* Department of Medicine, University of Sydney, NSW 2006, Australia;
Centenary Institute of Cancer Medicine and Cell Biology, Locked Bag No. 6, Newtown, NSW 2042, Australia; and
Peptech Ltd. North Ryde, New South Wales 2113, Australia
Correspondence: Dr. Helen Briscoe, Department of Medicine, Blackburn Building DO6, University of Sydney, NSW 2006, Australia. E-mail: hbriscoe{at}med.usyd.edu.au
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) to reduce bacterial
replication and to induce synthesis of bactericidal nitric oxide (NO)
in BCG-infected, bone marrow-derived murine macrophages. The
dose-dependent inhibitory effect on bacterial replication was blocked
by neutralizing anti-IFN- and anti-hTNF mAbs. Further,
n-monomethyl-L-arginine (n-MMA) and a
soluble TNF-receptor I (TNFRI-IgG) blocked bacterial growth and
NO synthesis. Therefore, the peptide acted with IFN- via induction
of NO synthase and signaled through TNFRI receptors. Concomitant
in vivo treatment with TNF70–80 or hTNF
prevented reactivation of chronic BCG infection in mice depleted of
CD4+ T cells by injecting anti-CD4 antibodies. Granuloma
number and bacterial load were comparable in treated, T cell-depleted
mice and in chronically infected, intact animals. Thus,
TNF70–80 and hTNF can modulate recrudescent BCG infection
in CD4+ T cell-deficient mice.
Key Words: mycobacteria • cytokine • nitric oxide • granuloma • macrophages
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T cells function through a complex pattern of cytokine production and
induction of bacterial killing mechanisms in macrophages. Interferon
(IFN)- produced by CD4+ T cells, CD8+ T
cells, and natural killer (NK) cells stimulates effective phagocytosis
and synthesis of bactericidal reactive nitrogen intermediates (RNI).
Mice deficient in IFN- fail to control M. tuberculosis
[10
, 11
], and administration of
neutralizing anti-IFN- antibodies exacerbated the infection
[12
]. Administration of recombinant IFN- protected
mice against lethal M. tuberculosis infection in some, but
not all, animal models [10
, 12
] and was a
useful adjunct to chemotherapy of MDR-TB in humans [13
].
IFN- alone, however, is insufficient to fully activate bactericidal
mechanisms in human and murine macrophages [4
]. Other
cytokines, including interleukin (IL)-4 [14
], IL-6
[15
], granulocyte-macrophage colony-stimulating factor
(GM-CSF) [16
], 1,25 dihydroxy-vitamin D3 in humans
[17
], and notably tumor necrosis factor (TNF)
[14
, 16
], synergize with IFN- to enhance
mycobacterial killing.
TNF is produced by macrophages after stimulation with IFN- and in
direct response to mycobacterial products. Murine bone marrow-derived
macrophages (BMM) activated by a combination of IFN- and TNF
inhibited intracellular growth of M. bovis bacillus
Calmette-Guerin (BCG) in vitro through RNI-dependent
processes [14
]. Mice that had disrupted genes for TNF
[18
] or TNFp55 receptor [19
] were
profoundly susceptible to infection with M. tuberculosis,
and neutralization of TNF decreased resistance to M.
tuberculosis [19
, 20
] and M.
bovis [21
] in vivo. Defective TNF
production may contribute to disseminated M. avium infection
in humans [22
]. Exogenous TNF treatment increased
clearance in experimental mycobacterial infections [12
,
23
], but systemic therapy is complicated by the short
half-life of TNF in vivo and its propensity to induce
cachexia [24
]. To overcome this, synthetic peptide
analogues of human TNF with prolonged half-life and in vivo
activity have been developed [25
, 26
].
