Materials and Methods

Mice.

Male C57BL/6ByJ mice, aged 6 to 8 weeks (Jackson Laboratories, Bar Harbor, ME), were used for all studies. The mice were given access to food and water ad libitum.

Murine Splenocyte Preparation.

Mice were sacrificed by CO₂ inhalation, and spleens were removed aseptically as described previously (Bidlack et al., 1996). Splenocytes were dissociated by gently pressing the spleens between the frosted ends of sterile microscope slides in ice-cold, HEPES-buffered balanced salt solution (HEPES-BSS), consisting of 15 mM HEPES, 3.4 mM K₂HPO₄, 0.6 mM KH₂PO₄, 150 mM NaCl, 5 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgSO₄, and 1% BSA, pH 7.4. Cell suspensions were passed over sterile, glass-wool columns to remove dead cells and debris. After centrifugation at 200g for 10 min at 4°C, cell suspensions were treated with a cold, isotonic ammonium chloride solution to lyse contaminating erythrocytes. Unfixed cells, washed twice by centrifugation at 200g for 10 min at 4°C and counted in a Coulter Z1 Counter, were resuspended at a final concentration of 2 × 10⁶ cells in 200 μl HEPES-BSS for optimal staining of the κ-opioid receptor on B cells as described below. Enrichment of mature, splenic T cells was achieved by further passage of the cell suspension over a nylon-wool column eluted with 37°C HEPES-BSS. After centrifugation at 200g for 10 min at 4°C, fluorescent opioid labeling of the cells was measured as described below.

Peritoneal Mφ.

Thioglycollate (sterile, 3% solution; aged 2–4 months before use; Sigma Chemical Co., St. Louis, MO) was injected i.p. (1.5 ml/mouse) to invoke the production of an inflammatory exudate rich in Mφ. Elicited Mφ were harvested and pooled from 4 to 5 mice, and resident peritoneal Mφ were harvested and pooled from 10 to 15 mice by peritoneal lavage. Ice-cold PBS (without Ca²⁺ and Mg²⁺) was used for all Mφ harvests, as well as in the κ-opioid receptor-labeling procedure for Mφ. All Mφ samples, in addition to 1% BSA, contained 10% rabbit (Rb) serum to decrease further nonspecific staining in the labeling procedure, which is similar to that described below for lymphocytes. For in vivo Mφ activation experiments, each mouse was injected i.p. with 10 μg of lipopolysaccharide (LPS) (20 μg/ml;Escherichia coli type 0111:B4; Sigma Chemical Co.) or with 0.5 ml of sterile, 0.9% sodium chloride solution, the LPS vehicle control. Various times after saline or LPS injection, mice were sacrificed and Mφ were harvested and pooled from 10 mice/group. Resident, peritoneal Mφ were used as an additional control group for the Mφ activation studies because, as noted by others (Fisker et al., 1992), the i.p. injection of sterile saline induced a mild inflammatory reaction as revealed by an increased polymorphonuclear cell immigration into the peritoneum.

Indirect Immunofluorescent and Phenotypic Labeling.

The FITC-AA labeling and amplification procedure were performed as described previously (Lawrence et al., 1995a; Ignatowski and Bidlack, 1998). In a final volume of 200 μl HEPES-BSS, cells were incubated with 30 μM FITC-AA for 30 min at 25°C for optimal staining. The κ-selective antagonist nor-BNI (500 μM) was included to measure nonspecific fluorescence. Samples were chilled on ice, diluted with 1 ml of HEPES-BSS, and centrifuged at 400g for 3 min at 4°C. After aspirating the supernatants, cells were washed twice and resuspended in a final volume of 100 μl of HEPES-BSS, containing 10 μl of biotinylated Rb antifluorescein IgG, except in FITC-AA- and PE-only controls, in which the Rb Ab was omitted. After incubation for 30 min at 4°C in the dark, samples were diluted with 1 ml of HEPES-BSS and centrifuged at 400g for 3 min at 4°C. The supernatants were aspirated, and the cells were washed again. Cells were resuspended in 40 μl of HEPES-BSS and 10 μl of PE for 15 min at 4°C in the dark. The cells were washed twice as described above and were resuspended in a final volume of 1 ml of HEPES-BSS for flow cytometric analysis.

