35'

THE ROLE OF LYMPHOID TISSUE AND THEADRENALS IN IMMUNITY

By MARY BARBER and ALBERT DELAUNAYFrom the Pasteur Institute, Paris, and St. Thomas's Hospital Medical School, London

I. INTRODUCTIONII. THE ROLE OF LYMPHOID TISSUE

I. Lymphocytes and the Production ofAntibodies

2. Plasma Cells and the Production ofAntibodies

3. Effect of Destruction of LymphoidTissue on Antibody Formation

I. INTRODUCTIONAccording to the theory of Ehrlich (I898, see

also Topley, I933) all cells have ' side chains' orreceptors ' for the attachment of nutrient or other

substances. In his view, antigens become attachedto these receptors which, since the antigen isforeign to the cell, then lose their normal function.This stimulates the production of new receptors,which are shed and constitute specific antibodies.This theory would seem to imply that any cell iscapable of producing antibodies if an antigenbecomes attached to it. The majority of bacterio-logists, however, have looked on antibody forma-tion as the property of only certain types of celland, until recently, the reticulo-endothelial systemwas regarded as the major site of antibodyformation.

Following the work of McMaster and Huddack(I935), attention has been turned to the role oflymphocytes in the production of antibodies. Ofthese cells, described by Yoffey (1950) as the' mystery cells of mammalian tissues,' we knowneither the function nor the fate. In the lastdecade, however, a great deal of work has appearedon the pituitary-adrenal control of lymphoid tis-sue and the possible part played by the lymphocytein bacterial infection and immunity.

Outstanding in this field has been the work ofDougherty, White and their colleagues. As theresult of their investigations, these workers.(Dougherty and White, I947) have concluded that

lymphocytes are a major source of antibodies andthat both lymphocytes and antibody formation areunder pituitary-adrenal control. Thus, in theopinion of these investigators, the injection ofadreno-cortical hormones is followed by involutionof lymphoid tissue, dissolution of lymphocytesand the release of normal or immune gammaglobulin.

These far-reaching conclusions are by nomeans universally accepted. Many investigators,approaching the problem from widely differentangles, have entered the field. The literature onthe subject has become vast and appears in adiversity of journals. The present review is anattempt to bring together the experimentalevidence and, where possible, assess its value.

II. LYMPHOID TISSUEI. Lymphocytes and the Production of

AntibodiesMcMaster and Huddack (1935) published the

first experimental evidence for the production ofantibodies in lymph nodes. These workers foundthat when heat-killed suspensions of bacteria wereinjected intradermally into the ears of mice,agglutinins appeared in the lymph nodes drainingthe corresponding ear about seven days later.At this time agglutinins had begun to make theirappearance in the blood stream, but usually at alower titre. No agglutinins were found in thelymph gland draining the opposite ear until aboutthe twelfth day. Amputation of the injected eartwo hours after injection did not affect the result.Two years later McMaster and Kidd (I937)showed a similar early production of antivacciniavirus principle in the lymph nodes draining theear after injection of this antigen.These observations have been greatly extended

by the work of Ehrich and Harris, T. N., and theircolleagues. In early studies Ehrich (Ehrich andWohltrab, 1934, and Ehrich and Voigt, 1934)

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

failed to find a correlation between reticulo-endothelial proliferation and antibody titres fol-lowing the injection of a staphylococcal vaccine,whereas the rise in antibody titre was found toparallel the activity of the Malphighian bodies ofthe spleen. Since the active cells in the latterwere thought to be lymphoblasts, this led to aseries of investigations into the production ofantibodies by lymph glands. Ehrich and Harris,T. N. (1942), working with rabbits, injectedtyphoid vaccine into the left hind-foot-pad andsheep's red blood cells into the right. The sub-sequent appearance of agglutinins and haemolysinsin the popliteal lymph node and afferent andefferent lymph was compared with histologicalchanges in the lymph node and the output oflymphocytes. Antibody tities in the foot tissue andafferent popliteal lymph never reached significantlevels. In the efferent lymph on the correspond-ing side antibodies began to appear in two to fourdays and reached their maximum level after sixdays. Their appearance was preceded by a sharprise in the output of lymphocytes through theefferent lymph and. by lymphatic hyperplasia inthe lymph gland. Harris, T. N., et al. (1945) gaverabbits similar injections and five, six or sevendays later collected lymph from the efferentlymphatic of the popliteal nodes. The antibodycontent was estimated for whole lymph and forlymph plasma and cell extract separately. Forthe latter, the lymphocytes were separated bycentrifuging, suspended in saline and extractedby freezing and thawing three times. In everycase the cell extract gave higher antibody titresthan the plasma, and the average difference wasfour-fold. When lymphocytes containing onetype of antibody were suspended in plasma withthe other type there was no uptake of antibodiesby the lymphocytes but a release of antibodies.from the cells into the suspension. The authors,therefore, concluded that the lymphocytes did notsimply absorb antibodies but were concerned intheir formation. Essentially similar results wereobtained by Harris; S., and Harris, T. N. (1949),when influenza virus was used as the antigen. Incontrast to the high antibody titres occurring inthe efferent lymph of the lymph, node drainingthe site of injection only insignificant titres werefound in the tissue at site of injection, althoughthis contained many macrophages (Ehrich et al.,1946).

It was assumed from the above experimentsthat the lymphocyte was concerned with the pro-duction of antibodies. Since, however, lympho-cytes are not phagocytic cells, it occurred toHarris, T. N., and Ehrich (1946) that particulateantigens must be converted to a soluble formbefore they are available to the lymphocytes.

They thus investigated the production of antigen-specific soluble material after the injection ofsheep's red blood corpuscles or dysentery vaccine.They identified such a substance in the tissue atsite of injection, the draining lymph node and theefferent lymph. The quantity diminished slowlyin the infected tissue, but disappearea rapidly intwo to three days in lymph nodes and efferentlymph, its disappearance being succeeded by theappearance-of antibodies.

In further experiments Harris, Tr. N., et. al.(1948) studied the role of the thymus and spleenin antibody production. It was found thatremoval of the thymus did not affect the produc-tion of antibodies, and that when present thethymus only contained very low titres. Thespleen also played little part in antibody formationafter subcutaneous injection of an antigen, but ifthe latter were injected intravenously, where thespleen probably received the bulk of the antigenadministered, comparatively high antibody titreswere found in the spleen

Considerable evidence has appeared to confirmthe work of Harris, T. N., and his colleagues.Thus Kass (1945) has demonstrated the presenceof gamma globulin in human lymphocytes obtainedfrom the mesenteric lymph nodes. Roberts et al.(1948) studied the antibody titres of various tissueminces taken from mice immunized against vac-cines and sheep's red blood corpuscles and in-cubated in vitro for three hours in serum fromnon-immunized mice. They found that pooledlymphoid tissue from the thymus and mesentericglands gave higher antibody titres than othertissues. Dougherty et al. (i944), working withmice, recorded the presence of immune globulinin the lymphocyte of lymph glands after immuniza-tion with sheep erythrocytes and in I945 recordedthe presence of antibodies to staphylococcal fil-trates in normal and malignant (lympho-sarcoma)lymphocytes. Furthermore, a very large amountof evidence on the lymphocyte in relation toimmunity has accumulated from a study of theaction of adreno-cortical hormones and will bediscussed later.Not all workers, howeyer, are in agreement with

the above -observations. Spear (I948) in a dis-cussion of technique pointed out how easily resultsin such experiments could be distorted by quitesmall carry-overs of fluid containing antibodies.Craddock et al. (I949) studied thoracic duct lymphin cats immunized with typhoid vaccine. Theyfailed to find any antibodies in. cell extracts,although supernatant lymph plasma and bloodserum of all animals had demonstrable titres.Habel et al. (1949), using a technique similar tothat of Ehrich and Harris, T. N. (1942) and asuspension of S. typhi-murium or Past.- tularensis

352 Yune I1952by copyright.

on Decem

ber 24, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.320.351 on 1 June 1952. D

ownloaded from

BARBER and DELAUNAY: The Role of Lymphoid Tissue

as antigen, found slightly higher antibody titresin the popliteal gland on the inoculated side thanthat on the control side, but they did not regardthe difference as significant. Harris, S., andHarris, T. N. (I950), in reply to Craddock et al.,point out that the thoracic duct is likely to containmany lymphocytes from nodes having no accessto the injected antigen and also that lymphocytesfrom antibody-forming nodes may have yieldedtheir antibodies into the lymph plasma by thetime they reach the thoracic duct. With regardto the work of Habel et al., Harris, S., and Harris,T. N. (I950) suggest that the dose of antigen ofthese workers was too large to get optimal dif-ferences. This paper (Harris, S., and Harris,T. N., I95o) also describes further experimentsshowing the importance of careful technique foraccurate results.

