Depletion of B cells rejuvenates the peripheral B‐cell compartment but is insufficient to restore immune competence in aging

Abstract Aging is associated with increasing prevalence and severity of infections caused by a decline in bone marrow (BM) lymphopoiesis and reduced B‐cell repertoire diversity. The current study proposes a strategy to enhance immune responsiveness in aged mice and humans, through rejuvenation of the B lineage upon B‐cell depletion. We used hCD20Tg mice to deplete peripheral B cells in old and young mice, analyzing B‐cell subsets, repertoire and cellular functions in vitro, and immune responsiveness in vivo. Additionally, elderly patients, previously treated with rituximab healthy elderly and young individuals, were vaccinated against hepatitis B (HBV) after undergoing a detailed analysis for B‐cell compartments. B‐cell depletion in old mice resulted in rejuvenated B‐cell population that was derived from de novo synthesis in the bone marrow. The rejuvenated B cells exhibited a "young"‐like repertoire and cellular responsiveness to immune stimuli in vitro. Yet, mice treated with B‐cell depletion did not mount enhanced antibody responses to immunization in vivo, nor did they survive longer than control mice in "dirty" environment. Consistent with these results, peripheral B cells from elderly depleted patients showed a "young"‐like repertoire, population dynamics, and cellular responsiveness to stimulus. Nevertheless, the response rate to HBV vaccination was similar between elderly depleted and nondepleted subjects, although antibody titers were higher in depleted patients. This study proposes a proof of principle to rejuvenate the peripheral B‐cell compartment in aging, through B‐cell depletion. Further studies are warranted in order to apply this approach for enhancing humoral immune responsiveness among the elderly population.


Antibody quantification
Quantification of mouse or human IgM in supernatants and quantification of IgG specific for OVA, HSA or BSA in mouse sera were performed by ELISA 3 .

Immunoglobulin heavy chain repertoire analysis
Mouse splenic B cell IgH repertoires were investigated by high throughput sequencing.
DNA was extracted from splenic B cells of 3 old (>20 month), 2 young and 2 B celldepleted Balb/C mice using DNA purification kit (QIAGEN). Mouse Ig heavy and light chains were amplified by semi-nested PCR and sequencing as detailed below.
Human peripheral blood repertoires were investigated by spectratyping using DNA extracted from peripheral blood mononuclear cells (Qiagen, Crawley) as described 4 . The CDR3 region of the rearranged IgH was amplified using the IgH Gene Clonality Assay (Invivoscribe) and products were run by capillary electrophoresis (ABI PRISM Genetic Analyzer) according to manufacturer instructions (Applied Biosystems). The results were analyzed using GeneScan (Applied Biosystems) to determine the peak sizes for each main peak in the spectratype. The peaks at 3-bp intervals were further analyzed as described below.
PCR, sequencing and mouse Ig repertoire determination IgH and IgL repertoires were determined by PCR and sequencing. For the mouse Ig heavy, kappa and lambda chains we used three separate primer sets of 8, 8, and 3 forward primers respectively within the FR1 region, and 2 reverse primers for each chain type 5 .
The 5′ sequence of the primer (IDT, HPLC grade) contained an adaptor that is specific for the 454-FLX emulsion PCR, followed by a 10bp multiplex identifier (MID) tag (supplementary Figure 2). PCR was performed with the XP Thermal Cycler (BioER). Each reaction contained 10-100ng of genomic DNA as a template, 0.6-1.2mM dNTP (Fermentas), 0.4uM primers, 1mM MgCl2 and 1-2U of proofreader Taq polymerase (AB gene or Finnzyme). According to the manufacturer, the polymerase error rate is of an order of magnitude of 10-7bp. Each round of PCR contained an initial denaturation step at 98°C for 3 minutes. Denaturation and extension were carried out at 94°C for 60s and 72°C for 90s throughout, respectively.
The annealing temperature in the first 5 cycles was 52.5°C for 60s, followed by 5 cycles at 52°C and 25 additional cycles at 51.5°C. The second PCR round was performed by the same cycling program as in the first round. PCR products were separated by electrophoresis on 2% agarose gel containing ethidium bromide. Sharp bands of around 400bp were cut out of the gel, and DNA was cleaned by QIAquick gel extraction columns (Qiagen) according to the manufacturer's protocol. Two distinct rounds of high-throughput sequencing were performed the 454 flex titanium instrument by Dyn Diagnostics, Israel.
For each round, DNA concentration and the quality of cleaned PCR products were measured by a NanoDrop 3300 Fluorospectrometer (Thermo Scientific) using Quant-iT™ PicoGreen® dsDNA Reagent (Invitrogen) according to the manufacturer's protocol.
Sequencing data were processed, annotated, assigned to clonally-related groups, and analyzed as described 6 .

Analysis of BrdU Incorporation
Mice were analyzed for BrdU incorporation 65 days after depletion, when reconstitution of the peripheral compartment was completed 7 . BrdU labeling in vivo, staining and analysis was as previously described 7 . The data on total B cell numbers, the fraction of cells in each B cell subset, and the fraction of BrdU-positive cells within each subset were recorded and analyzed, employing a unique mathematical model, as described below.
Mathematical modeling of B cell development population dynamics and B cell depletion The model (Supplementary Figure 3A We neglected BrdU toxicity (and thus assigned the same death rates to labeled and unlabeled cells), as in our previous work; we found that when the BrdU experiment is very shortas in this case, only 7 daysthe latter assumption holds.