Phenotypic characterization of lymphocytes possessing the κ-opioid receptor was undertaken as above with minor modifications. Fluorophore-conjugated, rat mAbs (Sigma Chemical Co.) directed against mouse cell surface markers FITC-CD4, Quantum Red (QR)-CD4 (clone H129.19), QR-CD8 (clone 53–6.7), QR-CD45R (clone RA3–6B2), as well as FITC-Mac-3 (clone M3/84), PE-NK1.1 (clone PK136), and FITC-IgD^b (clone 217–170) (PharMingen, San Diego, CA) were assayed for optimal titer before use. QR-CD4, -CD8, or -CD45R was added to appropriate tubes during the second 30-min incubation, followed by the subsequent steps listed above. For phenotypic analysis alone, lymphocytes (1 × 10⁶ cells/100 μl HEPES-BSS containing 10% Rb serum) were incubated with 2 to 4 μl (1 μg) of each mAb for single phenotypic staining or with 4 μl of QR-CD8/FITC-CD4 for double staining, and peritoneal exudate cells (PEC) were incubated with 2 μl of FITC-Mac-3 in PBS/BSA for 30 min at 4°C in the dark. Samples were diluted with 1 ml of HEPES-BSS or PBS and were centrifuged at 400g for 3 min at 4°C. After aspiration of the supernatants, cells were washed two additional times, resuspended in 1 ml of HEPES-BSS or PBS, and analyzed for fluorescence labeling. Controls consisted of unstained immune cells (autofluorescence controls), FITC-AA- and PE-only stained cells, and cells stained with appropriate fluorophore-conjugated, isotype-matched control mAbs (nonspecific staining controls).

Flow Cytometric Analysis.

Samples were analyzed on a Becton-Dickinson FACScan flow cytometer (San Jose, CA) equipped with a 15-mW argon-ion laser for excitation (488 nm) of FITC and PE using band-pass filters of 530 ± 15 nm and 585 ± 21 nm, respectively. The red fluorescence channel (FL3) for QR (which emits at 670 nm) uses a long-pass emission filter that transmits long-wavelength light beams above 650 nm. In each sample, 25,000 lymphocytes or Mφ were analyzed.

Data Analysis.

Data [forward angle and right angle (side) light scatter, as well as green (FITC), orange (PE), and red (QR) fluorescence] were measured, collected, and stored into list-mode data files using a Macintosh Quadra 650 computer system with CELLQuest software (Becton-Dickinson Immunocytometry Systems). Median peak values of relative fluorescence intensity distributions were used to compare the fluorescence among samples, which assumed that only a single population was labeled. Viabilities were >95% as assessed by propidium iodide (Sigma Chemical Co.) exclusion. To report a relative quantification for κ-opioid receptor labeling of subpopulations of immune cells identified by mAb labeling, gating and histogram subtraction were performed using CELLQuest software as described previously (Ignatowski and Bidlack, 1998).

Results

Phenotypic Characterization of Splenocytes as Assessed by Flow Cytometry.

To determine whether preferential κ-opioid receptor expression exists on splenocyte subpopulations, it was first necessary to phenotypically characterize the mouse immune cells of interest. Splenocytes, as compared with thymocytes, differ vastly in the phenotypic profile of their lymphocyte populations (Table1). Various FITC-, PE-, and QR-conjugated mAbs directed against mouse cell surface markers were used for phenotypic-labeling experiments to define the immune cells analyzed. As reported previously (Ignatowski and Bidlack, 1998), simultaneous labeling of thymocytes with the T-helper cell marker, CD4, and with the mAb used to identify T-cytotoxic cells, CD8, demonstrated that the majority of thymocytes labeled with both mAbs, identifying them as double-positive, immature T cells. Few mature T-helper or T-cytotoxic cells were located within the thymus. In contrast, the spleen virtually lacked double-positive, immature T cells (0.35 ± 0.1%). In single-labeling experiments, it was determined that approximately 40% of the splenocytes were T cells, either T-helper cells (CD4⁺) or T-cytotoxic cells (CD8⁺; Table 1), which was confirmed by double-labeling experiments (CD4⁺/CD8⁻, 23 ± 2%; CD8⁺/CD4⁻, 16 ± 2%). Using the pan B cell marker CD45R, the majority of remaining cells were identified as B cells, of which 76% were shown to be mature B cells by labeling with the IgD^b mAb (Table 1). The remaining cells within the spleen were identified as either natural killer (NK) cells or Mφ based on labeling using their respective phenotypic mAb markers (Table 1). These phenotypic splenocyte results are similar to those reported by Beavis and Pennline (1994).

