SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

The mechanism responsible for persistent synovial inflammation in rheumatoid arthritis (RA) is unknown. Previously, we demonstrated that expression of the cyclin-dependent kinase inhibitor p21 is reduced in synovial tissue from RA patients compared to osteoarthritis patients and that p21 is a novel suppressor of the inflammatory response in macrophages. The present study was undertaken to investigate the role and mechanism of p21-mediated suppression of experimental inflammatory arthritis.

Methods

Experimental arthritis was induced in wild-type or p21−/− (C57BL/6) mice, using the K/BxN serum–transfer model. Mice were administered p21 peptide mimetics as a prophylactic for arthritis development. Lipopolysaccharide-induced cytokine and signal transduction pathways in macrophages that were treated with p21 peptide mimetics were examined by Luminex-based assay, flow cytometry, or enzyme-linked immunosorbent assay.

Results

Enhanced and sustained development of experimental inflammatory arthritis, associated with markedly increased numbers of macrophages and severe articular destruction, was observed in p21−/− mice. Administration of a p21 peptide mimetic suppressed activation of macrophages and reduced the severity of experimental arthritis in p21-intact mice only. Mechanistically, treatment with the p21 peptide mimetic led to activation of the serine/threonine kinase Akt and subsequent reduction of the activated isoform of p38 MAPK in macrophages.

Conclusion

These are the first reported data to reveal that p21 has a key role in limiting the activation response of macrophages in an inflammatory disease such as RA. Thus, targeting p21 in macrophages may be crucial for suppressing the development and persistence of RA.

Macrophages play a central role in the pathogenesis of rheumatoid arthritis (RA). Conventional therapies, including methotrexate and cytokine inhibitors, block proinflammatory cytokines produced primarily by macrophages (1). Importantly, synovial macrophage infiltration correlates with subsequent radiographic joint destruction (2). In addition, reduction of RA synovial sublining macrophage numbers by various therapeutic strategies correlates with clinical improvement, making it a sensitive biomarker for disease activity (3, 4). Intensive research is ongoing to elucidate the mechanisms responsible for the increased synovial macrophage numbers in RA. Thus far, increased chemotaxis (5), reduced emigration (6), and decreased apoptosis of macrophages (7) have all been implicated in RA pathogenesis. In addition to their increased numbers, the response of RA synovial macrophages to Toll-like receptor (TLR) stimulation is amplified compared to macrophages from patients with other inflammatory joint diseases or normal circulating monocytes differentiated in vitro (8). As is the case in human patients, monocytes and macrophages are necessary for pathology to occur in various experimental models of RA, such as collagen-induced arthritis and K/BxN serum–transfer arthritis (9–11). Despite the abundant data supporting the crucial role of macrophages in RA, little is known about the factors that control their state of activation.

Cyclin-dependent kinase inhibitors are of central importance in suppressing cell cycle activity. To this end, the vast majority of studies on cell cycle machinery in RA have focused on the suppression of synovial fibroblast proliferation or production of inflammatory cytokines by cell cycle inhibitors such as retinoblastoma protein, or cyclin-dependent kinase inhibitors p16, p18, and p21 (12–17). Additionally, administration of a replication-defective adenovirus expressing p16 or p21 via multiple intraarticular injections leads to suppression of inflammatory arthritis in rodents (12–14, 16). While those studies focused on the role played by p16 or p21 in synovial fibroblasts, two recent studies have shown that the presence of p21 reduces serum cytokine levels and confers a protective effect on survival during lipopolysaccharide (LPS)–mediated endotoxic shock in mice (18, 19), a macrophage-dependent model. Expression of the activation markers CD40 and class II major histocompatibility complex and secretion of the proinflammatory cytokines interleukin-6 (IL-6), tumor necrosis factor α, and IL-1β are enhanced in p21-deficient macrophages following TLR ligation, even though these cells have terminally withdrawn from the cell cycle (18, 19). These data suggest that p21 may function as a novel suppressor of inflammation in macrophages.

