Numbers of circulating immune cells in the study cohort are representative for healthy aged adults
Upon aging, the composition of the immune system changes. The median age of the studied cohort was 72 years (range: 59–103 years), as most of the donors were children of a centenarian who carried the p.P522R variant, and only a few were centenarians or siblings from a centenarian (details provided in Methods section and Table 1). To assess to what extent this cohort represented the general (age-matched) population, we first compared absolute counts of multiple innate and adaptive cell populations with reference values from healthy age-matched cohorts.(24, 26, 27) (ref 22: van der Pan et al, manuscript in revision) Overall, innate, CD4 T- and B-cell counts in the donors were representative of their age category, with exception of IgG3, IgG4 and IgA2 plasma cells, which were slightly elevated in several donors. In three donors (carriers, one family), a subclinical B-cell expansion was observed. Additionally, two donors (non-carriers, different families) in whom we observed B-cell aberrancies were excluded from all further analyses and referred to a hemato-oncologist for further evaluations (the expanded B-cell phenotypes observed in the initial research-based screening were: CD19 + CD20 + CD5 + CD21 + CD27 + IgM + IgD-/dim and CD19 + CD20 + CD5 + CD21dimCD27 + IgG1dim). For CD8 T- and NK-cell subsets no age-matched reference values were available. No impact of sex on cell counts was observed. An impact of age on cell counts was found for several subpopulations. To reduce impact of sex or age on the comparison between p.P522R carriers and non-carriers, cohorts were sex- and age-matched as much as possible. Lastly, we observed a pronounced effect of pedigree, as in several subsets cell counts from all first-degree family members tended to cluster together, even when individuals were living at different locations for many years (Figure S1).
Several Parameters In P.p522r Carriers Resemble Younger Healthy Donors
To evaluate the impact of PLCγ2 carriership on cell counts in innate and adaptive immune cell subsets, we related this information to the carriership status. Overall, p.P522R carriers tended to have higher total and immature B-cell numbers, and significantly more CD20 + + CD21-CD24 + naive and IgG1 + MBCs (Fig. 1A). This pattern was observed in at least half of the families with mixed carriership (CD20 + + CD21-CD24 + naive B cells; 3/6, IgG1 + MBCs; 5/6 families). Interestingly, upon comparing median cell counts carriers and non-carriers with an internal reference cohort (n = 25, 9 females, aged 18–54, median age: 27 years), the carrier cell counts more closely resembled the counts of the younger cohort than the non-carriers (Fig. 1A). Interestingly, for the CD20 + + CD21-CD24 + naive B cells, the median cell count was slightly higher in carriers than in the younger cohort.
Next, we evaluated expression of several surface markers: HLA-DR, CD62L, CD16, CD33 and FcεRI on innate immune cells, CD45RA, CD27, CD28 and CD3 on T cells, and CD20 and CD21 on B cells. Interestingly, we found a reduced expression of FcεRI on CD62L- FcεRI + classical monocytes (CD62L- FcεRI + cMo), basophils and plasmacytoid dendritic cells (pDCs) in p.P522R carriers with a median age of 71 years (range 59–103 years) versus non-carriers with a median age of 74 years (range 61–83 years) (Fig. 1B). Median FcεRI expression in carriers was below, or in the lower range, of the young reference cohort (n.s.). In contrast to FcεRI expression, the expression of CD33 on CD62L- FcεRI- cMos was increased in the p.P522R carriers, and in the high range of the 95% confidence interval of the younger reference cohort, while non-carriers resided on the lower end of the 95% confidence interval of the younger reference cohort (n.s.) (Fig. 1B).
Higher Plcγ2 Expression In Immune Cells Of P.p522r Carriers
Then, we evaluated the PLCγ2 expression levels in different leukocyte populations in the total cohort (irrespective of carriership). The highest PLCγ2 expression was found in eosinophils, with high variation between individuals, and, more consistently, in antigen-experienced B cells (Fig. 2A). When comparing carriers with non-carriers, the levels of PLCγ2 in all evaluated B-cell subsets were consistently higher in p.P522R carriers than in non-carriers (Fig. 2BC). Likewise, innate cells tended to have slightly higher median values for PLCγ2 expression in p.P522R carriers compared to non-carriers (Fig. 2BC).