κ-Opioid Receptor Labeling of Mouse Splenocytes.

The extent of κ-opioid receptor expression on the phenotypically defined splenocyte subpopulations next was determined by using the amplification procedure for κ-opioid receptor labeling along with staining of cells by QR-conjugated mAbs to identify the particular cells of interest. Analysis of double-labeled splenocytes to identify whether T-helper cells (CD4⁺), T-cytotoxic cells (CD8⁺), or B cells (CD45R⁺) expressed the κ-opioid receptor was undertaken as described previously (Ignatowski and Bidlack, 1998). For example, in the histogram of QR-CD4 fluorescently labeled splenocytes (Fig.1A), a region was set to enclose the CD4⁺ cells. This region then was selected and analyzed in the PE fluorescence histogram for determination of κ-opioid receptor expression (Figs. 1B and 2A). Subtraction of background PE median peak fluorescence intensity demonstrated a reduction in the median fluorescence intensity values with FITC-AA/PE when nor-BNI was included as an inhibitor (Fig.2B and Table2). The percentage of specific labeling for the κ-opioid receptor was approximately the same regardless of subpopulation analyzed (Table 2), suggesting that one population of κ-opioid receptor was expressed on multiple lymphocyte subsets.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow cytometric analysis of CD4⁺splenocytes labeled for the κ-opioid receptor. Splenocytes were labeled via the amplification procedure for the κ-opioid receptor and/or anti-CD4-QR. A, CD4⁺ splenocytes were gated (R2) in the FL3 (QR) fluorescence histogram. These gated cells then were viewed in the FL2 (PE) fluorescence channel for subsequent κ-opioid receptor-labeling analysis. B, total labeling of CD4⁺splenocytes for the κ-opioid receptor is demonstrated in relation to the autofluorescence control for the whole splenocyte population. In the absence of added PE, the arbitrary fluorescence units were 0.3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

κ-opioid receptor labeling of CD4⁺splenic lymphocytes. CD4⁺ splenocytes were labeled via the amplification procedure for the κ-opioid receptor. Control samples were labeled with an isotype-matched (IgG2a) control mAb conjugated to QR. A, κ-opioid receptor labeling of CD4⁺splenocytes is demonstrated by overlaid histograms. B, specific labeling of the κ-opioid receptor on CD4⁺ splenocytes as demonstrated after histogram subtraction of the background PE fluorescence. Note the downward and leftward shift of the PE histogram in the presence of nor-BNI. Data represent four separate experiments.

Differential κ-Opioid Receptor Expression on Immature Lymphocytes Compared with Mature Lymphocytes.

Additional analysis of the double-labeled splenocyte subpopulations provided further insight into the distribution of the κ-opioid receptor on mouse immune cells. CELLQuest software was used to perform histogram (FL2 fluorescence channel for PE) subtraction of cells labeled for each mAb and the κ-opioid receptor in the presence of the κ-selective opioid nor-BNI (nonspecific labeling) from those cells labeled with each mAb and the κ-opioid receptor in the absence of nor-BNI (total labeling) on a channel-by-channel basis. This subtraction procedure yielded the resultant relative percentage of cells specifically labeled for both the mAb of interest and the κ-opioid receptor. These results are presented in Table 2. Previously, it was shown that whether thymocytes were identified using CD8 or CD4 mAbs, the relative percentage of cells expressing the κ-opioid receptor was similar—approximately 60% (Ignatowski and Bidlack, 1998). This finding was not surprising because the majority of all thymocytes were identified as CD4⁺/CD8⁺ immature T cells (Ignatowski and Bidlack, 1998). However, whereas B cells and mature T cells from the spleen also demonstrated specific labeling for the κ-opioid receptor, a much smaller percentage of cells within these subpopulations expressed the κ-opioid receptor. Less than 25% of the mature T cells and only 16% of B cells labeled for the κ-opioid receptor (Table 2).

