THE JOURNAL Of 1NVESTIGATIVE DERMATOLOGY, 6i:625 - 629. 1976 Vol. 6;, :'\0. S.

Part 2 of 2 parts
Copyri~ht © 1976 by The William, & Wilkin. Co. Printed in l.:.S.A .

LYMPHOKINES

ROBERT L. LUNDAK. PH.D .. JOHN LEWIS. PH.D ., AND GALE GRANGER. PH.D .

Department of Molecular Biology and Biochemistr)" Uniuersiry of California Iruine, Irvine, California, U. S. A.

Wh en a foreign material or an infectious agent, active or t ransformed blast cell . The morphologic
antigen(s), penetrates natural defense barriers and transformation may occur within the first 24 to 48
gain access into the tissues of a mammalian host, hr. After 36 to 48 hr, certain cells in the activated
specialized cells in the lymphoid tissue recognize population begin D A synthesis and. later, cell
and incite a response specifically directed against division ... Activated" lymphoid cells gain the abil-
the fo reign antigen . Thi s response is biphasic. ity to express certain effectur functions which can
composed of two sepa rate components . One culmi- be measured in vitro. such as specific or nonspe-
nates in the syn thesis and secretion . by B-Iym- cific cytolysis of ·· target " cells [3.4] and secretion
phoid cells. of specific antibodies which a re dis - into the culture med ia of a family of soluble
seminated into the lymph and blood vascular molecules. which have collectively been termed
systems. where they gain access to and react Iympho kines (LK s) [5,6 ]. The amount of informa-
specifically with the antigen(sl. The second phase tion available on each of the aforementioned step:
occurs simultaneously with the first and results in (a) cellular transformati o n. (b) D]\;A biosynthesis.
stimulation and expansion of a clone of reactive (c) cvtotoxicitv. and (dl LK secretion. is extensive.
T-Iymphoid cells which. when fully differentiated. and ~ach area- has been reviewed . yet the relation-
can migrate from the lymphoid tissue(s) to the s hip of these steps, one to anothe r. is still not clear.
antigen and t.here reac t against it. The latter It is not yet clear whether the same cells partici-
reactions have been termed cell-mediated immu- pate in each of the above steps or whether a
nit~- (CM II and are primaril~' associated with composite of man~· separate cell types and reac-
tissue destructi on. typified by allograft rejection. tions is iO\·olved in a progression of Telated events.
t.umor i mmunit~· , autoimmunit~·. and , ·a ri ous It is ob , ·io us. however. that the cells ho ld the key to
forms of delayed hypersensiti,·i iies . The effector ou r understanding of the mechani mIS' operati,·e
cells involved in these important immunologic in eVIl.
reactions in "i,·o appear to be primar i l~· lympho- The general purpose of thi s re,·iew is to briefly
cytes and monocytes or macrophag-es. Our under- disc uss salient featu res about the famil~· of soluble
standing of the basic mechanismls) that is opera- molecule released by acti,·ated lympho id cells.
ti,·e in t hi_ important group of reactions has co me and more specificallv. studies on two of them.
primarilv from recent studies performed b~· numer- leukoc~'te inhibiting factor IUF) and lymphotoxin
ous im·estig-ators. In ,·itro systems. which appear to (L Tl. and the role t ha t one of them, L T. ma~· play
be models of in vivo situations. ha'·e facilitated in c"tot.oxic CMl reactions. For more detailed
mo re defined studies and allowed us to dissect in for'mation on t he general lopic of LK. one shou Id
these Teactions into their various parts. consult any of se,-eral recent fe,'iew articles 16- 9].
eM! rea ction, in ,'itro are co mplex and take Table I contains an up-to-date list of the spectrum
several forms. One of the earliest indica tio ns of a of acti"it ies. collectively termed LKs. which have
reaction is ··actiyation·· of the lymphocyte. which been reported to be present in the cell-free super-
occu rs afte r the immune cell is cultu red with the natant media obtained from mitogen-stimulated.
specific ant ig-en . or "normar- cells are cultured antigen-acti,·a ted. and mixed lymphocyte cultures
with anyone of a number of mitogenic agents [1. of human and experimental animall~'mphoid cells
2]. This change is indicated by morpholog-i c and [10]. While the list of individual acti,·ities is exten-
biosynthetic transformation of t.he normall~' quies- si,-e. they appear to fall int o se,·eral g-ene ral cate-
cent small lymphocytes into a biosyntheticall~· gories dependent upon their effect on secondary
cell in culture . The first caiegory includes those
Reprint requesls to Dr. R. L. Lundak , 447 Stein house factors which exclusivel~· affecl monoeytes and
Hall. Unive rsity of Ca lifornia Irvine. In·ine. California macrophages . The various reported acti,·ities
92664 . .
AbbreviatioM: include: (a) inhibition of migration. (b) biosyn-
CM I: cell-mediated immunity thetic activation. and (c) facto rs whieh cause a
Con-A: concanavalin A - clumping or aggregation. The second categor~' of
LIF: leukocyte inhibi ting factor LK s exe rt the ir effect upon polymorphonuclear
LK: Lymphokine
LT: lymphotoxin (PM]\; ) leukocytes: (a) a facto r(s) which inhibits
M IF : migration inhib ition facto r migration. as well as (b) spec ific chemotactic
M LC: mixed leukocvte culture facto rs for eosinophils and neutrophils. A third
PHA: phytohemaggiutinin category of actiyities express their effects on vari -
PMN : polymorphon ucl ear
PPD : purified protein derivati,·e ous type of somatic cells: (8) inhibition of acti,·e
TF: transfer factor p rol iferation and DNA synthesis of various somatic
625
626 LUNDAK, LEW IS, AND GRANGER Vol. 67, No.5, Part 2 of 2 ports

