Introduction

Since the emergence of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in Wuhan, COVID-19 rapidly spread worldwide. At present, 754 million SARS-CoV-2-confirmed cases and more than 6.8 million of deaths by COVID-19 have been reported as of February 5, 2023, globally1,2,3,4.

From the initial stage of the global pandemic, massive efforts were made toward development of novel vaccines against SARS-CoV-2 around the world5,6,7 Currently, more than 50 vaccines have been approved by at least one country (https://covid19.trackvaccines.org/vaccines/approved/). The efficacy of vaccines against SARS-CoV-2 had been beyond expectation. As of February 2023, more than 13.3 billion of anti-SARS-CoV-2 vaccine doses have already been administered in the world8. Among various vaccines, two mRNA vaccines, BNT162b2 (Pfizer/BioNTech) and mRNA-1273 (Moderna), have previously shown ~ 95% efficacy in preventing symptomatic COVID-19 in early-phase of pandemic9,10,11, and these mRNA vaccines accounted for 90 and 97% of total administrated doses of COVID-19 vaccine in U.S. and European Union respectively, up to present8.

According to the recent emergence of immune-evasive Omicron variants, novel bivalent mRNA booster vaccines were developed by targeting the Spike protein of SARS-CoV-2Wuhan and Omicron BA.4/BA.5 sublineages and have been provided to individuals who had completed at least 2 doses of monovalent COVID-19 vaccination. Although the Morbidity and Mortality Weekly Report (MMWR) described that among immunocompetent adults, who were ≥ 65 years, a bivalent booster dose provided 73% additional protection against COVID-19 hospitalization compared with monovalent mRNA vaccination only12, there have been multiple results reporting antibody evasion profiles of new Omicron sublineages BQ.1.1 and XBB, posing further concerns on the efficacy of anti-SARS-CoV-2 vaccines13.

We have continuously evaluated the neutralizing activity of sera obtained from Pfizer/BioNTech monovalent BNT162b2 (mvBNT)-vaccinated health care workers (HCWs) in Japan14,15,16,17,18. In the present proactive cohort study, we focused on the vaccinated participants’ sera obtained pre- and post-5th-dose of Omicron BA.4/5-adapted bivalent BNT162b2 (bvBNT), and determined neutralization titers (NT50s) of sera against wild-type Wuhan SARS-CoV-2 strain (SCoV2Wuhan) and Omicron sublineages, including BA.2, BA.5, BQ.1.1, XBB, and XBB.1.5, vaccine-elicited S1-binding IgG levels. All the SARS-CoV-2 strains/variants used in this study were infectious viruses, isolated from individuals at airport quarantine stations or hospitals in Japan and were not recombinant- or pseudo-viruses.

Results

Effects of Omicron BA.4/5-adapted BNT162b2 (bvBNT) vaccination in sera obtained from health care workers (HCWs) with risk factors.

Firstly, we examined SARS-CoV-2 neutralization activity of sera post-bvBNT booster vaccine dose, obtained from 23 out of 225 HCWs in Kumamoto General Hospital, Japan (225 individuals were initially recruited in the primary clinical study14), who were either of ≥ 60-years of age and/or had pre-existing diseases/risk factors (see demographic characteristics in Table 1).

Table 1 Demographic characteristics of the participants who received Omicron BA.5-adapted 5th-dose of BNT162b2 vaccination.
Full size table

The SARS-CoV-2 neutralizing activity (NT50) of their sera against the Wuhan strain of SARS-CoV-2 (infectious SCoV2Wuhan) was determined over 650 days using sera consecutively collected on (1) 1 week pre-4th-dose (day-470 [HCWs with risk factors]/-530 [HCWs without a risk factor(s)] from 1st-dose), (2) 2 weeks post-4th-dose (day-490 [HCWs with a risk factor(s)]/-550 [HCWs without a risk factor(s)]), (3) 10 weeks post-4th-dose (day-550 [HCWs with a risk factor(s) only]), (4) 1 week pre-5th-dose (bvBNT; day-630), and (5) 2 weeks post-5th-dose (day-650)(see Methods and Fig. 1), representing a continuation of our previous studies14,15,16,17,18. We also evaluated the profile of S1-binding IgG levels following pre-/post-3rd, 4th, and 5th-dose (Table 2). In addition, we determined NT50s of the same sera using VeroE6TMPRSS cells against infectious Omicron BA.2, BA.5, BQ.1.1, XBB, and XBB.1.5 variant sublineages, whose emergence has been associated with the present explosive increases globally1. Results of samples obtained from participants whose swab PCR and/or anti-nucleocapsid-IgG in serum proved to be SARS-CoV-2 positive during this cohort study were excluded here.