2. Plasma Cells and the Production ofAntibodies

Recently a number of papers have appeared,notably from Scandinavian workers, indicatingthat the plasma cell or its precursor, not themature lymphocyte, is the main source of antibodyproduction. The ball was set rolling by Bingand Plum, who in 1937 advanced the hypothesisthat plasma cells were globulin producers. Thiswas based on clinical observations of the correlationbetween hyperglobulinaemia and plasma cell pro-liferation in diseases, not including myelomatosis,such as various infections and some cases ofleucopenia.

Bjorneboe and Gormsen (1943) injected rabbitsintravenously with massive doses of polyvalentvaccines from mixed bacteria or a mixture offoreign proteins, every other day for up to twomonths. The animals were killed at variousintervals for histological studies. From the 3rdto the 5th day of immunization, moderate plasmacell proliferation was seen in the spleen, lungsand liver. From the 6th day plasma cell pro-liferation was present in spleen, lungs, liver, kid-neys, suprarenals, lvmph glands and bone marrow.This proliferation increased until about the 13thday, after which it remained constant until be-tween the 3Ist and 68th day. In the spleen theinfiltration was massive, the pulp being densewith cells, almost all of which were plasma cells,interspersed with a few reticulum cells. Therewas no hyperplasia of follicles. Many ' plasmo-cellular reticulum' cells were seen, favouring thehypothesis that the plasma cells originated fromreticulum cells. Massive plasma cell proliferationoccurred, however, in the reticulum cell-freeadipose tissue. The relation between titre ofantibody formation and plasma cell proliferationwas proportional. Thus there was an almost

constant relation between antibody formation andsplenic weight. In conclusion, the authors suggestthat the failure of earlier workers to note plasmacell proliferation in immunity was because im-munization was not sufficiently massive. Infurther similar experiments, Bjorneboe et al. (I947)showed that during immunization plasma cellproliferation occurred in almost all organs exceptthe thymus, and that the latter showed only verysmall amounts of antibody. In a review of allthe work of these Danish authors Gormsen (1950)concludes that the plasma cells are antibody-producing functional stages of reticulo-endothelialcells.More recently, Ringertz and Adamson (I950)

and Marshall and White (1950) have shown asimilar plasma cell proliferation of the organs ofguinea-pigs and rabbits following the injectionsof a large variety of particulate and non-particulateantigens. Both these groups of workers alsonoted some hyperplasia of lymphoid tissue andthe former suggest that, under the influence of anantigen, immature lymphocytes may be convertedinto antibody-producing plasma cells, instead ofdeveloping into mature lymphocytes.'The role of the plasma cell in immunity has

been further amplified by the in vitro studies of'Fagraeus (I947, 1948 and 1950). During thesecondary response, elicited in rabbits by intra-venous injections of living S. typhi, this inves-tigator noted a great increase in plasma cells inthe red pulp of the spleen simultaneously with theincrease of circulating antibodies. The cellsappeared to arise from reticulum cells, passingthrough a chain of development from transitionalcell, to immature plasma cell to mature plasmacell. Fagraeus excised portions of spleen atvarious intervals and studied their capacity toform antibodies in vitro in tissue culture. Shefound that the red pulp, rich in plasma cells,produced far more antibodies than the lymphfollicles, and that antibody production was maxi-mum when transitional cells had been largelyreplaced tby immature plasma cells, which was thetime at which antibody formation in viivo wasrising or at its maximum, and receded whenmature plasma cells predominated. She alsonoted that the thymus, which contained no plasmacells, was incapable of forming antibodies in vitro.

This work is confirmed by Keuning and Stikke(1950), who in similar studies conclude that therole of the immature plasma cell in antibodyproduction is' beyond doubt.' The latter workers,however, although they repudiate the idea thatmature lymphocyte. produce antibodies, suggestthat certain lymphoblastic cells of the Malpighiancorpuscles may be involved in antibody produc-tion. These authors draw attention to the fact

Yune I1952 353

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

that Epstein in 1929, in a study of the histologicalpicture of immunity, noted that phagocyticrietculo-endothelial cells give rise to aggregates ofnon-phagocytic round cells resembling plasmacells and other lymphoid elements.

Since the observations of the Scandinavianworkers, Harris, T. N., and their colleagues havemade a number of histochemical studies in rabbitsof the lymph glands draining areas injected withvarious antigens (see Harris, T. N., and Harris,S., I948 and I949; Ehrich et al., I949; andForman et al., 1949). Following the observationof Casperson (I94) that cells actively formingnew protein are characterized by large amounts ofribonucleic acid in their cytoplasm, these workersstudied these lymph nodes stained with methyl-green and pyronine to demonstrate the presenceof desoxyribonucleoprotein (stained green) andribonucleoprotein (stained red).

Harris, T. N., and Harris, S. (1948 and 1949)found that in lymph nodes actively engaged inantibody formation, pyronine staining material(presumably ribonucleic acid) appeared first incells resembling reticulum cells, then in youngimmature lymphocytes and then later in cellswith progressively smaller and denser nuclei. Noother cell types contained such material. Whenthe concentration of ribonucleic acid was deter-mined in aqueous extracts of the lymph nodes itwas found that the peak concentration occurredjust before the appearance of the maximum con-centration of antibodies in the nodes. Theseworkers thus adhere to their view that the lympho-cyte is the primary source of antibodies. Theypoint out that the cells containing ribonucleicacid appear to form a continuous series fromtransitional reticulum cells to mature lymphocytesand suggest that antigenic stimulation offers ameans of studying lymphocytopoiesis.

Ehrich et al. (i949) and Forman et al. (I949)found that the cellular reaction of the lymphglands during the first six days after injection ofan antigen consisted chiefly of plasma cells, firstplasmoblasts and then mature plasma cells pre-dominatihg. Most of the ribonucleoprotein asshown by pyronine staining was contained inthese cells. Lymphocytes began to proliferatelater, showing their greatest activity on the gthday, when ribonucleoprotein and antibody for-mation had passed their peak. Thus these latterauthors now agree with the Scandinavian workersthat plasma cells and not lymphocytes areresponsible for antibody formation.

3. Effect of Destruction of Lymphoid Tissueon Antibody Formation

In 193I, Dustin demonstrated that the chromo-somes of certain cells of the animal body were,

peculiarly sensitive to a number of noxious agents.Among the most readily affected cells were corticalthymocytes, the lymphocytes occurring in ger-minal centres and the cells of the crypts of Lieber-kuhn in the small intestine. Destruction of thesecells or inhibition of mitosis was readily producedby such agents as colchicine, acridines or nitrogenmustards, or by irradiation.The effect of splenectomy on the action of these

poisons was studied by Dustin and Zylberszac(I93I), who found that such a procedure rendered,the thymocytes resistant to the action of trypa-flavine, whereas the lymphocytes and the cells ofthe small intestine remained equally sensitive tothis compound after removal of the spleen. Theseauthors, therefore, regarded thymocytes andlymphocytes as being different types of cell inspite of their structural similarity. This is animportant point in view of the observations ofHarris, T. N., et al. (I945), Bjorneboe et al. (1947)and Fagraeus (I947) that the thymus does notproduce antibodies.

