nature reviews immunology https://doi.org/10.

1038/s41577-024-01040-6

Review article Check for updates

A guide to adaptive immune

memory
Nora Lam , YoonSeung Lee
1,4 2,4
& Donna L. Farber 2,3

Abstract Sections

Immune memory — comprising T cells, B cells and plasma cells and Introduction

their secreted antibodies — is crucial for human survival. It enables Immune memory and
the rapid and effective clearance of a pathogen after re-exposure, to protection in humans

minimize damage to the host. When antigen-experienced, memory General principles of immune
memory
T cells become activated, they proliferate and produce effector
molecules at faster rates and in greater magnitudes than antigen- Studying memory cells in mice
and humans
inexperienced, naive cells. Similarly, memory B cells become activated
T cell memory
and differentiate into antibody-secreting cells more rapidly than
naive B cells, and they undergo processes that increase their affinity B cell memory

for antigen. The ability of T cells and B cells to form memory cells after Current challenges and
antigen exposure is the rationale behind vaccination. Understanding future prospects

immune memory not only is crucial for the design of more-efficacious

vaccines but also has important implications for immunotherapies
in infectious disease and cancer. This ‘guide to’ article provides an
overview of the current understanding of the phenotype, function,
location, and pathways f­or t­he generation, maintenance and protective
capacity of memory T cells and memory B cells.

Department of Pathology and Cell Biology, Columbia University Irving Medical Center, New York, NY, USA.
1

2
Department of Microbiology and Immunology, Columbia University Irving Medical Center, New York, NY,
USA. 3Department of Surgery, Columbia University Irving Medical Center, New York, NY, USA. 4These authors
contributed equally: Nora Lam, YoonSeung Lee. e-mail: [email protected]

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 810

Review article

Introduction immunosenescence. As an example, all infants and young children

The immune system has evolved to protect the host against tissue dam- become infected with respiratory syncytial virus (RSV) multiple times and
age caused by infectious pathogens and other insults. Both inverte- have markedly greater disease severity than older children and adults2–4
brates and vertebrates have innate defence mechanisms, including (Fig. 1a). The levels of RSV-specific antibodies in an individual are
specialized cells recognizing common features of infectious micro- inversely proportional to susceptibility; they increase substantially
organisms or signals of tissue damage, that trigger the secretion of during early childhood and are maintained at high levels in adults for
molecules and cellular functions to rid the host of pathogens and miti- many decades (Fig. 1a). Similar age-related trends in susceptibility and
gate damage. However, the longer lives of vertebrates necessitate an the acquisition of specific immunity are seen for influenza virus infec-
additional, adaptive form of host defence with two key features: (1) tion, although the increase in antibody titres occurs earlier, followed by
higher-order specificity for different types of antigens, and (2) the a slight decline and then stable maintenance in adults2–4 (Fig. 1b). In both
capacity to ‘learn’ from previous encounters and develop a type of cases, the development of stable stores of anti-pathogen antibodies
immune memory. Although innate immune cells of the myeloid line- correlates with long-term protection from these viruses. Because most
age such as macrophages can also undergo changes following antigen antibody production requires both T cell and B cell responses, it is often
exposure that result in a type of memory or ‘trained immunity’1, we used as a surrogate for the generation of adaptive immune memory.
focus here on the adaptive forms of antigen-specific immune memory These examples illustrate how early and cumulative exposures to
in the context of host protection from infection. infections (and vaccines) are formative for development of the memory
The specificity and memory potential of the adaptive immune sys- repertoire that provides protection for our lifespan. However, immune
tem are endowed within lymphocytes — T cells and B cells, which each memory is highly specific, and we remain vulnerable to new pathogens
express a specific, clonally derived antigen receptor (T cell receptor throughout our lives, as recently demonstrated by the COVID-19 pan-
(TCR) and B cell receptor (BCR), respectively). Antigen encounter results demic caused by the novel pathogen SARS-CoV-2. In early 2020, the
in lymphocyte activation and the generation of both effector cells to global population encountered a virus for which almost all individuals
protect against the immediate threat and long-lived memory cells of all ages lacked immune memory, resulting in widespread infection
that respond more effectively to pathogen re-encounter. Long-lived and illness across the population, and causing severe disease and death
memory cells comprise memory T cells, memory B cells (MBCs) and in susceptible individuals. For SARS-CoV-2 infection, the most vulner-
plasma cells — all of which are predominantly localized within tissues able individuals were not infants and children but adults over 40 years
but can also be found in the circulation. The key soluble mediators of of age, with the risk of severe disease increasing substantially with each
memory are antibodies, produced by MBCs and plasma cells, and these subsequent decade, such that individuals over 70 years of age were
are found mainly in blood and in bodily secretions. 80 times more likely to die from COVID-19 than children were5 (Fig. 1c).
The concept of immune memory was first recognized thousands Over time, as a result of multiple exposures, vaccination and other factors
of years ago during the plague of Athens (~430 bce) caused by a highly such as antiviral treatments and changes in the virus to less lethal forms,
contagious pathogen that killed nearly one-quarter of the population; mortality from SARS-CoV-2 infection was reduced (by more than 96% in
in records from the time, it was noted that survivors of the plague 2022 compared with 2020)6. In this way, the development of immune
had developed immunity and were able to care for the sick. However, memory has enabled increased resilience to COVID-19, emphasizing the
understanding of the mechanisms by which immune memory is formed, important role of immune memory generated throughout our lives in
protects and is maintained remains incomplete. Nevertheless, research our ability to survive in a highly dynamic and often hostile environment.
in the past few decades and in this current era of single-cell profiling
has revealed new insights into the complexity of immune memory. General principles of immune memory
In this ‘Guide to’ article, we present current evidence for the impor- Immune memory is generated following an initial immune response to
tance of immune memory in health and lifespan and a synthesis of the an antigen, referred to as the primary response, with a memory immune
known features of immune memory, including the key cells and mol- response persisting long after the antigen is cleared. The cellular and
ecules involved, and emerging evidence of the important role of tissue- molecular processes that mediate the generation, protective functions
resident memory populations in this process. We present the current and maintenance of immune memory have mainly been elucidated in
state-of-the-field for how memory is defined for each lymphocyte line- mouse models, in which the dynamics of an immune response can be
age, including phenotype, function, location, pathways for generation studied following a specific challenge with antigen or infectious patho-
and maintenance, and protective capacity. Finally, current unanswered gen (Fig. 2). A primary response involves the coordinated interactions
questions and challenges to address in the coming years are discussed in of antigen-presenting cells (APCs) with antigen-specific T cells and
the context of the potential to improve human health through targeting B cells in specialized lymphoid organs near the site of infection. For
this remarkable ability of the immune system to remember. example, infections by respiratory viruses trigger immediate innate
immune responses in the lungs, including the production of type I
Immune memory and protection in humans interferons, macrophage activation and the migration of virus-bearing
The functions of immune memory and its roles in health and protec- dendritic cells (DCs) to the lung-associated lymph nodes (Fig. 2a). In the
tion are readily apparent from looking at the impact of ubiquitous lymph nodes, DCs ‘present’ viral antigens on their surface in complex
pathogens over the human lifespan. The immune system of newborns with MHC molecules (known as MHC restriction) that ‘prime’ antigen-
lacks previous exposure and, therefore, immune memory to most specific, naive (antigen-inexperienced) T cells through engagement
pathogens, resulting in increased susceptibility during early life to of the antigen-specific TCR (Fig. 2b), resulting in T cell activation, IL-2
common respiratory, enteric and other types of pathogens. Over the production and subsequent differentiation to various types of effector
course of childhood, multiple exposures and, hence, immune memory cells (Fig. 2c). CD4+ effector T cells and CD8+ effector T cells migrate
result in reduced susceptibility to these pathogens for decades of adult back to the lungs to coordinate virus clearance through the secretion
life, with protection only declining at very advanced ages owing to of pro-inflammatory cytokines and direct killing of virus-infected cells,

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 811

 Review article

a immune response. However, specialized memory T cells, MBCs and

2,500
long-lived plasma cells (LLPCs) are generated that can persist in the
owing to RSV per 100,000
Estimated hospitalizations

2,000 IgG absence of antigen both in circulation and in certain tissues as resident
populations. For example, tissue-resident memory T (TRM) cells are the

Relative IgG level

1,500 main T cell population in mucosal and barrier sites such as the lungs,
people per year

intestines and skin (Fig. 2f), as well as exocrine tissues and the brain7–9.
1,000
Tissue-resident memory B (BRM) cells are found mainly in lymphoid
500
organs and are also present in mucosal sites10,11, whereas LLPCs popu-
Hospitalization rate late the bone marrow, intestines, skin and other sites12 (Fig. 2f,g). These
0 memory populations are poised to respond rapidly and effectively
<1 1 2–4 5–9 10–19 20–44 45–64 65–84 85
Age (years)
to repeat antigen encounters; memory T cells rapidly produce effec-
tor cytokines and cytolytic mediators, MBCs differentiate readily
b
to ASCs, and pre-existing antibodies produced by LLPCs bind directly to
400
influenza per 100,000 people

infectious pathogens and initiate neutralizing and non-neutralizing

clearance mechanisms.
Hospitalizations owing to

300

Relative IgG level Studying memory cells in mice and humans

No data
during flu season

200 IgG
available Several approaches have been implemented to assess the pathways,
processes and mechanisms for the development, functions and main-
100 Hospitalization rate tenance of immune memory; these methodologies continue to evolve
with the advent of new single-cell profiling approaches for transcrip-
0 tome, protein and DNA. In Table 1, we summarize approaches that
<1 1 2–4 5–9 10–19 20–44 45–64 65–84 85
Age (years) have been used to address the central questions in T cell and B cell
memory — to identify distinguishing features of memory T cells and
c
MBCs, to elucidate mechanisms for their generation and maintenance,
400,000
and to carry out in-depth profiling of memory cells in response to
Number of deaths in the
USA owing to COVID-19

their modulation. Immune memory has mainly been characterized in

as of February 2023

300,000
mouse models of infection or immunization with protein antigens to
200,000
generate antigen-specific, memory T cells and MBCs. The persisting
memory populations are measured using functional recall assays in
100,000 which lymphocytes are cultured ex vivo and stimulated with cognate
antigens, to measure cytokine or antibody production (Table 1). Given
0 that memory cell responses for specific antigens are of relatively low
17 9 9 9 4 4 4 + frequency, the use of mice with large proportions of T cells or B cells
0– –2 –3 –4 –6 –7 –8 85
18 30 40 50 65 75 expressing transgene-encoded antigen receptors — including rear-
Age (years)
ranged TCRs or BCRs specific for defined antigens — allows the isola-
Fig. 1 | Immune memory and pathogen susceptibility over the human tion of homogeneous populations of antigen-specific lymphocytes13,14.
lifespan. a,b, Hospitalization rates in the USA owing to infection with respiratory For BCR-transgenic mice, B cell memory development can be tracked
syncytial virus (RSV) (part a) or influenza virus (part b) according to age (dashed directly following immunization14, although the adoptive transfer of
line) and the corresponding level of virus-specific IgG antibodies in circulation B cells is used more commonly as it avoids the effects of pre-existing
(continuous line)2–4. c, The number of deaths owing to COVID-19 in the USA in the
antibodies on memory cell persistence. By contrast, direct immuniza-
indicated age groups up until February 2023. Source data from the US Centers for
tion does not activate TCR-transgenic T cells owing to a large number
Disease Control and Prevention.
of antigen-specific T cells; thus, it is necessary to adoptively transfer
much smaller numbers of TCR-transgenic T cells into congenic mice,
wherein their activation and memory formation are subsequently
respectively. In the lymph nodes, CD4+ T follicular helper cells (TFH cells) assessed following antigen immunization or infection15,16. In addition,
interact with antigen-specific B cells to promote their differentiation mice expressing reporter genes that encode fluorescent proteins or
in specialized structures known as germinal centres (GCs) (Fig. 2e). In surface markers inserted into the locus of a lymphocyte gene of inter-
the GC, antigen-specific B cells undergo extensive proliferation est, such as genes that encode cytokines or transcription factors, allow
and differentiation to antibody-secreting cells (ASCs; plasmablasts and spatial and dynamic tracking of gene expression after activation and
plasma cells), together with somatic hypermutation (SHM) and class- memory formation17–19.
switch recombination (CSR) (which can also occur at a pre-GC stage) to Human samples have been used for studies of memory formation
express alternative antibody isotypes. The resulting secreted antibodies by phenotype analysis and functional recall assays with peptide or
are found in blood plasma, body secretions and tissues (Fig. 2f) and have protein antigens. Although it is not possible to track individual memory
important roles in the clearance of free virus through direct binding and T cells over time in humans, longitudinal profiling of blood can enable
neutralization or by marking virus for destruction by macrophages antigen-specific antibodies and memory T cell and MBC persistence
and serum proteins (known as opsonization). to be assessed. In addition, the maintenance and turnover of human
Following pathogen clearance, most of the expanded immune memory T cells and MBCs can be assessed using labelled nucleotides
effector cell populations die during the contraction phase of the such as BrdU or deuterium that become incorporated into DNA, or with

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 812

Review article

high-dimensional and single-cell technologies (Table 1). Some of these T cell memory
approaches and the insights that they have revealed into both memory Memory T cells have traditionally been defined by their rapid recall
T cells and MBCs are discussed below. function, often manifested in their ability to produce effector cytokines

Lungs

Viral infection

Viral antigen
Lymph node a
Lymphatic circulation

CD4+ BRM cell

b TCM cell Virus
clearance
DC d Blood

TRM cell TEM cell Effector TEM cell CD4+

CD80/ MHC T cell TCM cell ASC
CD86
CD28 TCR
Proliferation
TEM cell IgA

c TRM cell
Naive T cell
(CD4+ or CD8+) Effector T cell
(CD4+ or CD8+)

CD40L CD40 BCR Bone marrow

CD4+ Naive g
e TFH cell B cell IgA IgG

TCR MHC Naive

Affinity B cell
maturation LLPC

Naive MBC
GC MBC T cell
B cell ASC
TRM cell

MBC
Germinal centre
ASC

BRM cell
TRM cell

IgA
f
LLPC

Intestines

Fig. 2 | The generation and localization of immune memory. The figure shows in humans are mainly CD4+, circulate and are found mainly in lymphoid tissue;
the key tissue sites, together with the T cell and B cell subsets, that are involved in effector memory T (TEM) cells, both CD4+ cells and CD8+ cells, are found in the
initiating adaptive immune responses, eliciting effector functions and establishing circulation and in lymphoid and mucosal tissues; tissue-resident memory T (TRM)
memory. a, During viral infection at mucosal sites such as the lung, dendritic cells cells localize mostly to mucosal tissues (lungs and intestines) and barrier sites and
(DCs) take up virus and ‘present’ viral fragments (viral antigens) on their surface are also found in lymphoid organs. e, Naive B cells in the lymph node interact with
in complex with MHC molecules. b, These DCs then traffic in the lymphatic CD4+ T follicular helper (TFH) cells to generate extrafollicular antibody-secreting
circulation to the tissue-draining lymph node, wherein they present viral antigens cells (ASCs), germinal centre (GC)-independent memory B cells (MBCs) and GC
to the T cell receptor (TCR) of naive T cells. c, Naive T cells differentiate into B cells. B cells that enter the GC undergo somatic hypermutation and selection for
effector T cells (both CD4+ cells and CD8+ cells), which rapidly proliferate and affinity maturation of the B cell receptor (BCR) and can differentiate into MBCs or
circulate in the blood throughout various tissue sites to clear the virus by ASCs. f, g, ASCs (both extrafollicular and GC dependent) circulate until they arrive
secretion of pro-inflammatory mediators or by direct killing of infected cells. in tissue niches, wherein they can become long-lived plasma cells (LLPCs). Both
d, Once the virus is cleared, most of these effector T cells die, but a small fraction MBCs and ASCs can be found in circulation, whereas tissue-resident memory B
are maintained as CD4+ memory T cells and CD8+ memory T cells with tissue- (BRM) cells can be found in lymphoid and mucosal tissues, and LLPCs are mostly
specific localizations and migration capacity. Central memory T (TCM) cells, which found in bone marrow and intestines.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 813