To summarize our findings in this first analysis (e.g. Cohort I): based on cell counts, Cohort I seemed representative for healthy individuals of this age group. Whenever differences in cell counts were found between p.P522R carriers and non-carriers, carriers more closely resembled the younger control cohort. Lastly, PLCγ2 expression tended to be higher in carriers, mostly on antigen-experienced B cells.
Higher Levels Of Phosphorylated Plcγ2 In Stimulated B Cells From P.p522r Carriers
In a second analysis (Cohort II), we evaluated the effect of p.P522R on PLCγ2 activity in several B-cell subsets derived from 14 healthy older adults (6 non-carriers, 8 carriers) (Table 1, Text S1). As PLCγ2 is located downstream of the BCR, we assessed whether specific elements of the signaling pathway were affected by the carriership status. The BCR was stimulated with IgM (stimulation with IgG Fabs was also tested, but results were considered unreliable due to high background signal). In the steady state, we observed no significant difference between the levels of phosphorylated PLCγ2 (pPLCγ2) in B cells from p.P522R carriers and non-carriers (Fig. 3A). Upon BCR stimulation with IgM Fabs, a similar percentage of unswitched MBCs was activated, but the integrated mean fluorescent intensity (iMFI) of PLCγ2 was almost two-fold higher in carriers than in non-carriers (3667 vs 2051, respectively), implying a stronger activation in carriers (Fig. 3A). Moreover, levels of pPLCγ2 tended to be higher in stimulated pre-GC B cells in carriers. According to expectations, no increase in pPLCγ2 was observed upon IgM stimulation in the (IgM-) class-switched MBCs in either group.
Higher Calcium Flux In B Cells Upon Bcr Stimulation In P.p522r Carriers
Next, we evaluated the effect of p.P522R further downstream of the BCR by measuring the calcium release (‘calcium flux’) upon BCR stimulation using IgM Fabs. After a pilot experiment using IgM and IgG Fab stimulation in 12 donors, we found IgM stimulation to be most robust (Figure S2). We observed no differences in calcium flux in pre-GC B cells derived from Cohort II donors (Fig. 3B), but there was a trend towards more robust calcium flux in p.P522R carriers vs non-carriers upon Fab stimulation in unswitched MBCs (‘Total flux’ and ‘Flux at peak’) (Fig. 3C). However, this trend was not observed when comparing the calcium flux of all donors included in cohort I and II (Figure S3). Again, stimulation with IgM Fabs did not result in calcium release in class-switched MBCs, thus confirming that the measured calcium flux is truly due to BCR-specific stimulation (Fig. 3D).
No Difference In Number Of Cell Divisions Between Carriers And Non-carriers
Stronger B-cell activation upon stimulation in p.P522R carriers may result in more robust proliferation upon antigen encounter. To test this hypothesis, we evaluated B-cell replication history with the KREC (kappa-recombination excision circle) assay. Despite generally low cell counts, we were able to determine that pre-GC B cells measured in all 14 individuals in Cohort II had undergone on average 1.58 cell divisions, while unswitched MBCs measured in 8/14 individuals, had undergone on average 7.82 cell divisions (Fig. 3E). These findings are in line with previous publications.(31) We observed no clear difference in replication history of B-cell subsets between p.P522R carriers and non-carriers, possibly due to the limited number of donors in the non-carrier cohort (replication history in pre-GC B cells: 1.43 vs 1.64; in unswitched MBCs: 7.12 vs 7.82 (Fig. 3F)). Lastly, although class-switched MBC samples were collected as well, the obtained cell numbers were generally too low to determine the number of undergone cell divisions.