Miller (1950) showed that adrenalectomy in-creased the resistance of thymocytes and lympho-cytes of rats to colchicine, podophyllotoxin andnitrogen mustard, although the animals them-selves were more sensitive to these compoundsafter adrenalectomy. For a detailed study of theeffect of nitrogen and sulphur mustards onlymphopoietic organs the reader is referred toKindred (I947)-

Zylberszac (193i) drew attention to the factthat most of the chemotherapeutic agents then in.use (e.g. acridines, arsenicals, Bismuth and mer-curial salts) as well as most bacterial toxins causedrapid destruction of thymocytes and lymphocytesand suggested that this action played a part inthe overcoming of an infective agent. Favour(I947) studied the effect of bacterial products -on.lymphocytes more carefully and showed thatlymphocytes from tuberculous mice were speci-cally lysed in vitro by tuberculin, whereas lympho-cytes from normal mice or mice infected withsalmonellae were not. Similarly, lymphocytesfrom mice infected with salmonellae were speci-fically lysed by a suspension of the same organisms.The work of later investigators, however, makes itclear that these cells can be destroyed in vivo bya wide variety of different factors and Selye (937)regards a caryoclastic crisis of thymus and lymphglands as one of the numerous non-specific lesionsproduced by almost any noxious agent. In 1947,Ickowicz demonstrated that such a crisis is anearly and sensitive response of the animal tosurgical injury.The idea that leucotoxic agents might affect

antibody response is not a new one. Thus Rusk(1915), Simmons and Jones (I9I5) and Hektoen

_7une I1952354

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

BARBER and DELAUNAY. The Role of Lymphoid Tissue

(I9I6) all showed' that intoxication with benzeneled to leucopenia and a depression of circulatingantibodies. In 1947, Philips et al., in some carefulstudies on goats immunized against ricin, demon-strated that the anamnestic response of theseanimals to a further injection of ricin was alwaysdelayed and sometimes decreased by leucopenicdoses of nitrogen mustards. Spurr (I947) foundthat similar treatment decreased the antibodyresponse of rabbits to typhoid vaccine. Marshalland White (I950), however, failed to demonstrateany such effect on the response of rabbits to theinjection of various vaccines and non-particulateantigens. Approaching the problem from anotherangle, Karnoffsky et al. (I948) studied the growthof Strep. pyogenes in the serum of normal rabbitsand rabbits who had previously been given intra-venous injections of a nitrogen mustard. Theyfound that the normal serum strongly inhibitedthe growth of these organisms, whereas serumtaken six or nine hours after injection of thenitrogen mustard was much more favourable totheir growth.Many studies have also appeared on the effect

of X-irradiation on antibody formation. Hektoen(I915) first demonstrated that total body exposureto ionizing irradiation suppressed the antibodyresponse to antigens injected shortly before orshortly after irradiation. Jacobson (I949) subjectedyoung adult rabbits to 8oo r. total body X-radiation(a dose which destroys nearly all the lymphoidtissue of the body) and 24 hours later gave them'an intravenous injection of sheep's red blood cells.Such animals produced either no haemolysins orvery low titres. If the spleen or appendix weresurgically mobilized and lead-protected prior tothe irradiation, essentially normal haemolysin-production occurred, even though lymphatic tissuein the unprotected areas of the body was largelydestroyed.

Evans (I948) studied the response of humanswith abnormal lymphocyte pictures to paratyphoidB vaccine. A group of 30 patients with radiation-induced lymphopenia had lower antibody titresand produced them more slowly than normalpeople. Very poor antibody responses occurredin cases of lymphatic leukaemia, Hodgkin's disease,lymphosarcoma and lymphofollicular reticulosis.In myeloid leukaemia, however, a disease in which-the lymph glands retain 'their morphologicalfunction to a greater extent, the antibody responsewas normal.

Recent work in connection with cortisone (see'below) has suggested that not only is lymphoidtissue concerned with the production of antibodiesbut that the destruction of adult lymphocytes inimmunized animals may lead to the liberation ofantibodies into the circulation. As early as '1925

Murphy and $turm found that if rabbits wereexposed to dry heat before and after injection ofan antigen, a marked lymphocytosis occurred, andwas associated with a temporary increase in serumantibody titre. In I945, Dougherty, Chase andWhite found that the administration of smalldoses of benzene increased the circulating anti-bodies in intact, but not in adrenalectomized, miceimmunized against sheep's red blood corpuscles,and attributed this to destruction of lymphocytesand consequent liberation of antibodies fromthese cells.

m. THE FOLE OF THE ADRENAL GLANDSx. Introduction2. Pituitary-Adrenal Control of

Lymphocytes(a) Lymphoid Tissue(b) Circulating Lymphocytes(c) Bone Marrow(d) Malignant Cells(e) Mode of Action and in vitro

studies(f) Effect of other Endocrine glands

3. Pituitary-Adrenal Control of Anti-body Formation

IV. CONCLUSIONIII. THE ROLE OF THE ADRENALS

z. IntroductionThe idea that the adrenal glands are affected by

bacterial diseases is no new one. Thus, Roux andYersin (I889) and many others have drawn atten-tion to the hyperaemia of the adrenals followingadministration of diphtheria toxin. Equally oldis the notion that abnormalities of the adrenalgland have an effect on lymphoid tissue. ThusAddison (i855) recorded lymphoid hyperplasia ina patient with suprarenal deficiency and Star(I95) observed that the thymus in a seventeen-year-old girl dying of Addison's disease was aslarge as that of a twelve-month-old child. Therecent isolation of cortico-steroids with adreno-cortical activity has led to a host of more detailedinvestigations of the effect of this gland on bothlymphoid tissue and antibody response.A relationship between the suprarenals, ' shock'

and lymphoid tissue has now been established.Thus Selye (I937) and Selye and Foglia (I939)studied the non-specific response of the organismto injury, i.'e. the so-called' alarm reaction.' Thisthey found was characterized by hypertrophy ofthe adrenal cortex, involution of the thymus andlymphoid tissue, degeneration and death of cellsin various tissues and ulcer formation in the

"In 1-952 355

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

digestive tract, together with various biochemicalchanges in the blood and urine. 'Adrenalectomyprior to the injury prevented the involution of thelymphoid tissue characteristic of the alarm re-action, but apparently increased some other signsof this reaction.

2. Pituitary-Adrenal Control of Lymphocytes(a) Lymphoid Tissue

Reinhardt and Holmes (I940) showed that inadrenalectomized male rats maintained on sodiumchloride for 45 days the thymus did not undergothe characteristic involution, and the spleen andglands were heavier than those of normalintact animals. Segal and Leblond (1938) foundthat if intact rats were subjected to X-irradiationof the abdomen all the lymphatic tissue of the bodyunderwent involution, but that if adrenalectomizedrats were similarly irradiated, the thymus andlymph glands other than the abdominal wereunaffected. Dougherty and White (I947) showedthat adrenalectomy prevented the lymphoid tissueinvolution usually produced by fasting.The injection of adreno-cortical extract (ACE)

or glucocorticoids in large doses has been shownto cause involution of the spleen and thymus inrats (Moon, i937 and I940; Ingle, I938 and 1940;and Aschkenasy et al., I950), mice (Doughertyand White, i944; Molumut et al., 1950; andAntopol, 1950) and rabbits (Dougherty and White,1944). Injections of pituitary adreno-trophichormone (ACTH) in large doses have similarlybeen shown to cause involution of thymus 'andlymph nodes of normal intact rats but notadrenalectomized animals (Crede and Moon, I940;Simpson et al., '943; and Dougherty and White,I944). Dougherty and White (I945b and I946b)made a careful study of the lymphoid tissue aftera single large injection of ACE or ACTH in miceand rabbits and recorded early dissolution ofmedium and small lymphocytes in thymus, lymphnodes, spleen and Peyer's patches. This wasassociated with an increased development of im-mature plasma cells and, as well as this, many ofthe smaller lymphocytes developed increasedlybasophilic cytoplasm. Aschkenasy et al. (1950)observed similar changes in the cells of the splenicfollicles. All these authors appear to agree thatthe changes in lymphoid tissue are maximal oneto six hours after injection and that within 24hours the picture has usually returned to normal.