TABLE 1. Activities associated with cell-free culture bodies by specific B-lymphocytes. This latter ac-
medium from mitogen-activated or antigen-activated tivity mayor may not be related to a CMI reaction.
lymphoid cells There are a number of important parameters
which affect the in vitro synthesis and secretion of
1. Factors active on lymphocytes LKs by activated lymphocytes. While ly mphoid
A. Histocom patibility antigen-HA cells from many animal species, including man,
B. Autostimulating factor-ASF secrete them when activated by co-cu lture with
C. Antigen requiring transforming facto r-ARTF antigens, mitogens. or histoincompatible cells, the
D. Blastogenic factors - BF actual levels present in a supernatant may differ
E. B-cell stimulating factor-BS greatly 15-7]. The cells from certain animal species
2. Factors active on polymorphonuclear leukocytes appear to secrete more than others in vitro.
A. Chemotactic factor for neutrophils-CFN Whether this effect. is due to inherent differences
B. Chemotactic factor for eosinophils- CFE between the capacity of the cells to activel y
C. Eosinophil stimulation promote r- ESP secrete LKs in vitro, or simply due to the su rvival
D. Leukocyte inhibition factor-LlF or maintenance of the activat.ed cells in tissue cul-
3. Factors active on monocytes and macrophages ture is not absolut.ely clear. A second important
A. Migration inhibition factor- MlF parameter to consider is the tissue(s) from which
B. Macrophage (aggregation) clumping factor-MCF the lymphoid cells are obtained. Lymph oi d cells
C. Macrophage activation fact.or-MAF derived from man~' diffe rent lymphoid organs have
D. Macrophage spreading inhibition facto r-MSlF been employed. They all appear to be com petent
E. Migration enhancement factor- MEF to secrete LKs. However, there is some evidence
4. Factors active on non lymp hoid mesenchyma l cells that cells removed directly from mouse tbymus
A. Lymphotoxin- LT cannot secrete L T, when they are mitogen stimu-
B. Proliferation inhibition factor-PlF lated 112.13 J. It also appears that lymphoid cells
C. Colony inhibition facto r- ClF from tissue closest to the local site of subcutaneous
5. Miscella neous factors found in supernatant medium deposition of antigen in a guinea pig release t.he
A. Skin reactive factor- SRF highest levels of migration inhibition facto r (MIF),
B. Antibodies- Ab when cultured with t.he specific ant.igen [14].
C. Interferon- lF Whether the increased le" els of LK are due to a
6. Factors extracted from intact lymphoid cells more "immune" or active secreting cell or a higher
A. Lymph node permeability factor-LNPF percentage of im'olved responding cells is not
B. Transfer factor- TF known; however. both situations may exist. We
have found that there appears to be some natural
segregation of mit.ogen-stimulatable. LT secret in g
cell type, and (b) a nonspecific cell toxin. The last cells within human lymphoid tis ·ues. The most
category of LKs to be cons idered here are those effective are cells from adenoids. followed by
t hat affect lymphocytes themsel ves . These lymph nodes. tonsils, then peripheral blood and
include: raj fact or (s) which specifically or nonspe- spleen cells . A very important considerat ion is that
cifically induces blastogenesis and transformation. almost any treatment or agent which induces
and (b) those which have helper properties in lymphocyte activation in vitro measured by DKA
T-cell- B-cell interactions. While the relative con- synthesis or blast transformation, a lso appears to
centrations of the agent(s ) responsible for the induce the release of LKs . As shown in Figure I ,
activities may vary. dependent upon the culture the level of LT secreted reaches peak values at the
conditions employed. it appears that all activities same dose of mitogen which induces maximum
can be secreted into the supernatant media D:\T A synthesis . However. these appea r to be
obtained from a sing le activated lymphocyte cul- independent events, because {a) LK secretion
ture [6]. There is one exception, transfer factor begins early. before Dt\A synthesis. and Ib) if
(TF ). which is an intracellular product obtained by DNA synt hesis is blocked. LT secretion is not
extracting cytoplasmic components from immune affected.
lymphocytes [1 1]. LKs are present in the supernatant fraction of
While no concrete evidence exists at t his time, a cultures activated b~' soluble, viral, fungal and
functional in vivo role for many of these mediators cellular antigens, mitogens. mixed lymphocyte
would be easy to envision. The activities associa ted cultures. and in supernatants from cert ain contin-
with the first , second . and fourth categoreis have uous lymphoid cell lines [9 ]. There are few or no
effects which would result in recruitment and data available to indica te whet.her one antigen is
activation of effector mechanisms and allow other- more competent than another to induce LKs.
wise uninvolved cell types to become participants However, there is evidence indicating t hat mito-
in t.he CMI reaction . The inhib ito ry activities gens. typified by phytohemaglutinin (PHA ) or
present in the third cat.egory, either cytostatic or concanavalin A (Con-A) induce high levels of LK.
cytotoxic, would permit the triggered lymphocyte The increased levels are presumed to be due to the
itself to be a direct and primary effector cell in the capacity of these agents to activate a higher
CMI reaction. Secreted helper factors for B-cells percent of the total secreting cell population than
would presumably facilitate the secretion of anti- that capable of activation by antigens . While the
Nov . 1976 LYMPHOKINES 627
1200 30 would indiscriminately release these highly active
biologic molecules out into the surrou nd ing tissue
r without some form of regulatory mechanism ( )
t: 1000
> operative on the sequence. am ou nt. and species of
f= I '" molecules released .
U ~
<f 20 a. ]\0 discussion of LKs is complete without a brief
:u
~I
considerat ion of some of the physica l properties of
(")
"U t his famil y of molecules. For more detailed infor -
3: mation, the reader is referred to other review
'0 )<
a rticles [5,6 ]. From Tab le II, it can be seen that
If) q the majority of hum a n lymphocyte effector mole-
t: 400 10
Z cules appear to be macromolecu la r in chara cter .
:::>