Serum Ig Responses To Sars-cov-2 Vaccination Comparable Between P.p522r Carriers And Non-carriers
To test whether the effect of p.P522R on B-cell activation as observed in our in vitro experiments also translates to an in vivo effect (e.g. effect on antibody production), we recognized a window of opportunity in the current global vaccination efforts against SARS-CoV-2. We collected serum samples from 22 individuals (Cohort III), 7–14 weeks after receiving a second vaccination against SARS-CoV-2. Donors were vaccinated with Comirnaty® (Pfizer/BioNTech, 9 carriers, 9 non-carriers); Spikevax® (Moderna, 1 carrier); Vaxzevria® (AstraZeneca, 1 non-carrier), or did not indicate vaccine type (2 non-carriers). All donors developed prominent IgG responses against the viral Receptor Binding Domain (RBD) and Spike (S)-protein. Antibody responses against the viral Nucleocapsid (N)-protein – indicative of viral infection- were predominantly IgM responses (Fig. 4A, B). No differences were observed between p.P522R carriers and non-carriers.
Summarizing our (B-cell related) findings in Cohort II/III: PLCγ2 in unswitched MBCs from p.P522R carriers is more strongly phosphorylated upon stimulation compared to non-carriers. Moreover, the calcium flux in p.P522R carriers was increased relative to non-carriers, at least in individuals from Cohort II, who were selected based on fewer comorbidities. Due to technical limitations, we were unable to obtain reliable results for class-switched memory cells, such that it is currently unclear to what extent these data can be extrapolated to other MBC subsets. Despite clear differences between p.P522R carriers and non-carriers in the multiple in vitro functional experiments, in vivo antibody responses raised against SARS-CoV-2 vaccine were comparable between carriers and non-carriers.
Increased Ros Production Upon Fcr-mediated Stimulation Of Innate Immune Cells In P.p522r Carriers
Besides being located downstream of the BCR, PLCγ2 is also located downstream of FcRs, which are present on many cells of the innate immune system, including the microglia of the brain. Therefore, we evaluated ROS production and phagocytic activity upon FcR-mediated stimulation in several cell types of the innate immune system: neutrophils, classical, intermediate, and non-classical monocytes (cMos, iMos and ncMos, and their subsets), and myeloid and plasmacytoid dendritic cells (mDCs, pDCs, and their subsets).
Although the percentage of phagocytosing cells did not differ between p.P522R carriers and non-carriers (Fig. 5A), the iMFI, reflecting the amount of phagocytosed E. coli particles per cell (phagocytic capacity), tended to be lower in carriers. This trend was especially prominent in classical and non-classical monocyte subsets, less prominent in CD14- mDC subsets and iMos, and not observed in CD14dim mDCs (Fig. 5A, Figure S4).
In addition to phagocytosis, we evaluated ROS production upon phagocytosis of opsonized E. coli particles (PhagoBURST assay). No difference between p.P522R carriers and non-carriers was observed in the amount of ROS produced (Fig. 5A, Figure S4). To combine information from both assays, we investigated ROS production relative to the number of phagocytosed particles. We observed that this ‘normalized ROS production’ was increased in carriers, which was especially evident in CD62L-cMos (Fig. 5A). Also, we observed that the ROS production relative to number of phagocytosed particles tended to be increased (n.s.) in carriers across all evaluated monocyte subsets and neutrophils (Fig. 5B, Figure S4A-C). Lastly, we observed increased normalized ROS production in CD14- mDCs, but not in CD14dim mDCs (Figure S4D-E). Thus, the ROS generation per particle tends to be somewhat higher in several cell populations in p.P522R carriers compared to non-carriers.
Additionally, we stimulated samples with Phorbol 12-Myristate 13-acetate (PMA), which results in FcR-independent ROS generation. Surprisingly, we observed a trend towards decreased ROS production in p.P522R carriers compared to non-carriers, which was seen for neutrophils, monocyte subsets and DC subsets (Figure S5). This effect was the strongest in monocytes, especially in CD62L + and CD62L- cMos, where p.P522R carriers showed a significantly lowered ROS production compared to non-carriers.
Thus, upon stimulation with opsonized E. coli, we observed that p.P522R carriers had a lower phagocytic capacity despite overall similar ROS production upon phagocytosis of E. coli and thus had increased ROS production relative to the number of opsonized particles. Nevertheless, the total FcR-independent ROS generation was lower in these same carriers. In general, differences were most prominent in the monocyte subpopulations.