A Promising Vaccination Strategy against COVID-19 on the Horizon: Heterologous Immunization

To overcome the ongoing COVID-19 pandemic, vaccination campaigns are the highest priority of majority of countries. Limited supply and worldwide disproportionate availability issues for the approved vaccines, together with concerns about rare side-effects have recently initiated the switch to heterologous vaccination, commonly known as mixing of vaccines. The COVID-19 vaccines are highly effective in the general population. However, none of the vaccines is 100% efficacious or effective, with variants posing more challenges, resulting in breakthrough cases. This review summarizes the current knowledge of immune responses to variants of concern (VOC) and breakthrough infections. Furthermore, we discuss the scope of heterologous vaccination and future strategies to tackle the COVID-19 pandemic, including fractionation of vaccine doses and alternative route of vaccination.


Fig.1. Proposed mechanism of action of COVID-19 vaccines. (A)
In mRNA-vaccines, the spike mRNA is modified in which uridine is replaced by pseudouridine in order to escape immune responses. Moreover, the mRNA is stabilized in its prefusion conformation by two consecutive proline substitutions at amino acid positions 986 and 987, at the top of the central helix in the S2 subunit. mRNA is loaded in the lipid nanoparticles, which interact with the cell membrane and release modified mRNA in the cytoplasm of the muscle cell or antigen-presenting cell (1) (2) (3). mRNA is translated into spike protein in the cytoplasm, later presented by MHC class I to CD8 + T cells. The spike antigens are also released in the extracellular environment where they migrate to the draining lymph nodes and are endocytosed by APCs within the germinal centers. APCs at the site of injection may also be involved. These endocytosed antigens are processed by the presented by MHC class II to CD4 + T cells. Activated CD4 + T cells help to activate CD8 + T cells and B cells. CD8 + T cells kill the infected cells (not investigated in the context of SARS-CoV-2). With the help of CD4 + T cells, B cells mature to plasma secreting cells and synthesize antibodies to combat SARS-CoV-2. (B) In viral vectored vaccines, the full-length spike gene (DNA) is inserted in a harmless adenovirus vector (rAdenovirus) (1). rAdenovirus latches to the host cell and releases DNA in the cytoplasm (2) (3), which later migrates to the nucleus of the cell and is transcribed to mRNA (4). (C) Inactivated pathogen vaccines are chemically inactivated by βpropiolactone. The active component, along with the alum, generates immune responses. APCs process the antigens by MHC I machinery and present antigens to CD4 + T cells. (D) Full-length stabilized spike gene is engineered into baculovirus (rbaculovirus). rbaculovirus delivers spike gene into the Sf9 (1) (2), migrates to the nucleus of the cell, and is transcribed to mRNA. mRNA further migrates to the cytoplasm and is translated to spike protein (3). Translated and glycosylated spike protein is eventually purified and mixed with adjuvant (5). The dotted line represents that the exact role of CD8 + T cells is not known. MHC = major histocompatibility complex. mRNA = messenger RNA. r = recombinant. Sf9 cell = insect cell line, a clonal isolate derived from the parental Spodoptera frugiperda cell line IPLB-Sf-21-AE. a role in combating SARS-CoV-2 [15]. SARS-CoV-2-specific CD4 + T cells commonly differentiate into Th1 and Tfh T cells [16][17][18]. Th1 cells have antiviral properties, and Tfh cells are specialized in providing help to B cells and are critical for the development of neutralizing antibodies, memory B cells, and long-term humoral memory. Mostly Th1 skewed responses with little to none Th2 cytokines were detected in mRNA [19] and adenovirus vector-based vaccines [20] while Tfh cells have also been detected in vaccinated individuals [19,21,22]. In addition, CD8 + T cells can directly kill infected cells, which are also induced after SARS-CoV-2 infection or vaccination [16,21]. The presence of virus-specific CD8 + T cells has been associated with better COVID-19 outcomes. SARS-COV-2-specific CD8 + T cells possess effector molecules, including IFN-γ, granzyme B, perforin, and CD107a [23][24][25]. In the case of inactivated COVID-19 vaccine, Th1 and Th2 T cell subsets were not defined although cellular immune responses in vaccinated individuals exhibited antigen-specific CD4 + and CD8 + T cells [26]. However, the role of T cells, especially CD8 + T cells, against SARS-CoV-2 remains to be elucidated. Although natural infection with SARS-CoV-2 and different vaccines induce more or less protective immunity, the ability of such immune responses to recognize and provide protection against variants of SARS-CoV-2 is a matter of concern.