κ-Opioid Receptor Labeling of Resident and TG-Elicited Peritoneal Mφ.

Mφ, another immunologically, well characterized cell population, were analyzed for κ-opioid receptor expression. Gating on the whole PEC population, which consisted of Mφ, lymphocytes, and polymorphonuclear cells (Eichner and Smeaton, 1983;Melnicoff et al., 1989), and resident and TG-elicited Mφ were identified for analysis in flow cytometry based on their distinctive forward scatter (FSC) versus side scatter (SSC) cellular characteristics (Ho and Springer, 1983; Hendrzak et al., 1994; Plasman and Vray, 1994) as well as on their labeling with the Mac-3 mAb, which was used for the detection of resident and inflammatory-elicited Mφ, but not monocytes, lymphocytes, or erythrocytes (Ho and Springer, 1983). Although the Mac-3 mAb also may label dendritic and endothelial cells, strict gating criteria were employed to exclude these other cell populations from analyses. Flow cytometric analysis of Mφ is demonstrated in Fig. 3. The dot plot of FSC versus SSC initially was used to depict the whole PEC population (Fig. 3A). Region R2 was drawn to enclose the larger-sized (based on FSC) cells of interest. Similar regions designated R3, R4, and R7 (Fig.3, B, C, and D, respectively) were drawn to enclose the cells of interest using dot plots of cellular size (FSC; Fig. 3B) or granularity (SSC; Fig. 3C) versus fluorescein fluorescence (FL1), as well as fluorescein versus PE fluorescence (Fig. 3D). Mutually exclusive addition of the cellular events located within regions R2, R3, R4, and R7 then was performed using CELLQuest software. The resultant group of cellular events located within region R4 was designated area G6 (Fig.3E). The cells within G6 preferentially labeled for the Mac-3-FITC mAb, further confirming their identity as Mφ. Once G6 was defined using Mac-3-FITC labeling of Mφ, all subsequent samples were analyzed using the same regions.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Flow cytometric analysis of PEC cells demonstrating selective Mφ gating. Dot plots of FSC versus SSC (A), FSC versus log-FITC fluorescence (B), SSC versus log-FITC fluorescence (C), and log-FITC versus log-PE fluorescence for PEC (D) are shown. Each dot represents a single cell; 25,000 cells are represented in each plot. The lower-end cutoff on the FSC plots indicates the threshold cutoff, which was set to eliminate lysed red blood cells and cellular debris from analysis. Mutually exclusive addition of the cellular events (E) located within regions R2, R3, R4, and R7 indicated in A–D above using CELLQuest software results in the cellular events viewed within R4 (E). Compare with R4 in C. This new region, G6 = R2+R3+R4+R7, was used to identify the Mφ populations for analysis of labeling, because cells within this region also labeled preferentially with the Mac-3 mAb.

TG elicitation initially was employed to increase the yield of Mφ harvested per mouse, thereby reducing the number of animals necessary per labeling experiment. TG injections resulted in an increase in the cell number washed from the peritoneal cavity, as expected (Eichner and Smeaton, 1983; Melnicoff et al., 1989; Fisker et al., 1992). In each case, the whole population of cells isolated from the peritoneal cavity was used in the labeling experiment because selective identification of Mφ was possible. As demonstrated in Table3, 3- or 4-day TG-induced inflammatory elicitation of Mφ resulted in an increased number of Mφ expressing the Mac-3 antigen as compared with the approximate same number of analyzed resident (nonelicited) peritoneal Mφ. Furthermore, the peak median FL1 fluorescence intensity values for Mac-3-FITC labeling were greater in TG-elicited Mφ (261 ± 19) as compared with resident Mφ (78 ± 9), suggesting an increase in the density of Mac-3 antigen per Mφ in the TG group. Interestingly, κ-opioid receptor expression also differed between the two Mφ populations. Resident Mφ demonstrated specific κ-opioid receptor labeling, with 54 ± 4% of the Mφ expressing the receptor (Table 3). In contrast, TG-elicited Mφ did not label for the κ-opioid receptor.