In this animal study we demonstrated that p21 is important not only for limiting the development of inflammatory arthritis, but also for induction of the resolution or wound healing phase that occurs after the arthritic stimulus is withdrawn. Similar to findings in RA patients, the severity of inflammation and destruction of bone correlates with the number of macrophages in the pannus. Further, a p21 peptide mimetic corresponding to the C-terminal domain of p21 is sufficient to reduce the severity of inflammatory arthritis and lower the number of macrophages in the pannus. Isolated macrophages treated with the p21 peptide mimetic displayed enhanced expression of phosphorylated Akt and reduced p38 activation. Taken together, these data are among the first to identify novel and essential roles of p21 in suppression of inflammation and to translate them into clinically significant information that may shed light on the pathogenesis of RA.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Mice.

Male KRN mice were kindly provided by Dr. Diane Mathis (Harvard Medical School, Boston, MA) and were crossed with female NOD mice purchased from Taconic; p21−/− mice were backcrossed onto the C57BL/6 background for at least 12 generations and tested for microsatellite markers of background contribution (20). All experiments on mice were approved by the Animal Care and Use Committee at Saint Louis University and/or Northwestern University.

K/BxN serum–transfer arthritis.

Serum was harvested via cardiac puncture from 8-week-old male and female progeny of KRN and NOD mice (K/BxN) and injected intraperitoneally (IP) into 6–8–week-old mice. Ankle circumference was calculated from measurements obtained with a caliper. Clinical scores were measured as follows: 0 = no swelling, 1 = mild swelling in 1 or 2 limbs, 2 = mild swelling in >2 limbs, 3 = moderate swelling in >2 limbs, 4 = severe swelling in >2 limbs, and 5 = compromised mobility. At 7, 14, or 25 days postinjection, mice were killed, serum was collected via cardiac puncture, and ankles were harvested and fixed in 10% formalin. Ankles were then subjected to microfocal computed tomography (micro-CT) analysis (performed at the core facility at Washington University School of Medicine) and/or prepared for immunohistochemistry analysis. For peptide studies, wild-type (WT) mice (The Jackson Laboratory) were injected IP with peptide (10 mg/kg) 30 minutes prior to K/BxN serum injection and daily throughout the experiment.

Immunohistochemistry.

Fixed ankles were decalcified in EDTA (Sigma-Aldrich) in 10% formalin, embedded in paraffin, and sectioned. Sections were stained with hematoxylin and eosin and with Safranin O–methyl green, and for tartrate-resistant acid phosphatase (TRAP), CD45, proliferating cell nuclear antigen (PCNA), and F4/80 antigens. All staining procedures were performed at the Saint Louis University core pathology facility except for TRAP staining, which was performed at Washington University School of Medicine Center for Musculoskeletal Biology and Medicine. Histopathologic scoring was performed, under blinded conditions, as previously described (21), using an Olympus BX40CY microscope. Photographs were taken on an Olympus BX41 microscope equipped with a DP20 Digital Camera (Olympus), at 100× magnification.

Cell culture.

Peritoneal cells from 6–8–week-old male or female mice were harvested via lavage 3 days after IP injection of 4% aged thioglycollate, adhered for 1 hour in serum-free media, and then maintained in complete Dulbecco's modified Eagle's medium and used within 2 days. For cytokine assays, peritoneal macrophages were incubated with peptide (50 μM) for 2 hours, followed by stimulation with LPS (10 ng/ml; Sigma-Aldrich) in the presence of peptide. For IL-1β secretion assays, macrophages were further treated with ATP (5 mM; Sigma-Aldrich) for up to 30 minutes to induce IL-1β release.

Luminex-based assays.

Cytokine levels in serum or cell supernatants were determined using Luminex-based assays according to the specifications of the manufacturer (Invitrogen). Data were collected on a Luminex 200 using xPONENT software version 3.0 (Luminex) and fitted to a weighted 5-point parameter log standard curve. For the peptide study (50 μM), only the Tat-Ctrl peptide was used as a control, since Tat and Tat-Ctrl showed no difference in arthritis development. For phosphorylation assays, phosphorylated Akt (BD Biosciences), phosphorylated p38 (Thr180/Tyr182), and total IκBα protein levels in whole cell lysates (prepared with a Cell Lysis Kit) were measured using a Luminex-based assay according to the instructions of the kit manufacturer (Bio-Rad). Data were collected on a Luminex 200 with IS 2.3 software (Luminex).

Enzyme-linked immunosorbent assay (ELISA).

For detection of IL-1β in cell supernatants, sandwich ELISAs were performed according to the instructions of the manufacturer (R&D Systems). ELISA results were quantitated by absorbance at 450 nm on a microplate reader (Bio-Rad) and normalized to the number of cells per well.