Antibody Responses
Planas et al. [27], evaluated the neutralizing potential of serum from BNT162b2, or ChAdOx1 nCoV-19 vaccinated individuals against D614G, B.1.1.7, B.1.351, and B.1.617.2 strains. After a single dose (post-3-weeks) of BNT162b2 vaccine, the levels of neutralizing antibodies were low against D614G and almost undetectable against the Alpha, Beta, and Delta variants. When evaluated 5-weeks after the booster, antibody titers significantly increased. However, in contrast to Alpha, 3-and 16-fold reductions in the neutralization titers against the Delta and the Beta variants, respectively, were observed. A similar pattern was observed with the ChAdOx1 nCoV-19. A single dose induced low levels of antibodies neutralizing the Delta and Beta variants (post-10-weeks) compared to the D614G and Alpha. Four weeks after the second dose, neutralizing titers were strongly increased. However, relative to the Alpha, 5-and 9-fold reductions in neutralization titers against the Delta and the Beta variants were observed. Further studies reported that BNT162b2 vaccinated individuals displayed 3.3-, 7.6-, 2.6-, and 2.5-fold reductions in the neutralization titer against Alpha, Beta, Gamma, and Delta variants, respectively, in contrast to Victoria strain [28][29][30][31]. Moreover, ChAdOx1 nCoV-19 vaccinated individuals showed a 2.33-, 9-, 2.9-, and 4.29fold loss in neutralization titer against Alpha, Beta, Gamma, and Delta variants, respectively, compared with Victoria [28][29][30][31]. In the following sections vaccination refers to full vaccination as recommended by respective manufacturers unless otherwise mentioned.
Although the VOC more or less escape neutralization by antibodies and there are reports of infection by variants in the vaccinated population, the vaccines effectively reduce the severity of the disease (Table 1). Currently, the prevention of severe disease and deaths is of utmost importance. However, the resilience of immune responses elicited by COVID-19 vaccines, especially against VOC, remains to be elucidated. Although, an 8month study in which 20 participants received the Ad26.COV2.S vaccine in 1 or 2 doses (either 5×10 10 viral  Table 2.

T Cell Responses
Currently, most of the vaccines contain spike [19,48,49], and mutations have been widely reported in the spike. Antibodies induced by spike of the prototypic strain of SARS-CoV-2 have less binding and neutralization abilities for newly emerging variants resulting in escape from the antibody responses. SARS-CoV-2 antibody responses have received a lot of attention. However, the arsenals of humoral and T cell responses may play diverse roles in different viral infections. In addition, T cells induced by vaccines are supposed to recognize SARS-CoV-2 variants [50,51]. For example, a study evaluated WT and variants of SARS-CoV-2 specific CD4 + and CD8 + T cell responses in BNT162b2 (n = 8) or mRNA-1273 (n = 11) vaccinees (samples were collected 2-4 weeks after the second dose of vaccination) [50]. Peptide mega pools (MPs) spanning the entire SARS-CoV-2 proteins or only spike were used to stimulate PBMCs, and the response was evaluated based on activation-induced markers (AIM) in CD4 + (OX40 + CD137 + ) and CD8 + (CD69 + CD137 + ) T cells. The CD4 + /CD8 + T cell reactivity in the vaccinees was not substantially affected by mutations in B.1.1.7 and P.1. However, decreases of 14% and 22% were observed with the B.1.351 spike-pools for CD4 + and CD8 + T cells, respectively. AIM T cell responses in COVID-19 vaccinees displayed a memory phenotype irrespective of the variant analyzed, with preferential enrichment for central memory (T cm ) and effector memory (T em ) for CD4 + and T em and terminally differentiated effector memory (T emra ) for CD8 + T cells. These provide evidence that donors primed by the ancestral strain spike