CD38 Is Robustly Induced in Human Macrophages and Monocytes in Inflammatory Conditions

Macrophages and their monocyte precursors mediate innate immune responses and can promote a spectrum of phenotypes from pro-inflammatory to pro-resolving. Currently, there are few markers that allow for robust dissection of macrophage phenotype. We recently identified CD38 as a marker of inflammatory macrophages in murine in vitro and in vivo models. However, it is unknown whether CD38 plays a similar marker and/or functional role in human macrophages and inflammatory diseases. Here, we establish that CD38 transcript and protein are robustly induced in human macrophages exposed to LPS (±IFN-γ) inflammatory stimuli, but not with the alternative stimulus, IL-4. Pharmacologic and/or genetic CD38 loss-of-function significantly reduced the secretion of inflammatory cytokines IL-6 and IL-12p40 and glycolytic activity in human primary macrophages. Finally, monocyte analyses in systemic lupus erythematosus patients revealed that, while all monocytes express CD38, high CD38 expression in the non-classical monocyte subpopulation is associated with disease. These data are consistent with an inflammatory marker role for CD38 in human macrophages and monocytes.

inTrODUcTiOn Chronic inflammatory diseases such as systemic lupus erythematosus (SLE) are commonly associated with functional deficits and a dramatically decreased quality of life. Innate immune responses mediated by macrophages and their monocyte precursors play a significant role as amplifiers and effectors of inflammation (1). Macrophages exposed to infectious and/or inflammatory stimuli adopt a phenotype characterized by high glycolytic activity and production of inflammatory cytokines and reactive oxygen species (ROS) [reviewed in Ref. (2)] aimed at pathogen destruction and immune effector recruitment. However, chronic activation of inflammatory macrophages may result in significant tissue and organ injury or dysfunction. Therefore, strategies that detect inflammatory macrophages may provide a means to diagnose inflammatory disease and/or follow treatment responsiveness.
CD38 was historically identified as a surface activation marker in T cells and later found to be expressed in additional immune and non-immune cell types, including macrophages (3). More recently, we found that CD38 is selectively upregulated in inflammatory murine bone marrow-derived macrophages (BMDM) (4). Other murine genes induced in inflammatory conditions are N-formyl peptide receptor (Fpr2) and G-protein receptor 18 (Gpr18), while Egr2 is induced in M(IL-4) conditions (4). CD38 labels most inducible nitric oxide synthase (iNOS) + macrophages stimulated in M(LPS + IFN-γ) conditions. By contrast, M(IL-4)stimulated BMDM fail to express CD38 (4). In vivo, CD38 +/ hi macrophages are also induced during infectious and sterile inflammatory processes such as murine models of Listeria infection (5,6), LPS-induced sepsis (4), DSS-induced colitis (7), and focal ischemia (8). This pattern of expression is consistent with macrophage CD38 serving as an inflammatory marker in both in vitro and in vivo murine models.
Although the exact functional role of CD38 in macrophages is unclear, CD38 may regulate cellular processes via dual receptor and enzymatic activities. As a receptor, CD38 can bind platelet and endothelial cell adhesion molecule/CD31 to promote cell adhesion and transport across endothelial cell layers (9). However, CD38 was first described as an enzyme capable of catalyzing the conversion of nicotinamide-adenine dinucleotide (NAD) to cyclic ADP ribose (cADPR) via cyclase activity and cADPR to ADPR via hydrolase activity. In acidic conditions, such as in endocytic compartments, CD38 can additionally convert NAD phosphate (NADP) to nicotinic acid adenine dinucleotide phosphate (NAADP) (10). cADPR signals the ryanodine receptor to induce Ca 2+ release from endoplasmic stores (11) and NAADP signals to Ca 2+ stores in lysosomes (12). The Ca 2+ signaling function of CD38 is evolutionarily conserved and is known to play crucial roles in infectious immunity, inflammation, and insulin signaling (3). Accordingly, CD38-deficient mice have increased susceptibility to infection as a consequence of reduced chemotactic activity and antigen presentation, modulation of bacterial uptake, and deficient T cell-dependent antibody and Th1 responses (5,6,(13)(14)(15)(16)(17)(18). Thus, murine animal models support a role for CD38 expression and activity in inflammatory disease processes.
CD38 is also associated with human inflammatory disease pathogenesis. In this context, both the presence and high percentage of CD38 + CD8 T cells are used as a biomarker to follow the progression of HIV infection into AIDS (19)(20)(21). Similarly, high percentages of CD38 + CD8 T cells are linked to anti-retroviral therapy unresponsiveness (19,22). CD38 is also used clinically as a poor prognosis biomarker in small B cell chronic lymphocytic leukemia (23,24). Moreover, CD38 expression is a prognostic biomarker in B cell acute lymphoblastic leukemia (25) and may play a role in multiple myeloma and acute promyelocytic leukemia (26). In addition, the anti-CD38 monoclonal antibodies daratumumab and isatuximab show therapeutic benefits in multiple myeloma (27)(28)(29) and are proposed as potential therapies for other conditions (30). However, while CD38 clearly plays a role in human disease and murine inflammation, little is known about its role in human inflammatory processes, particularly in macrophage-mediated innate immune responses.
Here, we ask whether CD38 can serve as an inflammatory marker in macrophages of human origin, similar to the murine system. In addition, we explore links between monocyte CD38 and disease activity in the human autoimmune-mediated inflammatory disease, SLE. Using both human monocytic cell line-derived and primary human monocyte-derived macrophages (MDMs), we report that surface CD38 is a selective marker of macrophages activated in M(LPS ± IFN-γ), but not M(IL-4) conditions. CD38 also contributes to maximal inflammatory cytokine secretion and glycolytic activity in human macrophages. Finally, we state that, while monocytes constitutively express CD38, high levels of CD38 expression in non-classical monocytes (NCMs) are linked to SLE disease and/or disease activity. Taken together, these data are consistent with an inflammatory marker role for macrophage/ monocyte CD38 in human inflammatory processes.

