CD5L is a canonical component of circulatory IgM

Abstract

Immunoglobulin M (IgM) is an evolutionary conserved key component of humoral immunity, and the first antibody isotype to emerge during an immune response. IgM is a large (1 MDa), multimeric protein, for which both hexameric and pentameric structures have been described, the latter additionally containing a joining (J) chain. Using a combination of single-particle mass spectrometry and mass photometry, proteomics, and immunochemical assays, we here demonstrate that circulatory (serum) IgM exclusively exists as a complex of J-chain-containing pentamers covalently bound to the small (36 kDa) protein CD5 antigen-like (CD5L, also called apoptosis inhibitor of macrophage). In sharp contrast, secretory IgM in saliva and milk is principally devoid of CD5L. Unlike IgM itself, CD5L is not produced by B cells, implying that it associates with IgM in the extracellular space. We demonstrate that CD5L integration has functional implications, i.e., it diminishes IgM binding to two of its receptors, the FcαµR and the polymeric Immunoglobulin receptor. On the other hand, binding to FcµR as well as complement activation via C1q seem unaffected by CD5L integration. Taken together, we redefine the composition of circulatory IgM as a J-chain containing pentamer, always in complex with CD5L.

Keywords: CD5L; IgM; immunoglobulin; mass spectrometry; structure.

Fig. 1.

All circulatory IgM is pentameric with J-chain and CD5L incorporated. (A) Total levels of CD5L and IgM heavy chain (IgµC) show a strong correlation in serum (r = 0.98). Absolute concentrations were assessed by label-free quantitation by proteomics. The molecular ratio between CD5L and IgµC was approximately 1:7. Considering that each IgM pentamer holds 10 copies of IgµC, this advocates for the incorporation of one CD5L molecule per IgM pentamer. The dotted line indicates a linear regression model fitted to logarithmically scaled data. (B) In complex-centric protein profiling of pooled serum, CD5L elutes exclusively (>99%) with IgµC and J-chain as a single complex of ca. 1 MDa. Shown are the elution profiles of the individual protein chains obtained from proteomics on individual size-exclusion chromatography (SEC) fractions. The secondary elution peak of the J-chain corresponds to incorporation in dimeric IgA. (C) Levels of IgM-CD5L complexes and total IgM ELISA similarly show a high correlation in serum (r = 0.99). The ELISA makes use of a mAb (5B5) that can recognize CD5L when bound to IgM. Quantification was based on a recombinantly produced IgM-J-CD5L complex standard (Fig. 2 and SI Appendix, Fig. S5). The ratio between IgM-CD5L complexes and total IgM was approximately 1:1, implying that all serum IgM incorporates a CD5L molecule. The grey line indicates this 1:1 molecular ratio. (D) The principal configuration of circulatory IgM is a pentamer with J-chain and CD5L. Combining mass measurement by CDMS (Left) with proteomics (Right) confirmed that recombinant IgM-J (targeting wall teichoic acid of Staphylococcus aureus) is a pentamer with J-chain [(IgM)₅:(J)₁] (Top). IgM purified from pooled serum was similarly homogeneous (Bottom), though the average mass was shifted by +38 kDa and CD5L was detected by proteomics. This matches the mass increase expected for the incorporation of one CD5L molecule to form [(IgM)₅:(J)₁:(CD5L)₁].

Fig. 2.