Review article

Table 1 | Technologies and approaches for studying immune concepts such as MHC restriction and protective immunity24. Early
memory studies have used the LCMV model to follow the priming of virus-
specific effector T cells and the subsequent development and persis-
Purpose Method Cell type
tence of memory T cells. Using functional recall by limiting dilution
Detection of Functional recall assay T cells and as a readout for virus-specific effector and memory T cell responses,
antigen-specific B cells these studies have shown robust expansion (by more than 1,000-fold)
memory cells
TCR or BCR transgenic mice or cells T cells and of virus-specific CD8+ T cells, differentiation into cytolytic effector
B cells cells coincident with viral clearance 1–2 weeks after infection, and
MHC-multimer technology T cells the persistence of memory T cells at high precursor frequencies
(between 1 in 200 and 1 in 1,000) for more than 1 year after infec-
Activation-induced marker (AIM) assay T cells
tion25 (Fig. 3). A subsequent analysis of virus-specific CD8+ T cells
Tetramer antigen assay B cells using MHC-tetramer reagents containing viral peptide epitopes has
Tracking fate, Adoptive transfer of antigen-specific T cells and shown that 70–80% of total CD8+ T cells were specific for LCMV during
turnover and lymphocytes into congenic hosts B cells the primary response, reducing to ~10% persisting as LCMV-specific
longevity of
memory cells Reporter mouse models for cytokines or T cells and memory CD8+ T cells26. These findings established that CD8+ T cells
transcription factors B cells have a remarkable capacity to develop into high-frequency memory
Molecular fate-mapping assay B cells cells that persist long term.
Following the dynamics of memory CD4+ T cells is more challeng­-
BrdU labelling T cells and
B cells ing owing to their lower frequencies overall. To detect antigen-
specific CD4+ T cells using MHC-tetramer reagents required addi-
Persistence of donor resident lymphocytes T cells and
in transplanted organs B cells tional technical and experimental innovations, including covalently
linking peptide epitopes to MHC molecules and magnetic enrich-
Deuterated water or glucose studies T cells and
B cells ment of tetramer-binding T cells 27. In response to bacterial
infection with Listeria monocytogenes, there was a 100-fold expan-
Retrospective 14C birth dating T cells and
B cells
sion of antigen-specific CD4+ T cells during the primary response,
persistence of memory CD4+ T cells at low frequency, and their
High- Multi-parameter flow cytometry T cells and
gradual decline in number after 1 year post-infection28. Adoptive
dimensional B cells
profiling of transfer of graded numbers of naive, TCR-transgenic CD4+ T cells into
memory cells TCR and BCR sequencing T cells and mouse congenic hosts, followed by priming with their specific (also
B cells
referred to as cognate) antigen (Table 1) to enable tracking of the
Single-cell sequencing T cells and antigen-specific T cell populations, resulted in very low frequencies
B cells
of memory T cells29,30. However, transferring effector T cells that were
CyTOF T cells and activated in vitro with the specific antigen into lymphocyte-deficient
B cells
or intact hosts enabled the persistence of a substantial population
BCR, B cell receptor; CyTOF, cytometry by time of flight; TCR, T cell receptor. of memory CD4+ T cells that could be isolated and studied31. This
approach resulted in the identification of memory T cell heterogene-
ity in terms of lymph node homing receptor (CD62L) expression32,33,
such as interferon-γ (IFNγ) after activation, which contrasts with naive and the testing of precursors and models for memory generation34
T cells that produce mainly IL-2. This functional distinction led to the (see below).
identification of specific phenotypic markers to distinguish putative Human antigen-specific T cells are more challenging to assess
memory T cells from naive T cells based on expression of the adhesion owing to low precursor frequencies and the variability in MHC hap-
molecule CD44 in mice20 and the CD45RO isoform (and lack of CD45RA lotypes in an outbred population. Quantifying human virus-specific
isoform) in humans21. Memory T cells are prevalent in circulation and memory T cells using a functional recall assay as used in mice, in which
also reside in diverse anatomical sites, including lymphoid tissues (the the readout is production of pro-inflammatory cytokines (IFNγ or
spleen, lymph nodes and bone marrow), mucosal and barrier sites tumour necrosis factor (TNF)), revealed that smallpox vaccination
(the lungs, intestines and skin), exocrine sites (pancreas and salivary generates memory CD4+ T cells and CD8+ T cells that can persist for
glands), the liver and the brain8,9,22,23. decades35. The use of MHC-multimer reagents for direct assessment of
However, phenotype and function alone are not sufficient to human antigen-specific T cells was initially implemented for HLA-A2
unambiguously identify memory T cells and study their generation haplotypes, given the significant frequency (30–40%) of this haplotype
and maintenance in response to specific antigens in vivo. New techno­ in the general population36. Using MHC-multimer reagents in combi-
logies for assessing antigen-specific memory T cells together with nation with mass spectrometry techniques such as cytometry by time
mouse infection models (Table 1) have enabled precise quantification of flight (CyTOF) (Table 1) has allowed more-comprehensive studies of
of the dynamics of memory T cell development, from the number of both CD4+ antigen-specific T cells and CD8+ antigen-specific T cells37.
antigen-specific T cells during the primary response to subsequent con- In humans, MHC-multimer reagents have enabled phenotypic and
traction of this population and the persistence of long-lived memory functional profiling of CD4+ memory T cells and CD8+ memory T cells
populations (Fig. 3). in peripheral blood generated from vaccination38,39 and in reservoirs
The lymphocytic choriomeningitis virus (LCMV) infection model of antigen-specific cells in tissues40.
in mice has been particularly informative, not only for understanding The use of MHC-multimer reagents requires knowledge of HLA
memory formation but also for the discovery of key immunological type and previously mapped peptide epitopes, and it is, therefore,

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 814

Review article

feasible only for well-characterized cohorts and pathogens. A method in the absence of co-stimulation through CD28 interacting with CD80
that enables the identification of T cells specific for any antigen with or CD86 on the APC — a pathway that is essential for activating naive
high sensitivity, independently of the HLA type and cytokine secretion T cells — and can be activated by a broader range of APC types54. How-
profiles, involves measuring the early upregulation of activation- ever, although CD28 co-stimulation may not be required for memory
induced markers (AIMs) on the surface of antigen-stimulated T cell activation, mouse memory CD4+ T cells have increased IL-2
T cells (Table 1). AIM assays detect antigen-specific T cells by their production and proliferation in the presence of CD28 engagement55, and
co-expression of multiple early activation markers, including CD69, human memory CD4+ T cells have increased transcriptional responses
OX40, CD40L and 4-1BB (also known as CD137 or TNFRSF9) for CD4+ in the context of CD28 co-stimulation56. During infection, memory
T cells, and OX40, CD69 and CD25 for CD8+ T cells41,42. For example, T cells can also be activated in response to certain innate cytokines
an AIM assay was used to study T cell responses to SARS-CoV-2 dur- (such as IL-12 and IL-18 for CD8+ memory T cells, and IL-1 for CD4+ mem-
ing the recent COVID-19 pandemic when little was known about the ory T cells)57–59, which indicates that there are distinct pathways for the
nature of SARS-CoV-2-specific T cells43. These studies have shown that activation of memory T cells compared with naive T cells. Memory
SARS-CoV-2-specific memory CD4+ T cells and CD8+ T cells could be T cells also have more efficient TCR-coupled signal transduction, in
consistently detected in blood following acute infection and decreased some cases requiring fewer signalling events to achieve a functional out-
in number over the months following infection43,44. Additional studies put60. Therefore, differentiation from naive to memory T cells involves
have shown the persistence of SARS-CoV-2-specific memory T cells at extensive molecular and cellular changes that together promote the
longer times (years) post-infection as a result of cumulative exposures to distinct functional features of the memory response.
infection and vaccination45, which suggests that repeated exposures
can help to establish immune memory. Memory T cell heterogeneity
Although memory T cells exhibit core features as described above, they
Functional properties of memory T cells also have remarkable heterogeneity in phenotype, function, prolifera-
Memory T cell differentiation is accompanied by changes in gene tive capacity, migration and tissue localization (Table 2). Memory T cell
expression, chromatin modifications46, and signalling and activation heterogeneity was originally identified based on differences in expres-
requirements. Differential gene expression analysis originally con- sion of the lymph node homing receptor (CCR7 in humans and CD62L in
ducted by microarrays has shown that antigen-specific memory CD4+ mice), denoting differences in their capacity to migrate to and localize
T cells and CD8+ T cells in mice and humans have distinct transcriptional in lymphoid organs61,62. Notably, the human CD45RA−CCR7+ subset was
profiles compared with naive T cells47,48. Interestingly, gene expression
differences between naive and memory T cells were manifested not in
the resting state but rather following TCR-mediated activation, includ-
Expansion Contraction Memory
ing differences in the upregulation of genes that encode effector and
cytolytic mediators47,48.
In the resting state, memory T cells maintain a type of epigenetic Effector
memory — marked by heritable changes in DNA and chromatin, includ- T cell
ing DNA methylation and histone acetylation or methylation — which Clonal
results in changes to the accessibility of genes for transcription expansion
(‘closed’ or ‘open’ chromatin conformations) that determine
whether a gene is expressed or silenced. Using methods such as DNA Viral load TEM cell TEMRA cell
footprinting and, more recently, ATAC-seq (assay for transposase-
accessible chromatin with sequencing) to measure chromatin acces- TRM cell CD103+
Magnitude

sibility, and ChIP-seq (chromatin immunoprecipitation followed by TRM cell

sequencing) to assess transcription factor binding to DNA promoter CD8+ T cells

regions, studies have shown epigenetic differences between naive
and memory T cells in mice and humans. Notably, in mouse and human
memory T cells, chromatin regions that encode effector molecules
are less methylated and more open than the same regions in naive TEM cell TCM cell
T cells, which allows the rapid upregulation of effector molecules after CD4+ T cells
Naive TRM cell Memory
memory T cell activation49–51. This chromatin remodelling occurs dur- Treg cell
T cell
ing the process of fate commitment to memory T cells52. In this way, key
Time
genes that encode functional recall for memory T cells are maintained
in an epigenetically ‘poised’ state, ready for rapid deployment after Fig. 3 | Memory T cell dynamics. After viral infection, T cells respond in three
phases known as expansion, contraction and memory phases. During the
pathogen encounter.
expansion phase, antigen-specific naive T cells differentiate into effector T cells,
In addition to molecular changes at the level of DNA, memory
which rapidly undergo clonal expansion and secrete effector molecules such
T cells also have cell-surface and intracellular changes that promote their
as pro-inflammatory cytokines and cytolytic effectors. CD8+ T cell populations
rapid responses. Memory T cells have increased expression compared expand to a greater magnitude than CD4+ T cell populations. Once the virus
with naive T cells of multiple integrins and other adhesion molecules is cleared, contraction occurs, whereby most of the effector T cells undergo
that promote interaction with APCs, effectively increasing the avidity of apoptosis. The remaining cells are then maintained as diverse memory T cell
TCR engagement53. This increased propensity for interacting with APCs subsets for long time periods. TCM cell, central memory T cell; TEM cell, effector
results in reduced activation requirements for memory T cells com- memory T cell; TEMRA cell, effector memory T cell re-expressing CD45RA; Treg cell,
pared with naive T cells. For example, memory T cells can be activated regulatory T cell; TRM cell, resident memory T cell.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 815

Review article

Table 2 | Memory T cell subsets

Subset Mouse phenotype Human phenotype Localization Longevity Functions

TCM cells CD44 CD62L

+ +
CD45RA CD45RO CCR7
− + +
Lymphoid tissue 69 days for CD4 +
Produce IL-2; migrate to
cells; 100 days for secondary lymphoid organs to
CD8+ cells coordinate effector responses
to secondary challenge
TEM cells CD44+CD62L− CD45RA−CD45RO+CCR7− Lymphoid and 21 days for CD4+ Produce more effector
mucosal tissues cells; 100 days for cytokines than TCM cells;
CD8+ cells rapidly and effectively migrate
to peripheral tissues to control
pathogen infection
TRM cells CD69+CD103+/− CD45RA−CD45RO+CCR7−CD69+CD103+/− Mucosal tissue Unknown Rapidly produce IFNγ, IL-2
and IL-17; provide localized
protection at the tissue site
TEMRA cells Unknown CD45RA+CCR7− Circulation (the lung, More than 36 years Produce inflammatory
blood and spleen) for CD8+ cells cytokines
TSCM cells CD44−CD62L+ CD45RA+CCR7+CD95+ Lymphoid tissue 500–2,500 days Can generate diverse
for CD4+ cells; memory T cell populations
500–4,500 days for owing to their enhanced ability
CD8+ cells to self-renew
For references, see main text. IFNγ, interferon-γ; TCM cells, central memory T cells; TEM cells, effector memory T cells; TEMRA cells, effector memory T cells re-expressing CD45RA; TRM cells,
resident memory T cells; TSCM cells, stem cell memory T cells.

designated as central memory T (TCM) cells and the CD45RA−CCR7− configuration followed by TCM cells and then TSCM cells72. These findings
subset as effector memory T (TEM) cells, and it was proposed that the suggested a mechanism by which TEM cells can more rapidly produce
lymphoid-homing TCM cell population replenishes the peripheral effector cytokines than other memory subsets. In addition, TCM and
TEM cell subset63. TEM cells differ in their expression of transcription factors regulating
Functionally, TCM cells and TEM cells both produce effector proliferative capacity; TCM cells express higher levels of EOMES and
cytokines, although greater frequencies of TCM cells tend to produce T cell factor 1 (TCF1) than TEM cells, which results in a greater potential
IL-2, and they have a greater proliferative capacity than TEM cells61. for self-renewal73,74.
CD4+ TCM cells are more prevalent than CD8+ TCM cells, and CD4+ CD4+ regulatory T (Treg) cells — which express the lineage-defining
TFH cells (which provide help to B cells for differentiation to ASCs) can transcription factor forkhead box P3 (FOXP3) and other regulatory
also persist as a type of TCM cell and have their own functional profiles, molecules for inhibiting T cell activation such as CTLA4, LAG3 and
including production of IL-4 and IL-21 (ref. 64). By contrast, TEM cells TIGIT — have immunosuppressive functions to minimize tissue dam-
are a major memory subset for both CD4+ lineages and CD8+ lineages, age resulting from infection and the ensuing immune response75,76.
and they can produce a range of effector cytokines, such as IFNγ, TNF, Similarly to the pro-inflammatory T cell subsets, Treg cells can also
IL-17, IL-4, IL-5 and IL-10, and cytolytic mediators such as granzymes have naive and memory phenotypes. Mouse infection models have
and perforin61. A more differentiated state of TEM cells, which express supported the existence of antigen-specific memory Treg cells, dem-
CD45RA (CD45RA+CCR7−) and have high levels of effector cytokine pro- onstrating their long-term persistence in the absence of cognate
duction and cytolytic functions, are TEMRA cells (effector memory T cells antigen77,78. In humans, memory-phenotype Treg cells have been iden-
re-expressing CD45RA). TEMRA cells have been associated with persis- tified in peripheral tissues such as the skin and other sites79,80; how-
tent cytomegalovirus infection in humans65,66, in which context they ever, their antigen specificity and function in protective responses
are found mainly in the blood as CD8+ T cells63. As there is no known remain unclear.
correlate for TEMRA cells in mice, their roles in protection are unclear.
Individual TEM cells also have the ability to produce multiple effector Tissue localization and residency
cytokines and mediators; these are referred to as ‘multifunctional’ The discovery of memory T cell subsets with distinct homing potentials
memory T cells and are associated with vaccine-mediated protection in prompted investigation of their tissue distribution; these studies have
humans67. Another type of memory subset, designated stem cell mem- revealed a wide distribution and persistence of CD4+ TEM cells and CD8+
ory T (TSCM) cells, resemble naive T cells in their CD45RA+CCR7+ phe- TEM cells in multiple lymphoid and non-lymphoid sites in mice30,81. This
notype, but express the memory T cell markers CD95 (also known as dissemination was initially interpreted to signify TEM cell surveillance
FAS) and CD122 (the β-subunit of the IL-2 receptor) and have increased of blood and tissues. However, studies in the past 15 years have estab-
proliferative and self-renewal capacity relative to TCM and TEM cells68. lished the existence of a memory T cell subset localized exclusively
The relative roles of memory T cell subsets have been extensively within tissues, termed TRM cells, that are designated as a separate subset
studied in adoptive transfer and infection models. Both TCM and TEM cell based on their tissue retention and distinct phenotypes and transcrip-
subsets are maintained in vivo and can mediate protective responses tional programmes (reviewed in refs. 9,23,82). TRM cells (either CD4+
to virus rechallenge, although phenotypic conversion between or CD8+) protect barrier sites and other entry points from pathogens.
TCM cells and TEM cells and vice versa can occur69–71. More recently, DNA The original studies identifying tissue residence of memory T cells
methylation analysis has revealed a graded modification of effec- used parabiosis experiments in which mice containing marked popula-
tor cytokine genes, with TEM cells having the most open chromatin tions of memory T cells were surgically conjoined to partner mice to