Figure 1
figure 1

Schedule of 4-times mvBNT doses and once bvBNT dose in this study. Administration schedule of 4-times mvBNT doses (red arrows), once bvBNT dose (blue arrows), and blood collections (yellow arrows) in this study are shown. HCWs with risk factors received 4th dose of mvBNT on day477, while HCWs without risk factor received it on day537. HCWs with risk factors had 11th blood collection (10-weeks post 4th dose) on day550.

Full size image
Table 2 S1-binding IgG level of pre/post 3rd, 4th, and 5th-doses sera.
Full size table

Neutralizing activity of sera against SCoV2Wuhan was seen elevated moderately on day-28 (1-week post-2nd-dose) samples (Supplemental Table S1, gMean-NT50 = 283), while remarkable elevations were observed in neutralizing activity in the same participants’ sera of day-300 (2-weeks post-3rd-dose), achieving gMean-NT50 of 2009 (Supplemental Table S1). On day-470 (1-week pre-4th-dose), the gMean-NT50 had remarkably decreased down to 390, by around 19% of the peak value of day-300 (2-weeks post-3rd-dose/Fig. 2A). However, by day-490 (2-weeks post-4th-dose), neutralization activity increased to 1820 (Fig. 2A). On day-550 (10-weeks post-4th-dose) and day-630 (1-week pre-5th-dose), the gMean-NT50 decreased down to 1121 and 477, around 60% and 15% and of the peak value of day-490 (2-weeks post-4th-dose/Fig. 2A). However, by day-650 (2-weeks post-5th-dose, bivalent), neutralization activity again increased to 1966, and was higher than the peak value of day-490 with no statistical significance (2-weeks post-4th-dose, p = 0.8199/Fig. 2A). Fold-changes of gMean-NT50 against Wuhan between pre/post-5th dose was similar with those between pre/post-4th dose (4.1 and 4.7, respectively, p = 0.6637/Fig. 2A). We also examined the profile of SARS-CoV-2 Spike S1-binding IgG levels of sera at six different time-points, on pre-/post-3rd, 4th, and 5th-doses (Table 2), showing that gMean S1-binding IgG level of sera taken 2-weeks post-5th-bvBNT dose (4688 BAU/ml; ranges 846–13,334) was similar with those of post-3rd-mvBNT dose (4788 BAU/ml; ranges 1344–16,634) and lower than those of post-4th-mvBNT dose sera with no statistical significance (6305 BAU/ml; ranges 1611–19,236; p = 0.1860/Table 2).

Figure 2
figure 2

Effect of Omicron BA.4/5-adapted BNT162b2 (bvBNT) vaccination in sera obtained from HCWs with risk factors. Temporal changes of neutralizing activity of sera obtained from HCWs with risk factors over 650 days post-1st dose of BNT162b2 are shown. 4th monovalent (mvBNT) and 5th bivalent BNT162b2 (bvBNT) doses were administered on days 477 and 637, respectively (n = 23). (A) The 50% neutralization titers (NT50) of participants’ sera against infection by SARS-CoV-2Wuhan strain (SCoV2Wuhan) were determined on days 470, 490, 550, 630, and 650 post-1st-dose using VeroE6TMPRSS2 cell-based neutralization assay. Solid circles denote NT50 titers of each participant’s serum and filled bars denote average NT50 titers of 23 participants’ sera at each time point. Geometric mean NT50 titers (gMean-NT50) and ranges of NT50 at each time point are shown at the bottom. (B) Temporal changes of neutralizing activity of participants’ sera at days 470, 490, 550, 630, and 650 post-1st dose against Omicrons BA.2, BA.5, BQ.1.1, XBB, and XBB.1.5 are shown. Solid circles denote NT50 titers of each participant’s serum and filled bars denote average NT50 titers of 23 participants’ sera at each time point. The circles and lines in same color indicate that the data were obtained from same participant’s sera.