Peptides.

A polycationic peptide derived from human immunodeficiency virus 1 (HIV-1) Tat (22) was fused to the p21 peptide mimetics, which were synthesized by and purchased from the Peptide Synthesis group at Tufts University (Boston, MA). The peptides were as follows: aa 15–40 (Ac-rkkrr-orn-rrr-SKACRRLFGPVDSEQLSRDCDALMAG), aa 46–65 (Ac-rkkrr-orn-rrr-RERWNFDFVTETPLEGDFAW-OH), aa 63–77 (Ac-rkkrr-orn-rrr-AWERVRGLGLPKLY), and aa 141–160 (Ac-rkkrr-orn-rrr-KRRQTSMTDFYHSKRRLIFS) (numbers indicate amino acids [aa]). A negative control peptide, Tat-Ctrl (Ac-rkkrr-orn-rrr-SKACRRLKKPVDSEQLSRDCDALMAG), was designed by incorporating F22K and G23K substitutions (23). A Tat (Ac-rkkrr-orn-rrr) peptide was also used as a control. Fluorescein isothiocyanate (FITC)–conjugated peptides were synthesized as above, with FITC replacing the acetyl group on the amino terminus.

Flow cytometry.

For evaluation of peptide entry, cells were incubated with FITC-conjugated peptides for 2 hours at 4°C or 37°C. Cells were trypsinized to remove surface-bound protein, and data were collected on a BD LSRII (BD Biosciences) using FACSDiva version 6.1.2 and analyzed using FlowJo version 5.5.5 (Tree Star).

Cell imaging.

Live cell imaging was performed at the Northwestern University Feinberg School of Medicine Cell Imaging Core Facility, on a Nikon C1Si laser scanning confocal microscope fitted on a PerfectFocus stand to actively maintain focal plane control. Cells were maintained at 37°C during imaging, using a Tokai HIT stage top incubator. Confocal imaging was performed using scan averaging of 12 and laser dwell time of 1.92 μsec/pixel.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Increased severity of K/BxN serum–induced arthritis in p21−/− mice.

In p21−/− mice backcrossed onto the C57BL/6 background and verified for >99% C57BL/6 background contribution (20), there were no obvious differences from WT mice in total leukocyte numbers or in various populations of leukocytes (data available from the corresponding author upon request). Since p21 has been associated with suppression of inflammatory disease, its role in the development of inflammatory arthritis was evaluated using the K/BxN serum–transfer model in WT and p21−/− mice. Compared to WT mice, p21−/− mice developed significantly worse ankle swelling, as measured by the change in ankle circumference following IP injection of K/BxN mouse serum (Figure 1A). Additionally, p21−/− mice displayed evidence of more severe disease as assessed by a significant elevation in clinical score, most pronounced on day 25 (3.0-fold increase), when the disease had resolved in WT mice but failed to fully resolve in p21−/− mice (Figure 1A).

thumbnail image

Figure 1. Increased and prolonged inflammatory arthritis and elevated cytokine levels in p21−/− mice. A, Change in ankle circumference and clinical score in arthritic wild-type (WT) and p21−/− mice (n = 15 per group on day 7, 10 per group on day 14, and 5 per group on day 25). Data shown are representative of at least 2 independent experiments. B, Representative day 7 ankle sections stained with hematoxylin and eosin (H&E). P = pannus; SL = synovial lining; C = cartilage; B = bone; BM = bone marrow. Original magnification × 100. C, Histologic scores of H&E-stained ankle sections. Inflam. = inflammation; Synov. = synovial; Cartil. = cartilage, Lymph. = lymphocytes; PMN = polymorphonuclear cells. Extraarticular refers to extraarticular inflammation. D, Representative day 25 ankle sections stained for tartrate-resistant acid phosphatase (TRAP). Original magnification × 100. E, Number of TRAP-positive cells per 100× field. Horizontal lines show the mean. F, Microfocal computed tomographic imaging of WT and p21−/− mouse ankles. AP = anteroposterior; ML = mediolateral. G and H, Levels of interleukin-6 (IL-6) (G) and IL-1α (H) in serum collected on day 7 from arthritic ankles, measured using a Luminex-based assay. Values in A, C, and G are the mean ± SEM. ∗ = P < 0.05 versus WT mice, by Student's t-test.

