Premature B-cell senescence as a consequence of chronic immune activation

Abstract

Similar features between the immune system of healthy elderly people and of younger individuals subjected to conditions of chronic immune activation are progressively being observed. This is raising the hypothesis that chronic immune activation may cause the premature aging of the immune system. Here we dissect this theory by comparing changes occurring to B-cells during healthy aging to the ones occurring during chronic immune activation in younger individuals. Moreover, we discuss how these changes may affect or predict response to vaccination in immune compromised individuals.

Keywords: :

• activation-induced deaminase (AID)
• aging
• double negative B-cells
• immunization
• immunosenescence
• inflammaging
• mature-activated B-cells
• predictive biomarkers

Introduction

Life expectancy has consistently increased during the last decades with a greater number of healthy aged individuals. In parallel, improved treatment of primary and secondary immune diseases has been achieved over the last years resulting in increased survival of individuals with chronic disease. A similar feature of these 2 populations of somewhat healthy individuals is the decreased ability of the immune system to undergo de novo or recall immune responses and to respond to vaccination. Interestingly, additional common features between the immune system of healthy elderly people and of younger individuals subjected to conditions of chronic immune activation are progressively being observed thus raising the hypothesis that chronic immune activation may cause the premature aging of the immune system. In particular, several defects of the B-cell compartment have been described.

Overview on B-Cell Development and Function

B-cells arise in the bone marrow (BM) from hematopoietic stem cells (HSC). Pro- and pre- B-cell precursors are characterized by the expression of CD10 and interleukin (IL) -7 receptor α. In the BM and in the presence of IL-7, B-cells rearrange the variable (V), diversity (D), and joining (J) regions of the immunoglobulin (Ig), or antibody (Ab), heavy (H) (VDJ), and light (L) (VJ) chain genes respectively, through a process involving RAG-1 and RAG-2 gene products.¹ After the Ig gene rearrangement is complete, RAG-1 and 2 gene expression is downregulated and immature B-cells in the BM express CD5, lose the expression of IL-7 receptor α and begin to express IgM as part of a not yet functional B-cell receptor (BCR).^1,2 Immature-transitional B-cells are found in peripheral blood. They still express CD10, co-express IgD and IgM and present an activated phenotype with high expression of CD24 and CD38.^3-5 These cells home mainly to the spleen where they differentiate into CD27+IgM+ memory B-cells.³ Immature-transitional B-cell frequencies are high in healthy children and slowly decrease with age until stable range levels are reached around 13 y of age.⁶ Their role, especially in the spleen marginal zone (MZ) is to provide a first line defense against quickly replicating pathogens, such as encapsulated bacteria. Mature B-cells can be found in the periphery as well as in secondary lymphoid organs. They are negative for CD10 and are able to recognize antigens (Ags). In particular, mature naïve B-cells brightly express surface IgD and poorly IgM while mature memory B-cells express CD27 and surface IgG, IgA, or IgE. The pool of resting memory B-cells increases over time upon encounter with Ags through vaccination or natural infection.^7,8 Resting mature B-cells, both naïve and memory, are characterized by high expression of CD21 while they downregulate this molecule after activation.⁹ In the absence of CD21 or Ig measurements, the frequency of CD27- naïve B-cells is reciprocal to the one of CD27+ memory B-cells.¹⁰ A summary of the phenotypic, molecular, and functional changes taking place during B-cell development is shown in Table 1.