MaTerials anD MeThODs human subjects irBs and ethics
To generate human MDM, peripheral blood mononuclear cells (PBMCs) were isolated from healthy donors (HD) by Ficoll gradient as previously described (31)

rna isolation
To examine RNA expression, macrophage RNA was isolated using the miRVana kit (Life Technologies) according to the manufacturer's specifications. RNA quality/concentration was quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific). Samples were stored at −80°C until analysis.

real-Time Pcr
Macrophage mRNA gene expression was determined using Taqman quantitative Real-Time PCR on cDNA templates. cDNA was generated from 500 ng RNA/sample using oligo(dT)12-18 primers and Superscript III (Life Technologies), according to the manufacturer's instructions. Product was amplified with commercially available Taqman primers (Life Technologies) and probe sets with Taqman Mastermix (Roche) on an Applied Biosystems 7300 Real-Time PCR or StepOnePlus thermocycler. The assay ID numbers for ABI sets were the following: CD38-Hs-1120071_ m1, FPR2-Hs02759175_s1, GPR18-Hs01921463_s1, EGR2-Hs00166165_m1, and c-MYC-Hs00153408_m1. Expression of target genes was normalized to βACTIN/HPRT as a loading control. Real-Time PCR data were analyzed using the comparative Ct (ΔΔCt) method.

cell Death assay
Cell death in drug-treated or transfected MDM was measured via release of intracellular lactate dehydrogenase (LDH) (Pierce™ LDH Cytotoxicity Assay Kit, Cat No 88953, Thermo Fisher Scientific). Supernatants were collected from established MDM monolayers just prior to cell processing on day 7 and processed according to the manufacturer's instructions. Absorbance was measured at 490 and 680 nm. LDH activity was determined by subtracting the 680 nm absorbance value (background) from the 490 nm absorbance value.

glycolysis assay
Extracellular l-lactate, the end product of glycolysis, was measured in either drug-treated or transfected MDM using the Glycolysis Cell-Based Assay Kit (Cat. No. 600450, Cayman Chemical, Ann Arbor, MI, USA). Supernatants were collected from established MDM monolayers just prior to cell processing on day 7 and processed according to the manufacturer's instructions. Absorbance was measured at 490 nm.
statistics Statistical significance was determined by unpaired or paired t tests (two-tailed, equal SD) or analysis of variance (ANOVA) with p values adjusted for multiple comparisons using Sidak's, Dunnett's, or Tukey's multiple comparisons test. p < 0.05 was considered to be significant. Correlation analyses were performed using linear regression with either Pearson's correlation analysis if the samples passed the D' Agostino and Pearson normality test or Spearman's correlation analysis if they did not. All analyses were completed using GraphPad Prism 6.0 or 7.0 software.