Full size image

When we evaluated neutralization activity in sera pre-/post-4th-dose of mvBNT vaccine using infectious Omicron variants, gMean-NT50 values against BA.5 on day-470 (1-week pre-4th-dose) and day-490 (2-weeks post-4th-dose) sera were 23 and 134, respectively (Fig. 2B). On day-550 (10-weeks post-4th-dose) and day-630 (1-week pre-5th-dose), the gMean-NT50 continuously decreased down to 63 and 44 after the 4th-dose. However, by day-650 (2-weeks post-5th-dose, bivalent), gMean-NT50 values against BA.5 was significantly elevated to 429 (ranges; 36–8184), 3.2-fold and 9.8-fold increases from the peak values of day-490 (2-weeks post-4th-dose, monovalent; p = 0.0021) and day-630 (1-week pre-5th-dose; p < 0.0001/Fig. 2B). Fold-changes of gMean-NT50 against BA.5 between pre/post-5th dose was higher than those between pre/post-4th dose with no statistical significance (9.8 and 5.8, respectively, p = 0.1032/Fig. 2B).

Similar profiles were observed when we examined neutralization activity against BA.2: gMean-NT50 values against BA.2 on day-650 (= 396; 2-weeks post-5th-dose, bivalent) was 1.6-fold higher than the peak value of day-490, but with no statistical significance (= 255; 2-weeks post-4th-dose, monovalent; p = 0.1698/Fig. 2B). We also evaluated neutralization activity against BQ.1.1 and XBB of day-650 (2-weeks post-5th-dose, bivalent) sera. gMean-NT50 values were 126 (ranges; < 20–513) and 114 (ranges; < 20–936) against BQ.1.1 and XBB, respectively, and were lower than those against BA.5 (p = 0.0008 and 0.0007, respectively) and against BA.2 (p = 0.0006 and 0.0006, respectively/Fig. 2B). Against XBB.1.5, day-650 sera showed the lowest gMean-NT50 values of 63 (ranges; < 20–433) among variants we tested (Fig. 2B).

Effects of Omicron BA.4/5-adapted BNT162b2 (bvBNT) vaccination in sera obtained from HCWs without risk factor.

Next, we examined SARS-CoV-2 neutralization activity of sera obtained from 90 out of 225 HCWs who were younger than 60 years of age and free from pre-existing diseases/risk factors. Results of samples obtained from swab PCR- or serum N-IgG-positive participants in this cohort were also excluded.

Against SCoV2wuhan, the gMean-NT50 value of day-530 (1-week pre-4th-dose) was 265 (ranges; < 20–2613), and on day-550 (2-weeks post-4th-dose), the value elevated to 2028 (ranges; 441–11,653/Fig. 3). After ~ 40% decrease of NT50 on day-630 (1195; 1-week pre-5th-dose) from the peak value of 4th-dose, day-650 (2-weeks post-5th-dose, bivalent) sera again had an elevated value up to 2091 (ranges; 265–34,681/Fig. 3), and the values were comparable (p = 0.7081) to the values of day-550 (2-weeks post-4th-dose, mvBNT/Fig. 3). Similar profiles were seen in sera obtained from HCWs with risk factor (Fig. 2A). Regarding fold-changes of gMean-NT50 against Wuhan, HCWs with risk factors’ sera between pre/post-5th dose was significantly higher than those of HCWs without risk factor (4.1 and 1.8, respectively, p = 0.001/Figs. 2A and 3).

Figure 3
figure 3

Effect of bvBNT vaccination in sera obtained from HCWs without risk factor. Temporal changes of neutralizing activity of sera obtained from HCWs without risk factor over 650 days post-1st dose of mvBNT are shown. 4th mvBNT and 5th bvBNT doses were administered on days 537 and 637, respectively (n = 90). NT50 of participants’ sera against infection by SCoV2Wuhan and Omicrons BA.2, BA.5, BQ.1.1, XBB, and XBB.1.5 were determined using VeroE6TMPRSS2 cell-based neutralization assay. Solid circles denote NT50 titers of each participant’s serum and filled bars denote average NT50 titers of 90 participants’ sera at each time point. gMean-NT50 and ranges of NT50 at each time point are shown at the bottom. The circles and lines in same color indicate that the data were obtained from same participant’s sera.