Yoffey and his colleagues (195I) examined thecervical lymph nodes and thymus in guinea-pigssix hours after the injection of ACTH or eschatinand found no evidence of marked degenerativechanges, although in some instances the lymphoidmasses were less densely packed than usual.Their findings, therefore, are in sharp contrast to

the degenerative changes described by Doughertyand White (I945a, I946b and I947) in lymphoidtissue of mice and rabbits. It must be borne inmind, however, that guinea-pigs differ from rabbitsand mice in vitamin C metabolism, which mightaffect their reaction to adrenal hormones.Most of the above workers were studying the

effect of a single injection. Yoffey and Baxter(1946) found that repeated injections of largedoses of ACE in rats lead to slight but definitehypertrophy of lymphoid tissue. Dougherty andWhite (I947), on the other hand, working withmice, found that daily injections of adrenotrophini n such a large dose as i mg. led to a significantand persistent decrease in the weight of lymphoidtissue. It is, however, stated in this latter paperthat a secondary stimulation of lymphoid tissuemay follow the primary lymphocyte dissolution;thus under certain conditions leading to lymphoidtissue hypertrophy.

Injections of adrenaline have also been foundto cause involution of lymphoid tissue in guinea-pigs (Delaunay et al., 1949), and conditions whichcause a discharge of endogenous adrenaline suchas haemorrhage, emotional excitement or insulinhypoglycaemia apparently have a similar effect inrats (Gellhorn and Frank, 1948 and 1949; andZwecker, 1948). Gellhorn and Frank concludefrom their observations that adrenaline does notact directly on lymphoid tissue, but stimulatesthe secretion of adreno-cortical hormones. Thelymphoid tissue involution regarded by Selye(I937) and Selye and Foglia (I939) as characteristicof the ' alarm reaction' may result from a similarmechanism.

Herland (1950) found that the influence ofcorticoids on the lymphoid tissue of adrenalec-tomized rats was conditioned by stress, beingonly very slight in the absence of the latter.Dougherty and White (1947), however, found thatthe effect of ACE on normal and adrenalectomizedmice was indistinguishable, wvhereas, of course,ACTH was inactive in the absence of the adrenalglands.

(b) Circulating LymphocytesA circulating lymphocytosis following adrenalec-

tomy has been shown to occur in rats (Dalton andMasson, 1940) and rabbits (Murphy and Sturm,I947).. Crafts (I.94I) failed to find such a changein rats and Thatcher et al. (I948), working withcats, found that lymphocyte counts showed widevariations which had no relation to the presenceor absence of the adrenal gland. Edmadjian andhis colleagues found that stress in mice was fol-lowed by lymphopenia in intact but not inadrenalectomized animals (Edmagdjian and Pineus,1945) and that glucose administration had a similar

June I952356

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

BARBER and DELAUNAY: The Role of Lymphoid Tissue

effect in rats (Edmadjian et al., I946). In bothcases activation of the adrenal gland was regardedas the mode of action.Lymphopenia following the injection of ACE

has been observed in rats (Dougherty and White,I944; Yoffey et al., 1946; and Netter and Mathe,1949), rabbits (Dougherty and White, I944) andhens (Shapiro, I949). A similar change afterACTH has been reported in rats (Dougherty andWhite, I944; and Reinhardt et al., I944), dogs(Reinhardt et al., 1944), rabbits (Dougherty andWhite, I944; and Fischel et al., I949), hens(Shapiro, 1949) and humans (Hills et al., I948;Forsham et al., I949; and Herbert and de Vries,I949). All authors agree that the phenomenon istransitory and inconstant. Mason et al. (1948)treated a woman of thirty-two years with ACTHand observed no significant change in circulatinglymphocytes until the 22nd day when there wasa lymphopenia of three hours' duration.

In a general review and discussion of theirresults, Dougherty and White (1947) concludethat the lymphopenia following pituitary-adrenalhormones is a reflection of the lymphocyte dis-solution in lymphoid tissue and results, therefore,from a failure of delivery. In their investigations,repeated injections of ACE did not necessarilycause a continued lymphopenia and sometimeseven resulted in lymphocytosis. They suggestthat a balance exists between lymphocyte dis-solution, lymphocyte production and the removalof lymphocytes from the peripheral circulationand that dissolution might be a stimulus forincreased synthesis.

Another possible cause of lymphocytosis afterACE is the effect of the latter on the spleen. ThusHechter (I948) has shown that the addition ofACE to the spleen of rabbits in vitro leads to adischarge of lymphocytes. Furthermore, Stoneand Hechter (I948) showed that the administrationof ACE to adrenalectomized rats subjected to'swim-stress' was followed by lymphocytosis,provided the spleen was in situ, but not insplenectomized animals.The effect of ACTH or ACE on the lympho-

cytes of the thoracic duct is disputed. ThusReinhardt and Li (I945) found that in rats injec-tion of ACTH led to a rapid decrease in theirnumber occurring within I5 to 30 minutes andlasting for four to ten hours. They concluded thatthis resulted from a greatly reduced outpouring ofthese cells from the lymph glands, thus supportingthe view of Dougherty and White that the lympho-penia caused by this hormone results from afailure of delivery. Yoffey, Reiss and Baxter(I946) report a drop in thoracic duct lymphocytesin cats after injection of ACTH. Valentine et al.(1948), on the other hand, found no change in

lymphocyte counts of the thoracic duct lymph ofcats after injection of ACE.

(c) Bone MarrowYoffey et al. (1951) investigated the effect of

adrenal hormones on the lymphocyte content ofthe bone marrow of the humerus of guinea-pigs.They record that six hours after the injectionof ACTH or eschatin there was no demonstrableincrease in damaged cells. Furthermore, theyfound a statistically significant increase in lympho-cyte content after eschatin and an increaseapproaching a significant level after ACTH. Indiscussion they suggest that this rise results froman increased uptake of lymphocytes from theblood and might thus be an important factor inthe development of lymphopenia following theinjection of cortico-steroids.

(d) Malignant CellsMurphy and Sturm (i944) recorded that rats

could be partially protected against inoculationswith lymphatic leukaemia cells by adrenal hor-mones. About 5 per cent. of untreated controlsescaped leukaemia; treatment with DOCA oreschatin increased this to 20 to 30 per cent.;ACE in oil gave 40 to 6o per cent. protection andACTH 40 to 75 per cent. Law and Spiers (I947)treated mice suffering from lymphatic leukaemiawith ACE and record a rapid fall in circulatinglymphocytes together with involution of thymusand lymphoid tissue. Bass and Feigelson (1948)observed a similar regression of malignant lympho-cytes in mice as a result of non-specific ' alarm'stimuli.

It would appear from the above results thatmalignant lymphocytes are as sensitive to adrenal-cortical stimuli as normal lymphocytes. Greenand Savigear (I948), however, from their studies ofmitosis in epithelial cells conclude that the epider-mis of mice can be rendered refractory to theantimitotic action of shock or cortisone by paintingwith a carcinogen.

(e) Mode of Action and in vitro StudiesIt must be emphasized that the effect of a single

injection of ACE or ACTH on the lymphoidtissue is agreed by all to be of very short duration,reversion to normal usually occurring within 24hours. If, therefore, it results from widespreaddestruction of lymphocytes in the body it mustbe followed by an extremely rapid regeneration.The life of the normal lymphocyte is, however,extremely short (see Adams et al., 1945; Rein-hardt and Holmes, 1946; and Drinker and Yoffey,194I). A very rapid synthesis must thereforeoccur under normal conditions. Holman et al.(I947) record that an injection of ACE leads to an

ju-ne 1952 357

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

increase in serum peptidase in mice and suggestthat this increase in enzyme activity reflects anacceleration of the' turnover ' of lymphoid tissue.A number of experiments on the direct effect

of ACE and cortico-steroids on lymphocytes invitro have appeared and the results are somewhatconflicting. The lymphocyte is, of course, adifficult cell to work with in vitro since it is an endcell and does not multiply. In any sample oflymphocytes, therefore, many cells showing vary-ing stages of degeneration will be present. Mostof the investigators based their studies on themethod of Schrek (I943), which consists essen-tially in making a suspension of lymphocytes bymincing thymus or lymph glands in normal, salineor some other physiological fluid and then takingsamples at various intervals for unstained andeosin-stained counts. Schrek (I943) found thatwith certain toxins, for example Mocassin venom,almost go per cent. of lymphocytes were stainedby eosin (thus indicating disintegration) withinten minutes and this was followed by lysis anddisappearance of the cells.Under similar conditions Schrek (I949) found

that cortico-steroids only had a very slight effecton lymphocytes, and Robertson (1948) found thateucortone had no significant action. Feldman(I950) found that lipo-adrenal cortical extract hada dramatic toxic action on lymphocytes, compoundA a significant but less marked effect and com-pound E no action. Hechter and Johnson (I949)found that ACE alone did not damage the lympho-cytes, but with the addition of a lymphoid tissueextract or a brain homogenate caused lysis oflymphocytes. He found cortico-steroids, on theother hand, to be inactive even in the presence oflymphoid tissue extracts.