resUlTs cD38 is Upregulated in human M(lPs + iFn-γ) Monocytic cell line-Derived Macrophages
Inflammatory macrophages contribute to a wide range of inflammatory diseases, so markers that allow robust and consistent detection of murine and human inflammatory macrophages are necessary. Since human inflammatory monocytes/macrophages inconsistently express the classical iNOS marker (35,36), current available markers are limited to CD40 detection or intracellular detection of IL-6, IL-12, and IL-1β cytokines requiring in vitro culture. Our laboratory previously discovered that CD38, an ectoenzyme and surface receptor, serves as a robust in vitro and in vivo marker of inflammatory murine macrophages (4).  Figure 1D).
cD38 is induced in M(lPs + iFn-γ) but not M(il-4) human MDM While THP-1 and U937 macrophages are often used to model human macrophages, primary human MDM provide a model that more closely reflects ex vivo macrophages. Therefore, we explored whether CD38 and FPR2 were also induced in primary human MDM activated with LPS + IFN-γ. We also explored the homologs of murine M(LPS + IFN-γ) marker Gpr18 and M(IL-4) markers Egr2 and c-Myc, which we had previously identified in BMDM murine macrophages (4). Primary human MDMs were activated with LPS + IFN-γ or IL-4 and analyzed by real-time PCR or flow cytometry. The results largely recapitulated our data generated from murine macrophage and human monocytic cell lines; CD38 ( CD38 flow cytometry analysis showed that, while the majority (three quarters) of M(LPS + IFN-γ) macrophages were CD38 + , fewer than 3% of M0 and M(IL-4) macrophages stained for CD38 protein (Figures 2F,G). These results indicate that CD38 is highly restricted to inflammatory M(LPS + IFN-γ) macrophages. To compare CD38 to the current human M(LPS + IFN-γ) marker CD40, we evaluated its expression by flow cytometry and observed that CD40 only labeled half of the M(LPS + IFN-γ) population and also labeled up to 12% of M0 or M(IL-4) macrophages ( Figure 2H) (37). These results indicate that CD38 labels more M(LPS + IFN-γ) macrophages than CD40 and fewer non-inflammatory M(0) and M(IL-4) macrophages than CD40. To determine which stimuli can induce CD38, we evaluated CD38 expression after exposure to various individual stimuli. We found that LPS alone resulted in levels of CD38 induction similar to those of M(LPS + IFN-γ) and that IFN-γ alone did not induce CD38 (Figure 2I). Among other cytokines tested, neither TNF-α nor IL-1β significantly induced CD38 (Figure 2I). Similar results were obtained with the TLR7 agonist imiquimod, while the TLR8 agonist R-848 mildly induced CD38 (Figure 2I). CD38 expression correlated with the secretion of the inflammatory cytokine IL-1β, further supporting the link between CD38 and inflammatory macrophages ( Figure 2J). In summary, these data indicate that CD38 is a robust and selective marker of human inflammatory M(LPS ± IFN-γ) macrophages. cD38 Promotes inflammatory cytokine secretion in human Macrophages The increased expression of CD38 in inflammatory macrophages raises the question of whether CD38 promotes inflammatory responses. To answer this question, loss-of-function experiments were performed. Since no exclusive CD38 inhibitors exist, we used rhein (38) and apigenin (39), two CD38 inhibitors that, respectively, originate from the structurally distinct anthraquinone and flavonoid families. This strategy was expected to reduce the likelihood of shared rhein and apigenin effects being mediated by any possible off-target effects (as off-target effects are expected to differ between rhein and apigenin). We  further confirmation via genetic method that CD38 was responsible for the observed effects ( Figure 3C). We induced CD38 knockdown in MDM with two CD38-targeting siRNAs and then differentiated to M(LPS + IFN-γ) conditions. Although CD38 knockdown reduced CD38 + cells by only ~35% (Figure 3D) and did not significantly reduce IL-6, significantly suppressive effects on IL-12p40 cytokine secretion (27% decrease vs. DMSO, results normalized and combined from six experiments, n = 14) were observed ( Figure 3C). In addition, we explored inflammatory IL-1β cytokine secretion and glycolysis metabolic activity, via secretion of the glycolysis end-product l-lactate, both of which increase in inflammatory macrophages (2). We observed a mild but significant decrease in both IL-1β (19% decrease) ( Figure 3E) and lactate secretion (15% decrease) ( Figure 3F) in CD38 siRNA conditions. Cytokine and lactate suppressive effects were not due to increased cell death, as no differences in cell viability between the control and CD38 siRNA were observed ( Figure S2 in Supplementary Material). Therefore, data are consistent with CD38 activity or expression contributing to maximal IL-12 and IL-1β inflammatory cytokine secretion and glycolytic activity in human inflammatory macrophages.

