|Home | About | Journals | Submit | Contact Us | Français|
The T cell-mediated immune response elicited by Pneumocystis plays a key role in pulmonary damage and dysfunction during Pneumocystis pneumonia (PcP). Mice depleted of CD4+ and CD8+ T cells prior to infection are markedly protected from PcP-related respiratory deficit and death despite progressive lung infection. However, the therapeutic effectiveness of antibody-mediated disruption of T cell function in mice already displaying clinical symptoms of disease has not been determined. Therefore, a murine model of PcP-related immune reconstitution inflammatory syndrome was used to assess whether antibody to the pan-T cell molecule CD3 is effective for reducing the severity of PcP when administered after the onset of disease. Mice that received anti-CD3 antibody exhibited a rapid and dramatic halt in the PcP-associated pulmonary function decline within one week post-treatment, and a striking enhancement of survival rate compared to mice receiving control antibody. Physiological improvement in anti-CD3 treated mice was associated with a significant reduction in the number of CD4+ and CD8+ T cells recovered in lung lavage fluid. This effectiveness of anti-CD3 was noted whether or not the mice also received antibiotic therapy with trimethoprim-sulfamethoxazole. These data suggest that monoclonal antibody-mediated disruption of T cell function may represent a specific and effective adjunctive therapy to rapidly reverse the ongoing pathological immune response occurring during active PcP. Thus, the anti-human CD3 monoclonal antibody OKT3, which is already in clinical use, has the potential to be developed as an adjunctive therapy for PcP.
Pneumocystis carinii pneumonia (PcP) remains a life-threatening disease process prevalent among immunosuppressed populations. Despite many improvements in our ability to care for critically ill patients, the mortality attributable to PcP has changed little and remains unacceptably high (1). Therefore new treatment modalities that specifically address the pathophysiology of PcP are needed. Mortality is particularly high, up to 80–90%, among those patients who require admission to an Intensive Care Unit. Recent analyses of two large cohorts of AIDS patients with PcP demonstrated that the need for intensive care or mortality is discernable at or soon after admission (2,3). Therefore, there is an identifiable group of patients with PcP who would benefit from an improved treatment regimen for PcP, were one available.
It is now becoming increasingly recognized that the host inflammatory response elicited by a microorganism can also produce tissue damage. Studies published by ourselves and others over the past several years have demonstrated that this is a prominent feature of PcP (4–10). Clinical observations in humans also support the link between inflammation and poor outcome in patients with PcP (11,12) and it has been postulated that the benefit of adjunctive corticosteroid therapy is related to anti-inflammatory effects (12,13).
The CD4+ T lymphocyte is the critical cell type required for both normal resistance to Pc infection, as well as for immune-mediated clearance of an existing infection (14,15). A consistent feature of animal model experiments is the finding that in the absence of sufficient CD4+ T cells to protect against PcP, the CD8+ T cell is a key cell in driving the injurious Pc-specific immune response (4,5,7,8,10,16). CD4+ T cells can also initiate inflammatory injury in response to Pc in the context of Immune Restitution Inflammatory Syndrome (IRIS), a clinical condition in which a period of immunosupression and infection is followed by immune recovery and a rapid onset of pulmonary inflammation and respiratory distress (4,17). Although the immunopathological role of T cells during PcP is a well-documented concept, effective, specific, and feasible therapeutic regimens to block PcP-related inflammatory processes in clinically relevant models have not been developed. Immunomodulatory therapies should provide a mechanism to improve the outcome of PcP when combined with effective antibiotics to eradicate the infection. Our working hypothesis is that the host’s T cell-mediated immune response to Pc infection is a major contributor to the morbidity and mortality of PcP, and that effective control of this response, combined with antibiotic treatment, will improve the outcome of patients presenting with active PcP. In mouse models we have provided support for this hypothesis by using antibody to specifically deplete T cells prior to infection (4,5,7,8,10). However, the effectiveness of antibody-mediated T cell depletion after the onset of PcP has not been determined.