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 816

Review article

establish a shared circulation. These experiments showed that memory tissue-specific transcriptional and functional adaptations108. For exam-
T cells in the intestines, lung and skin were preferentially maintained ple, TRM cells in human intestines express gut-homing molecules such
in the tissues of the host mice and did not circulate to partner mice83–85. as CCR9 and have increased expression of multiple residency markers
From these studies, canonical phenotypic markers for TRM cells were and IL-17-associated genes compared with TRM cells in the lung, which
defined, which include the activation and tissue-retention molecule express higher levels of regulatory molecules, and in the skin,
CD69 and the epithelium-binding integrin CD103, with CD8+ TRM cells which have a type 2 cytokine profile109.
in certain sites such as intestines and lung having higher levels of CD103 Although the protective capacity of TRM cells cannot be readily
expression than CD8+ TRM cells at other sites and CD4+ TRM cells. tested in humans, evidence from individuals receiving T cell-targeted
TRM cells also have unique transcriptional and epigenetic pro- therapies suggests that TRM cells have key roles in maintaining pro-
grammes that promote tissue residency. The transcription factors tective immunity. For example, patients treated with the potent
HOBIT, BLIMP1, RUNX3 and NOTCH have all been implicated in driv- lymphocyte-depleting drug alemtuzumab had extensive reduction
ing TRM cell generation, tissue retention and maintenance in mice86–88. of circulating T cells, whereas TRM cells in the skin were maintained
Other transcription factors such as T-bet and EOMES, which promote and the patients did not have increased infections in barrier sites110.
the generation of effector T cells, can inhibit TRM cell generation89–91. Mouse models also show that the drug FTY720, an S1PR1 antagonist
At the DNA level, ATAC-seq of mouse TRM cells showed decreased that inhibits the homing and egress of T cells from lymph nodes,
accessibility of genes controlling lymphoid homing (such as Sell, does not affect the TRM cell population111. Moreover, the ability of
which encodes l-selectin) and increased accessibility of genes con- transplant recipients to live relatively healthy lives despite being
trolling tissue residency (such as Itgae, which encodes CD103, and maintained on continuous, systemic immune suppression further
Ccr9, which encodes a chemokine receptor)91. Together, these distinct suggests that memory T cell responses in the tissues are maintained
features of TRM cells at the cellular and molecular levels suggest multiple in these individuals.
layers of control in their fate determination.
The protective efficacy of TRM cells for site-specific infections in Mechanisms for the generation of memory T cells
mice has been demonstrated in many different sites, including the Elucidating the differentiation pathways for memory T cell subsets after
lungs84,92, skin85, female reproductive tract93 and liver94, to diverse viral, naive T cell activation is an active area of discussion and research. His-
bacterial and eukaryotic pathogens95. For example, TRM cells in the lung torically, attempts to address this issue have been limited to studying
provide protection against a wide range of viruses, such as influenza memory T cells in bulk as tracking single precursor cells is challenging.
virus84,92, respiratory syncytial virus96 and SARS-CoV97, and diverse There is no consensus as to the factors that determine which effector
bacterial pathogens, including Bordatella pertussis and Mycobacterium T cells survive the contraction phase of the primary response and dif-
tuberculosis98,99. Similarly, in the skin, TRM cells protect against herpes ferentiate into memory T cells. In mouse models of LCMV infection,
simplex virus100, vesicular stomatitis virus93, and bacterial and fungal the CD8+ effector T cells that survive have increased expression of IL-7
infections101; in the liver, TRM cells protect against malaria and hepati- receptor-α (IL-7Rα) that enables them to respond to the important
tis94. The role of TRM cells in protection against diverse pathogens at survival cytokine IL-7 (ref. 112). Other survival factors for memory cell
multiple sites in mice strongly suggests that they are likely to have a generation that were first characterized in CD8+ T cells include CD27,
similar function in humans. the anti-apoptotic protein BCL2, and lack of expression of the killer
The study of human tissues from organ donors and surgical cell lectin-like receptor G1 (KLRG1)112–114. However, fixed expression of
explants has enabled the identification and characterization of human IL-7R or increased BCL2 expression is not sufficient to promote memory
TRM cells in multiple tissues, including cross-tissue comparisons of cell generation in CD8+ T cells115, which suggests that a combination
multiple sites from individual organ donors. Notably, most memory of intrinsic and extrinsic factors may determine memory T cell fate.
T cells in human mucosal tissues and skin express the TRM cell marker The lineage relationships between T cell subsets have been dif-
CD69, whereas memory T cells in the blood do not102,103. The levels ficult to assign and several models have been proposed. Results from
of CD69 and CD103 expression by human TRM cells are a feature of the adoptive transfer experiments and from in vivo fate-mapping models
tissue; the intestines contain the highest proportion of CD69+CD103+ show that memory T cells can develop from different types of acti-
TRM cells, followed by the lung, skin and spleen; even lymph nodes vated T cells, supporting multiple pathways for memory formation.
were found to have a proportion of CD69+ memory T cells with a The linear differentiation model proposes that TEM cell and TCM cell
TRM cell phenotype102–105. These findings indicate that the molecules subsets originate from effector T cells and that TEM cells can convert
controlling tissue localization and residency of memory T cells are to TCM cells over time (Fig. 4a). In support of the linear model, in vivo
highly conserved between mice and humans. fate-tracking studies in mice showed that memory CD4+ T cells and
Furthermore, human TRM cells have a core transcriptional, pheno- CD8+ T cells preferentially derived from IFNγ+ and effector cell precur-
typic and functional signature in lymphoid and mucosal tissues that is sors18,116, and that transfer of TEM cells resulted in eventual conversion to
homologous to that of mouse TRM cells106. Compared with circulating TCM cells71. The bifurcative differentiation model proposes that memory
TEM cells, CD69+ TRM-phenotype cells in humans have increased expres- and effector subsets are generated divergently from naive T cells117
sion of the integrins CD103 and CD49a and of the chemokine receptor (Fig. 4b). This model is supported by studies showing that the adoptive
CXCR6 to promote homing to tissue sites, but have decreased expres- transfer of antigen-specific T cells that are activated but lack effec-
sion of molecules that enable tissue egress such as the sphingosine- tor function can result in memory cell generation34,118 and that IFNγ−
1-phosphate receptor (S1PR1)106. Functionally, human TRM cells rapidly T cells can develop into memory T cells in fate-mapping models18. The
produce IL-2 and the pro-inflammatory cytokines IFNγ, TNF and IL-17 in divergent generation of effector T cells and memory T cells could be
response to activation, and they also express the immunoregulatory explained by asymmetric cell division (ACD) early after T cell activation,
molecules PD1, CD101 and IL-10, which may function to limit tissue a process whereby cytoplasmic proteins are unequally apportioned
damage during in situ responses106,107. Furthermore, TRM cells have to daughter cells, resulting in distinct fates119. ACD can be influenced

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 817

Review article

by the strength of the TCR stimulus and the localization of signalling to follow the generation of memory to an infection in vivo in humans,
intermediates, with increased signalling favouring effector cells and sampling tissues from the earliest life stages provides a unique oppor-
decreased signalling promoting memory cell fate and self-renewal tunity to observe the dynamics of TRM cell establishment. In lungs and
manifested by the maintenance of TCF1 expression120,121. These findings intestines obtained from infant and paediatric organ donors, the rapid
suggest that multiple cell fates can derive from T cell priming events accumulation of CD69+ TEM cells (both CD4+ cells and CD8+ cells) was
and are directed by the differential expression of fate-determining observed; however, these early memory T cells lacked effector cytokine
transcription factors. production, had reduced expression of other TRM cell markers, and had
In mice, a single, naive CD8+ T cell can give rise to heterogeneous a stem cell-like transcriptional programme127. Fully mature and func-
memory T cell populations after activation and clonal expansion, tionally protective TRM cell populations gradually developed in these
supporting a progressive differentiation model47,122,123 (Fig. 4c). Precur- mucosal sites over the first 4–5 years of life, suggesting that TRM cell
sor frequency can affect the generation of TCM cells versus TEM cells16, development occurs in progressive stages127. Analysis of the TCR rep-
which suggests that different antigen specificities may contribute to ertoire by deep-sequencing (TCR-seq) or single-cell profiling has
the heterogeneity of memory cell subsets. A differentiation hierarchy revealed tissue-specific clonal expansions within TRM cell populations
from TSCM cells to TCM cells and then TEM cells is supported by studies in the intestines, skin and lung, having little overlap with circulating
of human CD8+ T cells showing that TCM cells can adopt a TEM cell phe- memory T cell subsets109,128. Together, these findings provide evi-
notype when activated61, with epigenetic profiling showing increased dence for the tissue-directed generation of resident memory subsets,
chromatin accessibility72. The pathways and mechanisms for the together with the dissemination of circulating memory subsets to
generation of TRM cells are even less well-understood. Mouse studies surveil the body.
suggest that TRM cells are generated from activated effector T cells Memory T cell generation also involves changes in metabolic
and/or TEM cell precursors that migrate to tissue sites, with signalling profiles to meet the energy demands (in the form of ATP) for each
by certain cytokines, including transforming growth factor-β and IL-2 functional state (reviewed in ref. 129). Activated T cells use glycolysis for
(refs. 124,125), and expression of the transcription factors Hobit and the rapid mobilization of effector functions, whereas memory T cells
Blimp1 (refs. 87,126) having crucial roles in promoting tissue residency rely more on mitochondrial oxidative phosphorylation130. Memory
and adaptation in situ in mucosal and barrier sites. T cells have a greater mitochondrial mass than naive T cells, indica-
Memory T cell generation in humans and the lineage relationships tive of greater energy capacity131, and cytokines that promote the
between memory subsets remain difficult to study; however, advances maintenance of memory T cells, such as IL-15, increase mitochondrial
in sampling and in high-dimensional profiling of the transcriptome biogenesis132. In addition, different T cell subsets use distinct substrates
and TCR repertoire have provided new insights. Although it is difficult as energy sources; both TEM cells and TRM cells use fatty acid uptake to a

a Linear differentiation model b Bifurcative differentiation model

CD80/ Effector T cell TEMRA cell

CD86 CD28
Homoeostatic
TEMRA cell
proliferation

MHC TCR Naive Effector Naive TEM cell

APC T cell T cell APC T cell
Homoeostatic
proliferation
TEM cell TCM cell

TCM cell
c Progressive differentiation model
?

TRM cell ? CD103+

Homoeostatic TRM cell
proliferation

Naive Effector TSCM cell TCM cell TEM cell TEMRA cell
APC T cell T cell

Fig. 4 | Models for the generation of memory T cells. a, The linear naive T cell can differentiate into heterogeneous memory T cell populations
differentiation model proposes that effector memory T (TEM) cells and central and that memory T cell precursors (stem cell memory T (TSCM) cells) gradually
memory T (TCM) cells originate from effector T cells, with TEM cells being differentiate into longer-lived memory T cells, such as resident memory
precursors for TCM cells. b, The bifurcative differentiation model proposes T (TRM) cells and effector memory T cells re-expressing CD45RA (TEMRA cells).
that effector T cells, TEM cells and TCM cells are generated divergently from For references, see text. APC, antigen-presenting cell; TCR, T cell receptor.
naive T cells. c, The progressive differentiation model proposes that a single