Delaunay et al. (I949) prepared similar sus-pensions of lymphocytes and studied their capacityto take up neutral red. This the authors regardedas a more sensitive method of detecting earlydegeneration than eosin staining, but even so,the addition of ii-dehydroxorticosterone causedno significant change. Finally Heilman (I945)studied lymphocytes in tissue culture and recordedthat compound E in concentrations of 1.25 to 50,ug./ml. caused a moderate but significant decreasein migration of lymphocytes and an increased rateof degeneration of small and medium lympho-cytes. The latter effect was also noted withcompound A, but this did not inhibit migration.

(f) Effect of other Endocrine Glands on LymphoidTissue

Whilst the lymphocytes appear to be peculiarlysusceptible to pituitary-adrenal stimuli, certainother endocrine glands are not without their effecton these cells. The thyroid gland seems to have

an action which is the reverse of that of theadrenals. Thus Reinhardt and Wainman (I942)found that thyroidectomy decreased and injectionof thyroxin increased the weight of thymus, spleenand lymph glands in male rats. Marder (I949)showed a similar action of thyroxine on thelymphoid tissue of mice.

Chiodi (I940) found that male (testosterone)and female (estrone or estradiolbenzoate) sexhormones caused atrophy of thymus in rats andPersike (I940) recorded that pregnancy in youngmice led to thymus involution. Reinhardt andWainman (1942), working with male rats, foundthat castration increased, whilst injections oftestosterone decreased, the weight of thymus,lymph nodes and spleen.The spleen appears to play some part in the

lymphopenia induced by adrenal hormones. ThusDrury (I949) found that in rats adrenaline pro-duced lymphopenia only if the adrenal glandsand spleen were in situ. Ungar (I944 and 1945)has recorded that the 'alarm' reaction is -in-hibited by splenectomy. From his studies on thepart played by various glands in response to shock,he concluded that the pituitary responded toinjury by secretion of ACTH, which led to therelease of an adrenal product which stimulated thesecretion of a splenic substance.

3. Pituitary-Adrenal Control of AntibodyFormation

In I944, Swingle and Rimington, in a review ofthe role of the adrenal cortex in physiologicalprocesses, drew attention to the fact that adrenalec-tomy lowers the resistance of animals to theinjection of many antigenic substances. Thusadrenalectomized animals have been shown to behighly susceptible to small amounts of diphtheriatoxin (Belding and Wyman, I926), typhoid vaccine(Take and Marine, 1923), bacterial infection (Jaffe,H. L., and Marine, 1924; and Scott, I927),foreign cells (Perla and Marmorston-Gotterman,1928 and 1929) and foreign serum (Flashman,1926). Perla and Marmorston-Gotterman (1928and I929) further showed that this lowered resis-tance was associated with a reduced capacity toform antibodies. Later workers recorded thatresistance of adrenalectomized animals was re-stored to normal by injections of ACE (Hartmanand Scott, 1932; and Ettelson, 1941).Fox and Whitehead (1936) studied the effect

of injections ofACE on the production of haemoly-tic antibodies to sheep's red blood cells in rats andrabbits. In both groups of animals prolongedadministration ofACE at the time of the injectionsof erythrocytes led to an increase in the titre ofhaemolysins, especially in the early stages. Theyalso drew attention to the fact that animals not

358 Yune I952

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

BARBER and DELAUNAY: The Role of Lymphoid Tissue

treated with ACE.showed a loss in weight duringthe early stages of immunization, whereas thosereceiving ACE did not lose weight at any stage ofthe experiment.These observations have been amply extended

by the investigations of Dougherty and Whiteand their colleagues. White and Dougherty (I944and I945) reported an increase in serum proteinsoccurring at the time of maximum lymphoid tissueinvolution in mice and rabbits after injection ofACTH. In further studies (Dougherty, Chaseand White, i944) a group of I2 rabbits were givendaily injections of sheep erythrocytes intravenouslyand four of the rabbits were also given dailyinjections of ACE. Between the sixth and tenthweek all animals having ACE had higher titresof haemolysins than the controls, the averagedifference being about double. When nine rabbitswith high titres were given a single injection ofantigen alone or ACE alone, the former led to noincrease in antibodies, whereas six hours afterACE a transitory rise (rather less than double)occurred. Dougherty, Chase and White (I945)also recorded an anamnestic response three totwenty-five hours after a single injection of ACEor ACTH to mice and rabbits previously im-munized against sheep's red blood cells.These authors conclude from their observations

that three to six hours after an injection of ACEor ACTH there is involution of lymphoid tissuewhich leads to the release of serum proteins andantibodies into the circulation. Further experi-ments by these authors have confirmed theiropinion. Thus electrophoretic studies on extractsof washed lymphocytes from the thymus andlymph glands of rabbits revealed the presence ofa protein identical with the gamma globulin ofrabbit serum (White and Dougherty, I945).Total body X-radiation of mice previously im-munized against sheep's red blood cells led tolymphocyte destruction and an increase in cir-culating gamma globulin and antibodies (Doug-herty and White, 1946a). The rate of antibodyproduction to sheep erythrocytes was increasedin mice, rats and rabbits by the simultaneousadministraction of ACE (Chase, White andDougherty, 1946).

All authors, however, are not in agreement withthese workers. One obvious objection to thewholesale application of their conclusions is thattheir studies have been based on a single antigen,sheep's red blood corpuscles, the antibodies ofwhich do not lend themselves to precise quantita-tive measurement. Fischel et al. (i949) treatedrabbits previously immunized against crystallineovalbumin with ACTH or X-radiation. In, bothcases a fall in circulating lymphocytes occurred,but quantitative precipitin estimations, revealed

no anamnestic rise in antibodies. De Vries (I950),working with rabbits similarly immunized, failedto demonstrate an increase in antibody nitrogenafter ACTH. Craddock et al. (1949), workingwith cats inoculated with typhoid vaccine, failedto demonstrate any change in the antibody contentof thoracic duct lymph, following the injection ofACE.Murphy and Sturm (I947) found that adrenalec-

tomized rabbits with hypertrophied lymphoidtissue produced antibodies to horse serum far inexcess of intact animals. Reduction of lymphoidhyperplasia by administration of ACE did notreduce antibody formation. This may have beendue to a disruptive action of the hormone on-lymphocytes leading to a more rapid release ofimmune globulin, but the authors concluded thatit meant that antibody formation was not underthe control of the adrenals.Germuth and Ottinger (I950) studied the effect

of compound E and ACTH on the Arthus reactionand antibody formation to crystalline egg albumenin rabbits, and found that both substances in-hibited sensitization and suppressed antibodyformation.A series of papers by Stoerk and his colleagues

also attack the findings of Dougherty and White.'The first investigation (Eisen et al., I947) is con-'cerned with the effect of adrenalectomy with orwithout injection of ACE on the production ofantibodies by rats to a pneumococcal vaccine andsheep's red blood cells, both given to the same,animal. They conclude from their observationsthat ACE may exert a transitory influence on thedistribution of pre-existing antibodies, but that"adrenal cortical activity has no prolonged effecton the production or release of antibodies andgamma globulin. A further paper (Stoerk et al.,I947) records that the turnover of serum proteinin the rat is unaffected by adrenalectomy or the,administration of ACE. Finally, Stoerk andSolotorovsky (I950) found that the immune res-ponse of mice and rats including the anamnesticreaction was unimpaired by adrenalectomy, and'that a single injection of cortisone in previouslyimmunized rabbits was followed not by an anam-nestic rise but by a drop in circulating antibodies.