Table 1. Phenotypic, molecular, and functional changes during B-cell development

B-cell stage	Location	Immuno-phenotype	Ig expression	Description	Modulation during healthy aging	Modulation upon chronic immune activation
Pro/Pre	BM	CD10+ IL7Rα+	None	Immunoglobulin gene rearrangements ongoing	Probably maintained	Not known
Immature	BM	CD5+ CD10+	IgM+	VDJ rearranged/ Surrogate BCR	Probably maintained	Not known
Immature-transitional	Periphery/spleen MZ	CD10+ CD27- (CD24^highCD38^high)	IgD^low/high IgM^high	Co-express IgD and IgM	Decline physiologically	Increased and dysfunctional (with acute activation)
IgM memory	Periphery/LNs:	CD27+	IgD^low IgM^high	Respond to T-independent Ags and secrete natural antibodies	Decreased in numbers but not in frequencies	Lower frequencies are measured
Naïve	Periphery/ LNs:	CD10- CD21^high CD27-	IgD^high IgM^low	Able to recognize Ags	Higher frequencies are measured	Higher frequencies are measured
Resting memory	Periphery/LNs:	CD10- CD21^high CD27+	IgD- (Switched Igs)	CSR and SHM	Lower frequencies are measured	Lower frequencies are measured
Mature activated	Periphery/LNs:	CD10- CD21-	Any Ig	Activated naïve or memory B-cells	Not known	Increased in numbers and frequencies
Double negative	Periphery	CD27-	IgD- (Switched Igs)	Exhausted B-cells with low SHM	Increased in numbers and frequencies	Increased in numbers and frequencies
CD24-CD38-	Periphery	CD24- CD38-	Not known	Memory B-cells involved in inflammation	Higher frequencies are measured	Not known

The ability of mature B-cells to produce highly specific Abs in the germinal center (GC) is dependent on Ig affinity maturation.¹¹ This is a highly regulated process, controlled by the interaction between T-cell membrane CD40 ligand with the CD40 molecule on the surface of activated B-cells and by soluble T-cell cytokines in the GCs. Ligation of these cytokines, in particular of IL-4, IL-10, and IL-21, to cognate receptors on B-cells causes direct downstream activation of STAT6 and NFkB and particularly of activation-induced deaminase (AID) transcription.¹¹ The action of AID introduces double strand breaks in the Ig germline leading to the introduction of point mutations called somatic hyper-mutations (SHMs) in the V region of both naïve and memory B-cells, and to class switch recombination (CSR) in the C region of naïve B-cells. SHMs are meant to modulate the Ab-Ag affinity while CSR to allow the transcription of the selected V region in association with IgG, IgA, or IgE having improved effector functions than IgM.¹¹

Switching from IgM to IgG production and secretion is normally a phenomenon restricted to Ag triggered B-cells. Nonetheless, bacterial CpG and viral double stranded RNA can directly activate B-cells through TLR9 and TLR3 ligation respectively, inducing IgG and IgA CSR.^12,13 Most TLRs are localized in intracellular compartments and ligate viral nucleic acid in endosomes; this localization ensures that TLRs will not bind to the host nucleic acid which is not normally accessible in these compartments and it does not trigger TLRs.¹⁴ Human B-cells express TLR1, TLR3 and TLR6 through TLR10 and their levels increase upon B-cell activation, particularly those of TLR9 and 10.^13,15 Maintenance of serological memory has been shown to be performed by long-lived PCs recirculating between the periphery and the BM and by memory B-cells being continuously restimulated by bystander T-cell cytokines and microbial products triggering TLRs in the absence of specific Ags.¹⁶

B-Cells during Healthy Aging

The pure effect of aging on the immune system is not an easy task to assess in humans, rather than in aged mice models, as decline of immune functions may cause several age associated inflammatory pathologies having an impact on the different parameters analyzed.¹⁷ In view of this, below we will discuss the immune features of healthy aged individuals i.e., centenarians and those individuals over 65 y of age having a general healthy status.¹⁸

