Factor-Driven Macrophage Differentiation Macrophage-Colony ...

The Journal of Immunology

Failure of Monocytes of Trauma Patients to Convert to Immature Dendritic Cells is Related to Preferential Macrophage-Colony-Stimulating Factor-Driven Macrophage Differentiation1 Asit K. De,2* Krzysztof Laudanski,2* and Carol L. Miller-Graziano3*† Following trauma, increased inflammatory monokine activation and depressed APC function can occur simultaneously. These contradictory monocyte (M␾) dysfunctions could result if postinjury M␾ differentiation preferentially favored inflammatory macrophage (Mac) differentiation over development into the most potent APC, dendritic cells (DC). In this report, M␾ of trauma patients with a depressed MLR induction capacity are, for the first time, shown to be unable to differentiate in vitro to immature CD1aⴙ DC under the influence of GM-CSF and IL-4. Trauma patient M␾ that retained MLR-inducing capacity had a nonsignificant reduction in DC differentiation capacity. Only patient M␾ populations with depressed differentiation to immature DC (iDC) demonstrated depressed IL-12 and IL-15 production and a continued reduced MLR induction capacity. Neither increased IL-10 production nor decreased CD11cⴙ DC precursor numbers correlated with depressed M␾-to-DC differentiation. Instead, these patients’ APC-dysfunctional M␾ populations had increased expression of inflammatory Mac phenotypes (CD64ⴙ, CD86low, HLA-DRlow) and up-regulated secretion of M-CSF. M-CSF combined with IL-6 inhibits M␾-to-iDC differentiation and promotes M␾-to-Mac differentiation by down-regulating GM-CSFR expression and increasing DC apoptosis. Both depressed GM-CSFR expression and increased M␾ iDC apoptosis, as well as increased expression of CD126 (IL-6R) and CD115 (M-CSFR), were detected in APC-defective patient M␾. In vitro addition of anti-M-CSF enhanced the IL-4 plus GM-CSF-induced M␾-to-DC differentiation of these patients. This suggests that, in trauma patients, enhanced M␾-to-Mac differentiation with concomitant inhibited iDC development is partially due to increased circulating M␾ sensitivity to and production of M-CSF and contributes to postinjury immunoaberrations. The Journal of Immunology, 2003, 170: 6355– 6362. any studies have demonstrated that monocyte (M␾)4/ macrophage (Mac) dysfunctions are pivotal in the development of the dysregulated inflammatory cytokine production typical of trauma patients who develop multiple organ failure and infectious complications (1– 6). Simultaneous with the often-exaggerated inflammatory response, these patients’ M␾ show depressed HLA-DR expression and a failure of Ag-presenting functions, which appears to be correlated with the development of T cell immunosuppression (7–13). The most potent APC is the dendritic cell (DC) (14 –18). Consequently, a DC dysfunction might be causal or accompany some of these postinjury APC deficits. Human blood M␾ can differentiate to immature DC (iDC) under the influence of lymphokines like GM-CSF and IL-4 (17–

M

Departments of *Surgery and †Microbiology/Immunology, University of Rochester, 601 Elmwood Avenue, Rochester, NY 14642 Received for publication November 11, 2002. Accepted for publication April 8, 2003. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by Public Health Service Grant GM36214. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health. 2

A.K.D. and K.L. contributed equally to this work.

3

Address correspondence and reprint requests to Dr. Carol L. Miller-Graziano, Department of Surgery, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, NY 14642. E-mail address: [email protected] 4 Abbreviations used in this paper: M␾, monocyte; Mac, macrophage; DC, dendritic cell; iDC, immature DC; mDC, mature DC; ISS, injury severity score; MODS, multiple organ dysfunction syndrome; SEB, staphylococcal enterotoxin B; h, human; 4⫹GM, IL-4 plus GM-CSF treatment; MFI, mean fluorescence intensity; RPA, RNase protection assay.

Copyright © 2003 by The American Association of Immunologists, Inc.

20). These iDC require further maturation by T cell interactions or infectious stimuli to become potent immunostimulatory APC capable of activating naive T cells and required for memory T cell reactivation and maintenance (19 –22). Alternatively, under other cytokine influences and different in vivo environments, M␾ differentiate to Mac, which have higher potential for inflammatory cytokine production but lower expression of MHC class II and costimulatory molecules, are inefficient APC, and are unable to activate naive T cells (19, 23–25). A failure of M␾-to-iDC differentiation has been described in immunosuppressed Sezary syndrome and multiple myeloma patients and ascribed to their depressed levels of CD11c dendritic precursors (26 –27). The dysregulated M␾ function seen in some trauma patients could also be related to depressed iDC differentiation and increased Mac differentiation. M␾ differentiation to iDC can be redirected to Mac differentiation by IFN-␥, IL-6, M-CSF, LPS, or IL-10 (20, 23–25, 28). LPS and IFN-␥ are thought to mediate this M␾-to-Mac differentiation switch by increasing M␾ IL-6 and M-CSF production, and down-regulating GM-CSFR expression overbalancing any IL-4 plus GM-CSF (4⫹GM) effect (25). In this study, M␾ of trauma patients with diminished APC function (defined as depressed MLR-inducing capacity and decreased IL-12 production) were assessed for in vitro differentiation to iDC under the influence of 4⫹GM. Up-regulated expression of CD1a and CD86, increased IL-12 and IL-15 production, as well as DC maturation and MLR induction in allogenic T lymphocyte cultures were also assayed. M-CSF production by MLR-dysfunctional patient M␾ was assessed and found to be elevated. Increased M-CSFR and CD64 expression appeared simultaneously with decreased M␾ GMCSFR and IL-4R levels in APC-dysfunctional M␾. Exogenous 0022-1767/03/$02.00

6356

M␾ OF TRAUMA PATIENTS FAIL TO DIFFERENTIATE TO iDC

addition of Ab to M-CSF during DC generation was assessed for improved patient M␾-to-iDC differentiation. This study is the first investigation of APC-dysfunctional M␾ populations of trauma patients for an in vitro M␾-to-iDC conversion defect.