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 818

Review article

greater extent than TCM cells39. These different substrates and metabolic serum can treat individuals with diphtheria or tetanus. The discovery
pathways are important potential targets for memory modulation in a of antibodies as γ-globulins in 1939 and of plasma cells in 1948, and
subset- or tissue-specific manner133. decades later, studies in mice immunized with sheep red blood cells or
haptens, led to the discovery that MBCs bearing surface immunoglobu-
Maintenance of memory T cells lins are functional precursors to ASCs148,149. However, the study of B cell
One of the defining characteristics of immune memory is that it is main- memory has progressed in different ways to the study of T cell memory,
tained in the absence of antigen. The smallpox (vaccinia virus) vaccine owing to a previous lack of phenotypic markers and tracking tools for
allows us to study human immune memory in this context as smallpox B cell memory components, and the use of antigen-priming models
infection has been eradicated owing to vaccination. Vaccinia virus- rather than infection models. Moreover, the study of humoral immune
specific CD4+ T cells and CD8+ T cells have been detected in human blood memory resulting from vaccination or infection in humans has tended
75 years after vaccination35, indicating the potential of memory T cell to focus on antibodies and their neutralizing capacity rather than MBCs.
populations to persist for a lifetime. Maintenance of memory T cells in B cell-derived memory comprises both cellular and humoral com-
the absence of antigen is associated with homoeostatic turnover, which ponents, in the form of MBCs, LLPCs and antibodies found in plasma
can be driven by cytokines and cognate interactions. In mice, CD4+ mem- and extracellular fluids. Antibodies can directly recognize extracellular
ory T cells require cytokines such as IL-7 and IL-15 to survive in vivo134,135, pathogens and mediate protection by neutralization, opsonization and
and cognate interactions with MHC are dispensable for CD8+ memory complement activation. Immune memory to infection and vaccination
T cell persistence but required for CD4+ memory T cells136,137. CD4+ mem- as assessed by measuring antibody levels in serum can be long-lived;
ory T cells also require TCR signalling for homoeostasis138, suggesting measurements of antibody titres to many childhood vaccines show that
that active processes are involved in memory T cell persistence. humoral immunity can persist for decades and up to the lifetime of an
In humans, current understanding of the lifespan and turnover of individual150. During a primary response, antigen-specific activation of
memory T cell subsets is derived mostly from the sampling of periph- naive B cells leads to the generation of ASCs and MBCs. Short-lived ASCs
eral blood. Deuterium labelling of dividing cells can be used safely disappear once the pathogen is removed, but some ASCs can transfer
in humans139, analogous to BrdU labelling studies in mice (Table 1), to tissue niches, mostly in the bone marrow, gut and spleen, and dif-
and turnover dynamics can be inferred by fitting to mathematical ferentiate into LLPCs that continue to produce antibody long after
proliferation models. Deuterium labelling studies have shown that pathogen clearance (Fig. 5). Studies of MBCs and LLPCs using novel
memory T cell subsets are heterogeneous in their rates of turnover. reporter mice and techniques such as time-stamping or fate-mapping
CD4+ TCM cells (lifespan = 69 days) turnover less frequently than CD4+ have enabled the origin and cellular fates of these cells to be tracked,
TEM cells (lifespan = 21 days) when using the same mathematical showing that LLPCs are mainly derived from GC B cells but also can be
model140,141. Both CD8+ TCM cells and CD8+ TEM cells have a lifespan of differentiated from non-GC B cells151,152 (Table 1).
approximately 100 days141,142, whereas CD8+ TSCM cells have a longer
lifespan of 500–4,500 days, or up to 9 years143 (Table 2). These turno- Functional properties of B cell memory
ver data indicate that the less differentiated memory T cell subsets Traditionally, MBCs were distinguished from naive B cells by the anti-
(in other words, TSCM cells and TCM cells according to the progressive body isotype they express and their enhanced ability to respond to
differentiation model) have the longest lifespan. Deuterium labelling cognate antigens. MBCs were defined as class-switched B cells with high
has also been applied to longitudinal sampling of vaccine-specific affinity for antigen that can differentiate to ASCs after antigen encoun-
T cell memory generated by the live-attenuated yellow fever vaccine38. ter more rapidly than naive B cells153. However, subsequent studies
This analysis has shown the generation of long-lived, vaccine-specific have identified that somatically mutated MBCs expressing IgD and IgM
memory T cells with doubling times of 450 days, which arise from early isotypes (not yet having undergone CSR) can also be found in human
proliferation events 2 weeks post-vaccination38. peripheral blood and during infection in animals154–156. MBCs in mice
The lifespan of TRM cells is unknown as it is difficult to track indi- and humans can be quiescent and maintained for long periods157,158.
vidual TRM cells not only longitudinally but also spatially. In mice, main- Thus, currently, MBCs are mostly defined as long-lived, antigen-specific
tenance of TRM cells can be variable; in the lung, CD8+ TRM cells decline B cells that are generated during antigen exposure but remain quiescent
over time144, whereas CD4+ TRM cells are more stably maintained111, and and can respond to re-encounter of their cognate antigen159.
non-lymphoid TRM cells can migrate back to lymph nodes and estab- MBCs have a greater affinity for antigen and can respond to lower
lish lymphoid residency145. Studies of human TRM cells in the context antigen doses than naive B cells, owing to SHM and their increased
of organ transplantation (Table 1) provide compelling evidence for ability to receive signals from TFH cells. MBCs can also have broader
their stable maintenance. For example, donor-derived TRM cells were responses to antigen than naive B cells, enabling their direct differentia-
detected in the bronchoalveolar lavage of lung transplant recipients tion to ASCs in response to variant pathogen strains as shown in mouse
for more than 1 year post-transplant146, and they could be detected models152. In humans, MBCs begin proliferating earlier than naive B cells
in liver transplant recipients for as long as 11 years post-transplant147. in response to T cell-dependent and -independent stimuli. This rapid
TCR repertoire analysis has revealed the presence of expanded TRM cell response of MBCs was due to cell-intrinsic differences, such as higher
clones that are segregated within the lungs, intestines or skin109, levels of gene expression related to activation, co-stimulation and
which further suggests that TRM cells are maintained in situ. Elucidat- survival (such as the genes that encode the co-stimulatory molecules
ing the mechanisms by which memory T cells persist in tissues will be CD80 and CD86, TNFRSF13B (also known as TACI), and BCL2 family
important for targeting these resilient cells for immune protection. proteins (regulators of apoptosis)) and lower levels of gene expression
related to cellular quiescence (such as the genes that encode KLF4,
B cell memory KLF9 and PLZF)160–162. There are also epigenetic differences between
The notion of humoral immune memory dates back to 1890, when MBCs and naive B cells, as MBCs have more-accessible chromatin near
Emil von Behring and Kitasato Shibasaburō discovered that human to ASC-specific genes in both mice and humans163.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 819

Review article

Low affinity Re-infection

GC-independent MBC
FDC

CD40L CD40 GC
Antigen TFH cell
BCR Re-infection
IRF4+
BACH2+
BCL6+
Intermediate
TCR MHC affinity
TFH cell Naive GC B cell GC-dependent
B cell MBC

Re-infection

High affinity IRF4++

BLIMP1+
XBP1+

ASC

Antibody

LLPC

Fig. 5 | B cell memory differentiation. Once naive B cells become activated antigen that receive the weakest signals tend to differentiate to GC-independent
through their B cell receptor (BCR) by an antigen-presenting cell such as a MBCs. Activated B cells that do not differentiate into short-lived ASCs or GC-
follicular dendritic cell (FDC) and receive signals from T follicular helper (TFH) independent MBCs can seed GCs and differentiate to GC-dependent MBCs or
cells through CD40L–CD40 interactions, naive B cells differentiate along ASCs. After re-infection, GC-dependent and GC-independent MBCs can (re-)
three main pathways — to germinal centre (GC)-independent memory B cells enter GCs for affinity maturation or rapidly differentiate into ASCs. Some ASCs
(MBCs), extrafollicular antibody-secreting cells (ASCs), or GC B cells (which (both extrafollicular and GC-dependent) can become long-lived plasma cells
can themselves differentiate to GC-dependent MBCs or ASCs). B cells with (LLPCs) once they reach specific tissue niches such as the bone marrow.
high affinity for antigen and that receive strong signals from TFH cells tend TCR, T cell receptor.
to differentiate to extrafollicular ASCs, whereas B cells with low affinity for

Another distinct property of MBCs is they can re-enter the GC to class-switched, neutralizing antibodies to viral spike protein owing to
undergo further SHM and become ASCs that produce higher-affinity antibody feedback170. A potent and protective antibody response
antibodies, as shown using fine needle aspirates of human lymph nodes to the initial antigen encounter can also lead to a phenomenon known as
from individuals vaccinated against influenza164. However, experiments ‘original antigenic sin’, whereby the primary antibody response influences
in mice that used parabiont pairs to track the recruitment of naive subsequent responses to the boosting antigen (Box 1). LLPCs in the bone
B cells and MBCs into the GC, showed that after re-immunization, marrow of mice and non-human primates can persist long term following
re-entry of MBCs into GCs is infrequent and secondary GCs are mostly infection or vaccination without turnover and in the absence of MBCs167,171.
derived from newly primed naive B cells165. These findings suggest
either differences between humans and mice and/or that multiple Memory B cell phenotypes and heterogeneity
factors affect the fate of MBCs, such as antibody feedback (see below). Unlike for memory T cells, specific phenotypic markers that distinguish
In contrast to MBCs, LLPCs are not quiescent and continually pro- MBCs from naive B cells have been challenging to identify. Several
duce antibodies in a manner independent of antigen persistence166–168. canonical markers have been defined, such as isotype (IgD−) and CD38
These antibodies can provide so-called ‘sterile protection’ if their expression for mouse MBCs155,159,172 and CD27 expression for human
affinities and titres are sufficient, manifested as a complete preven- MBCs173, but additional markers are needed for the complete delinea-
tion of re-infection. These pre-existing antibodies can also influence tion of B cell memory. Recently, unbiased flow cytometry screens have
B cell responses in distinct ways depending on their affinities and titres. been used to discover markers for mouse and human MBCs by com-
As shown in mouse models of SARS-CoV-2 infection, high-affinity or paring naive B cells with antigen-specific mouse MBCs or CD19+CD27+
high-titre antibodies can inhibit the differentiation of naive B cells to splenic MBCs obtained from organ donors174. This study has defined a
become GC B cells, whereas low-affinity antibodies can enhance naive combination of two markers as distinguishing MBCs from naive B cells
B cell recruitment to the GC169. Similarly, individuals who received thera- in mice (CD180−CD267+) and humans (CD200−CD11a+), although there
peutic administration of monoclonal antibodies to SARS-CoV-2 gener- were some differences in expression levels of these markers between
ated lower-affinity antibodies to COVID-19 mRNA vaccines with fewer individuals and tissue sites174.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 820

Review article

Subtypes of mouse MBCs were originally classified by antibody and tonsil, but not in cord blood179,180 (Table 3). Also, CD27+IgD+ B cells
isotype, such as the expression of IgM, IgG or IgA. Newer studies have could produce antibodies after in vitro stimulation whereas CD27−IgD+
further found that for a specific antigen, the isotype expressed by the B cells could not181. Subdividing MBCs based on the level of CD27
MBCs determined the fate of the MBCs after recall or re-infection175,176. expression revealed that CD27hi MBCs are GC dependent, whereas
Subsequent studies have identified PDL2 (a ligand for the T cell CD27low MBCs are mostly independent of T cells and GCs, are more long-
inhibitory molecule PD1) and CD80 (a ligand for the co-stimulatory lived and frequent in infants, and have less mutated BCRs and more fate
receptor CD28 on T cells) as markers for functional subsets of MBCs177 plasticity than CD27hi MBCs182. Notably, vaccinia virus-specific MBCs
(Table 3). In particular, PDL2+CD80+ MBC subsets exhibit surface detected in the spleen of individuals who were vaccinated against small-
immunoglobulin having a higher number of mutations and with higher pox more than 40 years ago had a CD27+IgG+CD21+CD73+ phenotype
affinity, and more rapidly differentiate into ASCs after re-challenge, and were enriched in the CD20hiCD21hiIgG+ memory subset183.
whereas PDL2−CD80− MBC subsets have fewer or minimal mutations Although CD27 and CD21 expression can consistently demarcate
and more functional plasticity, and can seed GCs177. Interestingly, 50% of human classical MBCs, a subset of ‘atypical MBCs’ lacking CD27 and
PDL2+CD80+ MBCs were IgG+, whereas PDL2−CD80− MBCs were mostly CD21 expression and expressing the Fc receptor-like molecules, FCRL4
IgM+ (ref. 177), which indicates that isotype expression and interactions and FCRL5, has been identified in human blood184–186 (Table 3). Atypical
with T cells influence MBC function and capacity for differentiation to MBCs are found in individuals with chronic infection such as with HIV or
ASCs. Relatively recently, an ‘age-associated’ B cell subset was identi- hepatitis C virus, or with autoimmune diseases such as systemic lupus
fied based on expression of CD11c and downregulation of expression erythematosus, and these MBCs tend to be anergic and less activated or
of CD21 (a component of the B cell co-receptor complex), CD23, CD95 proliferative than classical MBCs187–189. More recently, single-cell trans­
and CD43 (ref. 178) (Table 3). B cells with this age-associated phenotype criptome analysis of human MBCs in peripheral blood from healthy
also tend to increase in number in the context of autoimmunity and individuals has also identified a CD19+CD27−IgD− subset, express-
chronic infection, and they can be distinguished from other types of ing TBX21 (which encodes T-bet) and ITGAX (which encodes CD11c)
MBC as they do not respond to BCR signalling178. but lacking CD21 expression, which is a property of age-associated
For classical MBCs in humans, CD27 expression reliably distin- B cells in mice; this subset also showed properties of precursors for
guishes antigen-experienced cells from naive B cells in peripheral blood GC-independent ASCs190. By contrast, another study has shown that

Box 1

Original antigenic sin

Original antigenic sin (OAS) — “the induction of a more robust that they derive from influenza-specific MBCs and naive B cells
immune response to the priming versus a boosting immunogen with lower affinity, which indicates that MBCs from an earlier
that itself binds poorly, if at all, to antibodies induced by the priming infectious challenge or vaccination can participate in the response to
immunogen” — was discovered by Thomas Francis in 1960 in his a new challenge, potentially inhibiting the generation of new clones
study of serum responses to influenza vaccine245. OAS is thought of MBCs or ASCs164. This process could be beneficial for pathogens
to occur owing to the presence of pre-existing antibodies that with minimal antigenic variations to maintain pre-existing memory. In
sequester or eliminate the boosting immunogen and, thus, suppress addition, OAS will not occur if the pathogen is highly mutable and the
the activation of naive B cells and differentiation of new memory mutated epitopes are sufficiently different from the original strain so
B cells (MBCs) or antibody-secreting cells (ASCs). This suppression as not to interact with pre-existing antibodies or MBCs.
of new responses could be a concern for vaccination against rapidly Long-lasting GCs may be able to modify OAS-induced
mutating viruses such as SARS-CoV-2 (ref. 246), but in this case, suppression of new clones, as the persistence of GCs correlates with
elucidating the mechanism of suppression has been difficult because the extent of somatic hypermutation (SHM) of MBCs and ASCs, which
it is not possible to determine the exact cellular origins of secreted increases the clonal diversity of these cells and, thus, the potential
antibodies. to respond to boosting immunogens. In non-human primates
Recently, a molecular fate-mapping method has been developed vaccinated with HIV envelope protein, GCs could last for more
that enables antibody production to be linked to a cohort of specific than 6 months with booster vaccinations, and MBCs and antibodies
B cells and, thus, provides insights into the mechanism of OAS247. generated by these conditions had high levels of SHM and could
The study showed that when the same antigen is used for both recognize non-immunodominant epitopes248. A similar scenario was
priming and boosting, the cohort of MBCs induced by priming is shown for SARS-CoV-2 vaccination, as a third vaccine dose both
the main source of antibodies produced following boosting. This expanded persisting MBC clones and generated newly detected
‘priming addiction’ effect decreases when there is greater antigenic clones, and these new clones could protect against newer Omicron
distance between priming and boosting antigens, and it was variants of the virus that were not included in the vaccine249–252.
shown that SARS-CoV-2 vaccination with a WH1 (ancestral) strain Designing vaccines to broaden the specificity of immune responses
mRNA vaccine followed by a BA.1 (variant) strain vaccine generated and to avoid OAS would be advantageous for generating protection
immune memory to both strains in mice. Analysis of human to rapidly mutating pathogens and is an important area for
germinal centres (GCs) generated after influenza vaccination found future research.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 821

