Repeated delayed-type hypersensitivity reactions against ...

Vol. 36, No. 2INFECTION AND IMMUNITY, May 1982, p. 768-7740019-9567/82/050768-07$02.O0/0

Repeated Delayed-Type Hypersensitivity Reactions AgainstMycobacterium lepraemurium Antigens at the Infection Site

Do Not Affect Bacillary Multiplication in C3H MiceMARTINUS LQVIK* AND OTTO CLOSS

University of Oslo, Institute for Experimental Medical Research, Ullevaal Hospital, Oslo 1, Norway

Delayed-type hypersensitivity was induced in cyclophosphamide-pretreatedC3H/TifBom mice by subcutaneous immunization in the thorax with ultrasonicat-ed Mycobacterium lepraemurium bacilli in Freund incomplete adjuvant. Sevenweeks after immunization, 2.5 x 107 acid-fast M. lepraemurium bacilli suspendedin diluted sonicate were injected into one hind footpad, and during the next 6weeks three additional injections of sonicate were given at intervals into theinfected footpad. After each injection a strong local reaction developed, whichafter the first three injections peaked at 24 h. The kinetics of the reaction was

accelerated after the repeat injections. Each time the reaction subsided within 1week. From 2 days to 11 weeks after the inoculation of bacilli there was a 10-foldincrease in bacillary numbers in the footpad and a 3,000-fold increase in thedraining popliteal lymph node. The degree of bacillary multiplication was thesame in animals which had experienced repeated local reactions and in controlanimals. Thus, repeated strong delayed-type hypersensitivity reactions to M.lepraemurium antigens apparently were without any measurable effect on thebacillary multiplication. This observation provides further evidence for a dissocia-tion in C3H/TifBom mice between delayed-type hypersensitivity to solublemycobacterial antigens and protective immunity against mycobacteria. Possibleexplanations for our findings are discussed.

Resistance against mycobacterial infections isthought to depend on cell-mediated immunity(41, 43). The expression of acquired protectivecell-mediated immunity requires activation ofmacrophages by T lymphocytes, but apart fromthis little is known about the mechanism ofresistance (43). Protective immunity against in-tracellular parasites is often accompanied bydelayed-type hypersensitivity (DTH) against mi-crobial antigens (27, 29), but the nature of therelationship between DTH and acquired protec-tive immunity is controversial (19, 20, 42).

Inbred C57BL mice are relatively resistant toinfection with Mycobacterium lepraemurium(MLM) (8, 15). The onset of the protectiveimmune response in this strain of mice coincideswith the development at the site of infection of alocal reaction with the features of a DTH granu-loma (6, 10). C3H mice are highly susceptible toinfection with MLM (8, 15). After inoculation ofMLM they do not develop DTH granulomas (6,10), and we have not been able to demonstrateDTH against bacillary antigens in infected ani-mals (M. Lovik and 0. Closs, manuscript inpreparation). However, in C3H as well as inC57BL mice DTH against soluble MLM anti-gens can be induced by a subcutaneous injectionof an ultrasonicate of MLM (MLMSon) in

Freund incomplete adjuvant (25). Cyclophos-phamide pretreatment has been found to have asmall enhancing effect on the development ofDTH in C3H mice immunized with MLMSon(L0vik and Closs, in preparation). Experimentswith C57BL mice indicated that immunizationwith MLMSon did not induce protective immu-nity (12). Because C3H mice do not develop aprotective immune response against MLM,MLMSon-immunized C3H mice are particularlywell suited for the study of the effect of DTHreactions on MLM infection. We therefore didexperiments to study the effect of repeated DTHreactions against MLMSon on bacillary multipli-cation and dissemination in C3H mice.(A preliminary experiment was reported at the

Symposium on Host-Parasite Interactions heldin Ume'a, Sweden, 6-8 June 1979 [25].)