Download figure to PowerPoint

Increased inflammatory cell numbers and articular destruction in the ankle joints of p21−/− mice.

To determine the extent of joint damage in WT and p21−/− mice, ankles were harvested 7, 14, or 25 days following injection of K/BxN mouse serum and examined histologically. Inflammation and development of pannus were observed in the joints of both WT and p21−/− mice, particularly on days 7 and 14 (Figures 1B and C). At all 3 time points, pannus was more extensive in p21−/− mice than in control mice (by a mean of 2.9-, 1.5-, and 1.8-fold on days 7, 14, and 25). The pannus formation was associated with enhanced destruction of cartilage, particularly on day 14 (3-fold), and of bone at all time points (3.7-, 1.5-, and 2.3-fold) in p21−/− mice as compared to WT mice (Figure 1C). Increased numbers of TRAP-positive cells, which were localized in pannus, were found at all time points in p21−/− mice compared to control mice (Figures 1D and E). Long-term bone destruction was further verified by micro-CT analysis (Figure 1F), which showed increased bone damage in p21−/− mice. Since articular destruction persisted in p21−/− mice even at the late time point (25 days), when there are clearly no arthritogenic antibodies remaining (11), these findings suggest that the deficiency in p21 leads to sustained inflammation and/or a failure of inflammation to resolve.

Because elevated cytokine production may contribute to the increased inflammation and destruction observed in the ankles of p21−/− mice, circulating cytokine levels were assessed. Seven days after the initiation of arthritis, the serum level of the inflammatory cytokine IL-6 was increased 3.0-fold in p21−/− mice compared to WT mice (Figure 1G). The level of IL-1α was significantly elevated in p21−/− mice as well (Figure 1H).

To further investigate the increased inflammation and destruction observed on hematoxylin and eosin–stained sections, infiltration of immune cells into arthritic joints was analyzed. Ankle sections were stained for CD45 (hematopoietic cells), F4/80 (macrophages), and PCNA (proliferation). On days 7 and 14, respectively, the number of hematopoietic cells (CD45+) was increased 2.8- and 1.4-fold in the pannus of p21−/− mouse ankles as compared to control mouse ankles (Figures 2A and D). On day 25, the number of CD45+ cells did not differ between WT and p21−/− mice. In addition, peak PCNA staining occurred on day 7 in the ankles of p21−/− mice and on day 14 in WT mice (Figures 2B and D), likely representing increased proliferation of synovial fibroblasts as macrophages are terminally differentiated. The number of PCNA+ cells did not differ between WT and p21−/− mice on day 25 (Figure 2D). This suggests that synoviocytes may be important at the early stages of disease development and that infiltration of inflammatory cells at later time points is central for disease progression. As such, macrophage numbers in the pannus were significantly increased in p21−/− mice as compared to WT mice on days 7 (3.7-fold), 14 (1.6-fold), and 25 (1.8-fold) (Figures 2C and D).

thumbnail image

Figure 2. Increased inflammatory cell infiltration in p21−/− mice. Arthritis was induced in wild-type (WT) and p21−/− mice, and ankles (n = 10 per group) were prepared for histologic analysis. A–C, Representative antigen-retrieved paraffin-embedded ankle sections (day 7), stained with antibody to CD45 (A), F4/80 (B), or proliferating cell nuclear antigen (PCNA) (C). Original magnification × 100. D, Number of positive cells in each of the indicated regions. Values are the mean ± SEM of at least 3 sections per ankle and 3 fields per section. ∗ = P < 0.05 versus WT mice, by Student's t-test. Macs = macrophages; Prolif. = proliferating cells.

Download figure to PowerPoint

Since there was no difference in total CD45+ cells on day 25 yet there were more macrophages, the above data suggest that the failure of arthritis to resolve in p21−/− mice may be due to the persistence of macrophages in the synovium. Thus, similar to findings in patients with RA (4, 24), increased numbers of macrophages in the arthritic p21-deficient mice are associated with more severe articular destruction.

Protection against K/BxN serum–transfer arthritis in vivo by a peptide mimetic corresponding to aa 141–160 of p21.

