Introduction
Hematopoietic stem cell transplantation (HSCT) is a therapeutic approach that utilizes hematopoietic stem cells to cure a variety of hematological diseases. These include acute and chronic leukemia, lymphoma, certain inherited disorders of the hematopoietic system, and metabolic conditions [
1,
2]. During HSCT, the recipient's immune system is replaced with an immune system derived from donor hematopoietic stem cells via allografting [
3,
4]. Consequently, immunity acquired through vaccination or natural infection prior to HSCT is believed to be significantly diminished or entirely lost. In the absence of reimmunization, immunity against many pathogens declines following HSCT. Accordingly, prophylactic or pre-emptive strategies to prevent infection in these patients are recommended, including the post-HSCT use of antibiotics, antivirals, antifungals, and revaccination. Notably, guidelines strongly advocate revaccination against vaccine-preventable diseases, such as influenza, pneumococcal infections, rubella, and measles [
5,
6]. mRNA vaccines introduced against SARS-CoV-2 in 2021 have demonstrated good immunogenicity even in HSCT recipients, as we previously reported, and revaccination after HSCT against COVID-19 is also recommended [
7,
8].
Although revaccination after HSCT is therefore necessary, the change in immune status following HSCT remains inadequately characterized. A major uncertainty is whether the immune memory is entirely erased following HSCT or partially preserved. This ambiguity complicates decision making regarding the optimal number of vaccine doses required post-HSCT. Furthermore, although vaccine efficacy is considered closely linked to immune status at the time of vaccination, the status of reconstructed acquired immunity, especially humoral immunity, after HSCT remains unclear. Previous studies have indicated that B-cell counts may recover within a few months following HSCT [
9]; however, many of these cells remain immature, potentially resulting in differences in the function and diversity (repertoire) of humoral immunity compared to healthy individuals [
10,
11]. We therefore considered that optimizing revaccination strategies after HSCT, including appropriate timing and methods, required a detailed understanding of the immune status following HSCT.
We previously established a novel method to assess the response to SARS-CoV-2 vaccination at the mRNA level by quantifying antigen-specific antibody sequences, named the “Quantification of Antigen-specific Antibody Sequence (QASAS) method” [
12‐
14]. Compared to conventional tests, which detect antibodies with a long half‐life, the QASAS method directly detects B-cell receptor (BCR) mRNA as short‐lived, allowing humoral immunological activity to be monitored in real time [
12‐
14]. We hypothesized that use of the QASAS method to analyze immune responses in detail following vaccination in post-transplant patients would allow us to assess immune status after HSCT.
Although the impact of HSCT on the immune system has been extensively studied, studies have primarily focused on T cells, and information on its effects on B-cell/humoral immunity is limited. Furthermore, while the genetic diversity of BCRs is a crucial indicator of humoral immune status, its alteration following HSCT has not been fully elucidated.
Here, to support the development of strategies to prevent post-transplant infection, we evaluated the loss of immunological memory and the status of humoral immunity after HSCT using various analytical approaches. We particularly focused on BCR repertoire analysis.
Materials and methods
Participants
For flow cytometry (fluorescence‑activated cell sorting; FACS) analysis, 10 HSCT patients (patients 1 to 10; cord blood,
n = 3; bone marrow,
n = 7) were enrolled (Table
S1). To investigate humoral responses to SARS‑CoV‑2 using the QASAS method, we collected 14 datasets (cases 1 to 14) of SARS-CoV-2 antigen exposure from nine individuals without hematologic disorders (Table
S2). Then, to examine vaccine‑induced immunity after HSCT, we further enrolled three HSCT patients (patients 11–13) who received cord blood transplantation with a history of exposure to SARS-CoV-2 antigen before HSCT (Table
S3).
The study protocol was approved by the Kobe University Hospital Ethics Committee (Protocol nos. B2056704 and 1481) and conducted in accordance with the Declaration of Helsinki. All samples were collected at Kobe University Hospital between July 2020 and January 2024. Written informed consent was obtained from all participants.
Sample collection and processing
Peripheral blood samples were collected using heparin-containing tubes. Peripheral blood mononuclear cells (PBMCs) were isolated from the blood by density gradient centrifugation using Lymphoprep (Serumwerk Bernburg AG, Bernburg, Germany) and SepMate-50 tubes (STEMCELL Technologies Inc., Vancouver, BC, Canada). PBMC samples were stored with CELLBANKER (Zenogen Pharma, Fukushima, Japan) at − 80 °C until analysis. Total RNA was extracted with TRIzol LS (Thermo Fisher Scientific, Waltham, MA, USA) and purified with an RNeasy Mini Kit (Qiagen, Hilden, Germany).
Flow cytometric analysis
PBMC were stained for 60 min at 4 °C with anti‑human antibodies for the B-cell lineage, namely CD19 BV510 (Biolegend, San Diego, CA, USA), CD27 BV421 (Biolegend), IgD FITC (Biolegend), CD24 PerCP-CyTM5.5 (BD Biosciences, Franklin Lakes, NJ, USA) and CD38 BV711 (Biolegend). Isotype-matched antibodies served as controls. Flow cytometric analysis was performed using a FACSAria III instrument (BD Biosciences) with established phenotypes used for B-cell analysis. CD19+ cells were defined as total B cells. The following subsets of B cells were then analyzed: transitional B cells (CD19+ CD24+ CD38+) and class-switched B cells (CD19+ CD27+ IgD−).