MATERIALS AND METHODSAnimals. Female mice of the inbred strain C3H/

TifBom were obtained as specific-pathogen-free ani-mals directly from the breeder (Gl. Bomholtgard Ltd.,Ry, Denmark). The mice were rested for about 3weeks in the laboratory and weighed 20 to 22 g whenthey were used in experiments. They were kept incages, five to six mice in each, and fed on pellets andtap water ad libitum.MLM bacilli. The Douglas strain of MLM was

768

RESISTANCE AT DTH REACTION SITE 769

maintained by repeated passage in C3H/TifBom mice.The bacilli were harvested from the spleens of miceinfected intraperitoneally with about 109 MLM 20weeks previously. The spleens were homogenized insaline in a Ten Broeck all-glass homogenizer (BellcoGlass, Inc., Vineland, N.J.), and purified suspensionsof bacilli were prepared by differential centrifugationas previously described (26). The bacilli were storedunder liquid nitrogen before inoculation. The bacilliwere counted by a slide technique (6). The prepara-tions were stained for acid fastness with auramine (21)and counted in a Leitz Ortholux microscope equippedwith incident illumination for fluorescence microsco-py-Preparation of MLMSon. MLM bacilli were purified

by differential centrifugation, washed twice in 0.9%NaCI, and finally resuspended in 0.9% NaCl at aconcentration of 4 x 1010 bacilli per ml. The suspen-sion was sonified for 15 min at a measured effect of 100W, using a Branson B12 Sonifier (Branson SonicPower Co., Danbury, Conn.). Insoluble material wasremoved by centrifugation at 20,000 x g for 20 min.The protein concentration of the supernatant to beused after appropriate dilution with saline for footpadinjection was 0.8 mg/ml as determined by the modifiedFolin-Ciocalteu method, with bovine serum albuminas a standard (24). The sonicate was stored in dilutionsof 1:1, 1:10, and 1:100 at -20°C. The same batch ofsonicate was used for all footpad injections throughoutthe experiment. A separate batch of sonicate to beused for immunization only was prepared in the samemanner, but before use it was reconstituted with theamount of insoluble material that had been removedby centrifugation.Cyclophosphamide pretreatment and immunization.

Cyclophosphamide (Sendoxan, batch BC 0117; Phar-macia, Uppsala, Sweden) was given intraperitoneallyin a dose of 100 mg/kg 3 days before immunizationwith MLMSon. For immunization, equal volumes ofMLMSon and Freund incomplete adjuvant were thor-oughly mixed and exposed to repeated brief pulses ofultrasound until a thick, stable emulsion was obtained.A single injection of 50 ,ul of this emulsion was givensubcutaneously in the right side of the thorax.Experimental infection and injection of MLMSon.

For all footpad injections a volume of 10 ,ul of theappropriate material was given through a 30-gaugeneedle from a 100-,ul syringe (Hamilton Bonaduz,Bonaduz, Switzerland). The needle path was sealedimmediately afterwards with a plastic spray (Nobecu-tan; AB Bofors, Nobel-Pharma, Molndal, Sweden).The bacillary suspension was made by mixing equalvolumes of MLMSon diluted 1:5 with saline and asuspension containing 5 x 109 MLM bacilli per ml insaline.Two days and 11 weeks after the inoculation of

bacilli, groups of five mice were killed, and the infect-ed footpad and the draining popliteal lymph node wereremoved. The bacilli were harvested from the footpadand the lymph node as previously described (6), andthe acid-fast bacilli were counted.Footpad reactions. Immediately before and usually

4, 24, 48, and 72 h after an injection, the thickness ofboth hind feet was measured with a modified dialgauge caliper (C. E. Johansson AB, Eskilstuna, Swe-den). Five successive measurements of each hind footwere recorded, and the swelling was expressed as the

difference between the median value of the injectedand control feet (6).

Histology. At 48 h after injection of bacilli inMLMSon or MLMSon without bacilli, two mice fromeach group were taken for histology. The footpadswere fixed for 24 h in 10% Formol-alcohol with 5%acetic acid and then transferred to 80% alcohol. Fur-ther processing was made by standard techniques, andthe sections were stained with hematoxylin and eosin.

Statistical methods. The Wilcoxon rank sum test forunpaired samples was used to test differences betweengroups (39). P = 0.05 was used as the limit of statisticalsignificance.