Materials and Methods Studied population A total of 46 patients admitted to either University of Massachusetts Medical Center or University of Rochester Medical Center were enrolled in the study. There were 40 patients with serious mechanical trauma (injury severity score (ISS), ⬎17; mean ISS, 38) and 6 patients with thermal trauma (total body surface burn area, ⬎25%). There were 29 males and 17 females. The mean age was 52.3 years. Patient samples were first collected at least 2 days after trauma to reduce variables related to the initial treatment. Then, sample collection was continued one to two times per week until the day the patient was released or died. Appearance of MLR dysfunction for two consecutive assays was required for classification as APC dysfunctional. Patients (n ⫽ 6) whose M␾ had a one-time transient depression were included in the APC-responsive group. Some patients’ M␾ data appeared as APC functional at first, and then dysfunctional, as they progressed to a defect over time postinjury. No more than two samples for each patient were included. The exclusion criteria included whole-blood transfusion, immunosuppressive therapy, a history of immunoproliferative disease, or other immunological aberrances. The clinical pathology of the patients was quantitatively compared with M␾ functions, using elevation in the Marshall multiple organ dysfunction syndrome (MODS) score during the 24-h period for which each blood sample was drawn (29). A Marshall MODS score of ⬎6 represents significant organ dysfunction and clinical pathology (29). A normal control sample was tested along with each patient’s samples. The control samples were collected from a panel of 48 normal donors (volunteers from hospital and laboratory staff) who had been repeatedly assayed over a 12-mo period to establish their immune response levels and variation parameters. Their mean age was 41 years. Informed consent was obtained from all subjects. The study was approved by the University of Rochester Institutional Review Board.

Isolation of T lymphocytes and M␾ Thirty milliliters of blood was drawn twice weekly from each patient using existing i.v. lines. Simultaneously, a similar amount of blood was drawn from the control’s antecubital vein. PBMC were isolated by centrifugation over a Ficoll-Hypaque gradient and then resuspended in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 10% FBS (HyClone, Logan, UT), 50 U/ml penicillin G, 50 ␮g/ml streptomycin, 50 ␮g/ml gentamicin, 2.5 ␮g/ml fungizone, 4 mmol/L L-glutamine, 1 mmol/L sodium pyruvate, and 1% of MEM nonessential amino acids (complete medium). All media were supplemented with 20 U/ml polymyxin B (Calbiochem, La Jolla, CA), and the endotoxin level was ⬍15 pg/ml. T cells were isolated from PBMC using neuraminidase-treated SRBC as described (30). M␾ were isolated from T cell-depleted PBMC by negative magnetic separation with anti-CD19 beads and anti-CD2 beads according to the manufacturer’s instructions (Dynal, Lake Success, NY). A subsequent flow cytometric analysis revealed ⬃83% CD33⫹ cells in M␾ samples and 95% CD3 positivity in separated T cells. Surface expression of CD126, CD115, IL-4R, GM-CSFR, CD64, and CD11c was evaluated with flow cytometry gating on fresh CD33⫹ M␾. CD33 was chosen as a M␾ marker based on its continued expression on Mac, M␾, and DC, as opposed to CD14, which is down-regulated in DC and may be shed from highly activated M␾ (31–33).

IL-4 and GM-CSF. The cells were harvested on the fifth day. Supernatants were used for IL-10 evaluation, whereas cells were analyzed with flow cytometry by staining with anti-CD1a, anti-HLA-DR, anti-CD83, and antiCD86. Apoptosis of DC was measured using Annexin V-FLUOS staining kits according to the manufacturer’s instructions (Roche Molecular Biochemicals, Indianapolis, IN). The Ag-presenting capabilities of cultured M␾ or DC were assessed by activity in MLR and production of immunostimulatory monokines. In the MLR, 1 ⫻ 104 M␾ or DC were cocultured with 1 ⫻ 105 cells/well allogenic T cells. The plates were incubated at 37°C under 5% CO2 for 5 days, and then pulsed with 1 ␮Ci/well of [H3]thymidine for 24 h and harvested. The monokine production was evaluated after stimulation with 100 IU/ml IFN-␥ and 0.5 ␮g/ml staphylococcal enterotoxin B (SEB) for 18 h. The supernatants were collected and assessed for IL-12 and IL-15 production by ELISA. Some iDC were cultured an additional 3 days in TNF-␣ (100 IU/ml; PeproTech), IL-4 (100 IU/ml; PeproTech), and GM-CSF (100 IU/ml; PeproTech). Cells were harvested, and mature DC (mDC) were enumerated with anti-CD83 (BD PharMingen). The DC of patients were defined as immuno-nonstimulatory (dysfunctional) when the MLR of IL4/GM-CSF-differentiated M␾ was ⬍33% of normal control value. In the initial analysis, MLR from unstimulated M␾ was calculated as a percentage of normal.

Flow cytometric analysis Mouse mAb recognizing human CD1a (clone BL6; Coulter, Brea, CA), CD11c (clone B-ly6; BD PharMingen), CD33 (clone P67.6; BD PharMingen), CD83 (clone HB15e; BD PharMingen), CD86 (clone 2331; BD PharMingen), HLA-DR (clone G46-6; BD PharMingen), and CD126 (clone M91; Immunotech, Marseille, France) were used. A rat anti-human CD115 (clone 7-7A3-17; Lab Vision, Fremont, CA) was detected by secondary anti-rat PE-conjugated Ab-defined CD115 (Serotec, Raleigh, NC). A total of 1 ⫻ 105 cells incubated with conjugated mAb or the appropriate isotype controls in recommended dilutions or in pretitrated amounts (CD115, CD126) was washed twice with PBS containing 2% FBS and resuspended in 500 ␮l for two- or four-color analysis after appropriate compensation adjustments. M␾ IL-4R and GM-CSFR␣ expression was tested by using specific fluorokine kits (R&D Systems) and analyzed on a Coulter XL or FACSCalibur.

RNase protection assay (RPA) A total of 1 ⫻ 106 4⫹GM-treated DC from controls and patients were treated for 8 –10 h with SEB (0.5 ␮g/ml) plus IFN-␥ (100 IU/ml). Total cytoplasmic RNA was isolated using Tri-reagent (Molecular Research Center, Cincinnati, OH). Antisense probes for quantification of RNA for hIL-10, hIL-12 p35 and p40, hIL-15, and hIL-6 were labeled with [32P]UTP using Riboquant in vitro transcription kit (BD PharMingen). The band intensities of RPA gels were quantified using NIH Image software and adjusted for loading irregularities by comparison to the ribosomal gene product L32.

ELISA-based cytokine detection assays Cytokine levels in culture supernatants were measured using cytokine-specific ELISA according to the manufacturer’s guidelines. The lower limits of detection for the assays were as follows: IL-12p70 plus p40, ⬍5 pg/ml; IFN-␥, ⬍2 pg/ml; IL-10, ⬍3 pg/ml (all from Endogen, Woburn, MA); IL-4, ⬍1.25 pg/ml; GM-CSF, ⬍0.27 pg/ml (BioSource International, Camarillo, CA); and M-CSF, ⬍25 pg/ml (R&D Systems).