Amino acids 70–80 were identified as essential for activity of human
TNF molecules, and the stability of this peptide in vitro
was enhanced by substitution of isoleucine for leucine at position 76
(unpublished results). This peptide, TNF70–80, activated
human and murine neutrophils in vitro, and in a murine
model, systemic treatment with TNF70–80 increased the rate
of recovery and clearance of Plasmodium chabaudi
[27
]. Here, we report the efficacy of
TNF70–80 in inducing bacteriostasis and preventing
recrudescence in a murine model of mycobacterial infection.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cytokines, peptide, and antibodies
The sequence of TNF70–80 peptide is
H-Pro-Ser-Thr-His-Val-Leu-Ileu-Thr-His-Thr-Ileu-OH. The peptide was
synthesized by the F-moc-polyamine method [28
] of
solid-phase peptide synthesis using PepSyn KA solid resin
[26
] and purified by high-pressure liquid chromatography
(HPLC). Murine IFN- (106 U/ml) was purchased from
Genzyme (Cambridge, MA). Recombinant human TNF (hTNF;
1.7x107 U/ml) was provided by Peptech Ltd. (Sydney,
Australia). Fresh dilutions of peptide and cytokines were prepared in
experimental medium. The relative potencies of hTNF and
TNF70–80 were compared for stimulation of nitric oxide
(NO) release. TNF70–80 (5.0 µg/ml) had activity
corresponding to 1000 U/ml of hTNF. Hamster anti-IFN- monoclonal
antibody (mAb) was purchased from Genzyme. Anti-hTNF mAb 054 was
purified from ascites fluid by ammonium sulphate precipitation
[25
]. Isotype control mAb L5 binds to the M.
leprae 18 kDa protein. A soluble form of the hTNF-receptor I (RI)
composed of recombinant 55 kDa TNFRI fused to human immunoglobulin G
(IgG) heavy-chain domain (TNFRI-IgG, designated p55-sf2)
[29
] was kindly provided by Dr. B. Scallon (Centocor,
Malvern, PA), along with control Ig. mAb to murine CD4+
[GK1.5, American Type Culture Collection (ATCC), Rockville, MD] was
purified from culture supernatant by affinity chromatography on
protein-G (Pharmacia Biotech, Uppsala, Sweden).
BCG culture in BMM
Bone marrow cells were flushed from the long bones of C57Bl/6
mice into experimental medium supplemented with 5% horse serum (Trace
Biosciences, Sydney, Australia) and 20% macrophage growth factors
(supernatant from cultured L929 fibroblasts) and cultured for 7 days in
suspension in six-well plates (Interpath Services, Sydney, Australia).
BMM were harvested into experimental medium at 106
cells/ml. Aliquots (100 µl, 105 cells) were distributed
into microtitre wells (Nunc, Roskilde, Denmark) and incubated for
6 h in 5% CO2 at 37°C. BMM were cultured with
varying concentrations of cytokines or the peptide for 16 h.
Cytokines were removed by washing in warm RPMI, and cells were infected
with 106 BCG organisms and cultured for 3 days. Culture
supernatants were tested immediately or stored at -70°C for nitrite
measurement. In some experiments, the soluble form of TNFRI (p55-sf2)
or normal IgG1 (control) was added to cultures together with cytokines
at concentrations of 0.5, 5.0, or 50.0 µg/ml. In other experiments
n-monomethyl-L-arginine (n-MMA; Calbiochem, San
Diego, CA) was added to the culture medium together with cytokines at
concentrations of 0.01–10 mM.
Assay for in vitro BCG growth
After 3 days of culture, supernatants were aspirated, and
infected BMM were washed twice with warm medium, then lysed in 100 µl
of 0.1% saponin (Sigma), and 1 µCi of 3H-uracil
(Amersham, Amersham, UK) was added in 100 µl experimental medium.
Further incubation for 24 h allowed incorporation of
3H-uracil into the RNA of viable mycobacteria.
3H-uracil incorporation was determined by liquid
scintillation spectroscopy. The percent inhibition of
3H-uracil into BCG was calculated as follows: (mean
incorporation of triplicate cultures without cytokines minus mean of
triplicate cultures with cytokines) divided by (mean of triplicate
cultures without cytokines).