Rabbit serum concentrations ranging from 10 to 50% were tested to reduce nonspecific labeling and, thereby, increase specific labeling for the κ-opioid receptor on TG-elicited Mφ. Although nonspecific labeling was reduced slightly in the presence of increased Rb serum concentrations, there was no specific labeling of the receptor (data not shown). As an attempt to induce κ-opioid receptor expression on TG-elicited Mφ, PEC were incubated in RPMI 1640 medium supplemented with 300 μg/ml glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% heat-inactivated FCS at 37°C for 1 h, washed free of unbound cells, and subsequently incubated at 37°C in the presence or absence of LPS (100 ng/ml) for either 1, 14, or 24 h. This in vitro activation was unable to induce detectable κ-opioid receptor expression (data not shown).

Magnitude of κ-Opioid Receptor Expression on Immune Cells.

κ-Opioid receptor labeling of resident, peritoneal Mφ next was compared with that of the primary lymphocytes, as well as with that of the R1EGO thymoma cell line. It was found that the magnitude of labeling was much greater on Mφ than on lymphocytes (Table4). Specific labeling was determined by subtracting nonspecific from total labeling. The subsequent differences in labeling are evident by comparing the median PE fluorescence intensity values for specific labeling among the immune cell populations of interest (Table 4). The greatest magnitude of labeling is demonstrated by the R1EGO positive control cell line, which is rich in κ-opioid receptors, followed by that demonstrated by resident Mφ. Compared with lymphocytes, Mφ demonstrate approximately 25-fold greater labeling for the κ-opioid receptor. Although PE arbitrary fluorescence intensity units cannot be used to specifically state the number of receptors per cell, using this amplification-labeling procedure, the difference in magnitude of labeling suggests that resident, peritoneal Mφ express more κ-opioid receptors per cell than either splenocytes or thymocytes.

In Vivo LPS Activation of Resident, Peritoneal Mφ.

Based on the findings that κ-opioid receptor expression was different on TG-elicited Mφ compared with resident Mφ and that in vitro LPS stimulation did not alter the lack of labeling on the TG-elicited Mφ, we next decided to investigate the effect of LPS activation on nonelicited, resident Mφ κ-opioid receptor expression. In vivo Mφ stimulation was chosen as an alternative to in vitro stimulation because of the large quantity of mice necessary to undertake the in vitro experiments. For these studies, the concentration of LPS (10 μg) chosen for i.p. injection into mice was based on reports of similar, nonlethal concentrations for in vivo Mφ activation experiments, where cytokine production, a hallmark activation-inflammatory response of Mφ, or thrombin receptor binding was monitored (Remick et al., 1989; Fisker et al., 1992; Wollenberg et al., 1993). Proper injection of LPS was determined based on the physiological observations of diarrhea and lethargy in LPS-injected mice compared with vehicle-injected mice as well as with noninjected mice. Determination of Mφ κ-opioid receptor expression was undertaken 1 h (to assess immediate activation effects) and 24 h (to allow for protein synthesis or turnover) after LPS or saline control injections, as well as on resident, peritoneal Mφ (noninjected). Comparison of κ-opioid receptor labeling on Mφ from the three groups revealed no apparent difference in the expression of this particular opioid receptor at 1 h (specific labeling: control, 64 ± 3%; saline, 59 ± 5%; LPS, 52 ± 9%; and the percentage of Mφ labeled for the κ-opioid receptor: control, 69 ± 6%; saline, 60 ± 6%; LPS, 55 ± 5%;n = 4) or at 24 h (data similar to that reported for 1-h post-LPS stimulation).