200 I They a re heterogeneous in size; the largest among

the human lymphocyte effector molecules appear s
to be LT. with a molecular weight of approxi-
mately 90,000 daltons, t he smallest TF a t 10.000
25 1.25 .625 daltons . Both human and guinea-pig LKs are also
CON A ()lg/ml) heterogeneous with regard to their elect rophoreti c
FIG . 1. Relationship observed between DKA synt hesis
mobilit y, a nd migra te all t he way from fast gamma
and Iymphot.oxin (LT ) sec ret ion when human ly mpho- glob ul ins to prea lbu min s. when compared to serum
cytes were cultured in the presence of va rious concentra- protei n markers 15,6.8.9 ]. Purifi catio n and study
ti ons of concanavali n A. The lymphocytes were cu lt ured of these molecu les is tedious. because the assavs
in MEM conta in ing 590 boiled seru m. A unit of L T are difficult. and all investigators agree that LKs
activi ty is defined as the reci procal of the dilution of
mediu m caus in g destruction of 50.000 larget L-cells in a are present in an active supernat a nt in very small
24-hr pe riod . quantities. This is furth er emphas ized b~' t he fact
that n ot one LK has yet been ver ified to be purified.
mit ogen a re termed nonspecifi c. beGau e they pri mari ly beca use the final fractio ns contain such
induce the ac t ivat ion of nonimmune cells. the a s ma ll amount of material.
evidence to date ind icates that MIF. released by
LEl'KOCYTE !\,HIBITORY FACTOR I LIF I
Con-A a nd a ntigen-a ctivated gu inea -pig cells. is
phys ically the same [5. 15 J, and LT released by H uman LIF is a soluble product produced by
PHA. Con-A. purified protei n derivative (PPD) , mitogen-stimulated or antigen-st imu lated lym-
and mixed leukocy te cu lture (?-.1LC) ac ti vated phocytes that inhibits the migration of human
human cell appears to be physica lly identica l PM:\, le ukocytes but does not in hibit t he migration
[16- 18J. These obsen'ations which. if true for the of human or guin ea- pig macrophages as does MIF
other LKs, indicate t hat no matter how the lymph- [19J. MlF' has an estimated molecular weight of
oid cell is ac tiva ted the pa rticula r LK released ma~' 23.000 daltons. LIF has an estimated molecula r
be the same molecule. Finall \'. there are. of course. wei ght of 68,000 daltons . Hum an LIF has been
various important te ch n ical ~onside rations related fou nd to migrate with album in in gel electrophore-
to the in vitro cu lture techn iques. such as the t~'pe sis. be resistant to ne uramin idase. inacti\'ated by
and amou nt of serum . the numbers of cel ls per ml c h~·motryp sin. be resistant to 56°C fo r 30 min. and
employed . the dosage of acti,-ating agen t (s) u ed. not inact iva ted by antibodies against Fa b or
all of which must be taken into account to induce huma n se rum albumin.
the "maximum" levels of a pa rt icular LK.
There is su rprisingly little kn own about the '\lODEL OF T HE BASIC 1:\ \lTRO SYSTDtS I\, WHICH
DIRECT A\,D I\'DIRECT LY:-1PHO CYTE ·I\,DlT ED
pa rt icula r cell type in voked and the event which CYTODEST Rl'CTl O\, OCC l 'RS
may "regulate" the secretion of LKs. Whil e th e
evidence is very co nvincing that the major LKs M an~' distinct in vit ro system s have been
a re lymphocyt.e prod ucts. it is not as clea r for the devised and stud ied which may sen -e as models of
more peripheral a nd lesse r kn own acti"it ies. It was the in vivo cvtodestructive react ions that occu r in
reported tha t T -cells were required to induce LT CM! 13.4]. While the results of these studies are
sec ret ion in PHA-activated mouse spleen cell cul- numerous a nd di'-erse. th ey can be grouped into
tures. yet it was no! co nv i n ci ngl~- demonstrated three major categories. whi ch are illustrated in
tha t B-cells could not pa rtic ipate afte r T -cells had Figure :2. The first class of react ions is t~-pified by
become acti"ated 11 2 J. Ident ificatio n of the cell the ituat ion where lymphoi d cell s are obta ined
types and regulatory process(es) im'oked in LK from experimenta l animals or human pa tients who
sec retion is im porta nt and an ex citi ng a rea about have been im mun ized wi t h cellu la r or t issue anti-
whi ch little is kn own. fo r if L1\:s ha ve a ce ntra l ro le gens. The im m une lymphoid cells are allowed to
in CM f reactions, t he mecha nis m (s) whi ch con trols interact with the imm un izing or donor "target
them regulates the degree and magnitude of th e cells" in culture. The first t ep is physical con tact
response itself. This is one of the most im portant of t he aggressor lymphoid ce lls with the a ntigen-
areas that requ ires study, for it seems high ly bearing target cell s [20,21 J. Thi is a recognit ion
improbable that upon activation the effector cell step. presumably fa cili ta ted b~' a s pecifi c receptor
628 LUNDAK , LEWIS. AND GRA. GEM Vol. 67. No.5. Part 2 of 2 parts

TABLE n. Physical properties of activities associated with hum.an lymphocyte effector molecules