To decipher the critical domain of p21 that plays a vital role in suppression of inflammation, we examined the impact of delivering established domains of p21 to arthritic mice. Previous studies have identified domains on p21 that are critical for interaction with cyclins, cyclin-dependent kinases, and PCNA, as well as modification of the activities of these peptides leading to alterations in cell cycle progression (21, 25). We took advantage of these studies and designed 6 sets of p21 peptide mimetics that were conjugated to a polycationic peptide derived from HIV-1 transactivator of transcription to allow cell entry (Tat, Tat-Ctrl, aa 15–40, aa 46–65, aa 63–77, and aa 141–160). We have previously shown that this approach is viable in this model of inflammatory arthritis using BH3 peptide mimetics (21).

The functionality of p21 peptide mimetics in suppression of K/BxN serum–transfer arthritis was evaluated. WT mice treated with aa 141–160 had a 3.6-fold reduction in ankle swelling on day 2, a 6-fold reduction on day 4, and a 4-fold reduction on day 7 as compared to mice treated with control peptide (Figure 3A). Additionally, there was marked improvement in the clinical score in WT mice treated with aa 141–160 as compared to Tat-Ctrl–treated mice (Figure 3B). While mild clinical improvement in ankle swelling and clinical score was observed with aa 15–40 and aa 63–77, these were not significant. Since aa 141–160 treatment led to reduced arthritis in p21-intact mice, we also examined its effect in mice lacking p21. There was no difference in ankle swelling or clinical score in p21−/− mice treated with Tat only or with aa 141–160 (Figures 3C and D).

thumbnail image

Figure 3. Reduced severity of arthritis in vivo after treatment with the p21 peptide mimetic aa 141–160. Arthritis was induced in 8 mice per group, and the mice were injected intraperitoneally with peptide (10 mg/kg) 30 minutes prior to injection of K/BxN serum and daily throughout the experiment. A and C, Change in ankle circumference in wild-type mice (A) and p21−/− mice (C). B and D, Clinical score in wild-type mice (B) and p21−/− mice (D). Values are the mean ± SEM. ∗ = P < 0.05, aa 141–160 versus Tat-Ctrl (control); † = P < 0.05, aa 141–160 versus Tat, by Student's t-test. Data shown are representative of at least 2 independent experiments.

Download figure to PowerPoint

Suppression of articular destruction and inflammatory cell infiltration by the p21 peptide mimetic aa 141–160.

Ankles were harvested and examined histologically 7 days following arthritis induction and treatment with peptide. Only treatment with the peptide corresponding to aa 141–160 consistently caused a significant decrease in pannus development, inflammation, synovial lining thickness, bone erosion, extraarticular inflammation, and infiltration of lymphocytes and polymorphonuclear cells as compared to treatment with control peptides (Figures 4A and B). No difference in cartilage destruction was detected in any of the groups of mice. Furthermore, decreased infiltration of all hematopoietic cells (CD45) and, in particular, macrophages (F4/80), was observed within the pannus of aa 141–160–treated mice (Figures 5A and B). No differences were noted in areas of normal synovium. PCNA staining also revealed reduced proliferation in the pannus, although this effect was elicited by several of the peptides (Figures 5A and B).

thumbnail image

Figure 4. Suppression of articular destruction by treatment with aa 141–160. Wild-type mice were treated with peptide, arthritis was induced, and ankles (16 per group) were prepared for histologic analysis. A, Representative ankle sections stained with hematoxylin and eosin (H&E). Original magnification × 100. B, Histologic scores of H&E-stained ankle sections. Scores were determined as previously described (11, 21, 45, 46). Values are the mean ± SEM of at least 3 sections per ankle and 3 fields per section. ∗ = P < 0.05 versus Tat-Ctrl (control); † = P < 0.05 versus Tat, by Student's t-test. Data shown are representative of at least 2 independent experiments.

Download figure to PowerPoint

thumbnail image

Figure 5. Reduced proliferation and inflammatory cell infiltration in ankles from aa 141–160–treated mice. Wild-type mice were treated with peptide, arthritis was induced, and ankles (16 per group) were prepared for histologic analysis. A, Representative antigen-retrieved paraffin-embedded ankle sections stained with antibody to CD45, F4/80, or proliferating cell nuclear antigen (PCNA). Original magnification × 100. B, Number of cells positive for CD45, F4/80, and PCNA. Values are the mean ± SEM of at least 3 sections per ankle and 3 fields per section. ∗ = P < 0.05 versus Tat-Ctrl (control); † = P < 0.05 versus Tat, by Student's t-test. Data shown are representative of at least 2 independent experiments.