Formation of IgM–CD5L complexes. (A) Isolated human peripheral blood B cells were cultured in vitro with both T-dependent (TD) and independent (TI) stimuli, after which supernatants were analyzed for secreted IgM and IgM–CD5L complexes by ELISA. For the T-dependent cultures, plasma cell differentiation was previously demonstrated (31). Only IgM devoid of CD5L was detected, irrespective of culture conditions. Data shown are from 2 and 3 different donors. (B) In vitro generation of recombinant IgM–CD5L complexes. Different molar ratios of CD5L to IgM [clone 2D5 (14)] were incubated in the presence of 0.1 mM reduced/oxidized glutathione (GSH/GSSG). Complex formation was assessed by ELISA. With a molar excess of CD5L, saturation was observed. Representative plots of n = 2 experiments. (C) Human CD5L contains two predicted unpaired cysteines, C191 and C300, that could potentially interact with IgM. Recombinant CD5L was produced with C191S or C300S mutations or both and was tested for their ability to form complexes with IgM as determined by ELISA. C191S disrupts complex formation, whereas C300S does not. Representative plots of n = 2 experiments. (D) Schematic representation of IgM–CD5L complex with highlighted C191 and C300 of CD5L and C414 of IgM-Fc. (E) Proposed structural model of IgM–CD5L complex. AlphaFold2 model of CD5L and J-chain was fitted into the structure of IgM core/Fc region with J-chain (PDB: 8ADY). (F) Detailed view on CD5L and J-chain within IgM.

Fig. 3.

CD5L in IgM decreases pIgR and FcαµR binding; complement activation and FcµR binding unaltered. (A) Complement activation by IgM was assessed by measuring C3b deposition or red blood cell (RBC) lysis upon binding of recombinant anti-biotin IgM ± CD5L to plates coated with biotinylated albumin, or biotinylated RBCs, respectively. Representative plots of n = 3 experiments. (B) Binding of recombinant IgM (2D5) ± CD5L and serum IgM to different IgM-binding receptors. Whereas binding to FcµR is unaffected by CD5L, binding to both pIgR and FcαµR is reduced for CD5L-containing IgM. Right panels: relative binding strength calculated in comparison to IgM without CD5L, data from n = 3 to 5 experiments. (C) Upon limited reduction of serum IgM with DTT, selective release of CD5L proved feasible (Left) which resulted in increased pIgR binding (Right). Data representative of n = 3 experiments. Bars and error bars represent mean and SE, respectively.

Fig. 4.

Secretory IgM is largely devoid of CD5L and contains instead the SC of pIgR. (A) Total levels of CD5L and IgM heavy chain (IgµC) show a remarkably strong correlation in serum (r = 0.98) but not in saliva and milk, as determined by label-free proteomics. CD5L levels in saliva and milk were very low, frequently even below the detection limit. These values are indicated at the Bottom as not determined (n.d.). The red line indicates a linear regression model fitted to logarithmically scaled serum data. (B) Levels of IgM–CD5L complexes and total IgM similarly show a high correlation in serum (r = 0.99) but not in saliva and milk, as determined by ELISA (Fig. 1). The gray line indicates a 1:1 molecular ratio. (C) The molecular ratio between CD5L and IgµC was approximately 1:7 in serum, closely matching the incorporation of one CD5L molecule per IgM pentamer. In saliva and milk, this ratio is much lower. (D) Similarly, while the ratio between IgM-bound CD5L and total IgM was approximately 1:1 in serum, IgM from saliva and milk was nearly devoid of CD5L. (E) Total levels of pIgR were very high in saliva and milk, frequently over an order of magnitude higher than IgµC levels. In serum, pIgR levels were much lower, sometimes even below the detection limit. These values are indicated as not determined (n.d.). (F) Consequently, levels of IgM-pIgR complexes and total IgM ELISA similarly show a good correlation in saliva and milk (r = 0.92) but not in serum. The gray line indicates a 1:1 molecular ratio. (G) The molecular ratio between pIgR and IgµC was in the order of 10:1 to 40:1 in saliva and milk, highlighting a large excess. In contrast, the ratio in serum was only 0.002:1. (H) While the ratio between IgM-bound pIgR and total IgM was approximately 1:1 in saliva in milk, serum IgM was principally devoid of pIgR.

Fig. 5.

Graphical summary. J-chain coupled IgM pentamers produced in B cells exclusively engage with CD5L when destined to be secreted into the bloodstream, whereas they attach to the SC of pIgR en route into secretory biofluids such as milk and saliva.