Full size image

When we examined the profile of S1-binding IgG (S1-IgG) levels of sera from HCWs without risk factor, S1-binding IgG of 2-weeks post-5th-bvBNT dose sera (4691 BAU/ml; ranges 1313–38,324) was lower than those of post-4th-mvBNT doses sera with statistical significance (6024 BAU/ml; ranges 1912–43,502; p = 0.0038/Table 2). When we compared fold-changes of pre/post 4th dose S1-IgG and pre/post 5th dose S1-IgG levels, fold-changes of pre/post 4th dose S1-IgG were significantly higher than those of pre/post 5th dose (p values were 0.0116 and < 0.0001, for HCWs with risk factors and HCWs without risk factor, respectively/Table 2). Also, when we compared fold-changes of S1-IgG levels in sera between HCWs with risk factors and HCWs without risk factor, fold-changes of S1-IgG levels for HCWs with risk factors pre/post 5th dose were significantly higher than those of HCWs without risk factor (p < 0.0001/Table 2), but no significant difference was observed in the fold-changes of S1-IgG levels pre/post 4th dose sera between HCWs with risk factors and HCWs without risk factor (p = 0.1029).

When we examined neutralization activity against BA.5 using sera from day-550 (2-weeks post-4th-dose) and day-650 (2-weeks post-5th-dose, bivalent), the day-650 sera showed the gMean-NT50 value of 368 (ranges; 32–9257), 3.0-fold higher value compared to that of day-550 sera (= 121/ranges; < 20–2424/Fig. 3). Also, against BA.2, day-650 sera showed gMean-NT50 value of 470 (ranges; 60–9260), 2.2-fold higher than that of day-550 sera (= 214/ranges; 42–2223/Fig. 3). When we evaluated NT50 against BQ.1.1 and XBB of day-650 sera, gMean-NT50 values were 127 (ranges; < 20–2586) and 111 (ranges; < 20–2091), respectively (Fig. 3). On the other hand, day-650 sera showed gMean-NT50 values of 61 (ranges; < 20–2175) against XBB.1.5 (Fig. 3).

Effects of bvBNT in sera from HCWs who had experienced symptomatic/asymptomatic breakthrough infection during Omicron wave

Among HCWs enrolled in the present study, 20 participants proved to be SARS-CoV-2 positive by swab PCR from April 2022 to August 2022, and 13 out of the 20 participants received the 5th-dose bvBNT vaccination after recovery. These 13 participants who had experienced symptomatic breakthrough (BT) infection and 5th-dose of vaccine were termed as “BT-Sym#1–13” (symptoms of each participant were indicated in Supplemental Table S2). Other 17 participants who had neither tested nor received any positive results for swab-PCR or antigen tests but proved to be positive for anti-SARS-CoV-2 nucleocapsid-specific-IgG in their sera obtained from August 2022 to December 2022. These 17 participants all received 5th-dose bvBNT vaccine and were termed as “BT-Asym#1–17”. Detailed information regarding infection date (PCR positivity/serum N-IgG positivity) and longitudinal changes of serum NT50 values against Omicron BA.5 are summarized in Fig. 4 (PCR positive cases [A]; serum N-IgG positive cases [B]).

Figure 4
figure 4

Neutralization activity against Omicron BA.5 of sera obtained from HCWs who had received booster dose of mvBNT and experienced breakthrough infection during Omicron wave. Detailed information of neutralization activity (NT50) of sera obtained from BT-infection experienced participants against Omicron BA.5, dates of swab PCR-positive, and positive periods of serum SARS-CoV-2 nucleocapsid-specific IgG (red colored columns) are shown. (A) shows the results of participants with risk factors, and (B) shows the results of participants without risk factor.

Full size image

As shown in Fig. 5, the 4th-dose mvBNT vaccination enhanced neutralization activity of sera against SCoV2Wuhan strain by around 3.5 folds [Fig. 5/gMean-NT50s; from 1114 (pre-4th-dose) to 3884 (post-4th-dose)], when ~ 43% (13 out of 30) BT-infected participants had not yet been SARS-CoVf-2-positive by that time (Fig. 4). At 1-week pre-bvBNT 5th-dose, NT50 value against SCoV2Wuhan strain remained high (Fig. 5/gMean-NT50 = 4021/ranges; 460–38,519), and 2-weeks post-bvBNT 5th-dose, NT50 value significantly increased up to 9037 (Fig. 5/ranges; 1896–36,758), 4.3-–4.6-fold higher than those of uninfected participants’ sera after 5th-dose bvBNT (Figs. 2, 3).