protein mount a memory T cell response that can cross-recognize the SARS-CoV-2 VOC. A limitation of this study is that overlapping peptide pools rather than individual peptides were used to evaluate the responses by which alterations in terms of antigen processing for either class I or class II MHC would be undetected. In another study, peptide pool (15-mers with 11 amino acids overlap) spanning mutated spike regions of B.1.1.7 and B.1.351 were used to detect the cross-reactivity of SARS-CoV-2 specific T cells with variants [51]. Blood samples were collected from COVID-19 naïve and recovered donors before and after the BNT162b2 vaccination. No differences in CD4 + T cell activation (based on AIM) were seen in response to variant antigens. However, in this study, the number of donors was limited to 20, and CD8 + T cells responses to VOC were not evaluated. In contrast, a study evaluated 747 SARS-CoV-2 virus isolates by deep sequencing and reported that MHC-I restricted mutant epitopes showed reduced (assessed based on melting temperature stabilizing capacity of wildtype or mutant peptides towards MHC-I) or even abrogated (HLA tetramers, loaded with WT or mutant peptide, were presented to expanded CD8 + T cells of HLA-matched COVID-19 patients) binding to MHC-I [52]. Moreover, CD8 + T cells stimulated with respective epitopes showed decreased proliferation and cytotoxicity. The tetramer-sorted CD8 + T cells revealed qualitative differences at the transcriptional level to mutant peptides. However, this approach should be extended to evaluate the T cell response after vaccination/immunization.

Breakthrough Cases
Breakthrough cases are people who get an infection even after complete immunization, meaning the pathogen breaks the protective barrier developed by vaccination. As already stated, none of the COVID-19 vaccines is 100% effective. Moreover, SARS-CoV-2 variant strains have emerged continuously. Reduced antibody responses in susceptible populations might render them prone to breakthrough infections [53]. Moreover, VOC may escape immune responses, so breakthrough cases are expected (Table 1). For example, a study including 1497 fully vaccinated health care workers reported 39 SARS-CoV-2 breakthrough infections. For 22 of the 39 workers with breakthrough infections, the results for peri-infection neutralizing antibodies were available. During the periinfection period, the neutralizing antibody titers in breakthrough cases were lower than those in matched uninfected vaccinated controls (n = 104) [54]. Although higher peri-infection neutralizing antibody titers were associated with lower infectivity, the levels of neutralizing antibodies in breakthrough cases were not significantly lower than matched uninfected vaccinated controls. Moreover, this analysis does not provide a specific level of antibodies that might be associated with protection [55]. Another study reported lower levels of antibodies (S-RBD IgG, 3.469 arbitrary units /ml, AU/ml) in a 41-year-old woman 34-days post complete vaccination [56] compared to a previous study [57]. This patient developed COVID-19 symptoms 40-days post-vaccination. Subsequently, 20-days post-symptom onset, the titer of the spike protein receptor-binding domain (S-RBD) IgG antibodies increased to 130 AU/ml. These results show that the vaccine failed to develop an effective immune response in the patient. In contrast, Hacisuleyman et al. [58] reported 2 breakthrough cases among 417 mRNA vaccinated individuals (19 and 36 days post-complete vaccination) [58]. One patient had extremely high titers of neutralizing antibodies. Moreover, the antibodies recognized the variants but were nonetheless insufficient to prevent a breakthrough infection. However, it can't be ruled out that the infection may have occurred before the booster shot took full effect.