Although anti-CD8 and anti-CD4 antibodies for use in humans are not available, the pan T cell antibody OKT3 (muromonab-CD3; Muromomab, Ortho Biotech Inc.) is currently used in the clinic. Importantly, OKT3 exerts its effects on both CD4+ and CD8+ T cells. Since some PcP patients may have residual CD4+ T cell function, administration of an OKT3-like antibody would have the combined benefit of interfering with the function of both CD8+ and CD4+ T cells that may be contributing to the pathological inflammatory response. This would make such an approach especially useful for the treatment of PcP in the setting of IRIS in which CD4+ T cells are known to play a role in immunopathogenesis. A final benefit of using an OKT3 or OKT3-like antibody is that there is extensive clinical experience using these molecules which would facilitate using them as adjunctive therapy for PcP should such an approach be validated. We therefore undertook a series of experiments to provide experimental animal data for the effect of anti-CD3 antibody on the outcome of PcP. We used an immune reconstitution model that mimics PcP-induced IRIS. Our findings suggest that this approach may have clinical utility in the management of patients with moderate to severe PcP.
Hybridoma cell line 145 2C11 that produces hamster anti-mouse CD3 antibody was obtained from ATCC (18). Antibody was purified by SAS precipitation and protein A purification of ascites fluid obtained from SCID mice. F(ab′)2 fragments were prepared from intact antibody by a previously published protocol (19) with following modifications: the antibody was digested with pepsin at 37 °C for four hours. The Fc fragments were removed by passing the antibody over a protein A column that allowed the F(ab′)2 to pass through. Whole hamster IgG was used to produce control F(ab′)2 using the same method. F(ab′)2 fragments were used for these experiments to avoid the potentially confounding effects of the generalized cytokine release sometimes seen when CD3 is cross-linked by intact immunoglobulin (19).
CB.17 scid/scid mice with heavy Pc infections were euthanized and their lungs removed aseptically. Pc organisms were isolated from the lung tissue as previously described (10,20). The final preparation was stained with ammoniacal silver to enumerate cysts, and Diff-Quick (Dade AG, Dudingen, Switzerland) to screen for bacterial contamination. In addition, the preps were routinely plated on commercially available chocolate blood agar plates to test for the presence of contaminating microorganisms.
Mice were infected by intranasal inoculation with 1 × 105 Pc cysts. Four weeks later, Pc-infected SCID mice were immune reconstituted with 5 × 107 splenocytes from syngeneic donor strains to mimic IRIS as previously described (4,8). Anti-CD3 F(ab′)2 treatment was begun when the mice were clearly symptomatic, which we defined as either the average body weight loss of more than 10% or average respiratory rate was more than 400 respirations per minute. Thus the experimental mice were monitored with body weight and respiratory rate determination and F(ab′)2 treatment was begun approximately a week after immune reconstitution [between day seven to day nine]. Experimental mice were given either 150 μg of 145 2C11 F(ab′)2 or control F(ab′)2 intraperitoneally every other day. In one experiment, a group of mice received anti-CD3 F(ab′)2 only on treatment days one and three in addition to the Trimethoprim-sulfamethoxazole (TMP-SMX) treatment. TMP-SMX (Sicor Pharmaceuticals, Irvine, CA) treatment was initiated one day after anti-CD3 treatment as a single daily dose until the end of the experiment. TMP-SMX was diluted in saline and given at 20 mg of TMP/Kg body weight as described (21).
Dynamic lung compliance and resistance was measured in live mice using a previously described method with modifications (8,10,22). Mice were anesthetized by intra-peritoneal injection of 0.13 mg of sodium pentobarbital per gram body weight. A tracheostomy was performed and a 20-gauge cannula was inserted 3 mm into an anterior nick in the exposed trachea. The thorax was then opened to equalize airway and transpulmonary pressure. To assure that the mice tolerated the procedure, they were examined for spontaneous respirations before proceeding further. Mice were immediately placed into a plethysmograph designed for anesthetized mice (Buxco Electronics Inc., Wilmington, NC), and connected to a Harvard rodent ventilator (Harvard Apparatus, Southnatick, MA). Mice were ventilated with a tidal volume of 0.01 ml per gram body weight at a rate of 150 breaths per minute. Respiratory flow and pressure were measured using transducers attached to the plethysmograph chamber. Data was collected and analyzed using the Biosystems XA software package (Buxco Electronics Inc., Wilmington, NC). Dynamic lung compliance was calculated in ml.cm−1 H2O from the flow and pressure signals using the method by Amdur and Mead (23), and then normalized for peak body weight. Lung resistance values were calculated in cm H2O.ml−1.sec−1 from the same input signals.
BAL and lung tissue samples were obtained following dynamic compliance measurements. The chest cavity was surgically opened to expose the lungs and trachea, and the left lung lobe was tied off securely at the bronchus with surgical silk and removed with sterile scissors. The remaining lung lobes were gently lavaged with four, one-mL aliquots of 1X Hank’s balanced salt solution (HBSS) via the tracheal cannula. Recovered lavage fluid (~3.5 ml per mouse) was centrifuged at 250 × g for 5 min to obtain the cellular fraction. The cells were resuspended in fresh HBSS, enumerated, centrifuged onto glass slides, and stained with Diff-Quick for differential counting. In addition, multiparameter flow cytometric analysis was performed on BAL cells following staining with fluorochrome-conjugated antibodies. Anti-CD4-Fluorescein (clone RM4-4), and anti-CD8a-Peridinin Chlorophyll-a Protein (clone 53–6.7) were purchased from BD Biosciences (San Diego, CA). At least 5,000 events per BAL sample were routinely analyzed on a FACSCalibur™ cell sorter (BD Biosciences, San Jose, CA).
Pc burden in the lungs of experimental mice was determined by real-time PCR as previously described (10). Briefly, the right lung lobes were homogenized with phosphate buffered saline (one mL of PBS per 150 mg of lung tissue) in a mechanical homogenizer. Homogenates were freeze-thawed three boiled for 15 minutes, vigorously vortexed for 2–3 minutes, and then centrifuged for 5 minutes at 12,000 × g. The supernatant was carefully removed and stored at −80°C for real-time PCR analysis. Boiled samples were assayed by quantitative PCR using TaqMan® primer/fluorogenic probe chemistry, and an Applied Biosystems Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). A primer/probe set specific for a 96 nucleotide region of the mouse-derived P. carinii kexin gene (24) was designed using the Primer Express software (Applied Biosystems). The sequences of the primers and probe used were as follows: forward primer, 5′-GCACGCATTTATACTACGGATGTT-3′; reverse primer, 5′-GAGCTATAACGCCTGCTGCAA-3′; fluorogenic probe, 5′-CAGCACTGTACATTCTGGATCTTCTGCTTCC-3′. Quantitation was determined by extrapolation against standard curves constructed from serial dilutions of known copy numbers of plasmid DNA containing the target kexin sequence. Data was analyzed using the ABI Prism 7000 SDS v1.0 software (Applied Biosystems), and is reported as total kexin DNA copies per right lung.
All values reported for each experimental group are mean ± 1 standard error of measurement. Analysis was performed using Sigma-Stat version 3.5 (Systat Software Inc., Point Richmond, CA). Different groups were compared by performing a one-way analysis of variance (ANOVA). In cases where the ANOVA was significant, individual groups were tested using pairwise t-test. Survival analysis was done using Kaplan-Maier Log Rank Test. Also, individual groups were compared in pairs using Fisher’s Exact Chi Square Test. A test was considered significant if the p-value was less than 0.05.