Recently two troups of investigators have con-firmed some of the findings of Dougherty andWhite. Thus Hammond and Noak (I950) wereable to demonstrate an anamnestic response inrabbits immunized with sheep erythrocytes bythe injection of ACTH. The rise in haemolysinswas temporary, being maximum about nine hoursafter the injection of ACTH. Bussard et al. (I950)found that in rats highly immunized against botha salmonella vaccine and sheep's red cells aninjection of ACE led to a temporary increase in

June I952 389by copyright.

on Decem

ber 24, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.320.351 on 1 June 1952. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

circulating antibodies, which was maximum abouteight hours after injection. Animals, however,immunized only against pneumococcal or sal-monella vaccines did not show this rise inantibodies after ACE injection.

Roberts and White (I950) studied the produc-tion of antibodies in vitro by splenic tissue fromrats which had had a single intravenous injectionof sheep's red blood cells five days earlier. Ifthe animals were normal the splenic tissue releasedlarge amounts of haemolysin; if animals had beenadrenalectomized two to four days previously therelease of antibodies was much smaller; if, on theother hand, the rats had been given an injectionof ACE two to four days earlier, the splenic tissueproduced an increased amount of antibodies invitro. The authors considered that ACE bothreleased antibodies from the spleen and increasedthe period of maximum antibody production.

In a more recent investigation (Roberts & White,1951) these authors have shown that the splenicproduction of antibodies in vitro may be signi-ficantly affected by adrenal cortical secretions with-out appreciably altering the level of circulatingantibody. They suggest that one explanation forthis discrepancy is that antibody removal tromthe blood by the tissues may also be dependent onadrenal secretion. Thus release and removal ofantibody may both be reduced in adrenalectomizedand accelerated in hormone-treated animals. Suchan explanation, if true, might also account for thedivergent results obtained by different authors inin vivo experiments.A few reports of the effect of ACTH and ACE

on antibody response in humans have nowappeared. Most of these are based on few experi-ments and it is, therefore, not surprising that theresults are conflicting. Forsham et al. (1948)found thatACTH injections revealed no detectableincrease in antibodies of two patients who hadlatent syphilis, one who had previously had T.A.B.vaccine and one who. had had mumps. Onepatient, however, who had previously been im-munized with a pure pneumococcal type I poly-saccharide showed a small rise in precipitin titrepersisting for one week after starting ACTHadministration. Mason et al. (1948) record nocirculating gamma globulin in one patient onACTH and Herbert and de Vries (1950) failed todemonstrate an increase in agglutinins in threepatients previously vaccinated with T.A.B. whowere given a single dose of ACTH, although allshowed a temporary fall in circulating lympho-cytes. Daughaday et al. (1948), on the other hand,gave a patient with a chromophobe adenoma ofthe pituitary three injections of a typhoid vaccineand then 40 cc. of ACE in two doses. This latterwas followed in about six hours by a temporary

fall in circulating lymphocytes plus a temporaryrise in antityphoid agglutinins.

Mirick (195i) has recently made a more esten-sive study in human beings. He vaccinated 59patients with a mixture of pneumococcal poly-saccharides and then I7 were treated with ACTHand I2 with cortisone. Serum globulin, measuredby zinc sulphate turbidity, was decreased in I3of the I7 treated with ACTH and 6 of the 12treated with cortisone. Production of mouse-protecting antibodies was certainly not depressedand seemed to be enhanced by treatment, butsince the treated and untreated groups were notabsolutely comparable as to age and diseases it isimpossible to be certain that this was the effect oftreatment. In ii treated cases the titre of iso-haemagglutinins was tested before and aftertreatment and in four instances there was a four-fold increase. Of five patients with a titre of i inio or more, of antityphoid H agglutinins twoshowed a four- to eight-fold drop. The authorsuggests in discussion that in the presence of apersisting antigenic stimulus such as that affordedby pneumococcal polysaccharide or the mech-anism governing isohaemagglutinin productionthese drugs may cause enhanced antibody for-mation, whereas when the stimulus is not per-sistent, as with previous exposure to typhoidbacilli, the result of treatment with ACTH orcortisone may be a fall.

CONCLUSIONThat lymphoid tissue, with the exception of

the thymus, plays a considerable part in theproduction of antibodies seems beyond doubt.The exact cell or cells concerned remains un-proved, although a study of the available evidencesuggests that immature cells resembling plasmacells are the most active in this respect. It must,however, be emphasized that the cells of manyother tissues of the body are also capable of anti-body formation and indeed recent work (Oakley,1950, I951) tends to favour the view of Ehrlichthat under certain conditions any cell is capableof reacting to an antigen by the production ofantibodies, although possibly not to the sameextent as certain lymphoid cells.

It is clear that the lymphoid tissue of the bodyis to some extent controlled by adrenal hormones.That the injection of ACE or ACTH causes atransitory circulating lymphopenia and decreasein weight of the lymphoid tissue of the body iscertain. The underlying mechanism is less clear.Thus Dougherty and White consider that thereis widespread lymphoid tissue degeneration anddissolution. Other workers, however, draw atten-tion to the transitory nature of the phenomenonand suggest that a temporary failure of production

360 3'une I1952by copyright.

on Decem

ber 24, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.320.351 on 1 June 1952. D

ownloaded from

June 1952 BARBER and DELAUNAY: The Role of Lymphoid Tissue 361

is more probable. In support of this is the factthat ACE has no dramatic effect on lymphocytesin vitro. The apparently increased number ofdegenerative pycnotic cells described by Doug-herty and his colleagues in lymph nodes afterinjection of ACE or ACTH might be apparentrather than real resulting from a diminution inthe number of healthy lymphocytes. In view ofthe rapidity with which the changes appear, how-ever, some degree of lymphocytic dissolutionseems probable. Excessive removal of lymphocytesfrom the blood, for example by the bone marrowas suggested by Yoffey et al. (I95I), may playa part. In this connection it is noteworthy thatgluco-corticoids may in large doses inhibit thereproduction of many cells of the body, par-ticularly fibroblasts (see Ragan et al., 1949;Aterman, I950; Coste, Basset and Delaunay,1951; and Barber and Delaunay, 195I). Greenand Ghadially (I95I) suggest that cortisone hasa general antimitotic action as a result of inter-ference with carbohydrate utilization.The role of the adrenals in the production of

antibodies is much more debatable. Since it isalmost certain that lymphoid tissue is both con-cerned with the production of antibodies and tosome extent under pituitary-adrenal control, someconnection between the adrenals and immunitymight be expected. lt is, however, difficult to

obtain unequivocal experimental proof of this.Theoretically, assuming that involution of lym-phoid tissue follows the injection of ACE orACTH, this might have one of two effects on thecirculating antibody titre. Already formed anti-bodies might be released into the circulation, thuscausing a rise in titre, or the production of furtherantibodies might be inhibited. If both mech-anisms occurred at the same time, circulatingantibodies might remain unchanged. The timeof injection of the hormone in relation to immunitystate is obviously all-important. In view of this,positive results are probably more important thannegative ones. The results of Dougherty andWhite, therefore, must be given considerableweight. It is, however, unfortunate that they con-fined themselves to a study of haemolysin pro-duction and, as Fischel and his colleagues (I949)pointed out, it remains true that so far no one hasdemonstrated a significant alteration in antibodytitre following injection of these hormones, withthe use of an antigen such as crystalline oval-bumen, which gives rise to antibodies vwhich lendthemselves to accurate quantitative measurement.Finally, it must be borne in mind that in all theexperiments referred to very large doses of hor-mone were used. lt cannot, therefore, be assumedthat the results obtained are comparable to whatoccurs under normal physiological conditions.

BIBLIOGRAPHY

ADAMS, W. S., SAUNDERS, R. H., and LAWRENCE, J. S.('945), Am. 3. Phys., 144, 297.