Download figure to PowerPoint

The p21 peptide mimetic aa 141–160 suppresses peritoneal macrophage production of cytokines following stimulation with TLR agonists.

We have shown that while Tat-conjugated peptides enter all cells, macrophages appear to preferentially uptake the peptides (21). To explore the mechanism by which aa 141–160 reduces the severity of arthritis, we focused on the effect of this peptide on innate immune responses induced by TLR agonists. Entry and cytoplasmic localization of FITC-conjugated Tat peptide into peritoneal macrophages was confirmed using confocal microscopy (Figure 6A). The percent and amount of incorporation was determined by flow cytometry. One hundred percent of the cells equally incorporated the Tat-Ctrl and Tat p21 peptide mimetics (Figures 6B and C), and only a minor reduction in cell survival was observed, even at 24 and 48 hours after administration (Mavers M, et al: unpublished observations). The capacity of aa 141–160 to inhibit cytokine production by peritoneal macrophages activated with TLR ligation (LPS) was also assessed. The aa 141–160 peptide, which protected against K/BxN serum–induced arthritis in vivo (Figures 3–5), suppressed production of the inflammatory cytokines tumor necrosis factor α, IL-6, and IL-1β in vitro (Figures 6D–F). There was no inhibitory effect of the p21 peptide mimetic on production of macrophage inflammatory protein 1α (MIP-1α), MIP-1β, or RANTES (Mavers M, et al: unpublished observations).

thumbnail image

Figure 6. The p21 peptide mimetic aa 141–160 enters into macrophages and reduces production of inflammatory cytokines following Toll-like receptor stimulation in vitro. A, Peritoneal macrophages were incubated for 2 hours with no peptide, Tat-Ctrl (control), or fluorescein isothiocyanate (FITC)–conjugated aa 141–160 and examined by confocal microscopy. Original magnification × 200. B and C, Peritoneal macrophages were incubated for 2 hours at 37°C (B) or at 4°C or 37°C (C) with no peptide or with an FITC-conjugated Tat peptide corresponding to various domains of p21. Shown in B is an overlay of results obtained with all of the cells treated with the Tat-conjugated, FITC-labeled peptides at 37°C. D–F, Levels of tumor necrosis factor α (TNFα) (D), interleukin-6 (IL-6) (E), and IL-1β (F) in supernatants from lipopolysaccharide-stimulated peritoneal macrophages incubated with aa 141–160 or control peptide were analyzed by enzyme-linked immunosorbent assay. G–I, Expression of phosphorylated Akt (G), phosphorylated p38 (H), and total IκB (I) in cell lysates was analyzed by Luminex-based assay. Data were normalized to cell number. Values are the mean ± SEM fold change relative to untreated cells. ∗ = P < 0.05 versus Tat-Ctrl, by Student's t-test. Data shown are representative of at least 2 independent experiments.

Download figure to PowerPoint

The p21 peptide mimetic aa 141–160 increases active Akt but reduces p38 activity in TLR-stimulated macrophages.

Because the p21 peptide mimetic aa 141–160 induced a dramatic reduction in the production of proinflammatory cytokines and development of arthritis, we examined its effect on upstream signaling events, using a Luminex-based assay. Treatment with aa 141–160 led to a marked increase in activation of the serine/threonine protein kinase Akt as compared to Tat-Ctrl–treated macrophages beginning at 15 minutes after stimulation with LPS, and Akt levels remained higher throughout 2 hours of stimulation (Figure 6G). In contrast, a reduction in the phosphorylation of p38 was observed at 30 minutes and 60 minutes following TLR ligation (Figure 6H). The peptide had little effect on the degradation of IκB as compared to control peptide–treated cells (Figure 6I).