These experiments were designed to test the hypothesis that anti-CD3 F(ab′)2 fragments could be used therapeutically to improve the morbidity and mortality associated with PcP. We chose to use an immune reconstitution mouse model of PcP because it allows for a vigorous inflammatory response to P. carinii which resembles the IRIS seen in patients with PcP. Pc-infected SCID mice were immune reconstituted with congenic splenocytes to induce IRIS. After nine days, when the mice were exhibiting obvious signs of PcP, treatment with anti-CD3 or control F(ab′)2 fragments. Our results demonstrated a clear cut advantage, both in terms of mortality and improved compliance among survivors, in those mice receiving anti-CD3 F(ab′)r fragments when compared to mice receiving control F(ab′)2 fragments. Overall survival was increased from 50% to 92% in mice receiving anti-CD3 (p = 0.003, Table I; group I vs. III). Three replicate experiments were performed and there was improved survival in all three trials. Only experiment 2 was large enough to demonstrate a statistically significant (p = 0.017) improved survival when individually analyzed. For these experiments no other intervention was performed other than use of antibody.
Because patients with PcP would receive TMP-SMX in addition to any adjunctive therapies, we included mice in each experiment that received TMP-SMX in addition to anti-CD3 or control F(ab′)2. As was observed in mice not receiving antibiotics, mice treated with TMP-SMX had an improved survival if they received anti-CD3 F(ab′)2 (96% vs 38% p < 0.001 as compared to the group that received control F(ab′)2; Table I; group IV vs. II). Again three replicate experiments were done and the improvement in survival was seen in all three trials.
To delineate the pattern of mortality over the approximately two week course of the experiment, the results from Table I were plotted as a Kaplan-Meier graph. As seen in Figure 1, mortality did not differ between anti-CD3 F(ab′)2 treated and control mice over the first several days of treatment. However, by the end of the first week after start of the treatment, the increased mortality in the control mice became apparent.
To determine the direct effect of the anti-CD3 therapeutic regimen on the progression of PcP, body weight loss was tracked non-invasively on all mice, and random mice were selected for physiologic measurement of dynamic lung compliance and resistance. By two days after the initiation of treatment, the PcP-related weight loss of the anti-CD3 F(ab′)2 treated groups had already begun to separate from the control F(ab′)2 treated groups (Figure 2). By four days after treatment the control F(ab′)2 treated groups with and without TMP-SMX had both lost nearly 10% more weight than the anti-CD3 treated groups. This relative difference was maintained out to one week after the onset of treatment, indicating that anti-CD3 treatment was able to rapidly halt the progression of PcP-related IRIS whether or not TMP-SMX was administered. Direct physiological assessment of pulmonary function demonstrated that after one week of treatment, pulmonary dynamic compliance was significantly improved in mice treated with anti-CD3 F(ab′)2 plus TMP-SMX (p = 0.004 as compared to mice treated with control F(ab′)2 plus TMP-SMX) (Figure 3A). Likewise, lung resistance was also significantly improved at this time in the anti-CD3 F(ab′)2 treated group (p = 0.002 as compared to mice treated with control F(ab′)2) (Figure 3B). In the case of antibody only treated groups [in the absence of any TMP-SMX treatment], compliance of Anti-CD3 F(ab′)2 treated mice was also improved compared to control F(ab′)2 (p=0.055). However, it is important to note that the differences between treated and control mice are likely underestimated due to the fact that so many of the control mice died an obvious respiratory death, thus in effect “removing” those mice with poorest lung function from the analysis. Overall, these results demonstrate that administration of anti-CD3 to mice already presenting with PcP produces a dramatic improvement in survival and preserves lung function.
Improved pulmonary function in mice receiving anti-CD3 F(ab′)2 was associated with decreased numbers of inflammatory cells into the lung (Table II). As would be expected, anti-CD3 F(ab′)2 produced a marked drop in both CD8+ and CD4+ T lymphocytes. In addition, both macrophages and neutrophils were also reduced, indicative of an overall down regulation of the pathological inflammatory response.