The number or total B-cells in peripheral blood declines with age but whether this is due to decreased B-cell generation or extravasation from the BM is not clear. Despite the frequencies of B-cell precursors and immature B-cells seem to be maintained throughout life in the BM,^19,20 the observation that the proportion of HSC relatively to BM stromal cells is lower with old age may point out to reduced B-cell generation in aged individuals.^21,22 Nonetheless, while no data are available for humans, stromal cell-derived IL-7 levels seem to be reduced in old mice thus suggesting that the aged BM may not provide adequate conditions for proper B-cell development.²³ As previously stated, the frequency of immature-transitional B-cells in peripheral blood is physiologically low in aged individuals,⁶ therefore their role during healthy aging is not easy to assess. Though, one study related increased systemic autoimmunity to impaired IL-10 production by immature-transitional B-cells (identified as CD19+CD24^highCD38^high) in aged individuals.²⁴ Nonetheless, decreased numbers but not frequencies of spleen IgM+ memory B-cells are observed in elderly as compared with young individuals.²⁵ Decreased frequencies of peripheral IgM+ memory B-cells have instead been observed in the elderly.²⁶ This may relate to weak responses to pneumococcal challenge in elderly subjects.²⁷ Peripheral mature B-cells seem to be the most affected by the aging process both in terms of altered composition and function.²⁸ Discordant data on the distribution of naïve/memory B-cells in the elderly have been reported. A dramatic decrease of CD27+ memory B-cells in parallel to an increase of CD27- naïve B-cells has been described.²⁹ Nonetheless, naïve B-cells did not differ significantly between young and elderly individuals in terms of absolute numbers.²⁹ On the contrary, a reduction of CD27- naïve B-cells, with no reciprocal increase of CD27+ memory B-cells, had previously been reported in older individuals and centenarians.³⁰ A possible explanation for these conflicting data has later been suggested to be the substantial age difference between the cohorts analyzed.³¹ Different levels of inflammation and immune activation may be observed between elderly individuals and centenarians.³² With this respect, discrimination of naïve and memory B-cells by surface CD27 alone does not in fact take into account different ratios between resting and activated cells (being CD21^high or CD21^low respectively) as well as different ratios between switched and IgM+ memory B-cells. Moreover, the measurement of resting memory B-cells in the elderly may be biased by another factor, i. e. the accumulation of a CD27-IgD- double negative (DN) B-cell population.³¹ These expanded B-cells in the elderly have in fact been suggested to be late or exhausted memory B-cells that have lost the expression of CD27.³¹ Interestingly, B-cell immunophenotyping based on the activation markers CD24 and CD38 also revealed another population of CD24-CD38- B-cells to be increased in elderly individuals.⁵ These cells produce the pro-inflammatory cytokine TNF-α which might contribute to the inflammatory status often observed in the elderly.⁵ A summary of the effect of healthy aging on the different B-cell subpopulations has been incorporated in Table 1.

As previously stated, the ability of B-cells to produce Ag-specific Abs relies on the upregulation of AID leading to SHM and CSR in the GC. Physiological age-related thymic involution and reduced T-cell generation may affect both the required number of interactions between T- and B-cells and the yield of soluble T-cell cytokines in the GCs.³³ As a result of this and of intrinsic B-cell defects, the ability of B-cells to upregulate AID expression and to undergo CSR decreases with age.²⁹ In addition, the Ig repertoire of different B-cell subtypes including DN B-cells, from elderly individuals presents a reduced number of SHMs.³⁴

Signs of Premature B-Cell Senescence are Observed during Human Immune Deficiency Virus (HIV)-1 Infection and in Other Diseases Characterized by Chronic Immune Activation

Among all immune diseases, human immune deficiency virus (HIV)-1 infection has been extensively studied. Intrinsic defects in B-cells have to date been described in chronically HIV-1 infected individuals.³⁵ These include hyperactivation and increased susceptibility to exhaustion and apoptosis in vitro.^36,37 Moreover, the B-cell repertoire of infected patients is affected by an increase of dysfunctional immature-transitional B-cells, in conditions of high viremia,^38,39 and by memory B-cell depletion which occurs during the first stages of primary infection resulting in loss of serologic memory throughout the later stages of chronic infection.^40,41 Initiation of antiretroviral therapy (ART) early during primary infection in adults or within the first year of life in HIV-1 vertically infected children is able to preserve high numbers and function of antigen-specific memory B-cells gained during vaccination.^40,42 Nonetheless, B-cell defects may only partially be reversed even upon early ART and suppression of viral replication as patients still experience impaired vaccination responses to influenza and other vaccines.⁴³ Residual chronic immune activation related to microbial products translocation between damaged tissue in the gut and other body compartments has been reported even in treated patients.⁴⁴