The effect of full-length p21 on intracellular signaling pathways was also explored in WT and p21−/− mouse peritoneal macrophages activated with LPS, using flow cytometry. Cells from p21−/− mice displayed decreased phosphorylation and activation of Akt 15 minutes after LPS stimulation, as compared to control cells; subsequently, at 30 minutes and 60 minutes following TLR ligation, an increase in phosphorylated p38 in p21−/− mouse cells as compared to WT mouse cells was observed (data available from the corresponding author upon request). Taken together, these results suggest that p21, via its C-terminal domain, suppresses macrophage function by enhancing the phosphorylation of Akt, thereby reducing p38 activation and subsequently limiting inflammatory cytokine production in TLR-stimulated macrophages.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Over the last several years, p21 has gained attention in the fields of inflammation and autoimmunity. Numerous studies have been performed using p21−/− mice to examine the role of p21 in murine models of sepsis, lupus, and RA. Recently, we and others have shown that p21−/− mouse macrophages, regardless of background, display enhanced activation in response to TLR agonists, as compared to control macrophages (18, 19). Further, p21−/− mice on either a mixed or an inbred background are more susceptible to LPS-induced endotoxic shock (18, 19). These data suggest that in sepsis, the background of the mice is not a contributing factor. However, in spontaneous development of lupus, the background of the mice may be more crucial (26, 27). The potentially conflicting results in the lupus studies may be attributed to the notion that hybrid (C57Bl/6:129) mice are more susceptible to spontaneous autoimmunity due to epistatic interactions between the two genomes (28).

In support of the argument that p21 may be considered a general inhibitor of autoimmune disease, studies by Salvador and colleagues showed that loss of p21 leads to early lethality due to lupus-like disease, which is enhanced by the concomitant loss of GADD45a (29). Further, genome-wide scanning studies have now shown that p21 is a susceptibility locus for SLE (30, 31). Taken together, these results suggest that p21 may be considered a negative regulator of spontaneous autoimmunity. However, its direct role in inhibiting macrophage function was not examined in those studies.

In the present study we have shown that p21−/− mice backcrossed onto a C57BL/6 background for at least 12 generations and screened for more than 150 loci (20) develop a markedly more severe experimental RA-like disease (Figure 1). The arthritis in p21−/− mice fails to resolve as compared to that in WT mice, with continuous articular destruction and a corresponding increase in macrophage numbers (Figures 1 and 2). While we previously obtained conflicting data with regard to the role of p21 in the development of K/BxN serum–transfer arthritis (11), these differences are attributable to the mouse background. After extensive phenotyping of the mice, we discovered that p21−/− mice on a mixed background, but not on an inbred background, develop significantly fewer inflammatory monocytes when compared to controls or even mice (op/op) lacking macrophage colony-stimulating factor (32). Further, injection of WT mouse macrophages into p21−/− mice restores their susceptibility to inflammatory arthritis (11).

Thus, we were the first to show that p21 cooperates with 129 loci to produce inflammatory monocytes and that inflammatory monocytes are crucial for K/BxN serum–induced arthritis (11). However, p21 is not required for the differentiation of mouse bone marrow–derived macrophages, splenic-derived macrophages, or thioglycollate-elicited peritoneal macrophages (17, 33, 34), regardless of the background. Further, p21−/− mice display similar numbers of tissue macrophages as compared to p21-intact mice, regardless of the background (C57Bl/6 versus C57BL/6:129). Collectively, these data are consistent with the results of immunohistochemical studies of p21 expression in RA synovium (17) and therapeutic studies using adenoviral vectors expressing p21 (12–14, 16), which demonstrate that p21 may be an important inhibitor of inflammatory arthritis.

Despite the general success of biologic therapy, many patients continue to experience the severely debilitating effects of RA. In addition, the mechanism behind the persistent production of proinflammatory cytokines targeted by these therapies remains to be fully elucidated. We have now demonstrated that a Tat-conjugated peptide mimetic corresponding to the C-terminus of p21 significantly reduces the severity of arthritis development in the K/BxN serum–transfer mouse model (Figures 3–5). Further, the aa 141–160 peptide requires the presence of p21 to reduce arthritis development. Previous studies have demonstrated the ability of Tat to carry peptides and other cargo into cells in vitro and in vivo (22, 35), and in the current study, entry into cells was confirmed by cell imaging, flow cytometry, and functional assays (Figure 6). WT mice were used for these peptide studies to more closely replicate the condition encountered in humans. Goulvestre et al have shown that a peptide mimetic corresponding to aa 141–160 was able to reduce the severity of lupus-like disease in mice (36). This effect was attributed to suppression of lymphocyte proliferation via inhibition of PCNA by the peptide, as this region at the C-terminus of p21 encompasses the PCNA binding domain (25).