enhanced cD38 expression in ncMs From active sle Patients
The association between exposure to an inflammatory stimulus such as LPS ± IFN-γ and expression of CD38 raises the question of whether myeloid CD38 plays a role in human inflammatory diseases. SLE is a systemic inflammatory autoimmune disease characterized by autoantibody generation and immune complexmediated tissue damage. Disease activity can vary widely and can be assessed clinically using the standardized SLE Disease Activity Index (SLEDAI) scoring system, which is based on clinical signs (such as fever, seizures, and muscle weakness) and laboratory values (including blood cell counts or anti-DNA antibody titers) (42). We analyzed CD38 expression in monocytes from HD, inactive SLE (SLE I ) patients with an SLEDAI score of 0-2, and active SLE (SLE A ) patients with SLEDAI scores over 4 (see Table 1 for patient demographics). There were no statistically significant differences in age, gender, or race among the three groups ( Table 1).
To exclude potential contaminating granulocytes, NK cells, T cells, and B cells from the CD45 + population, we gated out cells positive for CD66b, CD56, CD3, and CD19, respectively. We then examined CD14 and CD16 expression levels to differentiate various types of monocytes, the precursors of tissue macrophages ( Figure 4A, top row). CD14 ++ CD16 − cells correspond to classical monocytes (CM), while CD14 ++ CD16 + correspond to intermediate monocytes (IM) and CD14 + CD16 ++ to NCMs (Figure 4A, panels 5-6), as previously described (43,44). The inflammatory potential of each of these populations is controversial. While initial studies suggested that CM and IM were inflammatory and NCM were regulatory, attempts to functionally characterize the inflammatory potential of monocyte subsets have often been contradictory (45)(46)(47)(48)(49)(50)(51)(52)(53)(54)(55). Contrary to what was originally thought, several studies have now shown that NCM produce higher levels of inflammatory cytokines in response to inflammatory stimuli than CM (50,54,55). To identify links between SLE disease and monocyte CD38, we analyzed CD38 expression within CM, IM, and NCM. In contrast to the lack of CD38 expression on unstimulated macrophages, the majority of unstimulated CM and IM expressed CD38 on the surface while NCM had two distinct CD38 populations (Figures 4B,C). CD38 monocytic expression was not surprising as CD38 plays an important role in extravasation of blood monocytes into tissues (3). The % CD38 + cells did not change with SLE diagnosis or activity ( Figure 4C). We reasoned that CD38 level, rather than mere presence, may be associated with inflammatory disease activity. To address this question, we compared the mean fluorescence intensity (MFI) of CD38 + cells and the percentage of CD38 hi cells (as defined in Figure 4D) in HD and SLE patients (Figures 4E,F). The MFI of CD38 + NCM, a measure of CD38 protein load, was significantly higher in active SLE A patients (MFI = 9,646) than in inactive SLE I patients (MFI = 5,949) or HD (MFI = 4,667). It is possible that CD38 expression levels are associated with age, gender, or race. Instead, we found no significant correlation between age and NCM CD38 MFI ( Figure S3A in Supplementary Material) or significant differences between female and male ( Figure S3B in Supplementary Material) or Caucasian and African-American NCM CD38 MFI ( Figure S3C in Supplementary Material). The CD38 MFI of CM was also significantly increased in active SLE A (MFI = 3,543) vs. inactive SLE I patients (MFI = 2,888) ( Figure 4E). The increase in NCM MFI was also reflected by a small but significant increase in percent of CD38 hi NCM in active SLE patients compared with HD (93 vs. 78%) (Figure 4F). Although we found no significant correlation between age and NCM CD38 hi % ( Figure S3D in Supplementary Material), there was a non-significant trend toward increased NCM CD38 hi % in females vs. males ( Figure S3E in Supplementary Material), indicating that CD38 hi % may be subject to gender variation. We found no significant difference between races ( Figure S3F in Supplementary Material). To rule out that any non-monocyte CD14 − cells were responsible for increased CD38 in NCM analyses, we re-analyzed data using a more restrictive gating (named CD14 low CD16 ++ in Figure 4A, panel 6). This analysis, shown in Figures 4G,H, exactly reproduced the larger NCM gate data in Figures 4D,E. To further ascertain whether NCM CD38 MFI or % CD38 hi cells is linked to SLE disease activity, we performed correlation analyses in the SLE patient population, identifying significant positive correlations between NCM CD38 MFI and % CD38 hi and SLEDAI score (Figures 4I,J). Overall, these data indicate that CD38 is a marker of steady-state monocyte populations and that SLE disease activity is associated with CD38 MFI in a subset of NCMs, which have inflammatory potential.

DiscUssiOn
Here, we report the novel finding that surface CD38 protein is robustly and selectively induced in M(LPS + IFN-γ) human macrophages. We also observe selective induction of the FPR2 and GPR18 genes by LPS + IFN-γ and of EGR2 gene by IL-4 in primary human macrophages. We found that inhibition of CD38 activity or expression in primary inflammatory macrophages interfered with full inflammatory IL-12 and IL-1β cytokine production and glycolytic activity. In the chronic inflammatory disease SLE, we observed higher CD38 expression in NCM from active vs. inactive SLE patients or healthy controls, suggesting an association of CD38 to inflammatory disease activity. We also observed significant correlations between NCM CD38 MFI or CD38 hi % and SLEDAI disease scores. Overall, these findings are consistent with the idea that CD38 plays a conserved role in inflammatory macrophages. Since its early identification in human leukocytes, CD38 has served as a practical surface T and B cell differentiation and ]. The identification in 2015 of CD38 as a murine inflammatory macrophage marker that effectively dissects inflammatory from alternatively activated macrophages (4) raised the possibility that CD38 could also serve as an inflammatory marker in human macrophages. Here, we find that LPS + IFN-γ treatment induces CD38 at the transcriptional and protein level in multiple human macrophage models ranging from monocyte-derived cell lines to primary MDM. These results are consistent with CD38 representing a conserved response to inflammatory stimulation. Here, we induced inflammatory macrophages with a combination of LPS + IFN-γ, but other inflammatory stimuli also promote CD38. IFN-γ alone induces CD38 CD38 in Inflammatory Macrophages/Monocytes Frontiers in Immunology | www.frontiersin.org July 2018 | Volume 9 | Article 1593 in human monocytes (56), although we did not observe this effect in human macrophages (Figure 2I). Similarly, TNF-α (57) and TLR9 agonists (58) have also been reported to induce CD38 in smooth muscle cells and leukemic cells, respectively, but neither TNF-α, IL-1β nor TLR7 agonist imiquimod induced CD38 in human macrophages (Figure 2I). The only individual stimuli that induced CD38 were LPS and the TLR-8 agonist R-848, albeit to a much lesser extent (Figure 2I). CD38 expression is positively correlated with IL-1β expression and was present in a much higher proportion of M(LPS + IFN-γ) macrophages than the current M(LPS + IFN-γ) human macrophage marker CD40. In addition, it was practically absent from M0 and M(IL-4) macrophages, providing improved resolution over CD40. In summary, CD38 is a highly and consistently induced inflammatory marker conserved from mouse to human macrophage responses. Therefore, CD38 may serve as a marker to evaluate the effects of small molecules on inflammatory macrophage phenotype modulation or monitor inflammatory disease status. CD38 is a complex surface molecule that can play multiple roles, each one of which may have different functional effects. CD38 can act as an ectoenzyme on the cell surface, depleting extracellular NAD or NADP, and producing mediators cADPR, ADPR, and NAADP that promote intracellular Ca 2+ increase, proliferation, and insulin secretion (59). On the other hand, CD38 can act as a receptor with ability to bind CD31 on endothelial and other cells, thereby promoting endothelial adhesion as well as cell activation and proliferation (3,30). Here, we attempted to explore the functional role of CD38 in macrophages using CD38 inhibitors rhein and apigenin. Both these treatments suppressed IL-6 and IL-12, consistent with CD38 promoting inflammatory cytokine production. A caveat to this interpretation is that, while both rhein and apigenin inhibit CD38 activity, they can also impact other pathways. For example, besides targeting CD38, rhein also inhibits NF-κB and P2X7 (60) and IKKβ (61) and suppresses IL-6 and NO (61). Apigenin also affects other macrophage signaling pathways, including NF-κB. Apigenin can inhibit LPS-induced IL-6 and IL-1β production and caspase-1 and ERK1/2 inhibition has been observed with apigenin (62). Apigenin also reduces expression of miR-155 (63), a miRNA that promotes inflammatory macrophage phenotype (64) and activates PPARγ, promoting M(IL-4) phenotypes (65). Although it is unclear whether CD38 inhibition is occurring upstream or in parallel to these effects, it is possible that the inflammatory cytokine secretion effects are mediated by other pathways. To overcome these limitations, we attempted to knockdown CD38. Although only a suboptimal knockdown (35%) of CD38 was achieved, knockdown was mirrored by similar decreases in IL-12/IL-1β secretion and glycolytic activity. By contrast, no significant IL-6 decrease was observed, suggesting that rhein and apigenin's effects on IL-6 are mediated by pathways other than CD38.
Based on these intriguing results in macrophages, it will be interesting to further investigate the role of decreased vs. increased CD38 expression in monocyte function and activity. For example, baseline expression of CD38 in monocytes is expected to mediate endothelial cell adhesion, which is necessary for monocyte extravasation into tissues. However, high CD38 expression, as in the case of CM and NCM from active SLE patients, may promote inflammatory cytokine production or drive monocyte-to-macrophage differentiation. This hypothesis will be tested in future experiments. From studies in dendritic cells, neutrophils, and monocytes, CD38 is known to be essential for trafficking and chemotaxis to sites of infection (13,14,16,66,67). More recently, CD38 has been found to be necessary for resistance to Listeria infections via NAD depletion and actin cytoskeleton modulation (18). In this model, a significant reduction in Il1b, Il6, Il12b, and TNFa inflammatory cytokine transcripts was observed in CD38 KO macrophages, while Nos2 and Cox2 transcripts were unchanged (18). We now show that CD38 expression and activity promotes inflammatory cytokine production and glycolytic activity in human macrophages. Since NO is not produced by human inflammatory macrophages, it was not evaluated. However, it remains to be determined if CD38 also plays a role in generation of ROS.
Our results show increased CD38 expression (MFI and CD38 hi %) in NCM from SLE A patients compared with healthy controls. In addition, NCM MFI from SLE A patients was also significantly higher than that of SLE I patients, suggesting that the extent of CD38 expression is linked to disease activity in SLE. Indeed, correlation analyses between NCM CD38 hi % or MFI and SLEDAI score confirmed this prediction. Besides inflammatory monocytes/macrophages, activated T cells and B cells, plasma cells, and NK cells also express high levels of CD38 (3,30). In contrast to previous reports (68-70), we observed no significant changes in CD38 in SLE patients' T cells and B cells ( Figure S4 in Supplementary Material). The reason for this discrepancy is unclear but sample size is a likely factor as we observed a trend toward increases in both B cells and T cells. Our results are consistent with a role for CD38 in SLE, as previously suggested via links between CD38 and the Fas lpr murine models of SLE (71) and CD38 SNPs and severe discoid rashes in SLE patients (72). Since our data suggest that CD38 expression and activity are associated with SLE disease diagnosis and severity, future studies could further explore the marker and functional role of CD38 in larger cohorts, as well as whether genetic variation at this locus is linked to inflammatory pathology.
Therapeutic targeting of CD38 is currently used for multiple myeloma and other hematologic malignancies (30) and has also been proposed for SLE and other antibody-mediated autoimmune diseases (73). Among several anti-CD38 antibody therapies in development, daratumumab is the most advanced. Daratumumab binds an epitope of CD38, inhibiting its cyclase activity and efficiently promoting CD38 + cell lysis in vivo (30,74). Consequently, anti-CD38 antibody therapy has been proposed as a method to deplete pathogenic autoantibody-producing long-lived plasma cells in SLE (73). Our results raise the possibility that, in addition to effects on plasma cells, anti-CD38 therapy may modulate inflammatory CD38 + monocytes and macrophages. However, it is difficult to predict the effects of CD38 antibodies, as they may differentially impact the receptor function and the various enzymatic activities of CD38. For example, while daratumumab inhibits CD38 cyclase activity, it in turn enhances hydrolase enzymatic activity (30). It will be important to evaluate in detail what these effects are and how monocyte/ macrophage CD38 blockade modulates anti-cancer/anti-SLE therapeutic effects. Consistent with therapeutic effects of anti-CD38 therapy in SLE, anti-thymocyte antibody treatments that include CD38 antibodies suppress SLE and other autoimmune diseases (73). It is also interesting to note that SLE patients with anti-CD38 autoantibodies have lower levels of clinical activity and decreased titers of pathogenic anti-dsDNA antibodies (69). Perhaps the greatest challenge of targeting CD38 therapeutically stems from its broad expression pattern on different cell types and its multifunctional nature, as it acts both as a receptor and has various ectoenzyme activities. For instance, although CD38 deficiency impairs immune responses and disease pathogenesis in murine arthritis and asthma models (59,75), it can instead exacerbate murine models of lupus and diabetes (76,77). Since CD38 deficiency did not enhance pathogenic autoimmune responses, exacerbation effects appear to be due to non-immune effects on target tissue (59,76). Overall, these results highlight the need for better understanding the functional role of CD38 and its receptor and ectoenzyme activity in individual cell types, as well as of specific cellular targeting.
In conclusion, CD38 induction in inflammatory macrophages is conserved from mouse to humans. In addition, high CD38 expression in NCM is associated with active SLE disease and CD38 activity contributes to inflammatory cytokine release. These novel findings suggest that CD38 may be a useful marker of inflammatory macrophage/monocyte-mediated disease and warrant additional clinical and mechanistic studies to fully define its diagnostic and/or therapeutic targeting potential.

