Immunotherapy of a irborne tuberculosis in mice via the lung-specific delivery of cytokines

The immunotherapeutic potential of interleukin-2 (IL-2), tumour necrosis factor alpha (TNFα) and interferon gamma (IFN-γ) administered by aerosol was examined on mice infected with Mycobacterium tuberculosis by the aerogenic route. Infection of balb/c mice with 10(4) colony forming units (cfu) of M tuberculosis led to death of all mice at day 35 post infection after progressive microbial growth in the lungs. Aerosolization of IL-2 (100 μg per mouse) did not promote an increase in resistance to tuberculosis, as seen by growth of M tuberculosis in the lungs. Administration of IFN-γ or TNFα (100 μg) by the aerosol route led to a significant reduction in microbial growth in the lungs and a 100% survival of infected mice at day 60. Similarly, aerosolization of TNFα and IFN-γ combined led to a very high degree of tuberculostatic activity in the lungs of infected animals, but not superior to that seen with either cytokine alone. Administration of similar amounts of cytokines by repeated intraperitoneal infusions led to a very marginal improvement in mouse resistance. These results suggest that localized cytokine administration may be beneficial in the treatment of lung diseases.

I NFECTIONS WITH MYCOBACTERIA STILL POSE A FORMIDABLE health problem both in developed and developing countries (1).The recent acquired irmnune deficiency syndrome (AIDS) pandemic has exacerbated the problem, as irmnunosuppressed individuals become highly susceptible to infection with mycobacterial pathogens (2).Although chemotherapy with conventional antituberculous drugs leads to a beneficial effect in most cases (3) , drug-resistant strains may present serious problems for clinical treatments.Recent attention has been focused on the use of biological response modifiers (BRMs) for the treatment of infectious and malignant diseases (4) .Immunotherapy of mycobacterial diseases thus is still an attractive goal.Infusion ofBRMs often is considered in the treatment of infected individuals.However, it is apparent that systemic infusion of cytokines in vivo may lead to considerable toxicity (5).Moreover.distribution of cytokines administered systematically may be problematic to the activation of processes in infectious foci in organs such as the lungs.In that regard , irmnunotherapeutic measures aimed at treating mycobacterial diseases in the lungs are highly desirable , as mycobacteria thrive in the aerobic environment of the lungs (6).A strategy is described which allows the delivery of BRMs specifically to the lungs of tuberculous mice to modulate the infection positively.

MATERIALS AND METHODS
Pathogen-free BALB/c mice weighing between 18 and 25 g were bred in the authors' facilities.Mice were housed in plastic cages and were fed sterile Purina Chow and acidified water ad libitum.Mycobacterium tuberculosis H37Rv was grown in 7H9 broth (Difco Laboratories, Michigan) (7) .Dispersed cultures were obtained by the addition of 0.05% tween 80. Ampoules containing 1 mL of suspension were stored at -70°C.To produce lung infections, mice were exposed to aerosols of viable mycobacteria using a middlebrook airborne apparatus (Tri-R Instruments, New York).The nebulizer was filled with 10 mLof M tuberculosis H37Rv in phosphate buffered saline at 5x 10 7 colony forming units (cfu)/mL which leads to about 10 4 cfu deposited in the lungs of mice in a 30 min exposure.Mice were then exposed to cytokines by the aerosol route irmnediately after infection.
For this, mice were placed in a nose-only aerosol chamber (Intox Products, New Mexico).Cytokines were dissolved in 20 mL of buffer and aerosolized for 20 mins.Aerosols were generated by an Acorn 2 nebulizer (Marquest Products.Colorado) driven by compressed air at an airflow rate of 15 L/min; the nebulizer delivers a mean aerosol particle diameter of approximately 1. 0 mm under these conditions (8) .
Groups of mice were exposed to the following aerosolized agents: group 1, buffer control; group 2, 100 llg of recombinant mouse interleukin-2 (IL-2) (Cetus, California); group 3, 100 llg of recombinant mouse tumour necrosis factor alpha (TNFa) (Cetus); group 4, 100 llg of recombinant interferon gamma (IFN-y) (Genentech, California); group 5, 100 llg of TNFa followed by 100 llg ofiFN-y; and group 6, heat-inactivated cytokines (TNFa and IFN-y).(One hundred micrograms refers to the total amount placed in the nebulizer).Solutions were aerosolized to dryness.Other groups of mice were given repeated intraperitoneal injections of cytokines (5 llg per mouse daily) for 15 days irmnediately following infection.At predetermined intervals, numbers of viable bacteria in the lungs were determined by plating 10-fold serial dilutions of individual organ homogenate in saline on Middlebrook 7H 10 agar (Difco, Michigan) and counting cfu after incubation for 21 days at 37°C.For each time point, four to five animals were sacrificed.Data are expressed as the log mean number of viable organisms in the lungs.Colony counts were transformed to logw and subjected to a two-way analysis of variance.