NO measurements
Nitrite levels were measured with the Greiss reagent
[30
]. Briefly, Greiss reagent was freshly prepared by
mixing 3.0% phosphoric acid, 1% sulphanilamide, and 0.1%
n-(1-naphthyl)ethylenediamine (Sigma) in distilled water.
Culture supernatants, 100 µl, were incubated with 100 µl of Greiss
reagent in microtitre trays in triplicate for 10 min at room
temperature, and the optical density was measured at 540 nm. Nitrite
levels were calculated from a standard curve prepared with serial
dilutions of sodium nitrite (Sigma) in distilled water.
Cytokine and antibody treatment of mice
SPF C57Bl/6 female mice aged 6–10 weeks were purchased from The
University of Sydney Animal Facility. To modulate the course of
recrudescent infection, mice were infected intravenously with
106 cfu of BCG organisms 9 weeks prior to CD4+
T cell depletion and cytokine treatment: intraperitoneal injection of
300 µg anti-CD4 mAb daily for 3 days, then 4 days later, and
thenceforth, once weekly; and 0.5 µg TNF or 100 µg
TNF70–80 administered intraperitoneally every second day,
for 28 days.
Histological analysis
Liver tissues fixed in 10% neutral formalin, were set in
paraffin blocks and stained using haematoxylin and eosin. The average
number of granulomas in 10 randomly chosen high-powered (x400) fields
was determined. Sections of liver, frozen in OCT (Tissue-Tek, Sakura,
Torrance, CA), were incubated with rat mAb to mouse major
histocompatibility complex (MHC) class II (P7/7), then
biotin-conjugated goat anti-rat Ig (Caltag, San Francisco, CA),
streptavidin alkaline phosphatase (Amersham), and New Fuchsine (Sigma).
The average number of foci of MHC class II-positive tissue in 10
high-power fields (x400) was determined by image analysis.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
to inhibit
growth of BCG alone resulted in partial reduction of
3H-uracil uptake (Fig. 1 A
and C
). Addition of increasing doses of recombinant hTNF or
TNF70–80 enhanced bacteriostasis (decreased
3H-uracil uptake; Fig. 1A
and 1C
). This was associated
with dose-dependent production of NO, measured as nitrite in culture
supernatant 3 days after infection (Fig. 1B
and 1D)
. hTNF alone or
TNF70–80 alone had minimal effect on BCG replication and
NO production: The maximum nitrite release was <1.0 µg/ml at 1000
U/ml hTNF or 5.0 µg/ml TNF70–80, with 22% and 27%
inhibition of BCG growth, respectively. Therefore,
TNF70–80 mimics hTNF function in vitro and acts
with IFN- to induce NO synthesis and BCG growth inhibition. Murine
TNF had similar effects to hTNF and peptide TNF70–80, and
pretreatment with control peptide from another region of hTNF
(TNF6–18) had no synergistic effect with IFN- on BMM
activation (unpublished results). Addition of neutralizing antibodies
to IFN- or hTNF during the activation phase blocked BCG growth
inhibition, further emphasizing that the inhibitory effects of
macrophage activation were dependent on coactivation with IFN- and
hTNF or TNF70–80 (Table 1 ).
 View larger version (32K): [in a new window] |
Figure 1. Activation of BMM by IFN- and recombinant hTNF (A, B), and IFN-
and TNF70–80 (C, D). Activation was determined by
measuring the inhibition of BCG growth as reflected by
3H-uracil incorporation (A, C) and by nitrite release (B,
D) at 3 days post-infection. The cells were incubated with IFN-
alone ( ) or IFN- with increasing concentrations of hTNF (A, B)
( 64 U/ml, • 250 U/ml, or  1000 U/ml) or TNF70–80
(C, D) ( 0.31 µg/ml, • 1.25 µ/ml, or 5.00 µ/ml). Results
are mean values ± SEM in triplicate cultures from one
of four similar experiments. The maximum uptake of
3H-uracil into replicating BCG in the absence of
cytokines ± standard deviation was 38,826 ± 58.
|
|
View this table: [in a new window] |
Table 1. The Effect of Antimurine IFN- Antibodies and Antihuman TNF
Antibodies on the Activation of Antimycobacterial Activity of Bone
Marrow-Derived Macrophages by the Combination of IFN- with TNF or
Peptide TNF70–80 or by IFN-
Alonea
|
and increasing concentrations of
p55-sf2 blocked macrophage activation in a dose-dependent manner.