eThics sTaTeMenT
This study was carried out in accordance with the recommendations of Ohio State University Office of Responsible Research Practices Institutional Review Board with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by The Ohio State University Institutional Review Board.

acKnOWleDgMenTs
We would like to extend our appreciation to all the volunteers who participated in this study. Our research would not be possible without the time and effort they put forth. We also thank clinical research coordinators Holly Steigelman, Kayla Knoll, and Emily Sullivan and the Clinical Research Center for patient sample collection, Juhi Sharma, Ifeoma Okafor, Lindsay Webb, Julia Scordo, Juan Moliva, and Michael Duncan for sample processing. The project described was supported by Award Number Grant UL1TR001070 from the National Center for Advancing Translational Sciences and The Ohio State University College of Medicine-Health and Rehabilitation Sciences start-up funds. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health. MG is supported by NIH NIAID R01AI121405 and NIH NIAID 1R21AI127354 grants.

sUPPleMenTarY MaTerial
The Supplementary Material for this article can be found online at https://www.frontiersin.org/articles/10.3389/fimmu.2018.01593/ full#supplementary-material.  Figure 3. Human monocyte-derived macrophages (MDMs) were treated with 15 µM rhein or DMSO control (a), 25 µM apigenin or DMSO (B), or transfected with 100 µM CD38 siRNA cocktail or nonsense siRNA control (c) on day 5. On day 6, they were activated with LPS + IFN-γ for an additional 24 h prior to analysis. (a-c) IL-6 and IL-12p40 secretion was analyzed by ELISA from MDM supernatants. Graphs show data (relative to corresponding experiment vehicle control or nonsense siRNA) from a single experiment, one to two biological replicates with two technical replicates per sample. All data were analyzed by unpaired t tests. *p < 0.05, **p < 0.01, ***p < 0.001.  Figure 3. All graphs represent pooled data from three to five independent experiments with one to two biological replicates per experiment and two to three technical replicates per sample. All data were analyzed by unpaired t tests.
FigUre s3 | There are no significant associations between CD38 expression and age, gender, or race. CD38 mean fluorescence intensity (MFI) (a-c) and % CD38hi cells (D-F) from healthy donors (HD) were analyzed to determine if age, gender or race affected CD38 protein load or surface expression. (a,D) Correlation analyses of SLE Disease Activity Index score with non-classical monocytes (NCM) CD38 MFI (a) or % CD38 hi NCM (D) in HD. (a) CD38 MFI did not pass D'Agostino and Pearson normality test, so Spearman correlation analysis was performed, r = −0.1506, p = ns, n = 9. (D) Percent of CD38 hi NCM has Gaussian distribution as determined by D'Agostino and Pearson normality test, so Pearson correlation analysis was performed, r = −0.2617, p = ns, n = 9. (B,c,e,F) CD38 MFI (B,c) and % CD38 hi cells (e,F) in NCM from HD were analyzed by gender (B,D), n = 3 males (M) and 6 females (F); or race (c,F), n = 5 Caucasians; n = 3 African-Americans, n = 1 unknown. Data analyzed by unpaired t tests, p = ns.