Background
In March 2021, vaccinations with ChAdOx1 nCoV-19 were abruptly halted due to VITT [59,60]. The activation of platelet factor 4 (PF4) by antibodies might be amplified by booster vaccination with an adenoviral vector, which might induce and/or aggravate its adverse reactions. In addition, immune responses to the viral vector itself might compromise vaccine efficacy. Thus, boosting with an mRNA-based vaccine have instead been recommended [61]. Moreover, uneven availability issues for the approved vaccines around the world also compelled the switch to heterologous vaccination schedules [62]. A heterologous prime-boost vaccination (HtPBV) strategy could be an opportunity to make vaccination programs more flexible and reliable in response to fluctuations in supply or demand [61]. However, HtPBV has also been evaluated before COVID-19, and in many scenarios, heterologous vaccination has been more immunogenic than homologous prime-boost vaccination (HmPBV) [63,64]. In the context of COVID-19, some initial reports demonstrate that HtPBV is better or at least as immunogenic as HmPBV (Table 3).

Hybrid Vigor Immunity
Immunological memory induced by vaccines is a source of protection against infection. However, the vaccine effectiveness is more or less reduced against VOC [2][3][4]72]. On the other hand, natural infection by SARS-CoV-2 also induces memory immune responses. However, reinfections, especially with variants, including B.1.315 have been reported. What happens when previously infected individuals are vaccinated? The reports from several studies suggest that an impressive synergy results from a combination of natural immunity and vaccine-generated immunity called "hybrid vigor immunity" [73]. Natural immunity to SARS-CoV-1 or SARS-CoV-2, combined with vaccine-generated immunity, generates broad immune responses. For example, Tan et al. [74]  Stamatatos et al. [75] evaluated sera from 15 individuals who had previously been infected with SARS-CoV-2 and 13 individuals who had not been infected. The sera were collected before and after immunization with one of the mRNA vaccines (BNT162b2 or mRNA-1273). Prior to vaccination, sera from 12 of the 15 previously infected donors neutralized the Wuhan-Hu-1. However, the sera from these individuals showed weak and only sporadic neutralizing activity against the B.1.351. Interestingly, a single shot of vaccine in previously infected individuals with pre-existing virus-specific antibodies induced higher levels of virus-specific IgG and IgA than two vaccine doses in naive individuals. Compared to two vaccine doses in naïve individuals, a single dose of vaccine in previously infected individuals displayed 10-and 20-fold higher levels of neutralizing antibodies to the Wuhan-Hu-1 and B.1.351, respectively. Nevertheless, the serum of previously infected vaccinated individuals was 3 to 10fold less efficient in neutralizing the B.1.351 compared with Wuhan-Hu-1. Moreover, a second dose of the vaccine in the previously infected individuals within 3-4 weeks did not further boost neutralizing antibodies levels. Goel et al. [76] evaluated antibody and antigen-specific memory B cells in 33 SARS-CoV-2 naïve and 11 SARS-CoV-2 recovered subjects. Both groups received SARS-CoV-2 mRNA vaccines (BNT162b2 or mRNA-1273). SARS-CoV-2 naïve individuals required both vaccine doses for optimal increases in antibodies. Memory B cells specific for full-length spike protein and the RBD were also efficiently primed by mRNA vaccination and detectable in all SARS-CoV-2 naive subjects after the second vaccine dose. In SARS-CoV-2 recovered individuals, antibody and memory B cell responses were significantly boosted after the first vaccine dose. However, there was no increase in circulating antibodies, neutralizing titers, or antigen-specific memory B cells after the second dose. This robust boosting after the first vaccine dose strongly correlated with levels of pre-existing memory B cells in recovered individuals, identifying a key role for memory B cells in mounting recall responses to SARS-CoV-2 antigens.
In summary, hybrid vigor immunity is a potential field to explore the active components of COVID-19 vaccines. It is interesting to note that currently available vaccines mostly employ the spike protein as immunogen.