T cell proliferation and lymphokine production

Statistical analysis

Purified T cells (2 ⫻ 105 cell/200 ␮l/well) were cultured in a 96-well plate precoated with anti-CD3 and anti-CD4 (1 ␮g/well each; BD PharMingen, San Diego, CA). After 24 h, supernatants were collected for assessment of lymphokine production, whereas proliferation was measured in 72-h culture by [3H]thymidine incorporation assays, as described previously (30).

Due to the nonparametric distribution of the studied variables, the primary tests performed were the Kruskal-Wallis H and Mann-Whitney U tests for comparisons of the nonparametric variables. One-way ANOVA with least significant differences test and Student’s t test were performed when the effect of neutralizing Ab on DC generation was studied due to the parametric distribution of the studied variables confirmed by the Lavene test. A regression analysis was computed by the stepwise method. Statistical significance was set at a two-tailed value of p ⬍ 0.05. SPSS version 8.0 software was used for all calculations.

In vitro generation of iDC Isolated M␾ were cultured (3 ⫻ 106 cells/3 ml/well) in complete medium with 5 ⫻ 10⫺5 M 2-ME and 5 ⫻ 10⫺2 M L-asparagine, supplemented with 100 ng/ml both recombinant human GM-CSF and IL-4 (PeproTech, Rocky Hill, NJ) or left unstimulated in 6-well plates (17, 19). In some experiments, neutralizing Ab for human (h)M-CSF (1 ␮g/ml; goat polyclonal; R&D Systems, Minneapolis, MN) or a goat IgG control (Santa Cruz Biotechnology, Santa Cruz, CA) was added to DC cultures. Cells were incubated for 4 –5 days at 37°C under 5% CO2. After 3 days of culture, 1.5 ml of supernatant was removed and replenished with fresh medium containing

Results

Segregation of trauma patient M␾ data on the basis of altered M␾ functions Our data indicated that the M␾ of severely injured patients, as a group, have depressed MLR levels compared with M␾ of normal

The Journal of Immunology controls. However, the M␾ responses of the severely injured patient group could be segregated into those with essentially normal MLR induction capacity (high MLR inducing) and those with a dramatically reduced MLR immunostimulatory capacity which was ⬍33% of paired normal control (Fig. 1A). Patient M␾ with low MLR-inducing capacity also had significantly reduced IL-12 production (Fig. 1A) as described previously (7–9). However, patient M␾ retaining high MLR-inducing capacity also maintained IL-12 production (Fig. 1A). Patients whose M␾ had low MLRinducing capacity also developed T cell anergy as evidenced by a failure of their simultaneously isolated T lymphocytes to proliferate in response to direct TCR stimulation (immobilized anti-CD3 plus anti-CD4) (Table I). As previously described, this T cell anergy was also reflected in a failure to produce IL-4, IFN-␥, or

6357 Table I. Isolated T cell functions of lymphocytes assayed simultaneously with their M␾ MLR

Control

High M␾ MLR

Low M␾ MLR

17.85 T cell proliferation (DPM, ⫻103) IL-4 production (pg/ml) 12.4 IFN production (pg/ml) 2304.0 GM-CSF production (pg/ml) 355.5

17.69 9.3 1508.0 553.0

1.18 2.2 104.0 5.0

T Cell Functiona

a The cutoff point for patients’ samples was set up as 33% of the control M␾ MLR values.

GM-CSF (Table I) (30). Finally, the patients with APC-dysfunctional MLR did not show a significant correlation with their initial ISS. However, patients with low MLR-stimulatory M␾ had significantly increased incidence of clinical pathology as indicated by a mean Marshall MODS score of 8.5, where a score of ⬎6 indicates significant organ failure ( p ⬍ 0.05; Table II) (29). This suggested that postinjury M␾ APC dysfunctions may contribute to trauma patient pathology. Using these M␾ MLR-based segregation categories, patient M␾ with low or high MLR were assessed for M␾-to-iDC conversion. Patient M␾ differentiation to iDC Well-established M␾ APC dysfunction in trauma patients could partially result from failed M␾-to-DC differentiation. Other studies of immunosuppressed patients with APC defects identified a reduction in M␾ CD11c⫹ DC precursors numbers concurrent with increases in more Mac-like M␾ but normal in vitro 4⫹GM-induced differentiation of M␾ to iDC (26, 27). As seen in Fig. 1B, no decrease in CD11c precursor numbers was detected in populations of trauma patient M␾ that were identified as having low MLRinducing capacity (APC defect), indicating a normal level of released precursors. However, these low MLR-inducing M␾ did show increased expression of the Mac phenotype marker CD64 (Fig. 1B). This more Mac phenotype of circulating M␾ could indicate an altered M␾ differentiation potential. 4⫹GM-induced M␾-to-iDC differentiation was compared among M␾ of trauma patients with low MLR induction, M␾ of trauma patients with high MLR induction, and control M␾. The differentiation of M␾ to iDC indicated by the acquisition of CD1a expression was significantly depressed only in the low APC-capacity patient M␾ (Fig. 2A).

FIGURE 1. Characteristics of low MLR-inducing M␾ from trauma patients. A, M␾ data from normal controls and trauma patients were segregated based on their ability to induce proliferation of allogenic T cells in an MLR setup, as described in Materials and Methods. (Low MLR-inducing were ⱕ33% of control value, whereas high MLR inducing were ⬎33% of control value.) The low MLR-inducing M␾ (n ⫽ 30; skewness, 2.57 ⫾ 0.43) and high MLR-inducing M␾ (n ⫽ 58; skewness, 1.61 ⫾ 0.3) data were compared with control normal M␾ (n ⫽ 64; skewness, 1.97 ⫾ 0.3). M␾ from control normals (n ⫽ 23; skewness, 2.61 ⫾ 0.48) and trauma patients (n ⫽ 12; skewness, 0.3 ⫾ 0.61 for low MLR-inducing MLR; and n ⫽ 16; skewness, 2.31 ⫾ 0.56 for high MLR-inducing samples) were also stimulated with SEB (0.5 ␮g/ml) plus IFN-␥ (100 IU/ml) for 18 h, assessing IL-12 production (tested by ELISA). Median and range are presented. B, M␾ were also assessed for surface expression of CD64 (n ⫽ 30 (normal), 19 (high MLR), and 20 (low MLR)) and CD11c (n ⫽ 27 (normal control) and 6 and 23 for low MLR-inducing and high MLR-inducing patient M␾ samples, respectively) by flow cytometry and the data are individually expressed as the percentage of positive cells with the median indicated by the bar. ⴱ, p ⬍ 0.05, and ⴱⴱ, p ⬍ 0.001, compared with normal control values. #, p ⬍ 0.05, compared with immunostimulatory patient samples.