ADDISON, T. (I855), 'On the Constitutional and Local Effectsof Disease of the Suprarenal Capsules.' London, SamuelHighley.

ANTOPOL, W. (I950), Proc. Soc. exp. Biol., N.Y., 73, 262.ASCHKENASY, A., BUSSARD, A., CORVAZIER, P., and

GRABAR, P. (1950), Rev. d'Hematologie, 5, 107.ATERMAN, K. (I950), Lancet, fi, 517.BARBER, M., and DELAUNAY, A. (i9si), Ann. Inst. Pasteur. In

the press.BASS, A. D., and FEIGELSON, M. (1948), Proc. Soc. exp. Biol.,

N.Y., 69, 339.BELDING, D. L., and WYMAN, L. C. (1926), Am. _. Phys.,

78, 50.BING, J., and PLUM, P. (1937), Acta. Med. Scand., 103, 547.BJORNEBOE, M., and GORMSEN, H. (I943), Acta. Path. and

Micr. Scand., 20, 649.BJORNEBOE, M., GORMSEN, H., and LUNDQUIST, F.

(I947),.7. Immun0l., 55, 121.

BUSSARD, A., CORVAZIER, P., GRABER, P., and ASCH-KENASY, A. (I950), Rev. d'Hematologie, 5, 107.

CASPERSON, T. (I94I), Naturissenschaften, 29, 33.CHASE, J. H., WHITE, A., and DOUGHERTY, T. F. (I94s),

Endocrinol., 36, x6.CHIODI, H. (1940), Ibid., 26, 107.COSTE, H., BASSET, G., and DELAUNAY, A. (I95i), C.R

Soc. Biol., 145, 39.CRADDOCK, C. G., Jun., VALENTINE, W. N., and LAW-

RENCE, J. S. (I949), Y. Lab. Clin. Med., 34, I58.CRAFTS, R. C. (I94I), Endocrinol., 29, 596.CREDE, R. H., and MOON, H. D. (1940), Proc. Soc. exp. Biol.

N. Y., 43, 44.DALTON, A. J., and MASSON, G. (I940), Ibid., 43, 370.DAUGHADY, W. H., WILLIAMS, R. H., and DALAND, G. A.

(1948), Blood, 3, 1342.

DELAUNAY A., DELAUNAY M., and LEBRUN, J. (i949),Ann. Inst. Past., 76, i.

DOUGHERTY, T. F CHASE, J. H., and WHITE, A. (I944),Proc. Soc. exp. Bio., N.Y., s6, 78.

DOUGHERTY, T. F., CHASE, J. H, and WHITE, A. (I94S),Ibid., 58, I35.

DOUGHERTY, T. F., and WHITE, A. (I944), Endocrinol., 35, !.DOUGHERTY, T. F., and WHITE, A. (1945a), Ibid., 36, I6.DOUGHERTY, T. F., and WHITE, A. (Ig45b), Amer. Y. Anat.,

77, 8i.DOUGHERTY, T. F., and WHITE, A. (1946a), Endocrinol.,

39, 370.DOUGHERTY, T. F., and WHITE, A. (1946b), Anat. Rec.

Suppi., 94, 13.DOUGHERTY, T. F., and WHITE, A. (s947), 3. Lab. Cln. Aed.,

32, 584.DOUGHERTY, T. F WHITE, A., and CHASE, J. H. (i94S),

Proc. Soc. exp. Bio., 59, 172.DRINKER, C. K., and YOFFEY, J. M. (I94I), 'Lymphocytes,

Lymph and Lymphoid Tissue.' Harvard Univ. Press.DRURY, A. (I949), Fed. Proc., 8, 37.DUSTIN, A. P. (I931), C.R. Soc. Biol., Io8, II55.DUSTIN, A. P., and ZYLBERSZAC, S. (1931), lbid., I08, II62.EDMADJIAN, F., and PINEUS, G. (I94S), Ibid., 37, 47.EDMADJIAN, F., FREEMAN, H., and PINEUS, G. (1946),

Ibid., 39, 293.EHRICH, W. E., DRABKIN, D. L., and FORMAN, C. (I949),

J. exp. Med., 90, I57.EHRICH, W. E., and HARRIS, T. N. (1942), Ibid., 76, 335.EHRICH, W. E., HARRIS, T. N., and MERLENS, E. (I946),

Ibid., 83, 373.EHRICH, W. E., and VOIGT, W. (1934), Beit. path. Anat. u. illg.

Path., 93, 321.EHRICH, W. E., and WOHLTRAB, R. (934), Ibid., 93, 348.EHRLICH, P. (I898), Deuts. med. Wschr., 24, 597.

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from

362 POSTGRADUATE MEDICAL JOURNAL June 1952

EISEN, H N MAYER, M. M., MOORE, D. H., TARR, R. R.,and STOERK, H. C. (1947), Proc. Soc. exp. Biol., N.Y., 65,301.

EPSPEIN, E. (1929), Virchows. Arch. f. path. Anat., 273, 89.ETTELSON, L. N. (1941), Endocrinol., 27, 340.EVANS, R. W. (1948), Y. Path. Bact., 6o, 123.FAGRAEUS, A. K. (I947), Nature, 159, 499.FAGRAEUS, A. K. (1948), 7. Immunol., S8, I.FAGRAEUS, A. K. (I950), Le Sang, 21, 480.F-AVOUR, C. B. (i947), Proc. Soc. exp. Biol., N.Y., 65, 269.FELDMAN, J. D. (I950), Endocrinol., 46, S52.FISCHEL, E. C., LE MAY, M., and KABAT, E. A. (I949),.7.

Immunol., 6i, 89.FLASHMAN, D. H. J. (1926), 7. Inf. Dis., 38, 46I.FORMAN, C., EHRICH, W. E., and DRABKIN, D. L. (I949),

Fed. Proc., 8, 355.FORSHAM, P. H., THORN, G. W., PRUNTY, F. T. G., and

HILLS, A. G. (I948), Y. cin. Endocrinol., 8, I5.FOX, C. A., and WHITEHEAD, R. W, (1936), Y. Immunol., 30,

5'.GELLHORN, E., and FRANKS, S. (1948), Proc. Soc. exp. Biol.,

N. Y., 73, 399.GELLHORN, E., and FRANKS, S. (I949), Ibid., 71, 112.GERMUTH, F. G., Jun., and OTTINGER, B. (3950), lbid.,

74, 8I5.GORMSEN, H. (I950), Le Sang., 21, 483.GREEN, H., and GHADIALLY, F. N. (i9si), Brit. med. _.,

i, 496.GREEN, H., and SAVIGEAR, M. (i9Si), Ibid., , 498.HABEL, K., ENDICOTT, K. M., BELL, J. F., and SPEAR, F.

(I949), Inmunol;',6I,I3I.1HAMMOND, C. W., and NOAK, M. (I950), Proc. Soc. exp. Biol.,

N.Y., 74, 155.HARRIS, S., and HARRIS, T. N. (I949), 7. Immunol., 6i, I93.HARRIS, S., and HARRIS, T. N. (i950), Ibid., 64, 45.HARRIS, T. N., and EHRICH, W. E. (I946), 7. exp. Med., 84,157.HARRIS, T. N., GRIMM, E., MERLENS, E., and EHRICH

W. E. (I945), Ibid., 8I, 73.HARRIS, T. N., and HARRIS, S. (1948), Proc. Soc. exp. Biol.,

N.Y., 69, I8.HARRIS, T. N., and HARRIS, S. (7949),. exp. Med., 9o, 169.HARRIS, T. N., RHOADS, J., and STOKES, J. (I948), Y.