Effect of Molecular Enz.\'m~ treatment Electro- Molecular

Factor phorNic
heotinR wei:!:ht ~pecies
Protease Nuclease mubill1Y

Macrophage Stable to 56°G 25.000 Albumin

inhibition
Cytotoxic Inact-. 10° at 90,000 Resists Alpha glob. Protein
SO ± 5°C
Colony inhibi- Inaet. at 56°G Trypsin Resists Prot-ein (?)
tion &SO °C ensitive
Interfe ron Stable at 56°G IS- 20,000 T rypsin Resist-ant Prote in
sensitive
Macrophage Heterogeneous
clumping Alb .- Alpha
glob.
Histocompati- Inact. at 56°G Pelleted at Resistant Lipoprotein (?)
bility Ag I '- 10' x g
Autostimulat- Inact. 15" at Nondialyzable Resistant
ing factor 80°C
Antigen requir- Inact. 30" at 56 °C l'iondialyzable
ing transform-
ing facto r
T ransfer fae- Inact. 56°C 10.000 Idia- Resists Resistant Polynucleotide
tor (I) lyzablel trypsin peptide
Transfer fac- Inact. 90 °C 150.000 non- Protein lipid
tor (2) dialyzable carbohyd rate
Chemotactic 25.000

Steps in
Contact with AQ Lymphocyte Act ivat ion I CIII de,'ruc'lon

..
L Cellular Ags

IT~ ce lll

II. Soluble Ags

FIG. 2. Schemat ic diagram of tbe various t.ypes of possible mechanisms associated with lymphocyte-mediated
cytodestructi\,e reactions in vitro.