Development of mDC APC capacity during interaction with allogenic T cells in MLR Many of the described normal characteristics of in vitro M␾-to-DC differentiation were evident in our trauma patient M␾ culture systems. Five-day in vitro culture of human M␾ with 4⫹GM induces differentiation only to iDC, not to mDC (18 –22). Our 4⫹GMcultured M␾ populations showed in controls an iDC expression of CD83 (DC maturation marker) of 1.37% positive (mean) after 5 days of culture, whereas patient 4⫹GM-cultured M␾ were 5.36% CD83⫹ positive in MLR-high populations and 2% CD83⫹ positive in MLR-low populations. Human iDC in allogenic MLR cultures are stimulated by the control allogenic T cells to mature to potent

Table II. MODS score in patients

MODS a

High M␾ MLRa

Low M␾ MLR

5.0

8.5

High MLR vs low MLR, ⬎33% or ⬍33% of control values, respectively.

6358

M␾ OF TRAUMA PATIENTS FAIL TO DIFFERENTIATE TO iDC regated the patient M␾ data on the basis of MLR dysfunction after culture with 4⫹GM. This data analysis included both M␾ with a dysfunctional differentiation to iDC and those with iDC maturation defects. Using this more inclusive criteria, the difference in CD1a differentiation between the low- and high-MLR-inducing cells of patients was even more striking (Fig. 3A compared with Fig. 2A). A variation in 4⫹GM-induced CD1a differentiation was observed in the high MLR patient M␾ population, ranging from a low of 50% of control CD1a conversion (illustrated by the Fig. 3A flow cytometry profile) to a high of 110% of control. In contrast, CD1a levels of low MLR-inducing 4⫹GM-cultured M␾ were uniformly depressed, often almost negative (Fig. 3A). Similarly, the HLA-DR expression of low MLR-inducing iDC failed to increase (Fig. 3B), and the up-regulation of the costimulatory molecule CD86 by these cells was depressed (Fig. 3C) in both the total percentage of positive cells and the ligand expression density (mean fluorescence intensity (MFI)). These data explain some of the reduced MLR capacity of these 4⫹GM-cultured patient M␾. Other immunostimulatory defects of these 4⫹GM-induced patient M␾ included a reduced production of IL-12 even after additional stimulation with SEB plus IFN-␥, which should have matured any DC through Toll-like receptor 2 activation (Fig. 4A). IL-15 levels were also reduced in this system (data not shown). This reduced IL-12 and IL-15 production was also detectable at the mRNA level and was not the result of reduced viability, because the housekeeping ribosomal protein gene L32 was unaffected (Fig. 4B). Possible contributors to the failure of trauma patients to develop DC with potent APC function

FIGURE 2. Posttrauma low-MLR capacity M␾ fail to differentiate into iDC. A, Fresh M␾ of patients and controls were stimulated with 4⫹GM (100 ng/ml each) for 5 days and assessed for CD1a expression (n ⫽ 66, 61 (high MLR), and 31 (low MLR)) and for subsequent induction of allogenic T cell proliferation in an MLR (n ⫽ 60, 53 (high MLR), and 30 (low MLR)), as described in Materials and Methods. The M␾ of patients were again segregated into low MLR-inducing and high MLR-inducing M␾ based on their ability as fresh M␾ to stimulate an MLR (low MLR were ⱕ33% of control value, whereas immunostimulatory were ⬎33% of control value). Then, the 4⫹GM-treated M␾ were subsequently analyzed by MLR to induce allogenic T cells proliferation. Median is indicated by the bar. ⴱⴱ, p ⬍ 0.001, compared with control. #, p ⬍ 0.05, compared with immunostimulatory patient samples. B, Correlation of induction of MLR by M␾ to induction by subsequently differentiated iDC. Variance analysis was performed between MLR capacity of fresh M␾ and that of 4⫹GMtreated M␾ (iDC).

APC (34). Similarly, DC of controls induced a median T cell proliferation in the MLR of 22,845 dpm compared with 1,506 dpm for the MLR cultures of undifferentiated M␾ (Fig. 1A). The high MLR-inducing M␾ of patients that differentiated to iDC showed a similar allogenic MLR-induced maturation, stimulating 16,057 dpm of T cell proliferation vs 1,208 for the M␾. In contrast, patient M␾ with low-MLR capacity not only failed to differentiate to CD1a⫹ iDC but also showed only minimal increases in MLRinducing capacity when cultured with allogenic T cells (Fig. 2A). Although the MLR-inducing capacity of patient M␾ was highly correlated with the ability of their 4⫹GM-cultured M␾ to mature from iDC to highly competent APC in a subsequent allogenic MLR, there was a greater degree of MLR depression than reduced CD1a expression (Fig. 2B). Characteristics of APC-dysfunctional patient 4⫹GM-cultured M␾ Because the MLR defect of 4⫹GM-cultured M␾ was greater than the depressed CD1a expression of 4⫹GM-cultured M␾, we reseg-

4⫹GM-induced M␾-to-DC differentiation can be altered by numerous microenvironmental changes (28). Although increased IL-10 production inhibits M␾-to-DC differentiation, low-MLR capacity 4⫹GM-cultured patient M␾ had reduced IL-10 at both the mRNA (Fig. 4B) and protein levels (Fig. 4C). Increased DC apoptosis during differentiation could contribute to a DC differentiation defect. In fact, the low MLR-inducing 4⫹GM-cultured M␾ of patients did have significantly increased apoptosis compared with controls after a 5-day culture period. However, the patient M␾ with high MLR-inducing capacity also showed increased apoptosis during 5-day culture (Fig. 4C). Decreased GM-CSF stimulation can decrease M␾-to-iDC differentiation (25). Although all cultures had similar exogenous GM-CSF added, the expression of the GMCSFR on patient M␾ with low-MLR capacity was significantly reduced, perhaps reducing efficiency of the GM-CSF induction (Fig. 5). IL-4 is necessary to drive M␾ to iDC (19). A reduced ratio of IL-4 to GM-CSF can result in a failure of M␾ to convert to iDC in a timely fashion (25, 35, 36, 37). Again, although added IL-4 concentrations were constant, the levels of M␾ IL-4R expression were significantly reduced on low MLR-capacity M␾. Finally, the differentiation of human M␾ to iDC is finely balanced against their differentiation to activated Mac (21, 23–25). M-CSF drives M␾ differentiation to Mac and inhibits GM-CSFR expression and iDC differentiation (19, 23–25). Other cytokines like IL-6, which up-regulate the M-CSFR (CD115), cooperate with M-CSF to increase M␾-to-Mac differentiation (23–25). Patient M␾ with low or high DC MLR induction capacity were compared with control M␾ for production of M-CSF and expression of both the IL-6R (CD126) and the M-CSFR (CD115). The median expression of CD115 of control M␾ was 5.1% positive (Fig. 6A and Table III). Patients whose 4⫹GM M␾ had induced higher levels of proliferation in the MLR had a median CD115 expression of 3.8% positive. In contrast, 40% of low DC MLR M␾ were CD115 positive. Interestingly, the M␾ that showed low DC MLR function had similar CD126 expression (44.1% positive) to that of control