Discussion

In the present report, flow cytometry combined with indirect immunofluorescence was used successfully to identify the κ-opioid receptor on subpopulations of lymphocytes and Mφ. These findings substantiate further: 1) that κ-opioid receptors are not restricted to the central nervous system, 2) that this receptor type may be widely distributed on various immune cell populations, and 3) a role for κ-opioids in immunomodulation, as suggested by numerous functional studies (Foster and Moore, 1987; Taub et al., 1991; Ichinose et al., 1995; Radulovic et al., 1995).

Previous investigations in our laboratory have demonstrated specific κ-opioid receptor labeling of immature thymocytes, as demonstrated by the lack of inhibition for FITC-AA/PE labeling by both μ- and δ-selective opioids (Ignatowski and Bidlack, 1998). This specific labeling, which was inhibited by the κ-selective antagonist nor-BNI, also has been evidenced on the R1EGO thymoma cell line (Lawrence et al., 1995a; Ignatowski and Bidlack, 1998), as well as on human microglia (Chao et al., 1996). The present findings of κ-opioid receptor expression on mature, splenic lymphocytes and on resident, peritoneal Mφ substantiate further this sensitive methodology for the detection of cellular receptors, which may be expressed at low levels and are not detectable by other techniques, such as radioligand binding.

The relative number of mature, splenic T-helper, T-cytotoxic, and B lymphocytes expressing the κ-opioid receptor was minor (23, 17, and 16%, respectively) when compared with the number of immature thymocytes (60%), demonstrating labeling for this receptor (Table 2;Ignatowski and Bidlack, 1998). These findings substantiate our previous reports of a concomitant decrease in κ-opioid receptor expression on cellular and phenotypic maturation of immature thymocytes to those cells that have up-regulated the expression of the CD3 antigen (Ignatowski and Bidlack, 1998). Studies in progress are designed to assess the effect of various T cell stimulatory agents on κ-opioid receptor expression.

Resident, peritoneal Mφ, another immune effector cell population, similarly demonstrated specific κ-opioid receptor labeling (Table 3). Because the number of resident Mφ obtained from the mouse peritoneum was very small (i.e., approximately 1 × 10⁶cells/mouse) and the number of Mφ necessary for κ-opioid receptor labeling was large (i.e., approximately 1.5 × 10⁷ cells), an i.p. injection of 3% sterile, aged thioglycollate broth was used to induce an acute inflammatory response, thereby increasing the number of Mφ harvested from the mouse peritoneal cavity. However, TG-elicited Mφ differed vastly from resident, peritoneal Mφ. Mac-3 mAb labeling was more pronounced, both in the number of Mφ expressing the antigen and in the density of antigen present per cell, on TG-elicited Mφ as opposed to resident Mφ. This finding is similar to that demonstrated by other preferential Mφ mAb markers, such as the Mac-2 mAb. Although these mAbs do not label all resident Mφ, they label elicited Mφ to a greater extent, because these cells have up-regulated antigen expression (Ho and Springer, 1983; Melnicoff et al., 1989). Likewise, the two Mφ groups differed in their expression of the κ-opioid receptor. Resident, nonelicited Mφ demonstrated specific labeling for this receptor, similar to that demonstrated by lymphocytes (Table 4), whereas TG-elicited Mφ lacked detectable κ-opioid receptor labeling (Table 3). As an attempt to induce κ-opioid receptor expression, TG-elicited Mφ were stimulated in vitro with LPS at a concentration determined previously to activate Mφ (Ignatowski and Spengler, 1995). However, in vitro LPS stimulation was unable to induce receptor expression in TG-elicited Mφ after either 1, 14, or 24 h of incubation. Therefore, these findings suggest that: 1) the TG injection may have caused a down-regulation of the κ-opioid receptor on Mφ, possibly because of a component of the TG, such as agar, which has been observed to be taken up by Mφ harvested after TG injection (Eichner and Smeaton, 1983), or 2) the injection of TG elicited a population of Mφ lacking in κ-opioid receptor expression to the peritoneal cavity. These elicited Mφ may represent an immature Mφ population recruited from monocyte influx or from another possible peritoneal Mφ precursor population, such as omental Mφ (Melnicoff et al., 1989;Wijffels et al., 1992). The recruited Mφ have been shown in labeling and tracking studies to migrate to the site of acute, sterile-induced peritonitis on the departure of resident, peritoneal Mφ (Melnicoff et al., 1989). Caution, therefore, must be used when disclosing and comparing results from studies employing TG as an eliciting agent for Mφ enrichment, because elicited Mφ differ from resident Mφ in many ways, including metabolic (Morahan et al., 1982; Eichner and Smeaton, 1983) and bactericidal activity (Baker and Campbell, 1980).