RESULTSThirty-four days after immunization, a group

of five cyclophosphamide-pretreated sonicate-immunized mice and a group of five normal micewere injected in one hind footpad with 10 ,u1 ofMLMSon diluted 1:100. The resulting footpadswelling is given in Fig. 1. At 4 h, both groupshad a small swelling, with no significant differ-ence between immunized mice and controls.Thereafter, a swelling with the kinetics of a DTHreaction developed in the immunized mice. Theswelling peaked at 24 h, but the 48-h swellingwas almost equal in strength to the 24-h swell-ing. In the control mice no swelling developedafter the small 4-h reaction to the injection.Having determined that DTH against

MLMSon had been induced in the immunizedmice, we injected two other groups of immu-nized mice in one hind footpad with 10 ,u1 ofMLMSon (1:100) to boost local reactivity. Oneweek later, one of these groups was injected inthe same footpad with 2.5 x 107 MLM bacillisuspended in MLMSon diluted 1:10, and theother group was injected with MLMSon diluted1:10 without bacilli. A third group, normal mice,

Footpadswelling (mm)

0.30 ,

0.20 -

0.10-

0-I

To T

-rIT24 48 72II 2 A

4 24 48 72

hours

FIG. 1. Footpad swelling in mice immunized 34days previously with MLMSon in Freund incompleteadjuvant (0) and in normal mice (0) after injection of10 ,ul of MLMSon diluted 1:100. Median and range forgroups of five mice.

VOL. 36, 1982

770 L0VIK AND CLOSS

otpad The 4-h swelling was not measured after the first~elling (mm)0 injection in the present experiment, but in two

o0_ otherwise similar preliminary experiments the 4-0 l A h swelling reactions after the first injection were

o0. 0.16 and 0.20 mm, respectively (25; unpublished

~0I swelling was almost as strong as the 24-h swell-ing after the second and third injections, and

o\j1 It\ \after the fourth injection it was even strongero- t t t\l \, than the 24-h reaction. Whereas all four injec-ow 1vT*r ,,',j, tions induced a strong local reaction, a shift in

'o 1 1 t the kinetics towards a stronger 4-h reaction and0 a weaker 48-h reaction occurred after repeated::,B injections. Note that the swelling in the mice

given bacilli plus sonicate (Fig. 2A) and the

°01 ; } il tiswelling in the mice given sonicate only (Fig.go 2B) were very similar, both after the first injec-l4,4 ' ~~~~~tion and after the three next injections with

to]1\l\\|I\ sonicate.I- T + ; ,, The immunized mice had four repeated strong

o.,,, local reactions in the infected footpad overago- t lt period of 6 weeks. Generally we regard a foot-

o10 pad swelling of >0.10 mm as a significant reac-

o0 i tion, and the median footpad swelling was above14-''*-1 ^ ^ + v~this level for approximately 5 weeks. However,1346it was above 0.20 mm for four periods of approx-

weeks imately 3 days each, which gives a total of only

FIG.2. . . 12 days with strong reactions (Fig. 2A).7 acid-fast MLM bacilli in MLMSon diluted 1:10 Groups of five immunized mice and normaliick arrows) or after injection of MLMSon 1:10 controls inoculated with bacilli in MLMSonithout bacilli (thin arrows) into mice that had been were harvested for counting of bacilli in themunized subcutaneously with MLMSon in Freund infected footpad and the draining popliteal

incomplete adjuvant (A, B) or into normal mice (C).For each injection, the 4-, 24-, 48-, and 72-h intervalsare denoted on the abscissa. Median and range forgroups of five mice.

was given 2.5 X 107 MLM in MLMSon diluted1:10 to serve as controls. Thereafter, the twogroups of immunized mice were given threemore injections of MLMSon (1:10) into thepreviously injected footpad at the intervals givenin Fig. 2. After the first injection of MLMSon(1:10) with bacilli, a strong 24-h swelling devel-oped in the immunized group (Fig. 2A). In thenormal mice, the 24-h swelling was comparative-ly small, and the swelling rapidly disappeared,with no further swelling during the remainder ofthe experiment (Fig. 2C). After 1 week, a newinjection of MLMSon diluted 1:10 was given tothe immunized mice (Fig. 2A and B). The 24-hreaction now was somewhat increased com-pared with the reaction 1 week previously. Afterthe third and fourth injections there was a smallreduction in the 24-h reactivity. The 48-h reac-tion after the first injection was almost as strongas the 24-h reaction. The 48-h reaction after thethree last injections was also strong, but some-what weaker than the 24-h reaction. After allfour injections the reaction subsided rapidly.