Reduced NO production and concomitant increase in uptake of
3H-uracil into viable BCG organisms were observed
(Fig. 2
). Soluble TNFRI also blocked BMM activation by
TNF70–80 (Fig. 2)
, consistent with peptide signaling
occurring through the TNFRI. Control protein (irrelevant IgG) did not
block BMM activation (unpublished results). The modest effects of
IFN- alone (Fig. 1) were not affected by addition of soluble TNFRI.
n-MMA blocks production of NO from L-arginine via induction
of NO synthase (iNOS). Addition of n-MMA (1 µM) to cultures blocked
the induction of NO by IFN- and hTNF or TNF70–80, and
prevented the inhibition of BCG growth (Table 2
). Similar effects also occurred with the addition of a second iNOS
inhibitor, aminoguanidine (unpublished results). Therefore, the
inhibition of BCG growth by TNF70–80 is dependent on NO
production via the L-arginine pathway.
 View larger version (42K): [in a new window] |
Figure 2. Inhibitory effects of a soluble TNFRI-IgG fusion protein on the
activation of BMM by the combination of hTNF or peptide
TNF70–80 and IFN-. The inhibitory effect was determined
by measuring the suppression of antimycobacterial activity as reflected
by 3H-uracil incorporation into BCG (A) and by nitrite
release at 3 days post-infection (B). Results were calculated from
triplicate cultures and are representative of three similar
experiments. The cells were incubated with IFN- (100 U/ml) and hTNF
(250 U/ml; shaded bars) or IFN- (100 U/ml) and peptide
TNF70–80 (1.25 µ/ml; open bars) with increasing
concentrations of soluble TNFRI-IgG fusion protein (0.5, 5.0, or 50.0
µg/ml) for 24 h prior to BCG infection.
|
|
View this table: [in a new window] |
Table 2. The Inhibitory Effect of Monomethyl-L-Argigine (n-MMA)
on the Activation of Anti-Mycobacterial Activity of Bone
Marrow-Derived Macrophages by the Combination of IFN- with TNF
or Peptide TNF70–80a
|
1000 organisms per spleen [31
]. These organisms were
contained in small, discrete granulomas. Depleting CD4+ T
helper cells by antibody or corticosteroid hormone treatment causes
reactivation of infection in the liver and spleen [7
].
To determine if treatment with hTNF or TNF70–80 can
modulate this relapse, mice infected with 106 BCG organisms
intravenously were rested for 9 weeks, then treated with anti-CD4 mAb
or saline (control) on days 0, 1, and 2 to deplete CD4+ T
cells, then on days 7, 14, and 21, to maintain depletion. Concomitant
second daily treatment with hTNF (0.5 µg), TNF70–80 (100
µg), or saline commenced at day 0 and continued for the duration of
the experiment. This regime was shown by flow cytometric analysis to
cause a 97% reduction in the number of CD4+ T cells and no
reduction in CD8+ T cells or B cells in the spleens of
antibody-treated mice tested before antibody was injected on days 7,
14, and 21 (unpublished results). Figure 3
shows recrudescent BCG infection in CD4+ T
cell-depleted mice. Treatment with hTNF or TNF70–80
prevented relapse. Treatment with TNF70–80 or hTNF
decreased the bacterial load in the spleen (Fig. 3C)
and reduced the
numbers of liver granulomas, identified as accumulations of 
10
macrophages and lymphocytes (Fig. 3A)
and as foci of activated MHC
class II-positive macrophages (Fig. 3B)
. Cytokine treatment did not
affect the low-level chronic infection in intact mice (Fig. 3)
.