An important feature of these experiments was the finding that TMP-SMX did not improve survival in the absence of anti-CD3. The SCID mice used in these experiments were allowed to develop significant Pc infections before they were immune reconstituted, and consequently they developed severe PcP-related IRIS. Thus while the TMP-SMX did reduce Pc numbers compared to mice not treated with antibiotic, the exuberant inflammatory response produced a pulmonary insult that was not optimally responsive to antibiotics alone. Similarly, it is likely that anti-CD3 treatment alone would not be sufficient. Although mice treated with anti-CD3 in either the absence or presence of TMP-SMX were healthier than mice treated with control F(ab′)2 or control F(ab′)2 with TMP-SMX, the anti-CD3 treatment was found to interfere with Pc clearance. Mice treated for two weeks with anti-CD3 F(ab′)2 alone had a P. carinii burden of 1.5 ± 0.6 × 107 kexin copies compared to 3.2 ± 1.2 × 106 kexin copies in mice receiving both TMP-SMX and anti-CD3 F(ab′)2 (p= 0.02). This observation suggests that while anti-CD3 F(ab′)2 reduces inflammatory injury and improves lung function and survival of mice with PcP, it also prevents host-mediated clearance of Pc from the lungs. Thus, the combined effects of anti-CD3 and TMP-SMX appeared optimal for outcome of PcP.
To test whether a shorter treatment with anti-CD3 F(ab′)2 was also effective in improving morbidity and mortality of mice with PcP-related IRIS, we compared two doses of antibody given on days one and three to the every other day dosing used in our other experiments. TMP-SMX was continued until the end of the experiment in all groups. The results are summarized in Figure 4. At the conclusion of the experiment, all the mice in both the Anti-CD3 F(ab′)2 treatment groups survived, while only three and two out six mice in control F(ab′)2 treatment groups survived [either with or without TMP-SMX treatment respectively]. The mice that received of Anti-CD3 F(ab′)2 every other day showed weight gain until the end of the experiment. The mice that received only two doses of Anti-CD3 F(ab′)2 gained weight until day 12 and then slowly began losing weight probably because of restoration of lymphocyte mediated inflammation as the Anti-CD3 F(ab′)2 was halted. A number of mice that received control F(ab′)2 either with or without TMP-SMX died during the course of the experiment as denoted by black arrows in figure 4. The surviving mice in control F(ab′)2+TMP-SMX group had improved average body weight at the end of the experiment but that was mainly due to a single mouse out of six that had substantial improvement in body weight. The compliance of mice with only two Anti-CD3 F(ab′)2 doses was reduced as compared to the mice that received Anti-CD3 F(ab′)2 every other day, but this reduction was not statistically significant [0.81 and 0.99, respectively, p = 0.231]. This experiment suggests that the effect of limited Anti-CD3 F(ab′)2 treatment is temporary but still potentially useful in improving the morbidity and mortality in this model.
PcP is an infectious disease with significant morbidity and mortality attributable to the host inflammatory response as demonstrated in both patients (11,12) and animal models (4,5,8). T cells have been shown to be necessary to initiate the inflammatory immune response to Pc which in turn produces so called bystander injury to the lung. In most clinical scenarios patients lack CD4+ T cells, thereby leaving a predominantly CD8+ T cell-driven response which is ineffective in killing P. carinii but which produces critical lung injury (4,5,8). In other clinical situations, such as IRIS, CD4+ T cells also participate in the inflammatory response. In contrast to CD8+ T cells, the CD4+ T cell-driven immune response contributes to the eradication of Pc, although these responses may result in even greater lung damage (4).
We have shown, previously, that the inflammatory injury associated with PcP-related IRIS can be significantly prevented by depleting CD4+ and CD8+ T cells prior to immune reconstitution (4,8). While this is easily accomplished in mouse models, therapeutic agents to accomplish this in humans are not presently available. A pan T-cell antibody, OKT3 (Muromonab), is available for clinical use as an immunosuppressive agent especially in the control of transplant rejection. OKT3 and similar antibodies work by causing partial depletion of T cells and more importantly by disrupting the interaction of CD3 with the T cell receptor necessary for T cell antigen recognition and activation (25). Because both CD4+ and CD8+ T cells express CD3 we reasoned that using an antibody directed against CD3 would result in T cell inactivation similar to that observed with specific monoclonal antibodies in mice.