RESULTS AND DISCUSSION
The aerosol inoculum grew progressively in the lungs of untreated mice with no sign of a decrease in the growth until day 25 when mice started dying (Figure 1) .All mice were dead at day 35 after airborne infection.This growth pattern is similar to that described by Orme and Collins (9).IL-2 has been found to enhance resistance to infections with virulent mycobacteria, namely M avium (10).M lepraemurium (11) and M bovis (11) .Moreover, intradermal application of IL-2 in patients with leprosy led to an elimination of the bacilli  in the lesions (12).Also, IL-2 with its central role as aT cell growth factor ( 13) and a macrophage-activating molecule (14) has been shown to enhance significantly the resistance against various microbial pathogens (15).In the present study, 100 J-lg of IL-2 administered by the aerosol route did not lead to any significant increase in resistance of mice to M tuberculosis (Figure 1).This suggests that IL-2 may not increase resistance in this aerogenic model.This may be due to the poor expression of IL-2 receptors in normal pulmonary and bronchoalveolar lymphoid tissues.However, treatment with 100 J-lg of IL-2 three times at two day intervals of already infected mice at a time when mice should express immunity (16) had no beneficial effect on the progression of infection (Figure 1).TNFa is an important cytokine in host resistance against infections.Depletion of endogenous TNFa significantly increased susceptibility of mice to M bovis (17).Listeria monocytogenes (18) and Leishmania donovani (19).Also, TNFa has been shown to increase resistance against M avium strains in vivo.Mice infected with M tuberculosis were treated with TNFa shortly after infection.TNFa-treated mice did not show progressive growth of M tuberculosis: there was growth up to 7 log10 cfu after which there was bacteriostatic activity in the lungs of infected mice (Figure 2), indicating that local TN Fa application may increase resistance to tuberculosis in mice.
Another cytokine considered in this model was IFN-y.Recognized as a major macrophage-activating molecule.IFN-y has been shown to increase resistance to a large spectrum of microbial agents (20).In vitro, IFN-y enhances mouse macrophage resistance to M tuberculosis (21).In vivo.IFN-y may diminish growth of virulent mycobacteria in the organs of infected mice Differences in colony forming units counts are signifkant (P<O.05 student's t test) at days 20 and 30) (22), although some studies have shown that virulent strains of mycobacteria may resist IFN-y in vivo or in vitro (21).The possibility was tested that IFN-y may increase resistance to tuberculosis.Aerosolization of 100 J-lg of IFN-y in infected mice induced a level of resistance similar to that seen with TNFa-treated mice survived up to 60 days with a bacteriostatic activity in the lungs of mice (Figure 2).In another set of experiments, the possibility was investigated that IFN-y in combination with TNFa could increase resistance in an additive or synergistic fashion.Administration of cocktails of cytokines in vivo or in vitro has often been shown to result in greater increases in resistance to infectious agents than seen with individual factors (4).
Notably, combinations of IFN-y and TNFa enhance resistance to mycobacteria, listeria and Schistosoma mansoni in vitro to an optimal degree (23,24).Both IFN-y and TNFa were applied by aerosol sequentially after infection, and the progression of the infection followed as described above.Figure 3 shows results obtained with combined treatments.Mice treated with IFN-y and TNFa exhibited a high degree of antituberculous resistance, as seen by the reduced growth of the microbes in the lungs.M tuberculosis grew to only approximately 6 log10 cfu in the lungs of TNFa/IFN-ytreated mice.This enhancement of resistance was not superior to that seen with either cytokine used alone (Figure 2, P>0.1).The protection obtained with TNFa/ IFN-y was dependent upon the route of exposure.Infusion of cytokines, such as TNFa and IFN-y (5 f-lg each cytokine per mouse daily).by the intraperitoneal route for 15 days increased the resistance only by a marginal degree (P<0.05 at days 20 and 30); however, all mice were dead by day 40 (Figure 4).This set of experiments  shows that repeated intraperitoneal infusion with large doses of cytokines was ineffective at modi1)ring the resistance/ susceptibility of mice to tuberculosis.Previous results have shown that aerosolized IFN-y and/or TNFa induced a high level of alveolar macrophage activation.as seen by enhanced lA expression, tumour cytotoxicity and IL-l expression (25) present results also sh ow that aerosol infusion of s u ch cytokines may enhance antimicrobial activity in the lungs.Follow-up stu dies have shown that alveolar macrophages from cytokine-treated mice have elevated levels of s u peroxide anion release after p horbolmyristic acetate triggering, indicative of an increase in cellu lar effector functions (unpublished data).It is sill not clear what is the actu al distribu tion of the cytokines in the lungs of infected mice.Administration of potentially toxic agents by site-specific delivery has been proposed as a way to concentrate the cytokine at the infectious foci (8) and to reduce the toxicity of systemic administration (25).It remains to be determined how the cytokines were acting on the progression of the disease in the present study.As mentioned above, there is evidence that IFN-y and TNFa are involved in antimycobacterial resistance.TNFa may be involved in promoting granulomatous lesions which are bacteriostatic ( l 7).IFN -y may promote mouse macrophage antimycobacterial functions directly (21).In summary, the present results indicate that aerosolized cytokines may endow infected hosts with strong resistance against M tuberculosis airborne infection.The authors are currently investigating this system of lung-specific delivery of cytokines in other infectious models.