B-cell receptor repertoire analysis
BCR repertoire analysis was performed according to previous studies [
12‐
14]. For mutation analysis, paired-end reads were assembled into a complete sequence using the AssemblePairs.py command of the pRESTO package (
https://presto.readthedocs.io/en/stable/). The fasta-formatted assembled sequence was analyzed for immunoglobulin gene rearrangement using igblast-1.22.0 and output as adaptive immune receptor repertoire (AIRR)-formatted data. Sequence identity to the IMGT reference sequence (v_identity and j_identity) was used for mutation analysis. Probability of the generation (pGen) of amino acid sequences of CDR3 was computed from a generative model of V(D)J recombination using the OLGA packages [
15]. CDR3 sequences that failed to compute pGen were excluded from subsequent analysis.
Database and COVID-19-specific sequence search
Antibody sequences specific to COVID-19 were downloaded from CoV-AbDab (
http://opig.stats.ox.ac.uk/webapps/covabdab/). Data which had been updated on 20 December 2022 with 12,004 entries were used as reference. A total of 260,856,092 sequences were used to validate the method. IgG and IgM antibody sequences from 12 healthy volunteers before the COVID-19 pandemic were described previously [
16]. Sequences with V and J gene names identical to the query sequence and with CDR3 amino acid sequence differences (Levenshtein distance) of 0 to 2 were retrieved from the database.
Statistical analyses
Shannon–Weaver diversity index was calculated with the vegan package in R and tested for significant differences using the Wilcoxon test. Levels of somatic hypermutation between HSCT patients and healthy controls were compared with the Wilcoxon test. All p values were two sided and considered statistically significant at the < 0.05 level. Statistical analyses were carried out with GraphPad Prism 10.0 (GraphPad Software, San Diego, CA, USA).
Discussion
In this study, we used FACS, the QASAS method, and BCR repertoire analysis to assess the state of humoral immunity after HSCT. FACS analysis demonstrated that, in the early post-HSCT period, mature B cells were absent; instead, the compartment was dominated by highly immature cells that subsequently underwent gradual maturation. We interpret this change to represent the resetting of the immune system, resembling the establishment of a new immune system, similar to that of a newborn. Further, analysis using the QASAS method revealed that, despite antigen exposure prior to HSCT, the first vaccination after HSCT resulted in a primary immune pattern. We interpret this finding to mean that HSCT caused the loss of immune memory. Furthermore, when we compared the BCR repertoire of healthy controls and HSCT patients, we found that the repertoire of HSCT patients was more immature and closer to the initial state, which supports our interpretation.
B cells differentiate via a germinal center reaction, during which a high degree of somatic hypermutation is induced. IgG antibodies of healthy adults exhibited high somatic hypermutation levels and low pGen values, whereas post-transplant B cells showed low somatic hypermutation levels and high pGen values, a pattern characteristic of antigen-inexperienced B cells. These findings suggest that pre-transplant B-cell immune memory is lost following HSCT, and that the newly generated B cells remain functionally immature.
We propose that the response to vaccination after HSCT should be characterized by the fact that immune memory is reset by HSCT. To ensure the rapid acquisition of immunity, it may be appropriate to administer at least two doses of the primary vaccination series after HSCT. Tsoutsoukis et al. evaluated antibody responses following one to four doses of SARS-CoV-2 vaccine administered after HSCT and reported that antibody titers improved after the second and subsequent doses [
18]. Furthermore, the U.S. guidelines recommend a three-dose regimen as the primary vaccination series after HSCT, highlighting the importance of multiple (i.e., two or more) vaccine doses in this population [
19]. These reports support our suggestion.
Several limitations of our study warrant mention. First, immune reconstitution following HSCT is influenced by various distinct factors, and the results of this study may be affected by such variability. In our cohort, all three HSCT recipients underwent cord blood transplantation, and it is important to consider that immune reconstitution may differ when other hematopoietic stem cell sources are used. For example, previous studies have reported that the recovery of total T cells and natural killer cells after cord blood transplantation is comparable to that after bone marrow transplantation, whereas the reconstitution of CD4
+ T cells and B cells tends to proceed more rapidly than in bone marrow transplantation or peripheral blood stem cell transplantation [
20‐
24]. Therefore, further studies are needed to confirm whether similar results are observed in HSCT using other stem cell sources. In addition, factors such as the conditioning regimen and graft-versus-host disease prophylaxis may also influence immune reconstitution. Second, the impact of patient age on results should be considered. Previous reports have shown that the recovery of thymic function after HSCT affects immune reconstitution [
4], and that the recovery of thymic function after HSCT is faster in children than in adults [
25]. As our study included only three adult patients, the findings may not be generalizable to populations of different ages, including pediatric patients. Third, considering that some reports suggest immune reconstitution gradually matures over several years after transplantation [
3,
23], the timing of vaccination after HSCT may affect the observed immune responses. In our study, the timing of vaccination varied substantially among patients (ranging from Day 392 to Day 656 post-transplantation), and we cannot rule out the possibility that this variability influenced the results. Taken together, these considerations highlight the need for further validation in diverse patient populations to generalize our findings.
Even with pre-HSCT exposure to antigen, first vaccination after HSCT induced a primary immune response. Regardless of past infection or vaccination history, the first vaccination after HSCT should be considered to induce a primary immune response.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.