Immunol., 58, 27.HARTMAN, F. A., and SCOTT, W. J. M. (1932), 7. exp. Med.,

ss, 63.HECHTER, 0. (1948), Endocrinol., 42, 285.HECHTER, O., and JOHNSON,!S. (949), Ibid., 4S, 351.HEKTOEN, L. (I916), 7. Inf. Dis., ig, 69.HEILMAN, D. (I94S), Proc. Staff Meetings Mayo Clin., 20, 311.HERBERT, P. H., and DE VRIES, J. A. (I949), Endocrinol., 44, 259.HERLANT, M. (I950), Proc. Soc. exp. Biol., N.Y., 73, 399.HILLS, A. G., FORSHAM, P. H., and FINCH, C. A. (1948),

Blood, 3, 755.HOLMAN H. R., WHITE, A., and FRUTON, J. S. (x947),

Proc. goc. exp. Biol., 65, 196.ICKOWICZ, M. (I947), Radiologica Clinica, I6, 231.INGLE, D. J. (1938), Proc. Soc. exp. Biol., N.Y., 38, 443.INGLE, D. J. (1940), Ibid., 44, 174.JACOBSON, L. D., ROBSON M E., MARKS, E. K., and

GOLDMAN, B. S. (I949), Y. Lab. din. Med., 34, I6i2.JAFFE, H. L., and MARINE, D. (1924), 7. inf. Dis., 35, 334.KARNOFFSKY, D. A., GRAEFF, I., and SMITH, H. W. (1948),

Am. . Path., 24, 275.KASS, E. H. (1945), Science, IOI, 337.KEUNING, F. J., and SLIKKE, L. B. VANDER (I95o), . Lab.

cdn Med., 36, I67.KINDRED, J. E. (I947), Arch. Path., 43, 253.LAW, L. W., and SPIERS, R. (i947), Proc. Soc. exp. Biol., N.Y.,

66, 226.McMASTER, P. D., and HUDDACK, S. S. (I93S), . exp. Med.,

6I, 783.McMASTER, P. D., and KIDD, J. C. (1937), Ibid., 66, 73.MARDER, S. N. (I949), Proc. Soc. exp. Biol., N. Y., 72, 42.MARSHALL, A. H. E., and WHITE, R. G. (I950), Brit.J7. exp.

Path, 31, 1S7.MASON H. L POWER, M. H., RYNEARSON, E. H.,

CIAkAMELf.I, L, C., LI, C., pnd EVANS, H. M. (1948),.7. din, Endocrinol., 8,I.

MILLER, Z, B. (x950),. exp. Mad,, 92, itz.MIRICK, G S. (i951), Budl. 7. Hopkins Hosp., 18, 332.MOLOMUT, N SPAIN, D. M., and HABER, A. (1950), Pioc.

Soc. exp. Bid., N.Y.P.73, 416.

MOON, H. D. (I937) Ibid., 37, 34-MOON, H. D. (I940), Ibid., 43, 42.MURPHY, J. B., and STURM, E. (I92S), Y. exp. Med., 41, 245.MURPHY, J. B., and STURM, E. (1944), Science, 99, 303.MURPHY, J. B., and STURM, E. (i947), Proc. Soc. exp. Biol.,

N. Y., 66, 303.NETTER, A. and MATHE, G. (I949), Le Sang., 20, 442.OAKLEY, C. L., WARRACK, G. H., and BATTY, I. (1949),

7. Path. Bact., 6I, 179.OAKLEY, C. L., BATTY, I., and WARRACK, G. H. (1951),

Ibid., 63, 33.PERLA, D., and MARMORSTEN-GOTTERMAN, J. (I928),

.7. exp. Med., 47, 723.PERLA, D., and MARMORSTEN-GOTTERMAN, J. (1929),

Ibid., 50, 87.PERSICKE, E. C., Jun. (I940), Proc. Soc. exp. Biol.. N.Y., 45, 315.PHILLIPS, F. S., HOPKINS, F. H., and FREEMAN, M. L. H.

(I947), . Immunol., 55, 289.RAGAN, C., CROKOEST, A. W., and BOOTS, R. H. (I949),

Amer. _7. Med., I00, 741.RAGAN, C., HOWES, E. L., PLOTZ, C. M., MEYER, K., and

BLUNT, J. W. (I949), Proc. Soc. exp. Biol., N.Y., 72, 7i8.REINHARDT, W. O., ARON, H., and LI, C. (1944), Ibid., 57, 289.REINHARDT, W. O., and HOLMES, R. 0. (1940), lbid., 45. 267.REINHARDT, W. O., and HOLMES, R. 0. (1946), Anat. Rec.,

94, I97.REINHARDT, W. O., and LI, C. (I94S), Science, 202, 360.REINHARDT, W. O., and WAINMAN, P. (1942), Proc. Soc.

exp. Biol., N. Y., 49, 257.RINGERTZ, N., and ADAMSON, C. A. (I950), Acta. Path. et

Mia. Scand. Suppl., 86.ROBERTS, S., ADAMS, E., and WHITE, A. (1948), 7. Biol.

Chem., 274, 379.ROBERTS, S., and WHITE, A. (I950), Fed. Proc., 9, 220.ROBERTS, S., and WHITE,"A. (195s), Endocrinol, 48, 74I.ROBERTSON, J. S. (1948), Nature, I6I, 814.ROUX, E., and YERSIN, A. (1889), Ann. Inst. Pasteur, 3, 273.RUSK, G. Y. (Igis), Univ. Caliornia Pubi. Path., 2,139.SCHREK, R. (1943), Arch. Path., 35, 857.SCHREK, R. (1949).SCOTT, W. J M. (1927), .. exp. Med., 47, 185.SEGAL, G., and LEBLOND, C. P. (1938), C.R. Soc. Bios., 229,

279.SELYE, H. (1937), Endocrinol., 22, I6g.SELYE, H., and FOGLIA, V. G. (1939), Amer. 3'. Anat., 64, 133.SHAPIRO, k. B. (I9949), Proc. Soc. exp. Biol., N.Y., 70, 44o.SIMMONS, J. P., and JONES, H. M. (I95S), J. med. Res., 28. 197.SIMPSON, M. E., LI, C., REINHARDT, W. O., and EVANS,

H.rM. (I943), Proc. Soc. exp. Biol., N.Y., S4, I35.SPEAR, F. (1948),.7. Lab. clin. Med., 33, I563.SPURR, C. L. (I947), Proc. Soc. exp. Biol., N.Y., 64, 259.STAR, P. (I895), Lazncet, 1, 284.STOERK, H. C., JOHN, H. M., and EISEN, H. N. (1947), Proc.

Soc. exp. Biol., N. Y., 66, 25.STOERK, H. C., and SOLOTOROVSKY, M. (i9So), Am..7.

Path., 26, 708.STONE, D., and HECHTER, 0. (1948), Endocrinol., 42, 307.SWINGLE, W. M., and RIMINGTON, J. W. (I944), Phys. Rev.,

24. 89.TAKE, M. N., and MARINE, D.r(1923),.7. inf. Dis., 33, 217.THATCHER, J. S., HOUGHTON, B. C., and ZIEGLER, C. HI

(1948), Endocrinol., 43, 44o.TOPLEY, W. W. C. (1933), 'An Outline of Immunity.'UNGAR, G. (I944), Nature, 54, 736.UNGAR, G. (194S), Endocrinol., 37, 329.VALENTINE, W. N., CRADDOCK, C. G., and LAWRENCE

J. S. (1948), Blood, 3; 729.DE VRIES, J. A. H. (I9So), 7. Immunol., 6s, .

WHITE, A., and DOUGHERTY, T. F. (I944), Proc. Soc. exp.Biol., N.Y., 56, 26.

WHITE, A., and DOUGHERTY, T. F. (194S), Endocrinol., 36,207.WHITE, A., and DOUGHERTY, T. F. (1947), Ibid., 42, 230.YOFFEY, J. M. (i9So), Biol. Rev., 25, 314.YOFFEY, J. M., and BAXTER, J. S. (1946),.7. Anat., So, 232YOFFEY, J. M., METCALFE, W. K., HERDAN, G., and

NAVIN, V. (z9s ), Brit. med. Y.' L 66o.YOFFEY, J. M., REISS, M., and BAXTER, J. S. (r946), Nature,

I54, 368.ZWECKER, I. T. (1948), Am... PlPYs., IS2, 267.ZYLBERSZAC, S. (I93x), C.R. Soc. Biol., zoS, 22S6.

by copyright. on D

ecember 24, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.28.320.351 on 1 June 1952. Dow

nloaded from