Administration of control peptide (TNF6–18) in
vivo did not modulate the course of BCG infection (unpublished
results). Treatment with TNF70–80 or hTNF also resulted in
a decrease in the size of liver granulomas in CD4+ T
cell-depleted mice measured on stained histological sections
(Fig. 4
). These data suggest that treatment with
TNF70–80 or hTNF activated macrophages at sites of chronic
infection to inhibit BCG growth secondary to deficiency of
CD4+ T cells. This prevented resurgence in the bacterial
load and the consequent increase in granulomas.
 View larger version (27K): [in a new window] |
Figure 3. Reactivation of chronic BCG infection in CD4+ T
cell-depleted mice was prevented by in vivo treatment with
hTNF and TNF70–80. CD4+ T cell-depleted mice
(•) were compared to intact, control mice ( ). Reactivation was
determined by measuring the number of liver granulomas (A), the number
of foci of activated, MHC class II-positive macrophages in the liver
(B), and the bacterial load in the spleen (C). Each symbol represents
the value in an individual mouse. The columns represent the median
value for each treatment group in one of four similar experiments. The
number of liver granulomas in hTNF- and TNF70–80-treated,
CD4+ T cell-depleted mice was significantly less than that
in the CD4+ T cell-depleted, (PBS) phosphate-buffered
saline-treated, control group: p < 0.02 (hTNF), and
p < 0.05 (TNF70–80) using the
nonparametric Mann-Whitney U test. The number of foci in hTNF- and
TNF70–80-treated, CD4+ T cell-depleted mice
was significantly less than that in the CD4+ T
cell-depleted, PBS-treated, control group: p < 0.01
(hTNF), and p < 0.05 (TNF70–80) using the
same statistical analysis. The number of BCG cfus in hTNF- and
TNF70–80-treated, CD4+ T cell-depleted mice
was significantly less than that in the CD4+ T
cell-depleted, PBS-treated, control group: p < 0.02
(hTNF), and p < 0.02 (TNF70–80) using the
same statistical analysis. No significant difference was observed with
any treatment in intact mice.
|
 View larger version (184K): [in a new window] |
Figure 4. Histology of liver tissue showing treatment with hTNF and
TNF70-80 reduced the pathology associated with recrudescent
BCG infection. Compact granulomas, consisting of small accumulations of
lymphocytes and activated macrophages, were observed in
CD4+ T cell-depleted mice treated with hTNF (A) or
TNF70-80 (B). Similar lesions were seen in chronic BCG
infection (untreated, intact mice, 9 weeks post-infection, D). Larger
liver lesions containing higher numbers of lymphocytes and activated
macrophages were observed in reactivated infection (sections from
saline-treated, CD4+ T cell-depleted mice, C). Tissue was
fixed in buffered formalin, sectioned, and stained with haematoxylin
and eosin. Original magnification x 200.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our in vitro studies suggest mechanisms for the functional
activity of TNF70–80. Peptide TNF70–80, like
hTNF, synergized with IFN- to activate murine BMM and inhibit the
replication of M. bovis (BCG). BCG growth inhibition was
abrogated by addition of neutralizing anti-IFN- and anti-hTNF.
Similar observations were made in a murine macrophage cell line
infected with BCG and treated with peptide or soluble human or murine
TNF [33
]. TNF engages two receptors, the p55 TNFRI and
p75 TNFRII, on the surface of leukocytes, and the resulting signals
induce a wide range of biological effects, including apoptosis,
macrophage activation, and cell proliferation [34
]. A
soluble form of TNFRI blocks TNF-induced pathology in experimental
allergic encephalitis [35
] and protects mice against
endotoxic shock [36
]. This TNFRI-IgG fusion protein
abrogated the activity of TNF and peptide TNF70–80 on
macrophage activation, consistent with the signaling induced by
TNF70–80 occurring through the TNFRI. TNF is considered to
bind TNF receptors as a trimer [34
], and the mechanism
by which the 11-mer peptide TNF70–80 binds and signals
through the TNFRI is currently under investigation. The inhibitory
effect of TNF70–80 on BCG replication was dependent on
synthesis of RNI via induction of iNOS. hTNF and TNF70–80
synergized with IFN- to stimulate NO release. Addition of a
competitive inhibitor of iNOS function, n-MMA, to BCG-infected
macrophage cultures abrogated the mycobacterial inhibitory effects of
peptide TNF70–80 and hTNF.