on the membrane of the aggressor lymphocyte. sor and target cells , it is termed direct destruction
This poorly understood interaction leads to "acti- 13 j. The second class is where lymphoid cells are
vation" of the lymphocyte biosynthetic processes collected from donors who have been immunized
which result in induction of effector function and with soluble macromolecular antigens . This cyto-
which lead to specific destruction of the target cel l. toxic reaction in vitro take' two fo rms: (a) co-cul-
Since this phenomenon requ ires contact of aggres- ture of the immune lymphoid cell with the antigen
No v . 1976 LYMPHOKINES 629
results in contact and recognition of the ant igen, ph ocyte. Actual lysis in these systems may occur
resulting in activation of the ly mph ocyte and by two basi c mechani sms: (a) physical conta ct of
induction of nons pecifi c destruction of ta rget cell s the lym phocyte with the target cell induces a
at a distan ce from the aggressor lymphocyte destructive event wh ich continues afte r removal of
[22- 24]. This latter reactio n. since it does not the aggressor cells and res ults in target cell de-
require physical co ntact of th e lymph ocyte a nd structi o n [29]; (b ) activated ly mphocytes deposi t
target. cell. will be referred to as indirect cytolysis: L T or L T -li ke molecules on the target cell surface
(b) interacti on of the immu ne Iymph ocyt.e with and these actually cau. e dest ruction .
target cells that ha\'e t he ant igen chemically
attached to their s urface indu ces s pec ific trigge r- REfERE:\ CES
ing. ly mphocyte ac ti vation. and nonspecific de - 1. Oppenheim JJ : Fed Proc 27:21, 1968
struction of the target cell 125]. l t is not well 2. Ling :\,R: Ly mp hoc.l·te Stimulation. ;-';ew York .
substantiated but generally accepted tha t cell Wile\·. 1968
destru ction in the tirst three catego ries is mediated 3. H olm G: Ad \' Immunol 11:117 . 1969
4. Henney CS: Tra nsplant Re\' 1.:39. 1973
prim a rily by thy mus- depen dent cells . however. 5. Bl oom BR. Glade PR: In In Vitro Methods in
Ol her cell types can parti ci pate 126 j. Cell- Medi ated Im mu ni ty. Edited by BR Bl oom .
Wh ile th ese basi c categories of react ions ha \'e PR Glade . "'ew York. Aca demic. 1971. p ~ 73
fundamental differences. they do ha\'e certain 6. Granger GA: Ser H aema tol 4 : . 19. 2
7. Bloom B: Ad\' Immunol 13:102. 197 1
steps in co mmon . The~' all sha re some form of 8. Ruddle :\,H : Cu rr Top M icrobial Im m unol .';7:75.
recognit ion reaction ca using activation of th e 1972
lympho id cell . and they s hare the act ua l cytotoxic 9. Lawrence HS. La nd\' M ledsl: Mediators of Cellula r
Im muni t y. :\ew ) ;ork. Academic. 1969
step. 10. Du monde DC. Wolstencraft RA. Pana\' GS. Matthew
It has been determined t hat th ese reac tions are M , Morle\' J, Howson WT : :\ature <'Land) 224 :338.
co mplement i ndepend e nt. requi re "acti\'ated" "ia - 1969 '.