6359

FIGURE 3. Failure to increase MLR induction after 4⫹GM culture indicates a more severe M␾ differentiation defect. A, Fresh M␾ were stimulated with IL-4 plus GM-CSF (100 ng/ml each) for 5 days, and the cells were assessed for CD1a expression by flow cytometry as well as for their ability to induce proliferation of allogenic T cells in an MLR setup, as described in Materials and Methods. The patient 4⫹GM-treated M␾ data was resegregated based on the allogenic MLRinducing capacity of 4⫹GM-treated M␾ (iDC) (low MLR-inducing iDC were ⱕ33% of control value, whereas high MLR-inducing iDC were ⬎33% of control value). A, CD1a expression was reanalyzed based on this division. The patients’ individual iDC CD1a percent positive data from the low MLR group (n ⫽ 21), high MLR induction group (n ⫽ 16), and normals (n ⫽ 34) are presented with the median shown as a bar. ⴱ, p ⬍ 0.0001. One representative flow cytometry histogram is presented for each group (shaded curve, isotype control; open curve, specific CD1a staining). B and C, Simultaneous with measurement of the CD1a expression of the patient groups, the M␾ cultured with 4⫹GM (iDC) were also assessed for expression of HLA-DR (low DC MLR, n ⫽ 20; percentage of positive skewness, 0.186 ⫾ 0.51; MFI skewness, 3.51 ⫾ 0.52) (B) and CD86 (low MLR DC, n ⫽ 20; percentage of positive skewness, 1.48 ⫾ 0.51; MFI skewness, 2.02 ⫾ 0.51) (C), and the data are expressed as the median percentage of positive cells as well as MFI with the range indicated. ⴱⴱ, p ⬍ 0.0001; ⴱ, p ⬍ 0.001, compared with normal control value; whereas #, p ⬍ 0.05, compared with immunostimulatory patient samples.

M␾ (33.4% positive), but those trauma patient M␾ that could differentiate into iDC with high MLR-inducing capacity had downregulated their CD126 expression (19.5% positive). The difference in the M␾ M-CSF production of the two patient groups was the most dramatic. Normal and patient M␾ that could differentiate to potent DC APC produced significantly lower levels of M-CSF than did the patient M␾ that failed to differentiate to MLR-inducing DC (Fig. 6B). M␾ production of M-CSF and their response to M-CSF seemed to highly influence eventual DC differentiation. Addition of Ab to M-CSF into 4⫹GM patient M␾ cultures improved both their differentiation to CD1a⫹ iDC and their subsequent maturation to APC in allogenic MLR (Fig. 6C).

Discussion These data represent the first reexamination of the Ag-presenting function of M␾ of posttrauma patients in several years and the first identification of a DC differentiation defect in trauma patients. Our data indicate that trauma patient M␾ that are unable to support allogenic T cell proliferation in an MLR response are unable to differentiate into CD1a⫹ iDC under the influence of exogenous 4⫹GM. Unlike some previously detected patient DC defects, the posttrauma iDC defect does not result from a loss of circulating CD11c⫹ DC precursors, nor is it a universal result of injury. Patients whose M␾ can function as APC in allogenic MLR also possess the capacity to differentiate to iDC. The CD1a⫹ iDC pheno-

type was not developed in 4⫹GM-cultured M␾ of patients identified as MLR dysfunctional, nor was increased M␾ CD86 or HLA-DR expression seen. The 4⫹GM-cultured patient DC maturation in response to subsequent SEB (assessed by increased IL-12 and IL-15 production) also failed to occur in the MLRdysfunctional M␾ group, as did further DC maturation in an allogenic MLR subsequent to 4⫹GM cultures. Elevated in vivo IL-6 levels combined with high M-CSF levels favor M␾ differentiation to Mac over M␾ differentiation to DC (23, 25, 27, 38). In vitro culture of human M␾ with M-CSF differentiates them to Mac (19, 25, 39). The APC-competent and APC-dysfunctional trauma patient M␾ groups did not differ in their IL-6 levels (data not shown). However, elevated IL-6 levels are produced by numerous cell types in the postinjury microenvironment and would influence M␾-to-Mac differentiation in APC-dysfunctional patient M␾ that do not down-regulate their CD126 (IL-6R) expression (40, 41). The failure of APC-dysfunctional M␾ to decrease CD126 expression would increase their sensitivity to elevated circulating IL-6 levels postinjury. In this study, patient M␾ with low immunostimulatory capacity had increased Mac-like phenotype characteristics, such as increased CD64 expression, increased production of M-CSF, increased expression of the M-CSFR (CD115), and down-regulated GM-CSFR expression (31, 42, 43). A postinjury increase in myeloid M-CSFR expression has also been recently reported in a

6360


FIGURE 4. Altered functions of the 4⫹GM-treated low MLR M␾ of patients. A, 4⫹GM-induced M␾ from patients (n ⫽ 14, low-MLR capacity, and n ⫽ 21, highMLR capacity samples) and control samples (n ⫽ 36; two outliers are ⬎2500) were assessed for IL-12 production (tested by ELISA) after stimulation with SEB (0.5 ␮g/ml) plus IFN-␥ (100 IU/ml) for 18 h. Data represent individual patient values with the median shown as a bar. ⴱ, p ⬍ 0.05, compared with control. B, 4⫹GMtreated M␾ (1 ⫻ 106 cells) were stimulated for 8–10 h in the presence of SEB (0.5 ␮g/ml) plus IFN-␥ (100 IU/ml). Total cytoplasmic RNA was isolated and mRNA levels for IL-10, IL-12, and IL-15, as well as L32 (loading control), were assayed by multiprobe RPA. The data are representative of three experiments. C, Supernatants of 4⫹GM-cultured M␾ of patients (n ⫽ 18 (low MLR-inducing samples) and 28 (high MLR-inducing samples)) and controls (n ⫽ 47) were tested for IL-10 levels by ELISA, while the cells were assessed for apoptosis by staining with Annexin V-FITC and analyzed by flow cytometry. ⴱⴱ, p ⬍ 0.001, compared with normal control median value. Ranges are indicated. Skewness in low MLR IL-10 data, 2.41 ⫾ 0.54; skewness in AnnexinFITC samples, 1.14 ⫾ 0.69.