109 -

108-

0

Xn 107

0

c)

a) 1o6 -

z

o05

14

A

r2 11

days weeks

B

2 11days weeks

FIG. 3. Multiplication of M. lepraemurium in thefootpad (A) and popliteal lymph node (B) after footpadinoculation. Arrow indicates inoculum size. Symbols:[, immunized mice that had repeated DTH reactionsin the infected footpad; U, normal controls. Medianand range for groups of five mice.

Fo(Sw

0.7

0.3

0.2

0.1

10(thwiim

INFECT. IMMUN.

RESISTANCE AT DTH REACTION SITE 771

lymph node 2 days after inoculation. A newharvest was performed 11 weeks after the inocu-lation of bacilli, 6 weeks after the last injectionof MLMSon. There was a 10-fold multiplicationof the bacilli in the footpad in both groups after11 weeks, and the multiplication was the same inthe group of immunized mice with repeatedDTH reactions as in the normal mice (Fig. 3A).The doubling time was close to 3 weeks, whichis the usual doubling time in the footpad afterinfection with this dose of bacilli. In preliminaryexperiments, no indication was found that bacillisuspended in MLMSon had reduced growth innormal mice compared with bacilli injected insaline (data not given). In the popliteal lymphnode, the doubling time was 1 week, and therewas an approximately 3,000-fold multiplicationof the bacilli. At this site also there was nodifference between the mice with repeated reac-tions and normal mice (Fig. 3B). In a preliminary

experiment in which we similarly found no inhi-bition of bacillary multiplication by repeatedDTH reactions against MLMSon (25), the dis-semination of bacilli to the popliteal lymph nodewas significantly increased in immunized miceinoculated with bacilli in MLMSon as comparedwith immunized mice inoculated with bacilli insaline and normal mice given bacilli inMLMSon. Although slight and not statisticallysignificant, a similar tendency to increased dis-semination of the bacilli in mice with DTHreaction was observed in the present experiment(Fig. 3).

Histological examination of footpad tissue re-moved 48 h after injection of bacilli in MLMSonor MLMSon without bacilli showed in immu-nized mice a cellular infiltrate that was predomi-nantly mononuclear in composition and con-formed to that of a DTH reaction (Fig. 4).Deposits of a fibrillary fibrinoid substance were

t.., Si Xpi~Qo

X ~ ~ ~ ~ ~~~~~~::

* #

PS As0,

..sS X t:. ik.

::#:-F.dv ......... .;..............

t

|.*ShX.:1.*SS

+: }:|::Rs.

bsi..g.e.

|# *:|tsi X:

k .5 '

..

AO..

FIG. 4. Histological appearance of the 48-h footpad reaction elicited by 1%lO1 of MLMSon (dilution 1:10)containing 2.5 x 107 MLM bacilli injected into mice previously immunized with MLMSon. A preponderance ofmononuclear cells is seen infiltrating the subcutaneous tissue (hematoxylin and eosin; x800).

VOL. 36, 1982

Y.",I"

..w ::

-.10.1P7.-. - %,.,.

..A--

If--AmOll-

fall:-,

::t-- .,,i,,

772 L0VIK AND CLOSS

seen in the tissue. No difference was found inthe composition of the cellular infiltrate betweenmice given MLMSon with bacilli and mice givenMLMSon without bacilli.

DISCUSSIONWe have demonstrated that four repeated

DTH reactions elicited by MLM antigen injectedat the site of infection did not affect the multipli-cation of MLM bacilli at this site or in thedraining popliteal lymph node. The DTH reac-tions thus caused no detectable killing of thebacilli. Also, there was no detectable bacterio-static effect. Some local reactivity was presentfor 5 weeks, i.e., close to twice the doublingtime in the footpad and five times the doublingtime in the popliteal lymph node, and bacterio-stasis for this interval should have been detect-able. In C57BL mice, the footpad swelling thatcoincides with the onset of inhibition of bacillarymultiplication sometimes may be as low as 0.10to 0.20 mm (Livik and Closs, unpublished data),but as a rule the response to the dose of bacilliused in the present experiment is considerablystronger, long lasting, and not transient like thesonicate-induced reactions (12). If effective bac-teriostasis occurred for only 12 days during theacute, strong reactions, this would be difficult todetect in the footpad. In the popliteal lymphnode, where the doubling time was 1 week,bacteriostasis for 12 days would be detectable.Thus, we may conclude that there was no bacte-riostatic effect in the popliteal lymph node andprobably not in the footpad either.