Figure 2 )
Figure 2) Growth ojM tuberculosis in the lungs of infected mice.Mice were injected with M tuberculosis and were exposed to the following aerosolized agents: heat-inactivated cytokines only (o). 100 11g of recombinant interferon gamma (r IFN-y) (D) or 100 11g of recombinant tumour necrosis factor (r TNFa) (•) at day 0. Data expressed as in Figure 1.Data .fromjour experiments

Figure 3 )
Figure 3) Growth ojM tuberculosis in the lungs of infected mice.Mice were infected with M tuberculosis and were exposed to the following aerosolized agents: 100 11g of recombinant tumour necrosis factor (r TNFa) and 100 11g of recombinant interferon gamma (r IFN-y) (•) or the same cytokines heat-inactivated (o).Data expressed as in Figure 1.Data .fromjour experiments.

Figure 4 )
Figure 4) Growth ojM tuberculosis in the lungs of mice injected by the aerogenic route treated with cytokines by the intraperitoneal route.Mice were infused with 5 etg of interferon gamma (IFN-y) and 5 etg of tumour necrosis factor (TNFa) daily

Cytokines in tuberculosis
0% of the means).Results .fromjour combined experiments are shown.with .fivemice for each point fo r each experiment.X denotes death of mice.No signifr.cantdifference between any of the experimen• tal goups (ANOVA test) . The Mycobacterium tuberculosis infected lung.Dissociation between cells U1at passively transfer protective immunity and those that transfer delayed-type hypersensitivity to tuberculen.Cell Immunol 1984;84: 113-22.10.Bermudez LEM.Stevens P, Kolonoski P, We M. Young LS.Treatment of experimental clisseminated Mycobacterium CAN J INFECT DIS VOL 4 No 1 JANUARY/FEBRUARY 1993 Cytokines in tuberculosis