murine burn model (2). Increased M-CSFR and decreased GMCSFR expression are also characteristics of CD34⫹ stem cells that shifted from DC to Mac differentiation after M-CSF plus IL-6 culture (23). Increased circulating CD64⫹ inflammatory Macs have been previously described in injury, and these more inflammatory Mac have altered function (41, 44 – 46). We have previously reported that trauma patient M␾ with high CD64 expression produce exaggerated amounts of the 26-kDa, cell-associated form of TNF-␣ with increased TNF-␣ mRNA stability (1, 47). A recent report describes hM-CSF-differentiated PBMC Mac as producing higher TNF levels than M␾ because of their increased TNF mRNA stability. These data additionally suggest a shift toward a more Mac-like phenotype for circulating, aberrant M␾ populations of trauma patients (48). Finally, both M␾-to-iDC differentiation and the subsequent MLR-inducing capacity were improved if anti-MCSF was added to patient M␾ 4⫹GM cultures, supporting the contribution of M-CSF to the DC defect of the patients. However, additional M-CSF effects on DC function, besides blocking M␾to-iDC differentiation, have been described. Appropriate DC MHC class II expression is also altered by M-CSF exposure (38). Consequently, the iDC MLR increases after anti-M-CSF treatment

could reflect improved MHC class II levels as well as improved M␾-to-iDC differentiation. The data in this study do not address the full component of DC APC defects or mechanisms inducing APC dysfunction that could be occurring postinjury. The 4⫹GM M␾-derived iDC population represents both future interstitial DC and future Langerhans DC depending on received maturation signals (15, 19). Although human PBMC M␾-derived DC appear to be potent APC, the subsets may have different capacities for T lymphocyte activation (15, 18, 31, 34). Consequently, the detected difference in MLR-inducing capacity may be related to undetected shifts in patient DC subsets. TGF␤ matures Langerhans like DC. M␾ TGF␤ levels are known to be increased in these immunoaberrant trauma patients, possibly increasing Langerhans like DC (46). In two preliminary experiments, no increase in CD207 (the Langerhans marker) was detected on low MLR-capacity patient M␾, but these data are far from conclusive. A posttrauma increase in plasmacytoid or lymphoid-derived DC precursors could greatly affect overall DC APC capacity (15, 16, 49). The plasmacytoid DC population is suggested to preferably induce regulatory T cells (15, 16). However, our isolation technique selects only for myeloid DC, and therefore,

FIGURE 5. Altered IL-4R and GM-CSFR expression in MLR low-inducing patient M␾. M␾ from patients (low MLR (n ⫽ 10) and high MLR (n ⫽ 47) M␾ samples) and controls (n ⫽ 56) were assessed for expression of IL-4R and GM-CSF-R␣ chain using a specific fluorokine kit. Individual patient M␾ results are expressed as the percentage of positive cells with a bar at the median. ⴱⴱ, p ⬍ 0.01, compared with normal control value. #, p ⬍ 0.05, compared with immunostimulatory patient samples.


6361

FIGURE 6. The posttrauma DC differentiation defect in low MLR-inducing M␾ correlates with increased sensitivity to and production of M-CSF. A, Low and high MLR-inducing M␾ from patients and controls were assessed for expression of CD115 (M-CSFR) and CD126 (IL-6R) by flow cytometry. Flow cytometry gates were set according to the isotype controls. The data are representative of 10 experiments. B, Fresh M␾ from patients (n ⫽ 12 (low MLR-inducing) and 11 (high MLR-inducing samples)) and controls (n ⫽ 25) were stimulated with IL-4 plus GM-CSF (100 ng/ml each) for 5 days and the cells were assessed for M-CSF production (tested by ELISA) after stimulation with SEB (0.5 ␮g/ml) plus IFN-␥ (100 IU/ml) for 18 h. C, Neutralizing Ab to MCSF (1 ␮g/ml) or isotype control was added to patient M␾ (n ⫽ 12) at the beginning of 4⫹GM treatment (5 days) followed by assessment of CD1a expression as well as MLR activity. Data are expressed as mean percent increases in MLR activity or CD1a expression. ⴱ, p ⬍ 0.01, compared with normal control value.

the detected posttrauma defect is in the myeloid M␾ DC population. Our examination of isolated patient M␾ also does not address the in vivo T lymphocyte contribution to intensifying or prolonging a postinjury APC dysfunction. Diminished T cell IL-4 and/or IL-13 production due to postinjury T cell apoptosis would further intensify any in vivo M␾-to-iDC defect. A defect in iDC-to-mDC development may also be occurring. There was a small (⬍10%) population of CD1a⫹ DC generated in the 4⫹GM-cultured M␾ of some APC-dysfunctional patients. The APC defect was more apparent in the allogenic T cell MLR cultures of these patients’ iDC (Fig. 2B). Consequently, a DC maturation defect may also contribute to postinjury APC dysfunction. A recent report showed that septic but not simple trauma patients have increased in vivo levels of DC apoptosis in their lymph nodes, suggesting that the interdigitating DC may fail to respond to T cell maturation and/or survival signals (50). No difference in apoptosis levels between APC-competent and APC-dysfunctional patient M␾ during 4⫹GM cultures was detected. Both were increased. However, we did not assess apoptosis during the subsequent MLR. Increased apoptosis of the low MLR-capacity iDC populations during the MLR could have contributed to the more significant level of APC defects detected in this second culture period (Fig. 2B). The most significant difference seen in the low MLR-inducing patient M␾ population and the high MLR-inducing patient M␾ was their production of M-CSF and expression of M-CSFR. The failure of low MLR-inducing patient M␾ to differentiate to iDC most closely paralleled their expression of more Mac-like characteristics such as increased CD64⫹, high M-CSFR expression, re-

Table III. Expression of different surface markers on the surface of fresh M␾ M␾ Markersa

CD115 (%) CD126 (%) a b c

Control

High DC MLRb

Low DC MLRc

5.1 43.4

3.76 19.5

40.9 33.1

M␾ markers expressed as a median percentage of positive cells. High DC MLR values are defined as ⬎33% of MLR scores of controls. Low MLR DC defined are ⬍33% of normal DC MLR.

duced GM-CSFR expression, low APC function, low IL-12 production, and high M-CSF production (31, 42). Our previous publications, as well as those of others, suggest that posttrauma immunoaberrant M␾ have increased Mac-like characteristics (44 – 47). The partial amelioration of the M␾-to-iDC differentiation defects in these low APC-capacity M␾ of trauma patients after antiM-CSF treatment suggest preferential Mac differentiation. The known posttrauma microenvironment contains increased circulating IL-6, PGE2, and gut-released LPS that can alter M␾ receptor expression and preferentially induce M-CSF, thereby interfering with M␾-to-iDC differentiation while favoring inflammatory Mac development.