Increased resistance against infection at thesite of a DTH reaction elicited with homologousor heterologous antigen has been found in ex-periments in which Listeria monocytogenes wasused for challenge (17, 30, 32). To our knowl-edge similar experiments with mycobacteria asthe challenge organisms have not been reportedpreviously. The choice of challenge organismmay be of crucial importance for the outcome ofthe experiment. L. monocytogenes has beenfound to be very easily inactivated by activatedmacrophages (27, 28), whereas Mycobacteriumtuberculosis, MLM, and probably mycobacteriain general are relatively resistant to inactivationby cell-mediated immune mechanisms (23, 33).It has been found that whereas M. tuberculosisin macrophage cultures was inhibited after expo-sure of the cell culture to a lymphokine prepara-tion for 72 h, there was no inhibition of bacillarygrowth after exposure for 48 h (5). The suscepti-bility of M. tuberculosis to being killed by heat isdependent on its growth phase (38), and this mayapply also to being killed by an immune reac-tion. The generation time for L. monocytogenesis 4 to 8 h, that for M. tuberculosis is 24 h (27,28), and that for MLM probably is in the order of

1 to 2 weeks (6). It may well be that a microor-ganism with a long generation time requires acorrespondingly long exposure to the inhibitorymechanism to be affected. In our experiment,each of the four local reactions that were strongfor only 3 days may thus have failed to producemacrophages that were highly activated for along enough time to have an effect on themultiplication of MLM.The mechanisms in protective immunity

against MLM may even be qualitatively differentfrom those mediating protective immunityagainst L. monocytogenes. Mycobacteria tendto go into a dormant state instead of being killedby the host's immune apparatus (14), and thismay apply also to MLM. The mechanisms thatkill L. monocytogenes are not necessarily themechanisms that put mycobacteria into a dor-mant state. Various effector functions can bedissociated in the macrophage population, eitherbecause different lymphokines induce qualita-tive differences in the activated macrophages orbecause there are different subpopulations ofmacrophages having different functions (31, 40).Although Patel (32) found that resistance againstL. monocytogenes was always increased at theinjection site of a specific antigen into an immu-nized animal, he found no correlation betweenthe magnitude of the DTH reaction thus inducedand the level of resistance, and there was in-creased resistance even when no DTH could bedetected. Thus, there was a dissociation be-tween DTH reactivity and the degree of protec-tion against L. monocytogenes after injection ofspecific antigen. Furthermore, evidence hasbeen found that tuberculin hypersensitivity ismediated by different lymphokines than thosemediating tuberculoimmunity (16), and it hasbeen demonstrated that macrophages can at thesame time be responsive to lymphokines mediat-ing DTH and unresponsive to lymphokines me-diating inhibition of M. bovis BCG multiplica-tion (2). In our experiment there might havebeen a selective failure of the C3H macrophagesto respond to the lymphokines mediating resist-ance against MLM. There might also have beena selective failure of the lymphocytes to producethese lymphokines because C3H lymphocytesare genetically deficient or because the immuno-genic stimulus was inappropriate.C3H mice do not develop a local lymphocyte

reaction against MLM bacilli during infection,and they do not develop protective cell-mediat-ed immunity (6, 9, 10). They develop a humoralimmune response against MLM antigens (11),but it is not known whether those particularantigens are T dependent. However, if DTHagainst MLMSon is mediated by T lymphocytes,the presence of DTH shows that MLM antigen-reactive T lymphocytes are present in C3H

INFECT. IMMUN.