Including other viral genome components alongwith the spike in COVID-19 vaccines may mimic the natural virus more closely. And more importantly, development of replication-defective vaccines using the reverse genetics might pave way to better vaccines by inducing and mimicking the hybrid immunity described above. Idenfifying and deletion of viral factors [77][78][79][80][81][82][83][84][85], which modulate the host interferfon reponses, need to be considered for the development of next-generation COVID-19 vaccines.

Future Strategies: Fractional Dosing of Vaccines and Route of Vaccine Administration
In the context of COVID-19, various public health and social measures have been implemented to control the transmission of SARS-CoV-2. However, being emergency measures, they are difficult to sustain for longer periods [86]. Besides, a shortage in the supply of vaccines is a matter of concern, especially in low-income countries. However, if dose-sparing is effective in preventing symptomatic and severe disease, it would extend the limited supply of vaccines and will play a significant role in bringing the pandemic to an end. More importantly, vaccinating more people with lesser doses may reduce the transmission of the virus, which might reduce the incidence and occurrence of the disease [86]. Dose sparing in case of COVID-19 vaccines shall be evaluated to answer a number of questions [87]: Will dose sparing result in an abundant immune response to prevent symptomatic or severe disease and transmission of the virus; how effective will it be against VOC; how safe will it be to administer, including adverse reactions and emergence of new variants; will it be effective in different populations, including immunocompromised individuals? A primising example of successful vaccine dose fractionation is against yellow fever in Angola, the Democratic Republic of Congo. In 2015, in response to the yellow fever epidemic, emergency vaccination was required. However, due to the limited supply of vaccines, WHO's Strategic Advisory Group of Experts on Immunization reviewed the evidence on the immunogenicity and safety of fractional dosing of vaccines against yellow fever and recommended dose fractionation down to one-fifth of the standard dose [88,89]. Fractional dosing was predicted to substantially reduce population infection attack rates and save lives [88]. In the context of COVID-19 vaccines, a preliminary study comprising 600 individuals of different age groups evaluated 50 and 100 μg 2-dose regime (mRNA-1273) for safety and immunogenicity [90]. Anti-SARS-CoV-2 spike binding antibody levels increased substantially by day 14 after the second dose to geometric mean peak levels of 189 (173-207) and 239 (221-259) μg/ml at 50 and 100 μg dose respectively in younger participants (≥18 to <55-years age), and 153 (135-175) and 162 (142-185) μg/ml in older participants (≥55 years age). In addition, neutralizing antibody levels were increased to maximum geometric mean titers of 1733 (1611-1865) μg/ml at 50 μg dose and 1909 (1849-1971) μg/ml at 100 μg dose in younger adults, and 1827 (1722-1938) μg/ml at 50 μg and 1686 (1521-1869) μg/ml at 100 μg in older adults. Although no statistical evaluation was done for antibody levels in participants who received 50 or 100 μg doses, numerical antibody levels seem to be comparable which favors the feasibility of fractional dosing [90]. In an interim analysis of 4 randomized controlled trials, a subgroup of participants was primed with a half dose of ChAdOx1 nCoV-19 vaccine instead of a full dose, followed by a full-dose boost after a median of 12 weeks [91]. A vaccine efficacy of 90% (67-97%) was reported in this subgroup. Although only a small number of participants were included, the lower bound of 67% for the efficacy estimate is very reassuring [86]. However, fractional dosing of COVID-19 vaccines needs to be evaluated in larger populations especially because immune correlates of protection have not been established.