Figure 5
figure 5

Effect of bvBNT vaccination in sera from HCWs who had experienced breakthrough (BT) infection during Omicron wave. Temporal changes of neutralizing activity of sera obtained from HCWs who had experienced breakthrough (BT) infection over 650 days post-1st dose of BNT162b2 are shown (n = 30). NT50 of participants’ sera against infection by SCoV2Wuhan and Omicrons BA.2, BA.5, BQ.1.1, XBB, and XBB.1.5 were determined using VeroE6TMPRSS2 cell-based neutralization assay. Solid circles denote NT50 titers of each participant’s serum and filled bars denote average NT50 titers of 30 BT-infected participants’ sera at each time point. gMean-NT50 and ranges of NT50 at each time point are shown at the bottom. The circles and lines in same color indicate that the data were obtained from same participant’s sera.

Full size image

When we examined neutralization activity against BA.5 using sera of post-4th-dose and pre-5th-dose, they showed NT50 values of 554 and 625, respectively (Fig. 5), slightly greater than the peak-values of uninfected participants’ post-5th-dose sera (Figs. 2, 3). Surprisingly, post-5th-dose sera of BT-infected participants showed significantly high gMean-NT50 value of 2995 against BA.5 (Fig. 5/ranges; 576–16,031), which was 7.0–8.1-fold greater compared to the peak-values of uninfected participants’ sera post-5th-dose (Figs. 2, 3). Also, against BA.2, post-5th-dose sera showed good gMean-NT50 value of 2478 (ranges; 1163–15,105), which was 4.8-fold higher value compared to that of post-4th-dose sera (= 521/ranges; 74–19,158/Fig. 5), and was 5.1–6.3 folds greater value compared to those of uninfected participants’ sera post-5th-dose (Figs. 2, 3).

When we evaluated NT50 against BQ.1.1 using pre- and post-5th-dose sera of previously BT-infected participants, gMean-NT50 values were 180 (ranges; 35–1601) and 680 (ranges; 115–2668), respectively (Fig. 5), which showed 3.8-fold enhancement following the bvBNT vaccination. These data perhaps show the cross-neutralization elicited by the 5th-dose bvBNT between against BA.5 and against BQ.1.1 considering that BQ.1.1 emerged from BA.513.

Against XBB, gMean-NT50 values of pre- and post-5th-dose sera were 175 (ranges; 26–1601) and 512 (ranges; 122–2668) respectively (Fig. 5), showing 2.9-fold enhancement of neutralization activity after the 5th-bvBNT dose. Also, NT50 values against BQ.1.1 and XBB of BT-infected participants’ sera post-5th-dose were ~ 5.4- and ~ 4.5-fold higher than those of uninfected participants’ sera post-5th-dose (Figs. 2, 3). When we evaluated NT50 against XBB.1.5 using post-5th-dose sera of BT-infected participants, gMean-NT50 values were 228 (ranges; 85–832/Fig. 5). %Reduction of gMean-NT50 values of post 5th-dose sera against BQ.1.1, XBB, and XBB.1.5 compared with those against vaccine-strains (Wuhan and BA.5) are summarized in Table S3.

Discussion

In the present cohort study, we studied in detail the effectiveness of BA.4/5 adapted bivalent BNT162b2 (bvBNT) vaccine using various infectious SARS-CoV-2 s, by examining the participants’ sera obtained pre-/post-2nd–4th-doses of mvBNT, from participants with/without risk factors or who had experienced BT-infection during the Omicron-wave (January 2022 through December 2022) in Japan.