Review article

FCRL5+T-bet+ MBCs have effector-like properties in terms of influenza respiratory infections are prevalent. However, these structures tend to
vaccine-generated immunity191. In addition, CD27−IgA+ MBCs were be transient and may not support the long-term persistence of MBCs.
identified in individuals with CD40L deficiency (which disrupts B cell– BRM cells expressing CD69 that are analogous to TRM cells have been
TFH cell interactions), indicating that atypical MBCs could be generated identified in mouse lungs following influenza virus infection and are
by a T cell-independent and GC-independent pathway192. Recently, retained in the lungs in parabiosis experiments172. Influenza-specific
CD21lowCD27− MBCs were identified in individuals who had been BRM cells in the lungs required local antigen encounter and T cell help
infected with SARS-CoV-2 or received COVID-19 mRNA vaccines193,194. via CD40–CD40L interaction for their maintenance and they rapidly
It is unclear whether atypical B cells derive from aberrant or termi- differentiated into ASCs after antigen re-challenge and localized to the
nal differentiation, or whether they drive a form of immunopathol- infection site172,203. Mouse lung BRM cells generated in response to influ-
ogy, but it is suggested that atypical MBCs could be generated in an enza virus infection have distinct trans­criptional signatures compared
inflammatory milieu152. with MBCs in lymphoid organs, with lower levels of expression of Sell and
Markers for LLPCs are not as well-defined, although mouse ASCs S1pr1 (ref. 204), similar to TRM cells. BRM cells have important functional
are identified as B220loCD138+SCA1+BLIMP1+ cells195. Time-stamping roles in situ; for example, IgA+ BRM cells in the lung and local IgA secretion
mouse models have revealed new insights into LLPC phenotypes, were crucial for protection against secondary influenza virus infection205.
including the increased survival of antigen-specific B220loMHC-IIlo Optimal protection against site-specific infection, therefore, involves
plasma cells compared with B220hiMHC-IIhi plasma cells168, and both TRM cells and BRM cells.
more fine-grained phenotypic distinctions (reviewed in ref. 151). Human MBCs also persist as tissue-resident populations that
In multi-omics single-cell profiling, CD102 was suggested as a marker express CD69. Human BRM cells have been detected in lymphoid
of LLPCs in non-human primates196. Human LLPCs are mostly defined as organs (the spleen, lymph nodes and tonsils) and in mucosal and
CD27+CD138+ cells and, in the case of virus-specific LLPCs, as express- barrier sites, including the gut, lungs and skin11,199. Human BRM cells
ing high levels of CD38 but not expressing CD19 (ref. 197). Further share a core transcriptional signature with human TRM cells and express
high-dimensional profiling at the single-cell level, as well as within tissue-specific transcriptional profiles, particularly in the gut, lung
tissues, will reveal additional markers for studying the dynamics of and spleen11,206. Although MBCs can become resident and persist in
these populations. multiple sites, lymphoid organs seem to be the reservoirs for long-lived
vaccine-specific memory207, and the gut, spleen and bone marrow are
Tissue localization and residency reservoirs for LLPCs151,208. The LLPC niche is a crucial factor in determin-
Most MBCs are found in lymphoid and non-lymphoid peripheral ing the survival and maintenance of these cells. For example, plasma
tissues, including barrier and mucosal sites in mice and humans, cells in the bone marrow have longer half-lives than those in the spleen
wherein they can rapidly proliferate or differentiate into ASCs after or gut, indicating that location can affect the lifespan of LLPCs209. Thus,
antigen re-exposure11,198. MBCs in barrier sites such as the skin express B cell memory is stored in tissues and reservoirs of LLPCs maintain
CCR6 (receptor for CCL20), and gut MBCs express CCR9 and CCR10 circulating antibody levels. The precise role of human BRM cells in
(ref. 199). Circulating MBCs can also be recruited to infected barrier protection against infection and pathways for their generation and
sites and differentiate into ASCs in situ, but the majority of memory maintenance are unknown and are important areas for future study.
B cells reside in secondary lymphoid organs200. MBCs can also populate
tertiary lymphoid structures within mucosal tissues, such as the induc- Mechanisms for the generation of B cell memory
ible bronchus-associated lymphoid tissue (iBALT) that occurs in mice B cells develop in the bone marrow and seed the periphery with
following infection201 and in humans during the early years of life202 when transitional subsets that develop into naive B cells. Naive B cells

Table 3 | Memory B cell subsets

Subset Mouse phenotype Human phenotype Localization Longevity and turnover Functions

Classical PDL2+CD80+ CD19+CD27+IgD− or Lymphoid tissue (GC Doubling time of ~26 days Class-switched; rapidly
(human), CD19+CD27+IgD+/IgM+ dependent), also and half-life of ~18 days in become ASCs after
double- detected in blood blood (human); half-life re-infection (mice);
positive of 402 days in the spleen mostly IgG+ (mice)
(mouse) (mouse)
MBCs
Atypical PDL2−CD80− CD19+CD27−IgD− (CD21−T-bet+) Lymphoid tissue Doubling time of ~152 days Tend to seed GCs after
(human), (GC independent), also and half-life of ~22 days in re-infection (mice);
double- detected in blood blood (human) mostly IgM+ (mice);
negative some are age-associated
(mouse) or autoimmune-related
MBCs (human)
BRM cells CD69+CXCR3+CCR6+ CD69+ Lungs, intestine, Unknown Relocate to infected
lymphoid tissue, bone sites and rapidly
marrow, tonsil and differentiate into ASCs
thymus
LLPCs B220loMHC-IIlo CD19−/loCD138+CD27+CD38hi Bone marrow, intestine Decades (human) Gut LLPCs are mostly
(human) IgA+
ASC, antibody-secreting cell; BRM cells, resident memory B cells; GC, germinal centre; LLPCs, long-lived plasma cells; MBCs, memory B cells.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 822

Review article

Glossary

Antibody isotypes Complement MHC-tetramer T follicular helper cells

The designation of an immunoglobulin A set of plasma proteins that can kill A complex of four antigen- (TFH cells). A CD4+ T cell subset found in
chain in respect to its constant regions. extracellular pathogens directly or coat specific, peptide-bound MHC lymphoid follicles that promotes B cell
Light chains can be of either κ or λ (opsonize) them so that they can be molecules attached to a differentiation through interaction of the
isotype. Heavy chains can be of μ, δ, γ, α removed by phagocytes. streptavidin-coated bead. co-stimulatory molecule CD40L with
or ε isotype, which determines IgM, IgD, CD40 on B cells and by producing the
IgG, IgA or IgE antibodies, respectively. Cytometry by time of flight Neutralization cytokine IL-21.
(CyTOF). An application of mass Inhibition of the infectivity of
Antigen-presenting cells cytometry to study protein expression a virus or the toxicity of a toxin Time-stamping
(APCs). Immune cells such as dendritic at the single-cell level. molecule by the direct binding A technology of in situ lineage
cells, macrophages, Langerhans cells of antibodies. tracing using the tamoxifen-inducible
and B cells that process and present DNA footprinting Cre-recombinase system.
antigens on MHC molecules for A method to study whether a protein Opsonization
recognition by the T cell receptor or binds to a specific DNA region of The coating of the surface of a Unfolded protein response
B cell receptor of T cells or B cells, interest. pathogen by antibody and/or A cellular stress response that is
respectively. complement that makes it more activated when unfolded proteins
Fate-mapping readily ingested by phagocytes. accumulate in the endoplasmic
ATAC-seq Application of technologies to ‘mark’ reticulum in the cell.
Assay for transposase-accessible cells and track their fate and the fate Retrospective 14C birth dating
chromatin with sequencing, which is of their progeny. A process through which
a method to identify open, accessible levels of 14C in cellular DNA are
chromatin regions using mutated Germinal centres measured and their chronological
hyperactive transposase. (GCs). Sites of high-level B cell age calculated based on known
proliferation and differentiation that levels of atmospheric 14C, which
ChIP-seq develop in lymphoid follicles during increased in the 1960s and 1970s
Chromatin immunoprecipitation an adaptive immune response. because of nuclear weapons
followed by sequencing, which is a testing and declined thereafter.
genome-wide method to identify DNA Immunosenescence
binding sites for transcription factors Immune dysfunction that occurs with Somatic hypermutation
and other proteins. ageing, associated with defects in cell (SHM). A mechanism by which
function and quantity. mutations are induced in the
Class-switch recombination variable-region DNA of rearranged
(CSR). A somatic gene recombination MHC restriction immunoglobulin genes to produce
process in activated B cells that replaces A property of T cells whereby the variant immunoglobulins, some
the heavy-chain constant region of an T cell receptor recognizes antigen only of which bind antigen with
immunoglobulin with one of a different when bound to an MHC molecule on an a greater affinity.
isotype, for example inducing a switch antigen-presenting cell as a peptide–
from IgM or IgD to IgG, IgA or IgE. MHC complex.

circulate in the body until they encounter antigen, mostly in second- MBCs or ASCs after extensive clonal expansion and SHM. Most ASCs
ary lymphoid organs (Fig. 2e). Once these naive B cells are activated (both extrafollicular and GC-dependent) undergo apoptosis after a
by antigen through their BCR, they can migrate to the T–B border in few weeks, although a proportion can circulate and become LLPCs
the lymphoid organ, wherein they interact with TFH cells that were that home to survival niches, such as in the bone marrow, spleen
previously activated by cognate antigens on dendritic cells (Fig. 5). and gut168,215.
Depending on the signals received from TFH cells, activated B cells can A key transcription factor that determines the differentiation
differentiate along three distinct pathways: to extrafollicular ASCs, fate of activated B cells is IRF4, which is expressed in response to BCR
to GC-independent MBCs or to GC B cells, which later become GC- signalling and co-stimulation through the CD40–CD40L interaction
dependent MBCs or ASCs152,210. Different activation conditions favour with T cells216. Induction of IRF4 leads to expression of BCL6 (a key tran-
different B cell fates; for example, strong antigen reactivity and high scription factor for GC B cells); a strong BCR signal further maintains
antigen dose promote extrafollicular ASC formation211,212. By contrast, IRF4 expression leading to downstream induction of BLIMP1 (encoded
activated B cells with weak affinity for antigen that receive weak sig- by Prdm1), a transcription factor for ASCs216,217, whereas a weaker BCR
nals from TFH cells become GC-independent MBCs, which are mostly signal downmodulates IRF4 expression and enables BCL6-mediated
non-class switched with few somatic mutations213,214. Activated B cells differentiation to GC B cells. Cytokine signals from TFH cells (such as
that do not differentiate into extrafollicular ASCs or GC-independent IL-4 and IL-21) can also induce BCL6 (ref. 218), and BCL6 represses
MBCs can seed GCs, wherein they differentiate into GC-dependent expression of cell-migratory receptors such as S1PR1 and EBI2 to retain

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 823

Review article

B cells in the GC219. BCL6 and BLIMP1 repress expression of each other to be elucidated (reviewed in ref. 228). Resting B cells are mostly
and, thus, regulate GC versus ASC differentiation220. dependent on oxidative phosphorylation; once activated, they have
GC B cells undergo extensive clonal expansion and the fate of each marked increases in glucose and amino acid uptake and pyruvate
GC B cell is determined by affinity for antigen as they compete for signals generation through glycolysis229. GC B cells have unique metabolic
from GC TFH cells. GC B cells with lower affinity have a lower tendency to pathways to enable their rapid proliferation, predominantly using fatty
interact with TFH cells and subsequently undergo apoptosis221,222, whereas acid oxidation rather than glycolysis230. As for T cells, further study of
intermediate affinity B cells receive weak signals from GC TFH cells leading the distinct metabolic pathways associated with B cell differentiation
to induction of the transcription factor BACH2, which restrains BLIMP1 might lead to targeted strategies for its modulation.
activity, and differentiate into MBCs223. During this process of selection
and B cell differentiation, CSR and SHM are mediated by the enzyme Maintenance of B cell memory
activation-induced cytidine deaminase (AID)224,225, and GC B cells with Maintaining the cellular components of humoral immunity involves
high affinity are selected to become ASCs. CSR and SHM of MBCs are reg- the persistence of MBCs expressing surface immunoglobulin and
ulated through epigenetic mechanisms, including histone modifications, LLPCs actively secreting antibody in distinct niches. Although long-
DNA methylation and microRNAs, that modulate the expression of AID lived MBCs can be detected in humans, the lifespan, mechanisms
and BLIMP1 (ref. 226). XBP1, a mediator of the unfolded protein response, and requirements for maintenance of MBCs and LLPCs remain
is also an important transcription factor for ASC differentiation227. Finally, incompletely understood. The requirements for MBC longevity have
it has been shown that the accessibility of chromatin in the promoter been extensively studied in mouse models. It was originally found
regions of Irf4, Prdm1 (which encodes BLIMP1) and Xbp1 is greater in MBCs that B cell memory persistence required antigen that was bound
than in naive B cells163, which suggests that in a manner similar to that in to and presented by follicular dendritic cells (FDCs), the stromal
T cells, epigenetic changes are also involved in B cell differentiation and cell component of GCs231. However, subsequent development of
commitment to memory. novel mouse models to alter BCR expression in an inducible manner
Metabolic pathways also have roles in B cell differentiation to showed that MBCs could persist even in the absence of the priming
ASCs, MBCs and LLPCs, although their precise involvement remains antigen232. Later studies have suggested alternate roles for FDC-
bound antigen in the regulation of B cell memory233. However, there
is evidence that some type of BCR engagement is required for MBC
persistence because loss of surface BCR through immunoglobulin
Box 2 heavy-chain deletion or altered BCR signalling results in the apop-
tosis of MBCs234,235. Although CD4+ T cells are not required for the
maintenance of MBCs in mice236, certain T cell-derived cytokines,
Major unanswered questions such as IL-21, might have a role in MBC longevity and turnover237. In
humans, measuring the turnover of MBCs in blood in vitro revealed
relating to adaptive immune that classical MBCs had a doubling time of ~30 days and a half-life

memory of 2 weeks, with atypical MBCs having longer lifespans238 (Table 3).
The lifespan of other B cell memory subsets in different sites, such as
BRM cells in lymphoid organs, is unknown and is a subject for future
• What are the protective correlates of lasting immune memory investigations.
beyond neutralizing antibodies, how do these differ for various The lifespan and persistence of plasma cells depend on their tissue
pathogens and different vaccines, and what are their relative localization. The original discovery of LLPCs identified an unprec-
contributions? edented turnover time of 140 days for plasma cells in bone marrow
• Effective T cell and B cell coordination is crucial for the compared with 3 days for plasma cells in the spleen239,240. Moreover,
development of optimal antibody responses, but how do T cell LLPCs in the bone marrow could persist in the absence of MBCs167,
responses affect the antigen specificity of B cell responses? indicating that maintenance does not depend on the conversion of
• How can we monitor immune memory to vaccination and MBCs to LLPCs. In humans, plasma cells in the gut were found to persist
infection, and do responses that are measured in the blood for more than 1 year in intestinal transplant samples241. In addition,
predict protection at tissue sites? a technique known as retrospective 14C birth dating241,242 has been used
• What role does the tissue niche have in the persistence of to assess the lifespan of IgA+ plasma cells in the gut (Table 1), showing
memory T cells, memory B cells and plasma cells, and what are that they persisted for several decades241. These results suggest that
the specific niche factors for each? early life exposures may be formative for ‘setting’ the B cell memory
• How are memory T cell functions regulated and can we develop repertoire in tissues.
strategies to enhance or suppress their rapid effector responses? The requirements for maintenance of LLPCs are distinct from
• What are the major mechanisms for the functional persistence of those of MBCs, given that LLPCs lack surface expression of BCR or MHC.
immune memory and how can we monitor this process? The mechanisms for long-term persistence of LLPCs likely involve cell-
• What are the major mechanisms by which age influences the intrinsic properties but also the tissue niche in which they reside. For
generation and maintenance of memory responses? example, BAFF (B cell-activating factor of the TNF family) and APRIL
• Can we develop algorithms to predict the potential of an (a proliferation-inducing ligand) were found to be crucial for LLPC
individual for developing a memory immune response to survival in the bone marrow243, and APRIL was reported to support
a pathogen or vaccine based on age, sex, pathogen type the survival of LLPCs in the gut244. Determining the specific pathways
and other covariates? that target and maintain LLPCs in different sites is a crucial area for
future studies.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 824