A number of studies describe similar B-cell impairments for HIV-1 infected individuals and healthy elderly people.^31,45-47 In particular, altered B-cell subpopulations resembling DN B-cells in the elderly³¹ have also been described in HIV-1 infected individuals.⁴⁷ Moreover, AID expression is also impaired during HIV-1 infection.⁴⁷ This may be due to suboptimal T-cell cytokine concentration and T-cell help, to B-cell hyperactivation or to a direct effect of HIV-1 viral components on AID transcription.^47-49 However, this last hypothesis is less likely to be correct as different viral components were shown to have opposite effects on AID.^48,49 Recently, high frequencies of DN B-cells, in parallel to high frequencies of mature activated (MA) B-cells (CD19+CD10-CD21-),⁹ were confirmed in HIV-1 infected pediatric patients.⁵⁰ Interestingly, a longer time on ART related with lower DN while an earlier start of ART related with lower frequencies of MA in this population.⁵¹ This suggests that different levels of chronic immune activation during HIV-1 infection may actually relate with the appearance of these senescent and activated B-cells. The relation between high frequencies of both DN and MA with immune senescence during vertical HIV-1 infection was confirmed by direct correlation with the frequencies of terminal effector CD4+ T-cells.⁵⁰ Moreover, shorter lymphocyte telomere length was reported for viremic HIV-1 infected children as compared with healthy controls.⁵² Interestingly, increased frequencies of both DN and MA were also observed in the blood of kidney transplant pediatric patients.⁵³ In view of this, it is interesting to consider HIV-1 infected and kidney transplant patients as different models of secondary immune deficiency characterized by chronic immune activation. As during HIV-1 infection in fact, patients with solid organ transplantation may also experience chronic immune activation due to the persistence of a foreign antigen (the transplanted organ) and are immune deficient because of the immune suppressive therapies administered after transplantation. Previously, premature senescence of T-cells was observed in long-term survivors of kidney transplantation.⁵⁴ Increased DN-like B-cells associated with a high disease activity were also observed in the blood of patients affected by systemic lupus erythematous (SLE) but not in patients with chronic hepatitis C virus (HCV) infection.⁵⁵ During HCV infection, immune activation with high frequencies of MA B-cells has also been observed⁵⁶ whereas high frequencies of dysfunctional immature transitional B-cells have been observed during both SLE⁴ and HCV infection.⁵⁶ This may relate with different degrees of inflammation and immune activation in these pathologies. During HIV-1 infection in fact, dysfunctional immature transitional B-cells increase only during the acute phases of the infection and tail off during the chronic phase.⁴ In view of this, it can be hypothesized that chronic rather than acute inflammation and immune activation, may lead to premature immune senescence with high frequencies of DN B-cells. A summary of the effect of chronic immune activation on the different B-cell subpopulations has also been incorporated in Table 1.