Acknowledgments We thank Karen Kodys for her excellent technical assistance. The help of the technical and nursing staff at University of Massachusetts Medical School and University of Rochester Medical School is also sincerely appreciated.

References 1. Furse, R. K., K. Kodys, D. Zhu, and C. L. Miller-Graziano. 1997. Increased M␾ TNF␣ message stability contributes to trauma patients’ increased TNF production. J. Leukocyte Biol. 62:524. 2. Santangelo, S., R. Gamelli, and R. Shankar. 2001. Myeloid commitment shifts toward monocytopoiesis after thermal injury and sepsis. Ann. Surg. 233:97. 3. Wang, H., O. Bloom, M. Zhang, J. M. Vishnubhakat, M. Ombrellino, J. Che, A. Frazier, H. Yang, S. Ivanova, L. Borovikova, et al. 1999. HMG-1 as a late mediator of endotoxin lethality in mice. Science 285:248. 4. Abraham, E. 2000. Immune dysfunction following trauma-haemorrhage: influence of gender and age. Cytokine 12:69. 5. Jarrar, D., I. H. Chaudry, and P. Wang. 1999. Organ dysfunction following hemorrhage and sepsis: mechanisms and therapeutic approaches. Int. J. Mol. Med. 4:575. 6. Nast-Kolb, D., C. Waydhas, C. Gippner-Steppert, I. Schneider, A. Trupka, S. Ruchholtz, R. Zettl, L. Schweiberer, and M. Jochum. 1997. Indicators of the posttraumatic inflammatory response correlate with organ failure in patients with multiple injuries. J. Trauma 42:446. 7. De, A. K., K. Kodys, S. Fairfield, and C. L. Miller-Graziano. 1996. Relationship of post-trauma altered IL-12 and IL-10 to depressed patient mitogen responses. In The Immune Consequences of Trauma, Shock and Sepsis: Mechanisms and Therapeutic Approaches. Springer-Verlag, Berlin, p. 315. 8. Hensler, T., C.-D. Heidecke, H. Hecker, K. Heeg, H. Bartels, N. Zantl, H. Wagner, J.-R. Siewart, and B. Holzmann. 1998. Increased susceptibility to postoperative sepsis in patients with impaired monocyte IL-12 production. J. Immunol. 161:2655. 9. Goebel, A., E. Kavanagh, A. Lyons, I. B. Saporoschetz, C. Soberg, J. A. Lederer, J. A. Mannick, and M. L. Rodrick. 2000. Injury induces deficient interleukin-12

6362

10.

11.

12.

13. 14. 15. 16. 17.

18.

19.

20.

21.

22. 23.

24.

25.

26.

27.

28.

29.


production, but interleukin-12 therapy after injury restores resistance to infection. Ann. Surg. 231:253. Gibbons, R. A., O. M. Martinez, R. C. Lim, J. K. Horn, and M. R. Garovoy. 1989. Reduction in HLA-DR, HLA-DQ and HLA-DP expression by Leu-M3⫹ cells from the peripheral blood of patients with thermal injury. Clin. Exp. Immunol. 75:371. Ayala, A., W. Ertel, and I. H. Chaudry. 1996. Trauma-induced suppression of antigen presentation and expression of major histocompatibility class II antigen complex in leukocytes. Shock 5:79. Ditschkouski, M., E. Kreuzfeider, V. Rebmann, S. Ferencik, M. Majetschak, E. N. Schmid, U. Obertacke, H. Hirche, U. F. Schade, and H. Grosse-Wilde. 1999. HLA-DR expression and soluble HLA-DR levels in septic patients after trauma. Ann. Surg. 229:246. Ayala, A., M. A. Urbanich, C. D. Herdon, and I. H. Chaudry. 1996. Is sepsisinduced apoptosis associated with macrophage dysfunction? J. Trauma 40:568. Reid, S. D., G. Penna, and L. Adorini. 2000. The control of T cell responses by dendritic cell subsets. Curr. Opin. Immunol. 12:114. Kelsall, B. L., C. A. Biron, O. Sharma, and P. M. Kaye. 2002. Dendritic cells at the host-pathogen interface. Nat. Immunol. 3:699. Liu, Y.-J., V. S. Kanzler, and M. Gilliet. 2001. Dendritic cell lineage, plasticity and cross-regulation. Nat. Immunol. 2:585. Zhou, L. J., and T. F. Tedder. 1996. CD14⫹ blood monocytes can differentiate into functionally mature CD83⫹ dendritic cells. Proc. Natl. Acad. Sci. USA 93: 2588. Nagata, Y., S. Ono, M. Matsuo, S. Gnjatic, D. Valmori, G. Ritter, W. Garrett, L. J. Old, and I. Mellman. 2002. Differential presentation of a soluble exogenous tumor antigen, NY-ESO-1, by distinct human dendritic cell populations. Proc. Natl. Acad. Sci. USA 99:10629. Relloso, M., A. Puig-Kroger, O. M. Pello, J. L. Rodriguez-Fernandez, G. de la Rosa, N. Longo, J. Navarro, M. A. Munoz-Fernandez, P. Sanchez-Mateos, and A. L. Corbi. 2002. DC-SIGN (CD209) expression is IL-4 dependent and is negatively regulated by IFN, TGF-␤, and anti-inflammatory agents. J. Immunol. 168:2634. Palucka, K. A., N. Taquet, F. Sanchez-Chapuis, and J. C. Gluckman. 1999. Lipopolysaccharide can block the potential of monocytes to differentiate into dendritic cells. J. Leukocyte Biol. 65:232. Bancherau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. H. Li, B. Pulendran, and K. A. Palucka. 2000. Immunology of dendritic cells. Annu. Rev. Immunol. 18:767. Urban, B. C., N. Willcox, and D. J. Roberts. 2001. A role for CD36 in the regulation of dendritic cell function. Proc. Natl. Acad. Sci. USA 98:8750. Menetrier-Caux, C., G. Montmain, M. C. Dieu, C. Bain, M. C. Favrot, C. Caux, and J. Y. Blay. 1998. Inhibition of the differentiation of dendritic cells from CD34⫹ progenitors by tumor cells: role of interleukin-6 and macrophage colonystimulating factor. Blood 92:4778. Chomarat, P., J. Banchereau, J. Davoust, and A. K. Palucka. 2000. IL-6 switches the differentiation of monocytes from dendritic cells to macrophages. Nat. Immunol. 1:510. Delneste, Y., P. Charbonnier, N. Herbault, G. Magistrelli, G. Caron, J.-Y. Bonnefoy, and P. Jeannin. 2002. Interferon-␥ switches monocyte differentiation from dendritic cells to macrophages. Blood 4:1164. Wysocka, M., M. H. Zaki, L. E. French, J. Chehimi, M. Shapiro, S. E. Everetts, K. S. McGinnis, L. Montaner, and A. H. Rook. 2002. Sezary syndrome patients demonstrate a defect in dendritic cell populations: effects of CD40 ligand and treatment with GM-CSF on dendritic cell numbers and the production of cytokines. Blood 100:3287. Ratta, M., F. Fagnoni, A. Curti, R. Vescovini, P. Sansoni, B. Oliviero, M. Fogli, E. Ferri, G. R. Della Cuna, S. Tura, et al. 2002. Dendritic cells are functionally defective in multiple myeloma: the role of interleukin-6. Blood 100:230. Vicari, A. P., C. Chiodoni, C. Vaure, S. Ait-Yahia, C. Dercamp, F. Matsos, O. Reynard, C. Taverne, P. Merle, M. P. Colombo, et al. 2002. Reversal of tumor-induced dendritic cell paralysis by CpG immunostimulatory oligonucleotide and anti-interleukin 10 receptor antibody. J. Exp. Med. 196:541. Marshall, J. C., D. J. Cook, N. V. Christou, G. R. Bernard, C. L. Sprung, and W. J. Sibbald. 1995. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit. Care Med. 23:1638.