Review article

Current challenges and future prospects 4. Zheng, Z., Warren, J. L., Shapiro, E. D., Pitzer, V. E. & Weinberger, D. M. Estimated
incidence of respiratory hospitalizations attributable to RSV infections across age and
The study of immune memory has been an intriguing and active area of socioeconomic groups. Pneumonia 14, 6 (2022).
research since the earliest studies of the immune system owing to its fun- 5. O’Driscoll, M. et al. Age-specific mortality and immunity patterns of SARS-CoV-2. Nature
damental role in immune protection and survival. However, the recent 590, 140–145 (2021).
6. Horita, N. & Fukumoto, T. Global case fatality rate from COVID-19 has decreased by 96.8%
COVID-19 pandemic has highlighted major unanswered questions during 2.5 years of the pandemic. J. Med. Virol. 95, e28231 (2023).
about immune memory, the solution of which takes on a new urgency 7. Farber, D. L. Tissues, not blood, are where immune cells act. Nature 593, 506–509 (2021).
(Box 2). Understanding how immune memory is not only formed but 8. Farber, D. L., Yudanin, N. A. & Restifo, N. P. Human memory T cells: generation,
compartmentalization and homeostasis. Nat. Rev. Immunol. 14, 24–35 (2014).
also maintained, as well as its tissue specificity, is crucial for the devel- 9. Szabo, P. A., Miron, M. & Farber, D. L. Location, location, location: tissue resident memory
opment of vaccines to promote protective immunity at relevant sites T cells in mice and humans. Sci. Immunol. 4, eaas9673 (2019).
of pathogen entry. Similarly, strategies to modulate immune memory 10. Poon, M. M. L. et al. SARS-CoV-2 infection generates tissue-localized immunological
memory in humans. Sci. Immunol. 6, eabl9105 (2021).
in infectious diseases and immune pathologies require identifying 11. Weisel, N. M. et al. Comprehensive analyses of B-cell compartments across the human
the relevant subsets involved and their tissue reservoirs. body reveal novel subsets and a gut-resident memory phenotype. Blood 136, 2774–2785
(2020).
Extensive research has revealed the remarkable heterogeneity
12. Tarlinton, D. M., Ding, Z., Tellier, J. & Nutt, S. L. Making sense of plasma cell heterogeneity.
of T cell memory, although major challenges remain in dissecting Curr. Opin. Immunol. 81, 102297 (2023).
the complex pathways for its development, maintenance and func- 13. Kaye, J. et al. Selective development of CD4+ T cells in transgenic mice expressing a
class II MHC-restricted antigen receptor. Nature 341, 746–749 (1989).
tional regulation. A major consideration for T cell memory is that
14. Brink, R. et al. Immunoglobulin M and D antigen receptors are both capable of
for protection from site-specific pathogens and for memory T cell- mediating B lymphocyte activation, deletion, or anergy after interaction with specific
associated pathologies such as asthma and skin diseases, non-circulating antigen. J. Exp. Med. 176, 991–1005 (1992).
15. Pape, K. A., Khoruts, A., Mondino, A. & Jenkins, M. K. Inflammatory cytokines
TRM cells are the crucial subset, which cannot be monitored in the blood.
enhance the in vivo clonal expansion and differentiation of antigen-activated CD4+
Developing new approaches to monitor T cell memory that are directed T cells. J. Immunol. 159, 591–598 (1997).
towards the tissue site and/or biomarkers for assessing tissue responses 16. Marzo, A. L. et al. Initial T cell frequency dictates memory CD8+ T cell lineage
commitment. Nat. Immunol. 6, 793–799 (2005).
are key areas for future research. Memory T cells are functionally hetero-
17. Parga-Vidal, L. et al. Hobit identifies tissue-resident memory T cell precursors that are
geneous and the identification of molecular and epigenetic regulators regulated by Eomes. Sci. Immunol. 6, eabg3533 (2021).
of this process might enable strategies to reprogramme immune cells 18. Harrington, L. E., Janowski, K. M., Oliver, J. R., Zajac, A. J. & Weaver, C. T. Memory CD4
T cells emerge from effector T-cell progenitors. Nature 452, 356–360 (2008).
with the potential to restore or enhance regular immune function. 19. Luthje, K. et al. The development and fate of follicular helper T cells defined by an IL-21
For B cell-mediated memory, understanding the heterogeneity reporter mouse. Nat. Immunol. 13, 491–498 (2012).
of antibody responses, including their persistence and efficacy in 20. Budd, R. C. et al. Distinction of virgin and memory T lymphocytes. Stable acquisition
of the Pgp-1 glycoprotein concomitant with antigenic stimulation. J. Immunol. 138,
response to vaccination and infection, is a major goal for future research. 3120–3129 (1987).
CD4+ TFH cells are crucial for B cell development, but how TFH cell recog- 21. Smith, S. H., Brown, M. H., Rowe, D., Callard, R. E. & Beverley, P. C. Functional subsets of
nition and responses affect the efficacy of MBC and LLPC differentiation human helper-inducer cells defined by a new monoclonal antibody, UCHL1. Immunology
58, 63–70 (1986).
remains incompletely understood. Identifying tissue niches for BRM cells 22. Clark, R. A. Resident memory T cells in human health and disease. Sci. Transl. Med. 7,
and LLPCs can provide new insights into the mechanisms for varia- 269rv261 (2015).
tions in humoral immune memory, and new techniques for spatial pro- 23. Masopust, D. & Soerens, A. G. Tissue-resident T cells and other resident leukocytes.
Annu. Rev. Immunol. 37, 521–546 (2019).
filing will probably be transformative in this regard. For all memory 24. Zhou, X., Ramachandran, S., Mann, M. & Popkin, D. L. Role of lymphocytic
responses, both B cells and T cells, age is a determining factor; memory choriomeningitis virus (LCMV) in understanding viral immunology: past, present and
can be durably maintained over a lifetime, but early responses are future. Viruses 4, 2650–2669 (2012).
25. Lau, L. L., Jamieson, B. D., Somasundaram, T. & Ahmed, R. Cytotoxic T-cell memory
formative. Understanding how immune memory may be differen- without antigen. Nature 369, 648–652 (1994).
tially generated at distinct life stages could lead to the optimization of 26. Murali-Krishna, K. et al. Counting antigen-specific CD8 T cells: a reevaluation of
vaccines and immunotherapies for different age groups. bystander activation during viral infection. Immunity 8, 177–187 (1998).
27. Moon, J. J. et al. Naive CD4+ T cell frequency varies for different epitopes and predicts
The technological advances in multi-omics profiling over the past repertoire diversity and response magnitude. Immunity 27, 203–213 (2007).
decade have provided researchers with an unprecedented ability to study 28. Pepper, M. et al. Different routes of bacterial infection induce long-lived TH1 memory cells
the human immune response at high depth and dimension — from single and short-lived TH17 cells. Nat. Immunol. 11, 83–89 (2010).
29. Kearney, E. R., Pape, K. A., Loh, D. Y. & Jenkins, M. K. Visualization of peptide-specific
cells to tissues. Profiling human memory immune responses in different T cell immunity and peripheral tolerance induction in vivo. Immunity 1, 327–339
contexts and across diverse populations might enable the development (1994).
30. Reinhardt, R. L., Khoruts, A., Merica, R., Zell, T. & Jenkins, M. K. Visualizing the generation
of algorithms to predict how an individual will respond to a specific
of memory CD4 T cells in the whole body. Nature 410, 101–105 (2001).
vaccine or pathogen. With this improved understanding of immune 31. Swain, S. L. Generation and in vivo persistence of polarized Th1 and Th2 memory cells.
memory, we will be better equipped to respond to the many unknown Immunity 1, 543–552 (1994).
32. Ahmadzadeh, M., Hussain, S. F. & Farber, D. L. Effector CD4 T cells are biochemically
infectious challenges that we are likely to encounter in the years to come.
distinct from the memory subset: evidence for long-term persistence of effectors in vivo.
J. Immunol. 163, 3053–3063 (1999).
Published online: 3 June 2024 33. Ahmadzadeh, M., Hussain, S. F. & Farber, D. L. Heterogeneity of the memory CD4 T cell
response: persisting effectors and resting memory T cells. J. Immunol. 166, 926–935
(2001).
References 34. Moulton, V. R., Bushar, N. D., Leeser, D. B., Patke, D. S. & Farber, D. L. Divergent generation
1. Bekkering, S., Dominguez-Andres, J., Joosten, L. A. B., Riksen, N. P. & Netea, M. G. Trained of heterogeneous memory CD4 T cells. J. Immunol. 177, 869–876 (2006).
immunity: reprogramming innate immunity in health and disease. Annu. Rev. Immunol. 35. Hammarlund, E. et al. Duration of antiviral immunity after smallpox vaccination. Nat.
39, 667–693 (2021). Med. 9, 1131–1137 (2003).
2. Fuentes, S., Coyle, E. M., Beeler, J., Golding, H. & Khurana, S. Antigenic fingerprinting 36. Altman, J. D. et al. Phenotypic analysis of antigen-specific T lymphocytes. Science 274,
following primary RSV infection in young children identifies novel antigenic sites and 94–96 (1996).
reveals unlinked evolution of human antibody repertoires to fusion and attachment 37. Christophersen, A. Peptide-MHC class I and class II tetramers: from flow to mass
glycoproteins. PLoS Pathog. 12, e1005554 (2016). cytometry. HLA 95, 169–178 (2020).
3. Gouma, S. et al. Middle-aged individuals may be in a perpetual state of H3N2 influenza 38. Akondy, R. S. et al. Origin and differentiation of human memory CD8 T cells after
virus susceptibility. Nat. Commun. 11, 4566 (2020). vaccination. Nature 552, 362–367 (2017).

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 825

Review article

39. Pan, Y. G. et al. Vaccination reshapes the virus-specific T cell repertoire in unexposed 72. Abdelsamed, H. A. et al. Human memory CD8 T cell effector potential is epigenetically
adults. Immunity 54, 1245–1256.e5 (2021). preserved during in vivo homeostasis. J. Exp. Med. 214, 1593–1606 (2017).
40. Poon, M. M. L. et al. Heterogeneity of human anti-viral immunity shaped by virus, tissue, 73. Kaech, S. M. & Cui, W. Transcriptional control of effector and memory CD8+ T cell
age, and sex. Cell Rep. 37, 110071 (2021). differentiation. Nat. Rev. Immunol. 12, 749–761 (2012).
41. Dan, J. M. et al. A cytokine-independent approach to identify antigen-specific human 74. Kratchmarov, R., Magun, A. M. & Reiner, S. L. TCF1 expression marks self-renewing human
germinal center T follicular helper cells and rare antigen-specific CD4+ T cells in blood. CD8+ T cells. Blood Adv. 2, 1685–1690 (2018).
J. Immunol. 197, 983–993 (2016). 75. Rosenblum, M. D., Way, S. S. & Abbas, A. K. Regulatory T cell memory. Nat. Rev. Immunol.
42. Reiss, S. et al. Comparative analysis of activation induced marker (AIM) assays for 16, 90–101 (2016).
sensitive identification of antigen-specific CD4 T cells. PLoS ONE 12, e0186998 (2017). 76. Josefowicz, S. Z., Lu, L. F. & Rudensky, A. Y. Regulatory T cells: mechanisms of
43. Grifoni, A. et al. Targets of T cell responses to SARS-CoV-2 coronavirus in humans with differentiation and function. Annu. Rev. Immunol. 30, 531–564 (2012).
COVID-19 disease and unexposed individuals. Cell 181, 1489–1501.e15 (2020). 77. Brincks, E. L. et al. Antigen-specific memory regulatory CD4+Foxp3+ T cells control
44. Dan, J. M. et al. Immunological memory to SARS-CoV-2 assessed for up to 8 months after memory responses to influenza virus infection. J. Immunol. 190, 3438–3446 (2013).
infection. Science 371, eabf4063 (2021). 78. Sanchez, A. M., Zhu, J., Huang, X. & Yang, Y. The development and function of memory
45. Sette, A., Sidney, J. & Crotty, S. T cell responses to SARS-CoV-2. Annu. Rev. Immunol. 41, regulatory T cells after acute viral infections. J. Immunol. 189, 2805–2814 (2012).
343–373 (2023). 79. Sanchez Rodriguez, R. et al. Memory regulatory T cells reside in human skin. J. Clin.
46. Weng, N. P., Araki, Y. & Subedi, K. The molecular basis of the memory T cell response: Invest. 124, 1027–1036 (2014).
differential gene expression and its epigenetic regulation. Nat. Rev. Immunol. 12, 80. Thome, J. J. et al. Early-life compartmentalization of human T cell differentiation and
306–315 (2012). regulatory function in mucosal and lymphoid tissues. Nat. Med. 22, 72–77 (2016).
47. Kaech, S. M., Hemby, S., Kersh, E. & Ahmed, R. Molecular and functional profiling of 81. Masopust, D., Vezys, V., Marzo, A. L. & Lefrancois, L. Preferential localization of effector
memory CD8 T cell differentiation. Cell 111, 837–851 (2002). memory cells in nonlymphoid tissue. Science 291, 2413–2417 (2001).
48. Liu, K. et al. Augmentation in expression of activation-induced genes differentiates 82. Mackay, L. K. & Kallies, A. Transcriptional regulation of tissue-resident lymphocytes.
memory from naive CD4+ T cells and is a molecular mechanism for enhanced cellular Trends Immunol. 38, 94–103 (2017).
response of memory CD4+ T cells. J. Immunol. 166, 7335–7344 (2001). 83. Masopust, D. et al. Dynamic T cell migration program provides resident memory within
49. Araki, Y., Fann, M., Wersto, R. & Weng, N. P. Histone acetylation facilitates rapid and intestinal epithelium. J. Exp. Med. 207, 553–564 (2010).
robust memory CD8 T cell response through differential expression of effector 84. Teijaro, J. R. et al. Cutting edge: tissue-retentive lung memory CD4 T cells mediate
molecules (eomesodermin and its targets: perforin and granzyme B). J. Immunol. 180, optimal protection to respiratory virus infection. J. Immunol. 187, 5510–5514 (2011).
8102–8108 (2008). 85. Jiang, X. et al. Skin infection generates non-migratory memory CD8+ TRM cells providing
50. Araki, Y. et al. Genome-wide analysis of histone methylation reveals chromatin state- global skin immunity. Nature 483, 227–231 (2012).
based regulation of gene transcription and function of memory CD8+ T cells. Immunity 86. Milner, J. J. et al. Runx3 programs CD8+ T cell residency in non-lymphoid tissues and
30, 912–925 (2009). tumours. Nature 552, 253–257 (2017).
51. Youngblood, B., Davis, C. W. & Ahmed, R. Making memories that last a lifetime: heritable 87. Mackay, L. K. et al. Hobit and Blimp1 instruct a universal transcriptional program of tissue
functions of self-renewing memory CD8 T cells. Int. Immunol. 22, 797–803 (2010). residency in lymphocytes. Science 352, 459–463 (2016).
52. Henning, A. N., Roychoudhuri, R. & Restifo, N. P. Epigenetic control of CD8+ T cell 88. Hombrink, P. et al. Programs for the persistence, vigilance and control of human CD8+
differentiation. Nat. Rev. Immunol. 18, 340–356 (2018). lung-resident memory T cells. Nat. Immunol. 17, 1467–1478 (2016).
53. Fahmy, T. M., Bieler, J. G., Edidin, M., Schneck, J. P. & Increased, T. C. R. Avidity after 89. Milner, J. J. & Goldrath, A. W. Transcriptional programming of tissue-resident memory
T cell activation: a mechanism for sensing low-density antigen. Immunity 14, 135–143 CD8+ T cells. Curr. Opin. Immunol. 51, 162–169 (2018).
(2001). 90. Dominguez, C. X. et al. The transcription factors ZEB2 and T-bet cooperate to program
54. Croft, M., Bradley, L. M. & Swain, S. L. Naive versus memory CD4 T cell response to cytotoxic T cell terminal differentiation in response to LCMV viral infection. J. Exp. Med.
antigen. Memory cells are less dependent on accessory cell costimulation and can 212, 2041–2056 (2015).
respond to many antigen-presenting cell types including resting B cells. J. Immunol. 152, 91. Crowl, J. T. et al. Tissue-resident memory CD8+ T cells possess unique transcriptional,
2675–2685 (1994). epigenetic and functional adaptations to different tissue environments. Nat. Immunol.
55. Ndejembi, M. P. et al. Control of memory CD4 T cell recall by the CD28/B7 costimulatory 23, 1121–1131 (2022).
pathway. J. Immunol. 177, 7698–7706 (2006). 92. Wu, T. et al. Lung-resident memory CD8 T cells (TRM) are indispensable for optimal
56. Glinos, D. A. et al. Genomic profiling of T-cell activation suggests increased sensitivity of cross-protection against pulmonary virus infection. J. Leukoc. Biol. 95, 215–224
memory T cells to CD28 costimulation. Genes Immun. 21, 390–408 (2020). (2014).
57. Arkatkar, T. et al. Memory T cells possess an innate-like function in local protection from 93. Schenkel, J. M. et al. T cell memory. Resident memory CD8 T cells trigger protective
mucosal infection. J. Clin. Invest. 133, e162800 (2023). innate and adaptive immune responses. Science 346, 98–101 (2014).
58. Huang, W. & August, A. The signaling symphony: T cell receptor tunes cytokine-mediated 94. Hassert, M., Arumugam, S. & Harty, J. T. Memory CD8+ T cell-mediated protection against
T cell differentiation. J. Leukoc. Biol. 97, 477–485 (2015). liver-stage malaria. Immunol. Rev. 316, 84–103 (2023).
59. Freeman, B. E., Hammarlund, E., Raue, H. P. & Slifka, M. K. Regulation of innate CD8+ 95. Paik, D. H. & Farber, D. L. Anti-viral protective capacity of tissue resident memory T cells.
T-cell activation mediated by cytokines. Proc. Natl Acad. Sci. USA 109, 9971–9976 Curr. Opin. Virol. 46, 20–26 (2021).
(2012). 96. Luangrath, M. A., Schmidt, M. E., Hartwig, S. M. & Varga, S. M. Tissue-resident
60. Farber, D. L. Biochemical signaling pathways for memory T cell recall. Semin. Immunol. memory T cells in the lungs protect against acute respiratory syncytial virus infection.
21, 84–91 (2009). Immunohorizons 5, 59–69 (2021).
61. Sallusto, F., Geginat, J. & Lanzavecchia, A. Central memory and effector memory T cell 97. Zhao, J. et al. Airway memory CD4+ T cells mediate protective immunity against
subsets: function, generation, and maintenance. Annu. Rev. Immunol. 22, 745–763 emerging respiratory coronaviruses. Immunity 44, 1379–1391 (2016).
(2004). 98. Wilk, M. M. et al. Lung CD4 tissue-resident memory T cells mediate adaptive immunity
62. Farber, D. L. T cell memory: heterogeneity and mechanisms. Clin. Immunol. 95, 173–181 induced by previous infection of mice with Bordetella pertussis. J. Immunol. 199,
(2000). 233–243 (2017).
63. Sallusto, F., Lenig, D., Forster, R., Lipp, M. & Lanzavecchia, A. Two subsets of memory 99. Sakai, S. et al. Cutting edge: control of Mycobacterium tuberculosis infection by a subset
T lymphocytes with distinct homing potentials and effector functions. Nature 401, of lung parenchyma-homing CD4 T cells. J. Immunol. 192, 2965–2969 (2014).
708–712 (1999). 100. Gebhardt, T. et al. Memory T cells in nonlymphoid tissue that provide enhanced local
64. Crotty, S. Follicular helper CD4 T cells (TFH). Annu. Rev. Immunol. 29, 621–663 (2011). immunity during infection with herpes simplex virus. Nat. Immunol. 10, 524–530 (2009).
65. Derhovanessian, E. et al. Infection with cytomegalovirus but not herpes simplex virus 101. Strobl, J. & Haniffa, M. Functional heterogeneity of human skin-resident memory T cells
induces the accumulation of late-differentiated CD4+ and CD8+ T-cells in humans. J. Gen. in health and disease. Immunol. Rev. 316, 104–119 (2023).
Virol. 92, 2746–2756 (2011). 102. Sathaliyawala, T. et al. Distribution and compartmentalization of human circulating and
66. Gordon, C. L. et al. Tissue reservoirs of antiviral T cell immunity in persistent human CMV tissue-resident memory T cell subsets. Immunity 38, 187–197 (2013).
infection. J. Exp. Med. 214, 651–667 (2017). 103. Thome, J. J. et al. Spatial map of human T cell compartmentalization and maintenance
67. Darrah, P. A. et al. Multifunctional TH1 cells define a correlate of vaccine-mediated over decades of life. Cell 159, 814–828 (2014).
protection against Leishmania major. Nat. Med. 13, 843–850 (2007). 104. Clark, R. A. et al. The vast majority of CLA+ T cells are resident in normal skin. J. Immunol.
68. Gattinoni, L. et al. A human memory T cell subset with stem cell-like properties. 176, 4431–4439 (2006).
Nat. Med. 17, 1290–1297 (2011). 105. Watanabe, R. et al. Human skin is protected by four functionally and phenotypically
69. Bingaman, A. W. et al. Novel phenotypes and migratory properties distinguish memory discrete populations of resident and recirculating memory T cells. Sci. Transl. Med. 7,
CD4 T cell subsets in lymphoid and lung tissue. Eur. J. Immunol. 35, 3173–3186 (2005). 279ra239 (2015).
70. Patke, D. S. & Farber, D. L. Modulation of memory CD4 T cell function and survival 106. Kumar, B. V. et al. Human tissue-resident memory T cells are defined by core
potential by altering the strength of the recall stimulus. J. Immunol. 174, 5433–5443 transcriptional and functional signatures in lymphoid and mucosal sites. Cell Rep. 20,
(2005). 2921–2934 (2017).
71. Wherry, E. J. et al. Lineage relationship and protective immunity of memory CD8 T cell 107. Gray, J. I. & Farber, D. L. Tissue-resident immune cells in humans. Annu. Rev. Immunol.
subsets. Nat. Immunol. 4, 225–234 (2003). 40, 195–220 (2022).