Implications of B-Cell Senescence for Vaccination

Loss of protective immunity and impaired vaccination responses are common features between healthy elderly individuals and immune compromised patients. Due to high risk of complications in case of influenza infection, yearly immunization against seasonal influenza is highly recommended for both elderly and immune compromised individuals.⁵⁷ Therefore, data on vaccination response in these 2 populations is mostly available for influenza. The role of impaired AID expression on the success of such vaccination has previously been pointed out in both healthy elderly individuals and HIV-1 infected patients.^46,58 In both these studies, the ability of B-cells to upregulate AID upon in vitro polyclonal stimulation was assessed prior to A(H1N1)pdm09 vaccination and in both cases the distribution of the different levels of AID fold increase, measured in a mixed cohort of elderly and younger individuals in one case and of HIV-1 infected and age-matched healthy individuals in the other, related to the magnitude of antibody response measured 1 mo after vaccination.^46,58 Recently and in a similar manner, the frequencies of both DN and MA B-cells measured in a mixed cohort of vertically HIV-1 infected patients and age-matched healthy subjects prior to 2012–2013 seasonal influenza vaccination could predict the strength of humoral response to the single strains contained in the vaccine.⁵⁰ These data suggest that measuring the frequencies of DN and MA B-cells could be a useful tool for vaccination design, in terms of number of vaccine doses, in both elderly and immune compromised individuals. Concerns exist on whether vaccination against seasonal influenza should be performed with one or multiple vaccine doses in populations at high risk of complications due to influenza infection. A significant increase of polyclonal antibodies with low affinity and increased reactivity to lymphocytes (named anti-lymphocyte antibodies or ALA) was reported after 2012–2013 seasonal influenza vaccination in both HIV-1 and kidney transplant pediatric patients but not in healthy controls and this directly correlated with the frequencies of both DN and MA.⁵³ These additional data suggest that the quality of an immune response triggered by possibly any vaccination in immune compromised individuals may depend on the degree of B-cell senescence and on their activation status.⁵³ Another recent report also related impaired antibody response to influenza vaccination in both HIV-1 infected or uninfected aging women to immune activation and inflammation measured by high levels of plasma IL-21 and of TNFα respectively.⁵⁹

Conclusions and Future Perspectives

Several evidences indicate that chronic immune activation may cause the premature aging of the immune system and that this can negatively impact and even predict responses to influenza vaccination. However, whether upregulation of AID expression in vitro or the frequencies of MA and DN may be confirmed as predictive biomarkers of vaccination response in immune compromised individuals needs to be tested in large studies including different vaccine types and different models of immune disorders starting from SLE where high DN B-cells have previously been reported.⁵⁵ The need of such biomarkers is particularly critical for vaccination of severe immune compromised individuals as they may need multiple vaccine doses in order to achieve immune-protection.⁶⁰ Other populations experiencing suboptimal antibody production during vaccination and chronic immune activation include the large number of patients affected by inflammatory bowel diseases (IBD). Several B-cell disturbances, including B-cell hyperactivation and dysfunctional responses to TLR agonists, have in fact been described in patients affected by IBD.^61-63 Whether signs of premature aging of the immune system possibly related to suboptimal antibody production during vaccination exist in these patients has never been investigated.

In addition, the nature of DN B-cells must be further investigated. In the elderly, these cells have recently been characterized as having a tissue trafficking phenotype in terms of expression of different chemokine receptors and as being able to produce granzyme B.⁶⁴ Whether the pro-inflammatory environment observed during chronic immune activation in younger individuals could similarly influence the trafficking phenotype and function of DN B-cells also in this population is not known. Moreover, as previously stated, DN B-cells have been defined as late or exhausted memory B-cells.³¹ In view of this, a low number of SHM should not be expected in this population as SHMs would accumulate at each cell division in pre-established memory B-cells during specific recall or bystander activation. The possibility of increased circulation of microbial products between body compartments due to damaged and inflamed tissue cannot be excluded nor in immune compromised or in healthy aged individuals.^44,65 Increased microbial products would trigger non-specific B-cell activation and potentially cause CD27-IgD+ naïve B-cells to switch into CD27-IgD- B-cells.⁴⁷ With this respect, it would be interesting to assess the antigenic specificity of DN B-cells in order to verify whether they belong to the memory B-cell pool. These theories on the origin of DN B-cells have been summarized in Figure 1.

Figure 1. (A) DN B-cells may be late or exhausted memory B-cells which have lost the expression of CD27 carrying a decreased number of SHMs in the Ig V region. However, a low number of SHM should not be expected in this population as SHMs would accumulate at each cell division in pre-established memory B-cells during specific recall or bystander activation. (B) DN B-cells may be naïve B-cells being triggered by polyclonal stimuli due to increased circulation of microbial products between body compartments due to damaged and inflamed tissue. In this scenario, CD27-IgD+ naïve B-cells would switch to CD27-IgD- B-cells.

10.4161/hv.28698