RESISTANCE AT DTH REACTION SITE 773

mice. Therefore, if there is a defect in the T-cellantigen repertoire of C3H mice, it is limited to arestricted number of antigens, among which areto be found those responsible for the inductionof protective immunity.Our MLMSon has been shown in crossed

immunoelectrophoresis to contain more than 40antigenic components (7). Recognition of solu-ble protein antigens by the T lymphocytes isusually preceded by an initial uptake and proc-essing of antigen by antigen-presenting cells,and macrophage-like cells seem to have an oblig-atory role in antigen-specific T-cell activation(34, 37). The induction of DTH againstMLMSon shows that C3H macrophages canpresent at least some of the MLM antigens in thesonicate to the lymphocytes. The transient na-ture of the strong reactions in the present experi-ment, together with the lack of local reactivityafter inoculation of 2.5 x 107 MLM bacillisuspended in saline into mice similarly immu-nized with MLMSon (25), indicates that thelocal reactivity in the mice given bacilli in soni-cate was directed against antigens in the soni-cate and not against the bacilli. Infected C3Hmacrophages thus do not normally present theantigens that are active in sonicate-inducedDTH. It is not likely that the lack of responseagainst bacilli was caused by nonspecific sup-pressor cells: DTH against sonicate at the site ofinfection was not suppressed.The effector step in T-cell-induced macro-

phage killing is thought to be nonspecific: onceactivated, the macrophages kill microorganismsregardless of the specificity of the inducingimmunological reaction (4). However, the exactmechanisms for MLM inhibition and killing arenot known, and the participation of antigen-specific cells or soluble mediators in the lasteffector step cannot be completely ruled out.Experiments with C57BL mice indicate that theantigens responsible for the induction of protec-tive immunity are not functionally present in theMLMSon preparation (12). Therefore, there is apossibility that the DTH reaction in our experi-ment did not inhibit the growth ofMLM becauseit lacked certain specificities or that specificsuppressor cells were induced which eliminatedthe response to certain crucial antigens.The tuberculin-type DTH has a strong associ-

ation with protective immunity in tuberculosis,MLM infection, and challenge with neoplasticcells. Other types of DTH, namely, the Jones-Mote-type DTH and Jones-Mote-type DTHmodified by cyclophosphamide pretreatment,have been found to be not so strongly associatedwith protective immunity (1, 17, 18, 22). Sincethe MLMSon preparation we used for immuni-zation was reconstituted with the insoluble ma-terial after sonication, the emulsion we used for

immunization probably had the adjuvant proper-ties of Freund complete adjuvant rather than ofFreund incomplete adjuvant. MLMSon-inducedDTH was long-lived, and in the initial test (Fig.1) the kinetics of the reaction was very compati-ble with a tuberculin-type reaction (35). Also,the fibrinoid deposits seen in the histologicalsections indicate that the reaction after inocula-tion of bacilli in MLMSon was a tuberculin-typeDTH reaction, although the 24- to 48-h relation-ship was intermediate between tuberculin-typeand Jones-Mote-type DTH. It is important to beaware that the retest reactions of the tuberculintype are accelerated and have altered kinetics (3,13, 36). The kinetics as well as the histologicalpicture of the reaction after inoculation of bacilliin MLMSon argue against a significant contribu-tion of Arthus reactivity. It is possible that someArthus reactivity was later induced by the re-peated injections of MLMSon and occurred to-gether with the DTH (3), but the increased 4-hreactions seen after the three last injections canbe fully explained by the altered kinetics of thetuberculin retest reaction (3).

ACKNOWLEDGMENTSThis work was supported by the Heiser Fellowship Program

for Research in Leprosy, the Norwegian Research Council forScience and the Humanities, the Immunology of Leprosy(IMMLEP) component of the UNDP/World Bank/WHO Spe-cial Programme for Research and Training in Tropical Dis-eases, Professor Carl Semb's Medical Research Fund, andAnders Jahre's Fund for the Promotion of Science.The skillful technical assistance of Ellen S. Karlstrslm,

Grethe S. Fuglestad, and Ase E. Strutz is gratefully acknowl-edged. We thank Olav A. Haugen for help with the histology.

LITERATURE CITED1. Alexander, J., and J. Curtis. 1979. Development of de-

layed hypersensitivity responses in Mycobacterium le-praemurium infections in resistant and susceptible strainsof mice. Immunology 36:563-567.

2. Anderson, D. W., and A. J. Crowle. 1981. Evans blue dyeadjuvant enhances delayed hypersensitivity while block-ing immunity to Mycobacterium tuberculosis in mice.Infect. Immun. 31:413-418.