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 826

Review article

108. Szabo, P. A. Axes of heterogeneity in human tissue-resident memory T cells. Immunol. 144. Slutter, B. et al. Dynamics of influenza-induced lung-resident memory T cells underlie
Rev. 316, 23–37 (2023). waning heterosubtypic immunity. Sci. Immunol. 2, eaag2031 (2017).
109. Poon, M. M. L. et al. Tissue adaptation and clonal segregation of human memory T cells 145. Beura, L. K. et al. T cells in nonlymphoid tissues give rise to lymph-node-resident
in barrier sites. Nat. Immunol. 24, 309–319 (2023). memory T cells. Immunity 48, 327–338.e5 (2018).
110. Clark, R. A. et al. Skin effector memory T cells do not recirculate and provide immune 146. Snyder, M. E. et al. Generation and persistence of human tissue-resident memory T cells
protection in alemtuzumab-treated CTCL patients. Sci. Transl. Med. 4, 117ra117 (2012). in lung transplantation. Sci. Immunol. 4, eaav5581 (2019).
111. Turner, D. L. et al. Lung niches for the generation and maintenance of tissue-resident 147. Pallett, L. J. et al. Longevity and replenishment of human liver-resident memory T cells
memory T cells. Mucosal Immunol. 7, 501–510 (2014). and mononuclear phagocytes. J. Exp. Med. 217, e20200050 (2020).
112. Kaech, S. M. et al. Selective expression of the interleukin 7 receptor identifies effector 148. Mason, D. W. The class of surface immunoglobulin on cells carrying IgG memory in rat
CD8 T cells that give rise to long-lived memory cells. Nat. Immunol. 4, 1191–1198 (2003). thoracic duct lymph: the size of the subpopulation mediating IgG memory. J. Exp. Med.
113. Schluns, K. S., Kieper, W. C., Jameson, S. C. & Lefrancois, L. Interleukin-7 mediates the 143, 1122–1130 (1976).
homeostasis of naive and memory CD8 T cells in vivo. Nat. Immunol. 1, 426–432 (2000). 149. Okumura, K., Julius, M. H., Tsu, T., Herzenberg, L. A. & Herzenberg, L. A. Demonstration
114. Sarkar, S. et al. Functional and genomic profiling of effector CD8 T cell subsets with that IgG memory is carried by IgG-bearing cells. Eur. J. Immunol. 6, 467–472 (1976).
distinct memory fates. J. Exp. Med. 205, 625–640 (2008). 150. Amanna, I. J., Carlson, N. E. & Slifka, M. K. Duration of humoral immunity to common viral
115. Hand, T. W., Morre, M. & Kaech, S. M. Expression of IL-7 receptor α is necessary but not and vaccine antigens. N. Engl. J. Med. 357, 1903–1915 (2007).
sufficient for the formation of memory CD8 T cells during viral infection. Proc. Natl Acad. 151. Fooksman, D. R., Jing, Z. & Park, R. New insights into the ontogeny, diversity, maturation
Sci. USA 104, 11730–11735 (2007). and survival of long-lived plasma cells. Nat. Rev. Immunol. https://doi.org/10.1038/
116. Jacob, J. & Baltimore, D. Modelling T-cell memory by genetic marking of memory T cells s41577-024-00991-0 (2024).
in vivo. Nature 399, 593–597 (1999). 152. Inoue, T. & Kurosaki, T. Memory B cells. Nat. Rev. Immunol. 24, 5–17 (2024).
117. Ahmed, R., Bevan, M. J., Reiner, S. L. & Fearon, D. T. The precursors of memory: models 153. Gray, D. Immunological memory. Annu. Rev. Immunol. 11, 49–77 (1993).
and controversies. Nat. Rev. Immunol. 9, 662–668 (2009). 154. Klein, U., Rajewsky, K. & Kuppers, R. Human immunoglobulin (Ig)M+IgD+ peripheral blood
118. Manjunath, N. et al. Effector differentiation is not prerequisite for generation of memory B cells expressing the CD27 cell surface antigen carry somatically mutated variable
cytotoxic T lymphocytes. J. Clin. Invest. 108, 871–878 (2001). region genes: CD27 as a general marker for somatically mutated (memory) B cells. J. Exp.
119. Chang, J. T. et al. Asymmetric T lymphocyte division in the initiation of adaptive immune Med. 188, 1679–1689 (1998).
responses. Science 315, 1687–1691 (2007). 155. Krishnamurty, A. T. et al. Somatically hypermutated plasmodium-specific IgM+ memory
120. Grabnitz, F. et al. Asymmetric cell division safeguards memory CD8 T cell development. B cells are rapid, plastic, early responders upon malaria rechallenge. Immunity 45,
Cell Rep. 42, 112468 (2023). 402–414 (2016).
121. Wang, P. H. et al. Reciprocal transmission of activating and inhibitory signals and cell fate 156. Lyashchenko, K. P., Vordermeier, H. M. & Waters, W. R. Memory B cells and tuberculosis.
in regenerating T cells. Cell Rep. 42, 113155 (2023). Vet. Immunol. Immunopathol. 221, 110016 (2020).
122. Buchholz, V. R. et al. Disparate individual fates compose robust CD8+ T cell immunity. 157. Gil, D., Schamel, W. W., Montoya, M., Sanchez-Madrid, F. & Alarcon, B. Recruitment of Nck
Science 340, 630–635 (2013). by CD3ε reveals a ligand-induced conformational change essential for T cell receptor
123. Stemberger, C. et al. A single naive CD8+ T cell precursor can develop into diverse signaling and synapse formation. Cell 109, 901–912 (2002).
effector and memory subsets. Immunity 27, 985–997 (2007). 158. Schittek, B. & Rajewsky, K. Maintenance of B-cell memory by long-lived cells generated
124. Zhang, N. & Bevan, M. J. Transforming growth factor-β signaling controls the formation from proliferating precursors. Nature 346, 749–751 (1990).
and maintenance of gut-resident memory T cells by regulating migration and retention. 159. Weisel, F. & Shlomchik, M. Memory B cells of mice and humans. Annu. Rev. Immunol. 35,
Immunity 39, 687–696 (2013). 255–284 (2017).
125. Hondowicz, B. D. et al. Interleukin-2-dependent allergen-specific tissue-resident memory 160. Good, K. L., Avery, D. T. & Tangye, S. G. Resting human memory B cells are intrinsically
cells drive asthma. Immunity 44, 155–166 (2016). programmed for enhanced survival and responsiveness to diverse stimuli compared to
126. Mackay, L. K. et al. The developmental pathway for CD103+CD8+ tissue-resident memory naive B cells. J. Immunol. 182, 890–901 (2009).
T cells of skin. Nat. Immunol. 14, 1294–1301 (2013). 161. Good, K. L. & Tangye, S. G. Decreased expression of Kruppel-like factors in memory
127. Connors, T. J. et al. Site-specific development and progressive maturation of human B cells induces the rapid response typical of secondary antibody responses. Proc. Natl
tissue-resident memory T cells over infancy and childhood. Immunity 56, 1894–1909.e5 Acad. Sci. USA 104, 13420–13425 (2007).
(2023). 162. Tangye, S. G., Avery, D. T., Deenick, E. K. & Hodgkin, P. D. Intrinsic differences in the
128. Miron, M. et al. Maintenance of the human memory T cell repertoire by subset and tissue proliferation of naive and memory human B cells as a mechanism for enhanced
site. Genome Med. 13, 100 (2021). secondary immune responses. J. Immunol. 170, 686–694 (2003).
129. Geltink, R. I. K., Kyle, R. L. & Pearce, E. L. Unraveling the complex interplay between T cell 163. Price, M. J., Scharer, C. D., Kania, A. K., Randall, T. D. & Boss, J. M. Conserved epigenetic
metabolism and function. Annu. Rev. Immunol. 36, 461–488 (2018). programming and enhanced heme metabolism drive memory B cell reactivation.
130. Gubser, P. M. et al. Rapid effector function of memory CD8+ T cells requires an J. Immunol. 206, 1493–1504 (2021).
immediate-early glycolytic switch. Nat. Immunol. 14, 1064–1072 (2013). 164. Turner, J. S. et al. Human germinal centres engage memory and naive B cells after
131. van der Windt, G. J. et al. CD8 memory T cells have a bioenergetic advantage that influenza vaccination. Nature 586, 127–132 (2020).
underlies their rapid recall ability. Proc. Natl Acad. Sci. USA 110, 14336–14341 (2013). 165. Mesin, L. et al. Restricted clonality and limited germinal center reentry characterize
132. van der Windt, G. J. et al. Mitochondrial respiratory capacity is a critical regulator of CD8+ memory B cell reactivation by boosting. Cell 180, 92–106.e11 (2020).
T cell memory development. Immunity 36, 68–78 (2012). 166. Manz, R. A., Lohning, M., Cassese, G., Thiel, A. & Radbruch, A. Survival of long-lived
133. Corrado, M. & Pearce, E. L. Targeting memory T cell metabolism to improve immunity. plasma cells is independent of antigen. Int. Immunol. 10, 1703–1711 (1998).
J. Clin. Invest. 132, e148546 (2022). 167. Slifka, M. K., Antia, R., Whitmire, J. K. & Ahmed, R. Humoral immunity due to long-lived
134. Tan, J. T. et al. Interleukin (IL)-15 and IL-7 jointly regulate homeostatic proliferation of plasma cells. Immunity 8, 363–372 (1998).
memory phenotype CD8+ cells but are not required for memory phenotype CD4+ cells. 168. Koike, T. et al. Progressive differentiation toward the long-lived plasma cell compartment
J. Exp. Med. 195, 1523–1532 (2002). in the bone marrow. J. Exp. Med. 220, e20221717 (2023).
135. Purton, J. F. et al. Antiviral CD4+ memory T cells are IL-15 dependent. J. Exp. Med. 204, 169. Tas, J. M. J. et al. Antibodies from primary humoral responses modulate the recruitment
951–961 (2007). of naive B cells during secondary responses. Immunity 55, 1856–1871.e6 (2022).
136. Murali-Krishna, K. et al. Persistence of memory CD8 T cells in MHC class I-deficient mice. 170. Schaefer-Babajew, D. et al. Antibody feedback regulates immune memory after
Science 286, 1377–1381 (1999). SARS-CoV-2 mRNA vaccination. Nature 613, 735–742 (2023).
137. Kassiotis, G., Garcia, S., Simpson, E. & Stockinger, B. Impairment of immunological 171. Hammarlund, E. et al. Plasma cell survival in the absence of B cell memory. Nat. Commun.
memory in the absence of MHC despite survival of memory T cells. Nat. Immunol. 3, 8, 1781 (2017).
244–250 (2002). 172. Allie, S. R. et al. The establishment of resident memory B cells in the lung requires local
138. Bushar, N. D., Corbo, E., Schmidt, M., Maltzman, J. S. & Farber, D. L. Ablation of SLP-76 antigen encounter. Nat. Immunol. 20, 97–108 (2019).
signaling after T cell priming generates memory CD4 T cells impaired in steady-state and 173. Maurer, D., Holter, W., Majdic, O., Fischer, G. F. & Knapp, W. CD27 expression by a distinct
cytokine-driven homeostasis. Proc. Natl Acad. Sci. USA 107, 827–831 (2010). subpopulation of human B lymphocytes. Eur. J. Immunol. 20, 2679–2684 (1990).
139. De Boer, R. J. & Yates, A. J. Modeling T cell fate. Annu. Rev. Immunol. 41, 513–532 174. Weisel, N. M. et al. Surface phenotypes of naive and memory B cells in mouse and human
(2023). tissues. Nat. Immunol. 23, 135–145 (2022).
140. Macallan, D. C. et al. Rapid turnover of effector-memory CD4+ T cells in healthy humans. 175. Dogan, I. et al. Multiple layers of B cell memory with different effector functions.
J. Exp. Med. 200, 255–260 (2004). Nat. Immunol. 10, 1292–1299 (2009).
141. De Boer, R. J. & Perelson, A. S. Quantifying T lymphocyte turnover. J. Theor. Biol. 327, 176. Pape, K. A., Taylor, J. J., Maul, R. W., Gearhart, P. J. & Jenkins, M. K. Different B cell
45–87 (2013). populations mediate early and late memory during an endogenous immune response.
142. Ladell, K. et al. Central memory CD8+ T cells appear to have a shorter lifespan and Science 331, 1203–1207 (2011).
reduced abundance as a function of HIV disease progression. J. Immunol. 180, 177. Zuccarino-Catania, G. V. et al. CD80 and PD-L2 define functionally distinct memory
7907–7918 (2008). B cell subsets that are independent of antibody isotype. Nat. Immunol. 15, 631–637
143. Costa Del Amo, P. et al. Human TSCM cell dynamics in vivo are compatible with (2014).
long-lived immunological memory and stemness. PLoS Biol. 16, e2005523 (2018). 178. Cancro, M. P. Age-associated B cells. Annu. Rev. Immunol. 38, 315–340 (2020).

