Mononuclear phagocytes and marker modulation: when CD16 disappears, CD38 takes the stage

To the editor:

We have read with great interest the article from Döbel et al,¹  which shows the ability of freshly separated 6-sulfo LacNac (slan)–positive dendritic cells (slanDCs) to internalize immunoglobulin G (IgG)-opsonized particles and aggregated IgG via FcγRIIIA/CD16-dependent phagocytosis. SlanDCs are a subset of “nonclassical monocytes” expressing CD16.^2,3  Schäkel et al⁴  have previously elegantly demonstrated that these cells in vitro undergo rapid spontaneous maturation, becoming potent producers of interleukin-12, and that these events in vivo are restrained by CD47 expressed on the membranes of erythrocytes; now, the same group reveals that metalloprotease-dependent shedding of the low-affinity IgG receptor CD16 accompanies these changes.¹ 

We noted the downregulation of CD16 in all CD16-expressing mononuclear myeloid cells in experiments analyzing the cytokine response to Toll-like receptor (TLR) 4 and TLR7/8 agonists in lithium-heparinized whole blood (and thus in the presence of erythrocytes) (Figure 1A-C). As shown in Figure 1C, this modulation is concomitant with the production of tumor necrosis factor α (TNFα). Thus, microbial stimuli can bypass the restraining effect mediated by erythrocytes on the maturation of nonclassical monocytes.

Extending the observations of Belge et al⁵  who proposed HLA-DR as a stable marker for nonclassical monocytes, to track the loss of CD16 in monocyte subsets we developed a flow cytometric strategy which identifies nonclassical monocytes, together with other monocyte and myeloid DC subsets, taking advantage of the differential expression pattern of CD38, thus bypassing the need for CD16 as a critical selection marker. Nonclassical monocytes express CD38 at very low levels, unlike other monocytes and DC subsets (Figure 1D-E), consistent with their proposed developmental derivation from classical monocytes³  and their well-differentiated phenotype.

Because CD38 expression is stable, at least following short-term stimulation (Figure 1F), we have exploited this strategy to monitor CD16 downregulation in nonclassical monocytes and, as a control, CD62L shedding in classical monocytes⁶  after stimulation with TLRs. As shown in Figure 1G, the metalloprotease inhibitor Ilomastat (GM6001) dose-dependently prevents CD16 and CD62L shedding induced by the brief incubation with resiquimod (R848, a TLR7/8 agonist).

In conjunction with the results presented by Döbel et al,¹  our data thus demonstrate that the process of ectodomain shedding of 2 distinct surface receptors occurs in the same circumstances and by overlapping mechanisms in both nonclassical monocytes (and slanDCs) and classical monocytes, underscoring the similarities between the activation-dependent enzymatic machinery of 2 closely related monocyte family members. Furthermore, we show that TLR engagement can drive a response that overcomes the inhibition exerted by red blood cells in nonclassical monocytes. Our data extend the findings from Döbel et al not only to nonclassical monocytes other than slanDCs, but also to the whole monocyte family.

Nonclassical monocytes and circulating slanDCs are expanded during sepsis and HIV viremia.^7-9  These “patrolling monocytes”¹⁰  are physiologically present in the pool of blood leukocytes crawling along the luminal side of vessels, and their spontaneous activation is restrained by erythrocytes. Our evidence suggests that the systemic spreading of microbial products occurring in pathological conditions can bypass this constraint, and that the novel marker CD38 can be successfully used to track recently activated nonclassical monocytes that, irrespective of expression of slan, have lost their defining marker CD16.