ble lymphoid cells. and. once t ran sformed. the 11. Lawrence HS: Ad,' Im rnu nol 11:195. 1969
12. Sha cks SJ . Gran!,er GA: Cell Immunal l in press)
blast cell is a more effective killer r2 3 .2.5.~ 7]. The 1976
degree of destruction \'a r ies with time. t he t~' pe of 13. Williams TW . Grani(er GA. Kolb WP : In Cellular
target ce ll employed. and the nu m her of effector or Recogni tion. Edited by RT Smith. RA Good . "'ew
aggressor lym ph oid ce ll s pe r target cel l [3 J. The York . Appletan-Centu ry-C rofts. 1969. p 235
fi rst step in all of t he afor eme nti oned react ion. is 14. H alpern B . Storb Y . Fra~' A: "'ature fLondl 215:400.
1967
recogni tion . ln an unknown manner. the immun e 15. P ick E. Krejci J. TurkJL: Is r J Med Sci . :19•. 19. 1
ly mphoid cell spec ifi ca lly recognizes and interacts 16. G ran ger GA. Moo re GE. W hi te JG. Matzi nger P .
with the ant igen. Thi s reaction presumably is Sundsma JS. Shupe S. Kal b WP . Kramer J.J.
Glade PR: .J Immunal 104: 1416. 1970
in it iated by s pecific recepto rs on the aggressor cell 17. Boulos £:\. Rosenau W. Goldberg M: J Immunol
su rface. V,-het her this i, su ffic ien t to trigger activa- 1l2: 134'. 19'4
tion pe r se is still unknown. In rh e case of mitogen- 18. Ami no !\'. Linn ES . P\'sher TJ: Cell Immunol l in
induced C\·to(Ox ic reactions. th e mitogen appea rs press ) 19.6 .
19. Rockland RE . Remold HG . Da" id JR: Cell Immunol
to int.era~ t with the monosaccha rid e receptors 5:4~6. 1972
present on immune o r nonimm un e lym phoid cell 20. Rosena u W : In Cell Buund Anti bodi es. Edited b\' B
surfaces. Both mitogen and a nti gen indu ce activa- Am os. H Kop rowski. Phi la delphia. \\' istar In'sti-
t ion . which resu lts in transfo rm a ti on of t he nor - tu te Press. 1963. p . 5
21. Gm'aerts A.J: Immu nology 85:516. 1960
ma lly quiescent h'mphocyte int o t he functional 22. Ruddle H:\ . \\'a ksma n BH: Science 151:1060. 196,
effector cell. It has rece n t ly been determined in in 23. Gran ger GA. K olb WP:·J Imrnunoll01:1I1. 1968
vitro s ~·ste m s . emp l o~' in g immune mu rine lympho- 24. Heise ER. Weiser RS : J Immunol 103:570. 1969
cytes and mitogen- ac ti\'ated human Iymphoc~·tes . 25. Perl ma nn P. Holm G: In Imm unopathology. "01 5.
Ed ited b\' P Meisc her. P G ra bar. i'iew York. Grune
t hat the first and second steps can occur ver~' & Stratt on. 1967. p 32.5
rapi dly. within t he first hour of the interactio n ~6. Goldstein P . Schimmacher V. Rubin B. Wigzell H :
between lymphocyte an d target cell. and t hese Cell Im mu nol 9:2 11. 1973
have been termed t he lym phocy te dependent ste ps 27. Ginsberg H. Sachs L: J Cell Cam p Physi ol 66:199.
1965
128]. It also a ppea rs tha t the fi nal dest ru ctj\'e 28. M art z E . Benacerraf B: J Immunol 111:1538. 1973
phase. termed the Iymphocvte independent step. 29. Hellstro m !-I.E. Hellst rom I: Adv Cancer Res 12: 16, .
does not require the presen ce of t he \'iable I~' m - 1969