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 827

Review article

179. Agematsu, K. et al. B cell subpopulations separated by CD27 and crucial collaboration 212. Paus, D. et al. Antigen recognition strength regulates the choice between extrafollicular
of CD27+ B cells and helper T cells in immunoglobulin production. Eur. J. Immunol. 27, plasma cell and germinal center B cell differentiation. J. Exp. Med. 203, 1081–1091
2073–2079 (1997). (2006).
180. Pascual, V. et al. Analysis of somatic mutation in five B cell subsets of human tonsil. 213. Viant, C. et al. Antibody affinity shapes the choice between memory and germinal center
J. Exp. Med. 180, 329–339 (1994). B cell fates. Cell 183, 1298–1311.e11 (2020).
181. Maurer, D. et al. IgM and IgG but not cytokine secretion is restricted to the CD27+ B 214. Ise, W. et al. T follicular helper cell-germinal center B cell interaction strength regulates
lymphocyte subset. J. Immunol. 148, 3700–3705 (1992). entry into plasma cell or recycling germinal center cell fate. Immunity 48, 702–715.e4
182. Grimsholm, O. et al. The interplay between CD27dull and CD27bright B cells ensures the (2018).
flexibility, stability, and resilience of human B cell memory. Cell Rep. 30, 2963–2977.e6 215. Ribatti, D. The discovery of plasma cells: an historical note. Immunol. Lett. 188, 64–67
(2020). (2017).
183. Chappert, P. et al. Human anti-smallpox long-lived memory B cells are defined by 216. De Silva, N. S., Simonetti, G., Heise, N. & Klein, U. The diverse roles of IRF4 in late
dynamic interactions in the splenic niche and long-lasting germinal center imprinting. germinal center B-cell differentiation. Immunol. Rev. 247, 73–92 (2012).
Immunity 55, 1872–1890.e9 (2022). 217. Crotty, S., Johnston, R. J. & Schoenberger, S. P. Effectors and memories: Bcl-6 and Blimp-1
184. Isnardi, I. et al. Complement receptor 2/CD21− human naive B cells contain mostly in T and B lymphocyte differentiation. Nat. Immunol. 11, 114–120 (2010).
autoreactive unresponsive clones. Blood 115, 5026–5036 (2010). 218. Victora, G. D. & Nussenzweig, M. C. Germinal centers. Annu. Rev. Immunol. 30, 429–457
185. Sullivan, R. T. et al. FCRL5 delineates functionally impaired memory B cells associated (2012).
with plasmodium falciparum exposure. PLoS Pathog. 11, e1004894 (2015). 219. Huang, C. et al. The BCL6 RD2 domain governs commitment of activated B cells to form
186. Weiss, G. E. et al. Atypical memory B cells are greatly expanded in individuals living in a germinal centers. Cell Rep. 8, 1497–1508 (2014).
malaria-endemic area. J. Immunol. 183, 2176–2182 (2009). 220. Schebesta, M., Heavey, B. & Busslinger, M. Transcriptional control of B-cell development.
187. Jacobi, A. M. et al. Activated memory B cell subsets correlate with disease activity in Curr. Opin. Immunol. 14, 216–223 (2002).
systemic lupus erythematosus: delineation by expression of CD27, IgD, and CD95. 221. Henry, B. & Laidlaw, B. J. Functional heterogeneity in the memory B-cell response.
Arthritis Rheum. 58, 1762–1773 (2008). Curr. Opin. Immunol. 80, 102281 (2023).
188. Moir, S. et al. Evidence for HIV-associated B cell exhaustion in a dysfunctional memory 222. Mayer, C. T. et al. The microanatomic segregation of selection by apoptosis in the
B cell compartment in HIV-infected viremic individuals. J. Exp. Med. 205, 1797–1805 germinal center. Science 358, eaao2602 (2017).
(2008). 223. Shinnakasu, R. et al. Regulated selection of germinal-center cells into the memory B cell
189. Charles, E. D. et al. Clonal B cells in patients with hepatitis C virus-associated mixed compartment. Nat. Immunol. 17, 861–869 (2016).
cryoglobulinemia contain an expanded anergic CD21low B-cell subset. Blood 117, 224. Feng, Y., Seija, N., Di Noia, J. M. & Martin, A. AID in antibody diversification: there and back
5425–5437 (2011). again. Trends Immunol. 41, 586–600 (2020).
190. Stewart, A. et al. Single-cell transcriptomic analyses define distinct peripheral B cell 225. Muramatsu, M. et al. Class switch recombination and hypermutation require activation-
subsets and discrete development pathways. Front. Immunol. 12, 602539 (2021). induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553–563
191. Nellore, A. et al. A transcriptionally distinct subset of influenza-specific effector memory (2000).
B cells predicts long-lived antibody responses to vaccination in humans. Immunity 56, 226. Li, G., Zan, H., Xu, Z. & Casali, P. Epigenetics of the antibody response. Trends Immunol.
847–863.e8 (2023). 34, 460–470 (2013).
192. Berkowska, M. A. et al. Human memory B cells originate from three distinct germinal 227. Nutt, S. L., Hodgkin, P. D., Tarlinton, D. M. & Corcoran, L. M. The generation of antibody-
center-dependent and -independent maturation pathways. Blood 118, 2150–2158 secreting plasma cells. Nat. Rev. Immunol. 15, 160–171 (2015).
(2011). 228. Ripperger, T. J. & Bhattacharya, D. Transcriptional and metabolic control of memory
193. Lopes de Assis, F. et al. Tracking B cell responses to the SARS-CoV-2 mRNA-1273 vaccine. B cells and plasma cells. Annu. Rev. Immunol. 39, 345–368 (2021).
Cell Rep. 42, 112780 (2023). 229. Boothby, M. & Rickert, R. C. Metabolic regulation of the immune humoral response.
194. Steuten, J. et al. Distinct dynamics of antigen-specific induction and differentiation Immunity 46, 743–755 (2017).
of different CD11c+Tbet+ B-cell subsets. J. Allergy Clin. Immunol. 152, 689–699.e6 230. Weisel, F. J. et al. Germinal center B cells selectively oxidize fatty acids for energy while
(2023). conducting minimal glycolysis. Nat. Immunol. 21, 331–342 (2020).
195. Tellier, J. & Nutt, S. L. Standing out from the crowd: how to identify plasma cells. Eur. J. 231. Tew, J. G., Phipps, R. P. & Mandel, T. E. The maintenance and regulation of the humoral
Immunol. 47, 1276–1279 (2017). immune response: persisting antigen and the role of follicular antigen-binding dendritic
196. Staupe, R. P. et al. Single cell multi-omic reference atlases of non-human primate cells as accessory cells. Immunol. Rev. 53, 175–201 (1980).
immune tissues reveals CD102 as a biomarker for long-lived plasma cells. Commun. Biol. 232. Maruyama, M., Lam, K. P. & Rajewsky, K. Memory B-cell persistence is independent of
5, 1399 (2022). persisting immunizing antigen. Nature 407, 636–642 (2000).
197. Halliley, J. L. et al. Long-lived plasma cells are contained within the CD19−CD38hiCD138+ 233. Haberman, A. M. & Shlomchik, M. J. Reassessing the function of immune-complex
subset in human bone marrow. Immunity 43, 132–145 (2015). retention by follicular dendritic cells. Nat. Rev. Immunol. 3, 757–764 (2003).
198. Joo, H. M., He, Y. & Sangster, M. Y. Broad dispersion and lung localization of virus- 234. Kraus, M., Alimzhanov, M. B., Rajewsky, N. & Rajewsky, K. Survival of resting mature
specific memory B cells induced by influenza pneumonia. Proc. Natl Acad. Sci. USA 105, B lymphocytes depends on BCR signaling via the Igα/β heterodimer. Cell 117, 787–800
3485–3490 (2008). (2004).
199. Allie, S. R. & Randall, T. D. Resident memory B cells. Viral Immunol. 33, 282–293 235. Ackermann, J. A. et al. Syk tyrosine kinase is critical for B cell antibody responses and
(2020). memory B cell survival. J. Immunol. 194, 4650–4656 (2015).
200. Moran, I. et al. Memory B cells are reactivated in subcapsular proliferative foci of lymph 236. Vieira, P. & Rajewsky, K. Persistence of memory B cells in mice deprived of T cell help.
nodes. Nat. Commun. 9, 3372 (2018). Int. Immunol. 2, 487–494 (1990).
201. Moyron-Quiroz, J. E. et al. Role of inducible bronchus associated lymphoid tissue (iBALT) 237. Cancro, M. P. & Tomayko, M. M. Memory B cells and plasma cells: the differentiative
in respiratory immunity. Nat. Med. 10, 927–934 (2004). continuum of humoral immunity. Immunol. Rev. 303, 72–82 (2021).
202. Matsumoto, R. et al. Induction of bronchus-associated lymphoid tissue is an early life 238. Macallan, D. C. et al. B-cell kinetics in humans: rapid turnover of peripheral blood
adaptation for promoting human B cell immunity. Nat. Immunol. 24, 1370–1381 (2023). memory cells. Blood 105, 3633–3640 (2005).
203. MacLean, A. J. et al. Secondary influenza challenge triggers resident memory B cell 239. Manz, R. A., Thiel, A. & Radbruch, A. Lifetime of plasma cells in the bone marrow. Nature
migration and rapid relocation to boost antibody secretion at infected sites. Immunity 388, 133–134 (1997).
55, 718–733.e8 (2022). 240. Slifka, M. K. & Ahmed, R. Long-lived plasma cells: a mechanism for maintaining
204. Mathew, N. R. et al. Single-cell BCR and transcriptome analysis after influenza infection persistent antibody production. Curr. Opin. Immunol. 10, 252–258 (1998).
reveals spatiotemporal dynamics of antigen-specific B cells. Cell Rep. 35, 109286 (2021). 241. Landsverk, O. J. et al. Antibody-secreting plasma cells persist for decades in human
205. Oh, J. E. et al. Intranasal priming induces local lung-resident B cell populations that intestine. J. Exp. Med. 214, 309–317 (2017).
secrete protective mucosal antiviral IgA. Sci. Immunol. 6, eabj5129 (2021). 242. Spalding, K. L. et al. Dynamics of hippocampal neurogenesis in adult humans. Cell 153,
206. Tan, H. X. et al. Lung-resident memory B cells established after pulmonary influenza 1219–1227 (2013).
infection display distinct transcriptional and phenotypic profiles. Sci. Immunol. 7, 243. Benson, M. J. et al. Cutting edge: the dependence of plasma cells and independence of
eabf5314 (2022). memory B cells on BAFF and APRIL. J. Immunol. 180, 3655–3659 (2008).
207. Mamani-Matsuda, M. et al. The human spleen is a major reservoir for long-lived vaccinia 244. Chu, V. T. et al. Eosinophils promote generation and maintenance of immunoglobulin-
virus-specific memory B cells. Blood 111, 4653–4659 (2008). A-expressing plasma cells and contribute to gut immune homeostasis. Immunity 40,
208. Bemark, M. et al. Limited clonal relatedness between gut IgA plasma cells and memory 582–593 (2014).
B cells after oral immunization. Nat. Commun. 7, 12698 (2016). 245. Yewdell, J. W. et al. Antigenic sin: how original? How sinful? Cold Spring Harb. Perspect.
209. Robinson, M. J. et al. Intrinsically determined turnover underlies broad heterogeneity in Med. 11, a038786 (2021).
plasma-cell lifespan. Immunity 56, 1596–1612.e4 (2023). 246. Saggau, C. et al. The pre-exposure SARS-CoV-2-specific T cell repertoire determines
210. Kurosaki, T., Kometani, K. & Ise, W. Memory B cells. Nat. Rev. Immunol. 15, 149–159 the quality of the immune response to vaccination. Immunity 55, 1924–1939.e5
(2015). (2022).
211. Glaros, V. et al. Limited access to antigen drives generation of early B cell memory while 247. Schiepers, A. et al. Molecular fate-mapping of serum antibody responses to repeat
restraining the plasmablast response. Immunity 54, 2005–2023.e10 (2021). immunization. Nature 615, 482–489 (2023).

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 828

Review article

248. Lee, J. H. et al. Long-primed germinal centres with enduring affinity maturation and Additional information
clonal migration. Nature 609, 998–1004 (2022). Peer review information Nature Reviews Immunology thanks R. Rutishauser, who co-reviewed
249. Laidlaw, B. J. & Ellebedy, A. H. The germinal centre B cell response to SARS-CoV-2. Nat. with W. Hung, and the other, anonymous, reviewer(s) for their contribution to the peer review
Rev. Immunol. 22, 7–18 (2022). of this work.
250. Turner, J. S. et al. SARS-CoV-2 mRNA vaccines induce persistent human germinal centre
responses. Nature 596, 109–113 (2021). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
251. Goel, R. R. et al. Efficient recall of omicron-reactive B cell memory after a third dose of published maps and institutional affiliations.
SARS-CoV-2 mRNA vaccine. Cell 185, 1875–1887.e8 (2022).
252. Muecksch, F. et al. Increased memory B cell potency and breadth after a SARS-CoV-2 Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
mRNA boost. Nature 607, 128–134 (2022). article under a publishing agreement with the author(s) or other rightsholder(s); author self-
archiving of the accepted manuscript version of this article is solely governed by the terms
Author contributions of such publishing agreement and applicable law.
The authors contributed equally to all aspects of the article.

Competing interests © Springer Nature Limited 2024

The authors declare no competing interests.

Nature Reviews Immunology | Volume 24 | November 2024 | 810–829 829