AUTHORS
Christopher Maximilian Arends, MD, Rajiv Sharma, PhD2, Eugenia Wang3, Thais Pereira Aljentera4, Klara Kopp, MD5, Lisa Ma6, Daniel Nachun7, Frederik Damm, MD, PhD8, Irving Weissman, MD9, Siddhartha Jaiswal, MD/PhD10
AFFILIATIONS
1 Stanford University School of Medicine, Stanford, California, United States
2 Stanford University School of Medicine, Stanford, California, United States
3 Stanford University School of Medicine, Stanford, California, United States
4 Stanford University School of Medicine, Stanford, California, United States
5 Charite University Medicine Berlin, Berlin, Germany
6 Stanford University School of Medicine, Stanford, California, United States
7 Stanford University School of Medicine, Stanford, California, United States
8 Charite University Medicine Berlin, Berlin, Germany
9 Stanford University School of Medicine, Stanford, California, United States
10 Stanford University School of Medicine, Stanford, California, United States
ABSTRACT
With aging, hematopoietic stem cells (HSCs) tend to preferentially differentiate into myeloid cells at the expense of lymphoid cells. This myeloid differentiation bias occurs in mice and humans and is associated with chronic inflammation and impaired adaptive immunity. In aged mice, the antibody-mediated depletion of myeloid-biased HSCs was shown to rebalance lineage output and improve immune function (Ross et al. Nature 2024). In humans, however, the mechanisms and pathways underlying myeloid bias during aging remain unknown.
To address this we conducted genome-wide association studies (GWAS) on the ratio of myeloid to lymphoid cells in peripheral blood in 196,841 participants ≥ 60 years from the UK Biobank (UKB). This GWAS identified 177 genome-wide significant loci, many of which relate to lineage-specific cytokine signaling including IL7 and CSF3. To enrich for genes and pathways relevant in HSCs, we prioritized variants associated with opposite changes in myeloid and lymphoid cell count, associated with clonal hematopoiesis, and showing a significant age-interaction. This converged on two loci, tagging TET2 and CD164.
CD164 is a surface protein highly expressed on hematopoietic stem and progenitor cells (HSPCs) with putative functions in cell adhesion, migration and proliferation. Fine-mapping of the CD164 locus and co-localization with regions that are chromatin-accessible in HSCs identified two independent variants: rs2275652 is located at the transcription start site and rs1546752 in a distal enhancer region of CD164. To assess their effect on CD164 protein expression we performed flow cytometry on 40 BM samples from healthy donors. Among all marrow cells, CD164 was most highly expressed in the CD34+CD133+CD45RA-CD90+ long-term (LT-)HSC population. The myeloid bias increasing alleles rs2275652-C and rs1546752-A were independently associated with increased expression of CD164 in LT-HSCs. We therefore hypothesized that CD164 causally regulates lineage commitment in HSCs: increased CD164 expression promotes myeloid differentiation, whereas decreased CD164 expression promotes lymphoid differentiation.
To test this hypothesis, we established an in vitro co-culture system that promotes the development of myeloid and lymphoid progenitors from LT-HSCs. Mobilized peripheral blood (MPB) samples were CRISPR-edited to knock out CD164 (or edit AAVS1 as a control) and 500 sorted LT-HSCs per replicate were seeded on AC6.2.1 cells under external cytokine support with FLT3L, IL6, TPO and IL7 (n=6 replicates per condition). After three weeks, lymphoid and myeloid progenitors were quantified using flow cytometry and CITESeq. CD164 KO cultures had 2.4x higher numbers of CD34+CD33-CD10+ lymphoid progenitors (1.2 vs 0.5 per input HSC, p=0.002) compared to controls. Similarly, CITESeq on CD33low cells revealed significant enrichment of HSC, multipotent progenitor (MPP), and common lymphoid progenitor clusters in CD164 KO cultures (3 replicates per condition).
Two observations led us to propose a mechanism through which CD164 regulates differentiation. First, CD164 has a YHTL-motif in its cytoplasmic domain that mediates endocytosis and transport to the lysosomal compartment. In our assays, we observed that surface CD164 gets endocytosed within 4 hours after HSPCs are taken into culture and that endocytosis of CD164 occurs in THP-1 cells upon anti-CD164 antibody binding. Second, the rs2275652-C and rs1546752-A allele dosage was robustly associated with increased plasma levels of FLT3L (0.06 SD per allele, p=3.9 × 10-15, 0.06 SD per allele, p=1.3 ×10-17, respectively) in 50,760 individuals from UKB with available Olink proteomic data. We hypothesized that this reflects decreased surface abundance of the corresponding cytokine receptors on target cells. Indeed, CD164 protein expression was negatively correlated with FLT3 expression in CD34+CD133+ HSPCs across 14 MPB samples assessed by flow cytometry (R = -0.57, p = 0.034). Lentiviral overexpression of CD164 reduced FLT3 surface abundance in THP-1 cells. Together, these findings support a model in which CD164 regulates lineage-instructive cytokine receptors such as FLT3 in HSCs/MPPs, possibly by facilitating endocytosis and lysosomal degradation.
In summary, our data suggest that CD164 plays a role in tuning HSC differentiation. Blocking CD164 might provide a therapeutic strategy to increase lymphopoiesis and mitigate the consequences of myeloid bias in aged individuals.
DISCLOSURES
Christopher Maximilian Arends, MD: No relevant conflicts of interest to declare.