3. Arnason, B. G., and B. H. Waksman. 1963. The retestreaction in delayed sensitivity. Lab. Invest. 12:737-747.

4. Blanden, R. V., G. B. Mackaness, and F. M. Collins. 1966.Mechanisms of acquired resistance in mouse typhoid. J.Exp. Med. 124:585-600.

5. Cahall, D. L., and G. P. Youmans. 1975. Conditions forproduction, and some characteristics, of mycobacterialgrowth inhibitory factor produced by spleen cells frommice immunized with viable cells of the attenuated H37Rastrain of Mycobacterium tuberculosis. Infect. Immun.12:833-840.

6. Closs, 0. 1975. Experimental murine leprosy: growth ofMycobacterium lepraemurium in C3H and C57/BL miceafter footpad inoculation. Infect. Immun. 12:480-489.

7. Closs, O., M. Harboe, and A. M. Wassum. 1975. Cross-reactions between mycobacteria. I. Crossed immunoelec-trophoresis of soluble antigens of Mycobacterium leprae-murium and comparison with BCG. Scand. J. Immunol.4(Suppl. 2):173-185.

8. Closs, O., and 0. A. Haugen. 1974. Experimental murineleprosy. 2. Further evidence for varying susceptibility of

VOL. 36, 1982

774 L0VIK AND CLOSS

outbred mice and evaluation of the response of 5 inbredmouse strains to infection with Mycobacterium leprae-murium. Acta Pathol. Microbiol. Scand. Sect. A 82:459-474.

9. Closs, O., and 0. H. Haugen. 1975. Experimental murineleprosy. 3. Early local reaction to Mycobacterium leprae-murium in C3H and C57/BL mice. Acta Pathol. Microbiol.Scand. Sect. A 83:51-58.

10. Closs, O., and 0. A. Haugen. 1975. Experimental murineleprosy. 4. The gross appearance and microscopic fea-tures of the local infiltrate after subcutaneous inoculationofC3H and C57/BL mice with Mycobacterium lepraemur-ium. Acta Pathol. Microbiol. Scand. Sect. A 83:59-68.

11. Closs, O., and G. Kronvall. 1975. Experimental murineleprosy. IX. Antibodies against Mycobacterium leprae-murium in C3H and C57BL mice with murine leprosy andin patients with lepromatous leprosy. Scand. J. Immunol.4:735-740.

12. Closs, O., and M. Llvik. 1980. Protective immunity anddelayed-type hypersensitivity in C57BL mice after immu-nization with live Mycobacterium lepraemurium and soni-cated bacilli. Infect. Immun. 29:17-23.

13. Gogi, R., A. H. S. Rahi, and A. Garner. 1979. The natureof retest reaction in delayed hypersensitivity. I. Lightmicroscopic changes in the skin of the eyelid. Histopa-thology 3:51-67.

14. Hart, P. D., and R. J. W. Rees. 1960. Effect of macrocy-clon in acute and chronic pulmonary tuberculous infectionin mice as shown by viable and total bacterial counts. Br.J. Exp. Pathol. 41:414-421.

15. Kawaguchi, Y. 1959. Classification of mouse leprosy. Jpn.J. Exp. Med. 29:651-663.

16. Klun, C. L., R. G. Neiburger, and G. P. Youmans. 1973.Relationship between mouse mycobacterial growth-in-hibitory factor and mouse migration-inhibitory factor insupernatant fluids from mouse lymphocyte cultures. RESJ. Reticuloendothel. Soc. 13:310-318.

17. Lagrange, P. H. 1979. Role of the macrophages in immuni-ty to microbial infection. Transplant. Clin. Immunol.10:255-261.

18. Lagrange, P. H., and P. M. Thickstun. 1979. In vivoantitumor activity of various forms of delayed-type hyper-sensitivity in mice. J. Natl. Cancer Inst. 62:429-436.

19. Lefford, M. J. 1975. Delayed hypersensitivity and immu-nity in tuberculosis. Am. Rev. Respir. Dis. 111:243-246.

20. Lefford, M. J., P. J. Patel, L. W. Poulter, and G. B.Mackaness. 1977. Induction of cell-mediated immunity toMycobacterium lepraemurium in susceptible mice. Infect.Immun. 18:654-659.