In the UK, a decision was made in December 2020 to delay the second vaccine dose to 12-weeks post-first dose, which aimed to vaccinate more people to develope at least some protection against SARS-CoV-2. A third wave of COVID-19 caused by a highly transmissible Delta variant has led to considerations of the potential need and optimal timing for a second booster shot for vaccinated populations [92]. However, vaccinating more people appears more tempting. Two doses of COVID-19 vaccines are efficient in controlling severe disease, even those caused by VOC [2][3][4]. Although there are concerns about waning antibody responses, however, the declining antibody responses do not necessarily mean reduced vaccine efficacy because the effect against disease is not only mediated by antibodies that might be relatively short-lived for some vaccines but also by long-living memory and cellular immune responses [93]. For influenza, each annual vaccine is based on the most current data about circulating strains, increasing the likelihood that the vaccine will remain effective even if there is further strain evolution [94]. In the sense of COVID-19, there is an opportunity now to study variant-based boosters before there is a widespread need for them [95]. In this context, Moderna has started clinical trials (NCT04785144) for mRNA-1273.351, targeting novel B.1.351 VOC. The study is divided into 2 cohorts. Cohort 1 who received two vaccinations of mRNA-1273 at dosages of 50 μg, 100 μg, or 250 μg in the Phase 1 clinical trial (DMID 20-0003) will be given a single intramuscular (IM) booster of mRNA-1273.351. Cohort 2, who have never received a COVID-19 vaccine, will be given 2 or 3 IM doses of mRNA-1273.351. Moreover, a multivalent booster candidate mRNA-1273.211 (Combines mRNA-1273 and mRNA-B.1.351) to adult participants who previously received 2 doses of mRNA-1273 (NCT04470427) is currently in Phase 2 and 3 (NCT04927065).
SARS-CoV-2 specific T cells have been detected even in asymptomatic individuals [96] and those who don't seroconvert [97]. T cells can be especially important in convalescents who don't seroconvert or immunocompromised individuals who are less likely to develop an effective antibody response. Sterilizing immunity completely stops viral replication in the host, which can be achieved by antibodies. Among T cells, CD8 + T RM could come closest to sterilizing immunity by eliminating the pathogens at the portal of entry [98]. The route of COVID-19 vaccine administration shall be given more attention as both route and vaccine formulation are key determinants for T RM formation [99,100]. For example, the parenteral route of administration is unable to efficiently induce IgA and T RM in the lungs [101,102] in comparison to intranasal (IN) vaccination. A single IN dose of Chimpanzee adenoviral vaccine encoding stabilized S in mice almost entirely prevented SARS-CoV-2 infection in both the upper and lower respiratory tracts by inducing a mucosal immune response, including high levels of SARS-CoV-2 S specific IgA in serum and lung. Of note, CD103 + CD69 + CD8 + T cells, likely of a resident memory phenotype, were induced by IN route and not by the IM route [102]. These results depict that intramuscular vaccination does not confer sterilizing immunity. Eventually, Hassan et al. [102] extended their strategy to non-human primates and found that a single dose of IN adenoviral vectored vaccine protects rhesus macaques against SARS-CoV-2. However, in this study, IM and IN routes were not compared. Currently, 7 vaccines are in clinical phase trials which will be administrated by IN route. However, how effective IN vaccination will be, primarily in the long run, need to be evaluated in a more controlled and strict manner. A typical exemplary to understand the immune kinetics of IN immunization is vaccination against Influenza A virus (IAV) [reviewed by [103]]. IAV specific lung T RM provides potent protection against heterosubtypic influenza challenge. However, this protection is transient because of increased apoptosis of T RM in the lung and airways, unlike populations in the skin, nasal tissue, and intestinal mucosae. In this regard, COVID-19 vaccines effectively inducing and stabilizing T RM in the lungs will be an exciting field to explore.
In conclusion, VOC, especially B.1.315 and B.1.617.2, escape the antibody responses. The failure to generate sufficient immune responses might lead to breakthrough cases. However, recommended doses of vaccines are effective against severe diseases and deaths that are of utmost importance in the present scenario.
The uneven availability of COVID-19 vaccines can be tackled by heterologous vaccination, which generates better or at least comparable immune responses. The reports about the adverse reactions of heterologous vaccination are rare and shall be evaluated in larger populations. An emerging concept of hybrid vigor immunity shall be given prime attention. In this context, the inclusion of different SARS-CoV-2 proteins along with spike may provide broader protection against SARS-CoV-2 variants.