From these in vitro data, we infer that, as the BCG
infection starts to resurge following CD4+ T cell
depletion, early therapy with TNF70–80 activates
macrophages to contain the bacterial replication through synthesis of
RNI. The stimulation of RNI production is also dependent on local
levels of other cytokines, notably IFN-. Possible sources of IFN-
include CD8+ T cells [37
], 
T cells
[38
], or NK cells [39
]. Characteristic
MHC class II expression on activated macrophages is present in the
granulomas in CD4+ T cell-depleted,
TNF70–80-treated mice (Fig. 3B)
. This is associated with
increased iNOS expression in liver granulomas in mice treated with
TNF70–80 during primary BCG infection (unpublished
results). Therefore, TNF70–80 functions in vivo
by stimulating bactericidal RNI in infected macrophages. The low
bacterial load in intact mice was not further reduced by treatment with
hTNF or TNF70–80, perhaps because a balance has been
established between cytokine-induced macrophage activation by IFN-
and TNF and down-regulatory cytokines, for example TGF-ß and IL-10,
that restrict IFN- synthesis and macrophage antimycobacterial
functions [40
, 41
]. The effect of TNF and
TNF70–80 was apparent when this balance was perturbed.
This TNF70–80 peptide is effective in vivo in other models of infection. Significant reduction in parasitaemia occurred following TNF70–80 treatment of mice infected with P. chabaudi [27 ]. In chronic Pseudomonas aeruginosa infection of mice, TNF70–80 therapy resulted in reduced weight loss and enhanced clearance of organisms (unpublished results). Intrapulmonary administration of TNF70–80 in the lungs of mice prior to infection with Klebsiella pneumoniae or Aspergillus conidia resulted in improved survival and clearance of the microorganisms [42 , 43 ]. TNF70–80 stimulated and primed cultured human neutrophils for increased respiratory burst, release of their granular contents, and consequent enhanced killing of P. falciparum in vitro [27 ]. In this respect TNF70–80 closely resembles the previously described properties of TNF [44 ]. Interestingly, unlike intact TNF, the peptide did not promote expression of adhesion molecules [intercellular adhesion molecule 1 (ICAM-1), endothelial leukocyte adhesion molecule 1 (ELAM-1), and vascular cell adhesion molecule 1 (VCAM-1)] on endothelial cell membranes [27 ]. Also, in contrast to intact TNF, exogenous TNF70–80 was shown to be free of toxicity in P. chabaudi-infected mice in concentrations up to 400 mg/Kg [27 ].
In summary TNF70–80 stimulated macrophage activation and modulated recrudescent mycobacterial infection, leading to a reduction in pathology and bacterial load. Thus, the peptide has clinical therapeutic potential as an agent to modify pathology caused by intracellular pathogens and as an immunostimulatory adjunct to antimicrobial therapy, for example, in the treatment of chronic, drug-resistant M. avium complex infection in advanced HIV/AIDS or treatment of multidrug resistant tuberculosis.
Received December 30, 1999; revised April 25, 2000; accepted April 26, 2000.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in resistance to Mycobacterium tuberculosis infection J. Exp. Med. 178,2249-2254 gene disrupted mice J. Exp. Med. 178,2243-2247 via aerosol Lancet 349,1513-1515[Medline]
is required in the protective immune response against Mycobacterium tuberculosis in mice Immunity 2,561-572[Medline] by different tumour cell receptors Mol. Immunol. 28,79-86[Medline] peptides on tumour cytotoxicity, neutrophil activation and endothelial cell procoagulant Immunology 80,293-299[Medline] agonist peptide enhances human polymorphonuclear leukocyte-mediated killing of Plasmodium faciparum in vitro and suppresses Plasmodium chabaudi infection in mice J. Clin. Invest. 95,2315-2323
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