SCID mice with PcP that were immune reconstituted and treated with anti-CD3 F(ab′)2 fragments showed a marked physiological benefit. Not only were these improvements statistically significant, they were also clinically relevant. Treated mice demonstrated reduced pulmonary inflammation, improved dynamic compliance, improved lung resistance and improved survival. Increased neutrophil numbers in pulmonary lavage fluid have been correlated with increased severity of PCP both in patients as well as in mouse models (12,26). Using anti-CD3 F(ab′)2 in our model resulted in a significant reduction in neutrophil numbers in lavage fluid that was associated with less severe disease. Interestingly, macrophage numbers were also decreased although the significance of this observation with regards to the injury resulting from the Pc-driven inflammatory response is unknown.
The experimental approach described in this report differs from that described in our earlier publications in one important respect. In those prior experiments CD4+ or CD8+ T cell depletion was started before any signs of inflammation or pulmonary compromise were noted. That is, mice were treated at a point in time when they had PcP but were asymptomatic. While this approach has proved useful to identify mechanisms of immunopathogenesis in PcP, it may not be an ideal approach for modeling clinically-relevant therapeutic interventions. In the present study the disease process was allowed to progress to the point where the mice experienced a 10% acute weight loss and were tachypnic. That the mice had advanced PcP was evidenced by the fact that TMP-SMX did not consistently improve survival. We did not measure arterial oxygen tension in the current study but we have previously shown that mice in this condition would be hypoxic as well (8). Thus, we feel the experimental approach used for these studies is a reasonable approximation of a patient presenting for treatment with moderate PcP. The fact that anti-CD3 F(ab′)2 administration could be delayed until after symptoms of PcP were obvious and still improve the course of disease, makes these studies highly clinically relevant.
The approach outlined in these experiments of further immunosupressing an immunosuppressed patient could be construed as counterproductive. However, it is important to keep in mind that suppressing the pulmonary inflammatory response would be done in conjunction with the administration of antibiotics. Importantly, TMP-SMX reduced the Pc burden when given with anti-CD3 F(ab′)2 fragments. For these experiments, anti-CD3 F(ab′)2 was given every other day until the termination of the experiments. This schedule was chosen because OKT3 is typically administered daily for 5–10 days when used to treat transplant rejection. However, the biologic effects of OKT3 are demonstrable within minutes of infusion (25). This rapid onset of action may be why anti-CD3 treatment was effective in already symptomatic mice. Furthermore, as the experiment described in Figure 4 demonstrated, it may be possible to use fewer doses of anti-CD3 to suppress inflammation while TMP-SMX takes effect. Short term use of OKT3 would reduce its side effects and shorten the period of drug induced immunosuppression.
Although we chose to use an OKT3-like antibody our results suggest that other therapies which disrupt T cell function could be effective adjunctive therapy for PcP. For example, monoclonal antibody to the IL-2α receptor (CD25; basiliximab, Novartis Pharmaceuticals, Basel, Switzerland) may also be effective in this setting. In addition to global suppression of T cell function, it may be possible to identify and inhibit specific molecular pathways that lead to lung injury during PcP.
In summary, the pathogenesis of PcP includes an immune mediated inflammatory response which causes compromised lung function. T cells, especially CD8+ T cells, are a critical component of this inflammatory response. We have shown that inactivation of T cells with antibody directed against the CD3 molecule results in a marked physiologic benefit to mice with symptomatic PcP. Over the short term, this improvement is seen even in mice that are not treated with antibiotics to kill P. carinii, supporting the conclusion that lung injury during PcP is an immune mediated phenomenon. The clinical utility of using anti-CD3 as adjunctive therapy in patients with PcP will require clinical trials. Our results lend support to such an approach.
The authors wish to thank Margaret Chovaniec, Nabilah Khan and Jane Malone for technical support. The authors wish to thank Dr. Jing Wang for the critical reading of the manuscript.
This work was supported by NIH grant 1R01HL092797.
This work was presented in part as a poster at the Annual Meeting of American Thoracic Society [San Francisco, CA] in May 2007.
The authors have no conflict of interest to report.