy of using conserved epitopes has also been used in the development of preventative vaccines for HIV, dengue virus, Lassa fever virus (LASV), hepatitis virus and Kaposi’s sarcoma-associated herpesvirus (KSHV) ^{[52][53][54][55][56][57][58][59][60]}[105,106,107,108,109,110,111,112,113]. Therefore, identifying the conserved regions in the SP of SARS-CoV variants will be of great importance in developing a universal vaccine against CoVs. Although mutations in SARS-CoV-2 had a great impact on the SP, studies have revealed that certain conserved epitopes in S1 can induce neutralizing antibodies ^[61][114]. Interestingly, some monoclonal antibodies (including 7B11, 18F3, S309 and its Fab, S315, 154C, S304, 240C and VHH-72) that recognize SARS-CoV and MERS-CoV could cross-react and cross-neutralize SARS-CoV-2 by recognizing the ACE2 binding sites on the SARS-CoV-2 RBD ^[62][115].

Jaiswal et al. mathematically (in-house developed PERL scripts) revealed sets of epitopes on the S1 subunit around 453 to 538 that interact with the ACE and are 99% conserved among the variants of SARS-CoV-2 ^[63][116]. Their study further identified conserved common neutralizing epitopes on the SARS-CoV-2, including YLTPGDSSSGWTAGAAAYYV (247–267 aa), TFKCYGVSPTKLNDL (376–390 aa) on S1 and LNEVAKNLNESLIDLQELGK (1186–1205 aa) on the S2 ^[63][116]. A recent immunoinformatic study predicted conserved and highly immunogenic CTL-induced epitopes on S1 VRFPNITNL (327–335 aa) and PYRVVVLSF (507–515 aa) (Table 1), while the CTL-induced epitopes on S2 were identified to be VVFLHVTYV (1060–1068 aa) and GVVFLHVTY (1059–1067 aa) (Table 2) ^[64][117]. Based on the conservancy, antigenicity, allergenicity, population coverage and transmembrane location, another study chose potential conserved epitopes from the S1 (FNATRFASVYAWNRK, 342–356 aa) (Table 1), S2 (FLHVTYVPAQEKNFT, 1062–1072 aa) (Table 2) and the E/M protein to construct a SARS-CoV-2 vaccine and revealed that it had high immunogenicity and broad neutralizing activity against the SARS-CoV-2 RBD (Figure 2) ^[65][118]. Jiang et al., with web-based analytic tools, also predicted potential T cell epitopes induced by the SARS-CoV-2 SP and narrowed them down to CD4 or CD8 T cell epitopes using ELIspot and a cytolytic assay ^[66][119]. In their observation, YYVGYLQPRTFLLKY (264–278 aa), located at the end of the NTD and upstream of the RBD, is highly conserved among 11 variants of SARS-CoV-2 VOCs and variants of interest (VOIs) and could induce the T cells. Further studies also suggested that these epitopes could be well recognized by most HLA alleles globally ^[66][119].

Considering the role played by the S2 subunit during SARS-CoV-2 infection, it could also be targeted to induce immune responses against SARS-CoV-2. Interestingly, Ladner et al. generated an epitope-resolved analysis of IgG cross-reactivity among all CoVs in COVID-19-negative and recovered patients using a highly multiplexed peptide assay (PepSeq) and discovered that the epitopes at the FP, which is 93% similar among strains of betacoronaviruses and alphacoronaviruses, produced broadly neutralizing antibodies against the endemic coronaviruses, including SARS-CoV, MERS-CoV and SARS-CoV-2 ^[41][72][41,125]. Likewise, the HR2 region of the S2 is 100% conserved among the variants of the SARS-CoV-2 ^[41][75][41,128]. The S2 subunit proteins could also induce cross-reactive antibodies against the SARS-CoV SP and the endemic CoVs ^[77][130]. The S2 has been reported to induce neutralizing antibodies or T cell responses targeting the FP proximal region and HR2 domain of S2 in COVID-19 patients or animals vaccinated or infected with different CoVs ^{[78][79][80][81]}[131,132,133,134]. Interestingly, due to prior population exposure to common cold coronaviruses, nAbs and memory B and T cells against SARS-CoV-2 were found in some individuals who have never been infected by SARS-CoV-2 ^[82][83][135,136]. Pre-existing antibodies against conserved epitopes of S2, such as residues 901–906, 810–816, 851–856, 1040–1044 and 1205–1212, showed the greatest cross-reactivity and hindered SARS-CoV-2 entry into cells ^[82][135]. Other identified regions that are most widely recognized among SARS-CoV-2 linear epitopes in convalescent donors are EELDKYF (1150–1156 aa) within the stem helix of the HR2 terminal and EDLLFN (819–824 aa), which overlaps the FP and is adjacent to the S2 cleavage ^[72][73][125,126]. Moreover, a study conducted by Song et al. revealed that monoclonal antibody CC9.3 isolated from individuals before SARS-CoV-2 infection was characterized to recognize the S2 subunit of the SARS-CoV-2 and other huCoVs ^[84][137].

From the sera of patients recovered from SARS-CoV-2, Pinto et al. isolated five monoclonal antibodies that could recognize the stem-helix (SH) of the S2 subunits of other betacoronaviruses including the OC43 strain ^[73][126]. In addition, Lu et al. identified and crystallized T cell follicle helper cells (cTfh) among patients who recovered from the mild symptoms of COVID-19 and revealed that these cTfh could recognize SARS-CoV-2 S2 subunit epitopes (864–882 aa) that are conserved among the emerging variants ^[85][57]. In a recent publication, Wu et al. identified a monoclonal antibody hMab5.17 that could recognize the SARS-CoV-2 HR2 domain that is adjacent to TM (SPDVDLGDISGINAS; 1161–1175 aa) and could protect against SARS-CoV-2 in the Syrian hamster. They further cloned the mAb hMab5.17 and demonstrated that it could neutralize SARS-CoV-2 variants ^[75][128].

Another highly conserved region located at the S2 region is the SARS-CoV-2 MPER-like region (MPER) (GKYEQYIK; 1204–1211 aa), which has great potential for being used as an antigen for broadly neutralizing antibodies (bnAbs) ^[34][86][87][34,76,138]. Yu et al. have demonstrated that a cholesterol-conjugated lipopeptide containing SARS-CoV-2 MPER prepared by using cholesteryl succinate monoester could inhibit viral entry, indicating the importance of MPER in viral entry and fusion [34]. Like the MPER of HIV-1, the MPER of SARS-CoV-2 could possibly induce bnAbs against the variants of SARS-CoV-2. Studies have shown that bnAbs 4E10, 2F5, 10E8 and LN01 could interact with the MPER of the HIV-1 gp41 to prevent infection ^[88][89][90][139,140,141]. Likewise, the SARS-CoV-2 MPER could be a suitable immunogen for inducing neutralizing antibodies ^[34][87][34,138].

The above findings suggested that the S2 subunit of the SARS-CoV-2 is conserved among the previous strains of human coronaviruses and the SARS-CoV-2 variants and can induce cross-reactive antibodies.