In our previous data with mvBNT vaccinations between post-3rd and 4th-doses, neutralization activity against SCoV2Wuhan elicited by 4th-dose of mvBNT were not greater than those after 3rd-dose of mvBNT17, 18 (Supplemental Table S1). Similarly, the magnitudes of neutralizing activity against SCoV2Wuhan after the 5th-bvBNT dose were not greater compared to the significantly boosted response elicited by the 3rd-mvBNT in participants with/without risk factors (Supplemental Table S1). These limited restoration regarding neutralization activity against SCoV2Wuhan by 5th-bvBNT seems to reflect the difference the amounts of mRNA containing against original SCoV2Wuhan between 3rd-dose and 5th-dose (30 μg of mRNA for 3rd-dose of mvBNT, and 15 μg of mRNA for 5th-dose of mvBNT). However, 4th dose-mvBNT also contains 30 μg of mRNA against original SCoV2Wuhan, but 5th-bvBNT elicited higher neutralization activity against SCoV2Wuhan than those of 4th-mvBNT (Supplemental Table S1), indicating bvBNT may have different property from that of mvBNT against SCoV2Wuhan.

We also evaluated sera post-5th-bvBNT dose against not only BA.5 but also BA.2. All the post-5th-dose sera examined in the current study demonstrated significantly more robust neutralization activity against BA.5 and showed favorable neutralization activity also against BA.2 (p < 0.0001 for both BA.5 and BA.2 in Figs. 2 and 3).

In the present study, we also focused on the groups of participants who experienced BT-infection during the Omicron wave period (Fig. 4 and 5). BT-infected participants showed significant enhancement of neutralization activity after bvBNT dose against SCoV2Wuhan, BA.2, and BA.5, as well as BQ.1.1 and XBB. These results suggest that repeated stimulation caused by the exposure to Omicron’s Spike protein elicited broad and stronger neutralization activity against multiple SARS-CoV-2 variants. If it is the case and if further infection waves by SARS-CoV-2 variants arrive, booster bvBNT doses may have to be considered, although further data on the range of neutralization elicited by bvBNT have to be carefully examined.

When we compared gMean-S1-binding IgG levels and gMean-NT50 against BA.5 in sera obtained 2 weeks-post 3rd-mvBNT dose between symptomatic BT-infection (Sym-BTI) group and asymptomatic BT-infection (Asym-BTI) group, significant difference was observed in S1-binding IgG levels. Sym-BTI group showed gMean-S1-binding IgG of 3834 (ranges 1401–9843), while Asym-BTI group showed 1.9-fold higher gMean-S1-binding IgG value of 7272 (ranges 2668–15,963; p = 0.007/Table S4). Similarly, post 3rd-dose of Asym-BTI group’s sera showed 2.3-fold higher gMean-NT50 value against BA.5 compared to those of Sym-BTI group with no statistical significance (191.4 and 83.4, respectively; p = 0.0523/Table S4). These results might be an explanation why these groups showed different disease severity after SARS-CoV-2 BT-infection post 3rd-dose of original mvBNT vaccination.

The present data show that the 5th-bvBNT dose elicits greater levels of SARS-CoV-2-neutralizing activities against various SARS-CoV-2 variants including Omicron sublineages such as BA.2 and BA.5 although elicitation of neutralization against BQ.1.1, XBB, and XBB.1.5 is limited, indicating that more improved anti-SARS-CoV-2 vaccines capable of eliciting further broader and stronger neutralization are required to further better respond to the current COVID-19 pandemic. It was also suggested that individuals who previously experienced SARS-CoV-2 infection (mostly with Omicron variants) may have more robust neutralization against Omicron variants, which endorses vaccination with bvBNT dose following vaccination with mvBNT. However, further evaluation should be required for the administration of booster bivalent mRNA vaccination to the individuals who experienced recent BT-infection.

Methods

Participants and serum specimens.

The vaccination (on days 0, 21, 287, and 537, 30 μg of mRNA/each dose for mvBNT, and on day 637, 15 μg of mRNA against original strain and 15 μg of mRNA for bvBNT) and serum collection (on day-7, -28, -60, -90, -150, -280, -300, -360, -470, -490, -530, -550, -630, and -650 post-1st-dose) were carried out. Samples were collected from vaccinated health care workers at Japan Community Health Care Organization (JCHO), Kumamoto General Hospital (Kumamoto, Japan). In this report, 23 participants had previously received 4th-dose of monovalent BNT162b2 vaccine (mvBNT) 2 months earlier than other participants since they had a risk(s) of developing severe COVID-19. In contrast, 90 participants who were younger than 60 years of age and free from pre-existing diseases/risk factors received 4th-dose of mvBNT vaccination 2 months later than participants with risk factors, so that the interval between the 4th-dose and 5th-dose were 2 months shorter than participants with risk factors (See Fig. 1).