21. Lempert, H. 1944. Fluorescence microscopy in the detec-tion of tubercle bacilli. Lancet 247:818-822.

22. Lenzini, L., P. Rottoli, and L. Rottoli. 1977. The spectrumof human tuberculosis. Clin. Exp. Immunol. 27:230-237.

23. Levy, F. M., G. A. Conge, J. F. Pasquier, H. Mauss, R. J.Dubos, and R. W. Schaedler. 1961. The effect of BCGvaccination on the fate of virulent tubercle bacilli in mice.Am. Rev. Respir. Dis. 84:28-36.

24. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folin phenolreagent. J. Biol. Chem. 193:265-275.

25. L$vik, M., and 0. Closs. 1980. Delayed type hypersensi-tivity to mycobacterial antigens without protective immu-

nity: a failure to produce the right specificity or the righttype of immune reaction? Scand. J. Infect. Dis. Suppl.24:224-227.

26. LwEvik, M., and 0. Closs. 1981. Effect of BCG vaccinationon Mycobacterium lepraemurium infection in a highlysusceptible inbred mouse strain. Acta Pathol. Microbiol.Scand. Sect. C 89:133-138.

27. Mackaness, G. B. 1962. Cellular resistance to infection. J.Exp. Med. 116:381-406.

28. Mackaness, G. B. 1964. The behaviour of microbial para-sites in relation to phagocytic cells in vitro and in vivo.Symp. Soc. Gen. Microbiol. 14:213-240.

29. Mackaness, G. B. 1969. The influence of immunologicallycommitted lymphoid cells on macrophage activity in vivo.J. Exp. Med. 129:973-992.

30. Mitsuyama, M., K. Nomoto, H. Akeda, and K. Takeya.1980. Enhanced elimination of Listeria monocytogenesat the site of delayed footpad reaction. Infect. Immun.30:1-4.

31. Nacy, C. A., E. J. Leonard, and M. S. Meltzer. 1981.Macrophages in resistance to rickettsial infections: char-acterization of lymphokines that induce rickettsiacidalactivity in macrophages. J. Immunol. 126:204-207.

32. Patel, P. J. 1980. Expression of antibacterial resistance atthe site of a delayed hypersensitivity reaction. Infect.Immun. 29:59-65.

33. Rees, R. J. W., and P. D. Hart. 1961. Analysis of the host-parasite equilibrium in chronic murine tuberculosis bytotal and viable bacillary counts. Br. J. Exp. Pathol.42:83-88.

34. Rosenwasser, L. J., and A. S. Rosenthal. 1978. Adherentcell function in murine T lymphocyte antigen recognition.I. A macrophage-dependent T cell proliferation assay inthe mouse. J. Immunol. 120:1991-1995.

35. Turk, J. L., L. Polak, and D. Parker. 1976. Controlmechanisms in delayed-type hypersensitivity. Br. Med.Bull. 32:165-170.

36. von Pirquet, C. F. 1909. Quantitative experiments with thecutaneous tuberculin reaction. J. Pharmacol. Exp. Ther.1:151-174.

37. Waldron, J. A., R. G. Horn, and A. S. Rosenthal. 1973.Antigen-induced proliferation of guinea pig lymphocytesin vitro: obligatory role of macrophages in the recognitionof antigen by immune T-lymphocytes. J. Immunol.111:58-64.

38. Wallace, J. G. 1961. The heat resistance of tubercle bacilliin the lungs of infected mice. Am. Rev. Respir. Dis.83:866-871.

39. White, C. 1952. The use of ranks in a test of significancefor comparing two treatments. Biometrics 8:33-41.

40. Wing, E. J., I. D. Gardner, F. W. Ryning, and J. S.Remington. 1977. Dissociation of effector functions inpopulations of activated macrophages. Nature (London)268:642-644.

41. World Health Organization. 1973. Cell-mediated immunityand resistance to infection. W. H. 0. Technical ReportSeries no. 519. World Health Organization, Geneva.

42. Youmans, G. P. 1975. Relation between hypersensitivityand immunity in tuberculosis. Am. Rev. Respir. Dis.111:109-118.

43. Youmans, G. P. 1979. Tuberculosis, p. 285-301. W. B.Saunders Co., Philadelphia.

INFECT. IMMUN.