30. De, A. K., K. Kodys, J. Pellegrini, B. Yeh, R. K. Furse, P. Bankey, and C. L. Miller-Graziano. 2000. Induction of global anergy rather than inhibitory Th2 lymphokines mediates posttrauma T cell immunodepression. Clin. Immunol. 96:52. 31. MacDonald, K. P. A., D. J. Munster, G. J. Clark, A. Dzionek, J. Schmitz, and D. N. J. Hart. 2002. Characterization of human blood dendritic cell subsets. Blood 11:97. 32. Brade, H., S. Opal, S. N. Vogel, and D. C. Morrison. 1999. Endotoxin in Health and Disease. Marcel Decker, New York. 33. Dobrovolskaia, M. A., and S. N. Vogel. 2002. Toll receptors, CD14, and macrophage activation and deactivation by LPS. Microbes Infect. 4:903. 34. Osugi, Y., S. Vuckovic, and D. N. J. Hart. 2002. Myeloid blood CD11c⫹ dendritic cells and monocyte-derived dendritic cells differ in their ability to stimulate T lymphocytes. Blood 100:2858. 35. Mangan, D. F., B. Robertson, and S. M. Wahl. 1992. IL-4 enhances programmed cell death (apoptosis) in stimulated human monocytes. J. Immunol. 148:1812. 36. Kohrgruber, N., N. Halanek, M. Groger, D. Winter, K. Rappersberger, M. Schmitt-Egenolf, G. Stingl, and D. Maurer. 1999. Survival, maturation, and function of CD11c⫺ and CD11c⫹ peripheral blood dendritic cells are differentially regulated by cytokines. J. Immunol. 163:3250. 37. de Vries, I. J. M., A. A. O. Eggert, N. M. Scharenborg, J. L. M. Vissers, W. J. Lesterhuis, O. C. Boerman, C. J. A. Punt, G. J. Adema, and C. G. Figdor. 2002. Phenotypical and functional characterization of clinical grade dendritic cells. J. Immunother. 25:429. 38. Baron, C., G. Raposo, S. Scholl, H. Bausinger, D. Tenza, A. Bohbot, P. Pouillart, B. Goud, D. Hanau, and J. Salamero. 2001. Modulation of MHC class II transport and lysosome distribution by macrophage-colony stimulating factor in human dendritic cells derived from monocytes. J. Cell Sci. 114:999. 39. Sordet, O., C. Rebe, S. Plenchette, Y. Zermati, O. Hermine, W. Vainchenker, C. Garrido, E. Solary, and L. Dubrez-Daloz. 2002. Specific involvement of caspases in the differentiation of monocytes into macrophages. Blood 6:1778. 40. Abraham, E. 1996. Alterations in transcriptional regulation of proinflammatory and immunoregulatory cytokine expression by hemorrhage, injury, and critical illness. New Horiz. 4:184. 41. Szabo, G., K. Kodys, and C. L. Miller-Graziano. 1991. Elevated monocyte interleukin-6 (IL-6) production in immunosuppressed trauma patients. I. Role of Fc␥RI crosslinking stimulation. J. Clin. Immunol. 11:326. 42. Zeigler-Heitbrock, H. W. L. 2000. Definition of human blood monocyte. J. Leukocyte Biol. 67:603. 43. Macey, M. G., D. A. McCarthy, D. Vogiatzi, K. A. Brown, and A. C. Newland. 1998. Rapid flow cytometric identification of putative CD14⫺ and CD64⫺ dendritic cells in whole blood. Cytometry 31:199. 44. Schinkel, C., R. Sendtner, S. Zimmer, and E. Faist. 1998. Functional analysis of monocyte subsets in surgical sepsis. J. Trauma 44:743. 45. Miller-Graziano, C. L., G. Szabo, K. Kodys, and K. Griffey. 1990. Aberrations in post-trauma monocyte subpopulation: role in septic shock syndrome. J. Trauma 30:S86. 46. Miller-Graziano, C. L., G. Szabo, K. Griffey, B. Mehta, K. Kodys, and D. Catalano. 1991. Role of elevated monocyte transforming growth factor ␤ (TGF␤) production in post-trauma immunosuppression. J. Clin. Immunol. 11:95. 47. Szabo, G., C. L. Miller-Graziano, J.-Y. Wu, T. Takayama, and K. Kodys. 1990. Differential tumor necrosis factor production by human monocyte subsets. J. Leukocyte Biol. 47:206. 48. MacKenzie, S., N. Fernandez-Troy, and E. Espel. 2002. Post-transcriptional regulation of TNF-␣ during in vitro differentiation of human monocytes/macrophages in primary culture. J. Leukocyte Biol. 71:1026. 49. Matsuda, H., T. Suda, H. Hashizume, K. Yokomura, K. Asada, K. Suzuki, K. Chida, and H. Nakamura. 2002. Alteration of balance between myeloid dendritic cells and plasmacytoid dendritic cells in peripheral blood of patients with asthma. Am. J. Respir. Crit. Care Med. 166:1050. 50. Hotchkiss, R. S., K. W. Tinsley, P. E. Swanson, M. H. Grayson, D. F. Osborne, T. H. Wagner, J. P. Cobb, C. Coopersmith, and I. E. Karl. 2002. Depletion of dendritic cells, but not macrophages, in patients with sepsis. J. Immunol. 168:2493.