Samples were analyzed at Kumamoto University in Kumamoto and the National Center for Global Health and Medicine (NCGM) in Tokyo. The Ethics Committee from the Kumamoto General Hospital, NCGM, and Kumamoto university approved this study (Kumamoto General Hospital No. 180, NCGM-G-004176-00, and Kumamoto university No 2643). Each participant provided a written informed consent, and this study abided by the Declaration of Helsinki principles. The infection by a series of Omicron variants was dominant in Japan largely from January, 2022 through December, 2022. We defined the period of SARS-CoV-2 infection in Japan as “Omicron wave period”.

Cells and viruses

VeroE6TMPRSS2 cells19 were obtained from Japanese Collection of Research Bioresources (JCRB) Cell Bank (Osaka, Japan). VeroE6TMPRSS2 cells were maintained in DMEM supplemented with 10% FCS, 100 µg/ml of penicillin, 100 µg/ml of streptomycin, and 1 mg/ml of G418.

SARS-CoV-2 NCGM-05-2N strain (SCoV205-2N) was isolated from nasopharyngeal swabs of a patient with COVID-19, who was admitted to the NCGM hospital20. hCoV-19/Japan/TKYS02037/2022 (Omicron/BA.2; SARS-CoV-22037, GISAID Accession ID; EPI_ISL_9397331), hCoV-19/Japan/TKYS14631/2022 (Omicron/BA.5; SARS-CoV-2TKYS14631, GISAID Accession ID: EPI_ISL_12812500.1), hCoV-19/Japan/TY41-796/2022 (BQ.1.1; SARS-CoV-2TY41-796, GISAID Accession ID: EPI_ISL_15579783), and hCoV-19/Japan/TY41-795/2022 (XBB; SARS-CoV-2TY41-795, GISAID Accession ID: EPI_ISL_15669344) were provided from Tokyo Metropolitan Institute of public Health, Japan. hCoV-19/Japan/23-018-P1/2022 (XBB.1.5; SARS-CoV-223-018, GISAID Accession ID: EPI_ISL_16889601) was provided by National Institute of Infectious Diseases, Japan. Each variant was confirmed to contain each variant-specific amino acid substitutions.

Neutralization assay procedure.

The neutralizing activity of sera from vaccinated individuals was determined by quantifying the serum-mediated viral suppression in SARS-CoV-2-infected VeroE6TMPRSS2 cells as previously described with minor modifications14, 21. In brief, each serum was serially diluted in culture medium. The diluted sera were incubated with 100 TCID50 of viruses at 37 °C for 20 min (final serum dilutions were 1:20, 1:62.5, 1:250, 1:600, 1;1,000, 1:4000, 1:16,000, and 1:64,000), after which the serum-virus mixtures were inoculated to VeroE6TMPRSS2 cells (1.0 × 104/well) in 96-well microtiter culture plates. SARS-CoV-2 strains used in this assay were as follows: a wild-type Wuhan strain SCoV205-2N, Omicron strains SARS-CoV-22037 (BA.2; contains K417N/T478K/E484A/N501Y/D614G mutations in Spike), SARS-CoV-2TKYS14631 (BA.5; contains K417N/L452R/T478K/E484A/F486V/N501Y/D614G mutations in Spike), SARS-CoV-2TY41-796 (BQ.1.1; contains R346T/K417N/K444T/L452R/N460K/T478K/E484A/F486V/N501Y/D614G.

mutations in Spike), SARS-CoV-2TY41-795 (XBB; contains R346T/K417N/T478K/V445P/G446S/N460K/E484A/F486S/N501Y/D614G mutations in Spike), and SARS-CoV-223-018 (XBB.1.5; contains R346T/K417N/T478K/V445P/G446S/N460K/E484A/F486P/N501Y/D614G mutations in Spike). After culturing the cells for 3 days, the levels of virally caused cytopathic effect (CPE) observed in SARS-CoV-2-exposed cells were determined using the WST-8 assay employing Cell Counting Kit-8 (Dojindo, Kumamoto, Japan). The serum dilution that gave 50% inhibition of CPE was defined as 50% neutralization titer (NT50s). Each serum was tested in duplicates. All p values presented in the Figures and Results were calculated using the t-test.