In vivo administration of LPS was chosen to compare the effects of Mφ elicitation (TG injection) with Mφ activation (LPS stimulation) for κ-opioid receptor expression. Because TG is not considered a classic activator of Mφ, but only an elicitor of Mφ, a partial activator of Mφ, or a priming agent for subsequent Mφ activation (Karnovsky and Lazdins, 1978; North, 1978), it was of interest to determine whether LPS, a Mφ-activating agent (Doe and Henson, 1978; Doe et al., 1978;Karnovsky and Lazdins, 1978), administered in vivo would affect resident Mφ κ-opioid receptor expression in a manner similar to TG administration. Resident Mφ stimulated in vivo with LPS for either 1 or 24 h did not appreciably alter their expression of the κ-opioid receptor. The complexity of interactions occurring during the in vivo response to an inflammatory stimulus, such as LPS, may account for this seemingly lack of response. It has been established that Mφ and lymphocytes, in an area of localized inflammation, release a multitude of mediators including cytokines and endogenous opioids such as β-endorphin, enkephalin, and dynorphin (Schafer et al., 1994). These opioids then may interact with multiple opioid receptors to induce analgesia at nerve terminals and to either enhance or suppress various immune responses. The lack in change of κ-opioid receptor expression on Mφ after low-dose LPS stimulation in vivo suggests that the potential interplay among the various mediators (Schafer et al., 1994; Alicea et al., 1996) allows for the continual expression of this receptor throughout the 24-h time period assessed. Studies have demonstrated the beneficial, immunoenhancing regulatory effects of κ-opioid agonists on Mφ functions, such as phagocytosis (Ichinose et al., 1995), as well as suppressive regulatory effects, such as inhibition of cytokine production by LPS-stimulated Mφ (Alicea et al., 1996). Because cytokines have been shown to induce the release of endogenous opioids from Mφ (Schafer et al., 1994), the latter findings support a possible feedback regulatory loop allowing for the maintenance of κ-opioid receptor expression during an acute inflammatory response.

Finally, as depicted in Table 4, the magnitude of specific κ-opioid receptor labeling on resident Mφ was much greater than that determined for lymphocytes. This finding suggests that per cell, Mφ express more κ-opioid receptors than lymphocytes. The endogenous κ-opioid peptide dynorphin suppresses certain T cell activities (Prete et al., 1986) while enhancing various Mφ functions (Foster and Moore, 1987; Ichinose et al., 1995). These findings illustrate the preferential modulatory effect of this endogenous opioid on lymphocytes and Mφ. This disparity in κ-opioid-mediated regulation of immune cells may relate directly to the differential expression of this receptor population on these cells.

In conclusion, we demonstrated the differential expression of the κ-opioid receptor on various immune cell populations. Interestingly, the population of primary immune cells that demonstrated the greatest magnitude of labeling for the κ-opioid receptor was the resident Mφ. Numerous studies have demonstrated κ-opioid-induced regulation of cytokine release (Alicea et al., 1996), phagocytosis (Ichinose et al., 1995), and tumoricidal activity (Foster and Moore, 1987) by Mφ, corroborating a functional role for this opioid receptor in mediating immunocompetence. Further understanding of the immune cells expressing the various opioid receptors will help to improve our understanding of the immune and inflammatory responses regulated, both directly and indirectly, by opiates.