While we found that the aa 141–160 peptide mimetic reduces proliferation in ankles from WT mice 7 days after injection with K/BxN mouse serum (Figure 5), the other p21 peptide mimetics also reduced proliferation of synovial fibroblasts in vivo, but had no effect on clinical outcome. In vivo, macrophages are terminally differentiated and also appear to preferentially take up Tat-conjugated peptides over other immune cells (21). Therefore, the cells most likely affected by aa 141–160–mediated suppression of proliferation are synovial fibroblasts. Our study also demonstrated a significant decrease in hematopoietic cell infiltration, particularly by macrophages, in the ankles of aa 141–160–treated mice (Figure 5). Furthermore, we showed that aa 141–160 suppresses inflammatory cytokine production in macrophages (Figure 6). These data are consistent with the results of previous studies showing p21-mediated reduction in macrophage activation following TLR ligation (18, 19). Goulvestre and colleagues further attributed the decreased lupus development in aa 141–160–treated mice to significant proapoptotic effects on lymphocytes (36). However, we observed no difference in apoptosis between WT and p21−/− mouse macrophages, or following treatment with any of the peptides (Mavers M, et al: unpublished observations).

In pursuing the mechanism by which the aa 141–160 peptide mimetic mediates a reduction in arthritis severity and suppression of cytokine production, we found that it enhances Akt phosphorylation and inhibits p38 activation (Figure 6). Similarly, p21−/− mouse cells display reduced Akt phosphorylation and increased p38 activation as compared to WT mouse cells (data available from the corresponding author upon request). These data are consistent with our previous work demonstrating increased levels of IL-6 and IL-1β messenger RNA in p21−/− mouse macrophages (18), which suggests that the mechanism by which p21 inhibits inflammatory cytokine production is likely to be suppression of intracellular signaling pathways, leading to a reduction in transcription of these cytokines.

This effect of p21 on activation of Akt and suppression of p38, and subsequent reduction in secretion of proinflammatory cytokines, is not surprising given previous studies showing that sustained Akt activation leads to decreased production of these cytokines (37, 38). Furthermore, enhanced signaling through the MAPK pathway leads to increased production of inflammatory cytokines and has been shown to play a role in RA pathogenesis (39). In fact, the targeting of intracellular signaling pathways is a prominent focus of studies exploring novel mechanisms for the treatment of arthritis (40). Our data suggesting that the peptide mimetic corresponding to the C-terminus of p21 is sufficient to enhance Akt activation and suppress p38 activation (Figure 6) are supported by previous studies demonstrating that p21 interacts with Akt at the C-terminus (41). Interestingly, Akt has been shown to induce retention of p21 in the cytoplasm and enhance its stability by phosphorylating p21 on its C-terminus (42). Thus, one could envision a feedback loop in which p21-mediated Akt activation further activates p21 cytoplasmic functions, thereby perpetuating p38 suppression in order to turn off inflammatory reactions.

Additional work is needed in order to determine whether p21 also regulates other intracellular signaling pathways. Previous studies have shown that NF-κB DNA binding is increased in p21-deficient mouse macrophages, along with increased IκB degradation and IκB kinase complex activity (19). Other studies have shown that p21 may in fact promote transcriptional activation by NF-κB via activation of p300 through derepression of a repression motif in a promoter-dependent manner; however, those studies were not conducted in macrophages (43). We found that the aa 141–160 peptide mimetic has little to no effect on IκBα degradation (Figure 6I). In addition, direct inhibition of JNK by p21 has been previously demonstrated in cell-free systems, 293 cells, and synovial fibroblasts (13, 17, 44).

Further research is necessary to determine the exact role of p21 in regulating intracellular signaling, particularly in the context of macrophage activation. Elucidating the mechanisms by which macrophages are activated or inhibited remains crucial to promoting the development of new treatments for inflammatory disease.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Perlman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Mavers, Cuda, Misharin, Agrawal, Balomenos, Perlman.

Acquisition of data. Mavers, Cuda, Misharin, Gierut, Agrawal, Weber, Novack, Haines, Balomenos, Perlman.

Analysis and interpretation of data. Mavers, Cuda, Misharin, Gierut, Agrawal, Novack, Haines, Balomenos, Perlman.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

The authors are grateful for the assistance of staff of the Washington University School of Medicine Center for Musculoskeletal Biology and Medicine and the Northwestern University Feinberg School of Medicine Cell Imaging Core Facility and the Methodology and Data Management Core.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES