Proc. Nat. Acad. Sci.
USA
Vol. 71, No. 2, pp. 498-502, February 1974
Chemotactic Activity in Dialyzable Transfer Factor
(granulocytes/monocytes/immune response/lymphokines/primate/delayed hypersensitivity)
JOHN I. GALLIN AND CHARLES H. KIRKPATRICK
Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Diseases, National Institutes of Health,
Bethesda, Maryland 20014
Communicated by H. Sherwood Lawrence, October 10, 1973
ABSTRACT Dialyzable transfer factor from human
leukocytes was found to be strongly chemotactic for
granulocytes and weakly chemotactic for monocytes in
vitro. Chemotactic properties were also demonstrated in
vivo in rhesus monkey skin. Initial purification of the
dialyzable transfer factor by Sephadex G-25 chroma-
tography revealed multiple fractions containing material
with 255-nm absorbance. The fractions containing chemo-
tactic activity were also capable of transferring delayed
hypersensitivity to rhesus monkeys. This chemotactic
material has an apparent molecular weight of 5000 daltons
or less and was not inactivated by goat antibody to com-
ponents C3 or C5 of human complement; chemotactic
activity was lost after storage for two weeks at 40 and the
activity of two of three preparations was decreased by
heating for 30 min at 560. This previously undescribed
chemotactic activity of dialyzable transfer factor may have
significance in relation to cell-mediated immune re-
sponses.
In 1955, Lawrence demonstrated that material from dis-
rupted human leukocytes was capable of transferring antigen-
specific delayed hypersensitivity to previously unresponsive
recipients (1). This "transfer factor," subsequently shown to
be dialyzable (2), has recently received increased attention
because of its potential application in the therapy of diseases
accompanied by defective cell-mediated immunity. The
studies summarized in this report demonstrate that dialyzable
transfer factor from human leukocytes contains a potent
chemotactic agent which resides in chromatographic fractions
that also contain transfer-factor activity. This previously un-
described activity in transfer-factor preparations may be im-
portant for optimal expression of immunologically mediated
inflammatory responses.
MATERIALS AND METHODS
Preparation and Activity of Transfer Factor. Donors for
transfer factor were selected as previously described (3) by
evaluating their responses to skin tests with mumps skin-test
antigen (Lilly, Indianapolis, Ind.), streptokinase strepto-
dornase (SK-SD) (Varidase, Lederle, Pearl River, N.Y.),
tuberculin (intermediate strength PPD from Connaught,
Toronto, Ontario), and Candida albicans (Hollister-Stier,
Seattle, Wash.). The cutaneous responses were assessed 15
min after injection of skin-test material for immediate wheal
and flare reactions and after 8, 24, and 48 hr for induration.
Abbreviations: SK-SD, streptokinase-streptodornase; PBS, phos-
phate buffered saline; hpf, high-power field; CPM LF, counts per
min lower filter; C3a and C5a, cleavage products of the third and
fifth components of complement.
498
Positive delayed responses produced indurations of 0.5 cm
or greater at 24 or 48 hr.
Peripheral blood leukocytes were collected from the healthy
volunteers by leukapheresis with a continuous-flow blood-cell
separator (4). By using a bowl speed of 1700 rpm, leukocyte
preparations containing over 80% lymphocytes were ob-
tained. Blood leukocytes were also obtained from a 53-year-
old male with the Sezary syndrome, a variant of chronic
lymphocytic leukemia in which a high proportion of the circu-
lating cells are "T"-lymphocytes (5), at times when his white
blood count was in excess of 100,000 cells per mm3 with 98%
small lymphocytes. The leukocytes were washed three times
with 10 volumes of phosphate-buffered saline (PBS) [0.0067
M P04 buffer (pH 7.4) containing 0.85 M NaCl] to remove
plasma components. The dialyzable transfer factor was pre-
pared by our modification (3) of the method of Lawrence (6)
which involves disruption of cells by freezing and thawing,
digestion with pancreatic deoxyribonuclease (Worthington
Biochemicals, Freehold, N.J.), and then dialysis for 48 hr
at 40 against sterile, pyrogen-free, glass-distilled water (NIH
Media Unit). The dialysate was lyophilized and reconstituted
with distilled water so that the extract from 3 X 108 lym-
phocytes was contained in 1 ml. The solution was then
sterilized by millipore filtration and stored at -40°. A "re-
agent blank" was prepared by carrying PBS through the
entire process of freeze-thawing, enzymatic digestion, di-
alysis, lyophilization, and reconstitution.
The activity of transfer factor from normal donors had
been documented in patients with chronic mucocutaneous
candidiasis, as previously described (3). To facilitate studies
of the ability of transfer factor to confer delayed hypersensi-
tivity, a model was developed in rhesus monkeys in which the
test preparations (2 ml) Were injected subcutaneously into
the lower extremities and 5-7 days later intradermal tests
with antigens and PBS were placed across the anterior chest.
The test sites were examined 24 hr later, and, under phency-
clidine (Parke-Davis, Detroit, Mich.) anesthesia, full-thick-
ness skin biopsies were removed with a 4-mm circular punch.
The specimens were fixed for 24 hr in 10% buffered neutral
formalin and then in 70% ethanol. Histologic examinations
were made of 4-jAm sections stained with hematoxylin and
eosin or Giemsa's solution. The slides were coded and inter-
preted independently by two observers.
Characterization of Transfer Factor. Sephadex (G-25, Phar-
macia Fine Chemicals, Piscataway, N.J.) gel filtration was
performed using 0.04 M PBS (pH 7.4) and a 5.0 X 90-cm
column with upward flow. All gel filtration was conducted at
 Proc. Nat. Acad. Sci. USA 71 (1974) Chemotactic Activity in Transfer Factor 499

TABLE 1. Specificity of transfer of delayed hypersensitivity TABLE 2. Granulocyte chemotactic activity of dialyzable
to Rhesus monkeys with transfer factor transfer factor, C5a, and chemotactic lymphokine

Donor Chemotactic Stimulus Chemotaxis

Origin response Monkey 1866 i 178t
of transfer (cm in- re-
Transfer factor from normal donor (10)*
factor Antigen duration) sponse*
Transfer factor from Sezary's syndrome (2), 100 i1 331 ± 26
200 ,x 326 + 18
Normal Mumps 1.7 ++ "Reagent blank" (1) 154 i 8
donors SK-SD 0.8 + C5a (10) 2211 i 125
Tuberculin - - Lymphokinel
Candida No antigen (7) 979 i 221
PBS - - Antigen Negative (7) 1002 i 69
Normal Antigen Positive (9) 2003 i 168
donors Mumps 1.5 ++
SK-SD 1.0 - * Number of different preparations studied is indicated in
Tuberculin parentheses.
Candida 1.0 ++ t Mean i SEM, expressed as corrected counts per minute
PBS - lower filter.
Patient with t No antigen indicates fluid from lymphocytes cultured with-
Sezary out antigen; antigen negative indicates that the lymphocyte
syndrome Mumps 1.0 + donor was unreactive to the skin test; and antigen positive indi-
SK-SD cates that the donor had delayed hypersensitivity to the skin
Tuberculin test.
Candida 0.6
PBS -

number of neutrophils, mononuclear cells, and connective

* +, perivascular infiltration with mononuclear cells; ++,
perivascular and diffuse infiltration with mononuclear cells;
-, no inflammatory response.
4° with sterile, pyrogen-free buffers. Column fractions were
lyphilized, dissolved in minimal volumes of distilled water,
sterilized by millipore filtration, and stored at -40°.
In vitro Assay of Chemotactic Activity. Granulocyte chemo-
taxis was measured using the recently described radioassay
(7) which employs "Cr-labeled granulocytes (5"Cr, Amer-
sham-Searle, Arlington Heights, Ill.) and double micropore-
filter chemotactic chambers. Chemotactic activity was ex-
pressed as corrected counts per minute lower filter. The
chemotactic stimuli were either 100 4l of transfer factor in
2.0 ml Hank's balanced salt solution (NIH Media Unit),
2.0 ml of the cleavage product of the fifth component of com-
plement (C5a), containing 10 jsg of protein per ml [purified
from endotoxin-activated human serum (8)1, or 2.0 ml of a
1:20 dilution of chemotactic lymphokine from antigen-
stimulated peripheral blood lymphocytes (9).
Mononuclear-cell chemotaxis was evaluated utilizing a
previously described morphologic assay (10) employing
mononuclear cells obtained by Hypaque-Ficoll separation
(11) of heparinized whole blood. Mononuclear-cell chemo-
tactic activity was expressed as the average number of mono-
nuclear cells per high-power field (hpf) that migrated to the
lower surface of the micropore filter.
In vivo Assays of Chemotactic Activity. Rhesus monkeys re-
ceived 0.1 and 0.05 ml of transfer factor (3 X 108 lymphocyte
equivalents per ml) at intradermal sites on the anterior chest
wall. Control injections contained the PBS "reagent blank."
With the monkeys under phencyclidine anesthesia, biopsies
were taken at 2, 4, and 24 hr. Additional biopsies were ob-
tained from areas that had not been injected; the rows from
which biopsies were obtained were randomized. Sections (4
Mm) were stained with hematoxylin and eosin. The intensity
of the inflammatory response was scored by counting the
tissue-vascular endothelial cells in 10 oil immersion fields,
and the data were expressed as the mean number of cells.
RESULTS
Transfer of Delayed Cutaneous Hypersensitivity with Trans-
fer Factor. Delayed allergic reactions in monkey skin were not
grossly apparent, therefore microscopic examination of the
test sites was required. The three experiments summarized in
Table 1 illustrate the specificity of the transferred reactions.
In no case did reactions which were not present in the donor
appear in a recipient monkey. However, reactivity to SK-SD
was transferred in only one of two experiments and the re-
sponse was of moderate intensity. The transfer factor prep-
aration from the patient with Sezary syndrome did not trans-
fer reactivity to Candida and the response to the mumps skin
test was weak, even though the donor had significant delayed
allergic responses to these antigens.
In Vitro Chemotactic Activity of Dialyzable Transfer Factor.
Dialyzable transfer factor contained potent chemotactic
activity for human granulocytes. As shown in Table 2, the
mean chemotactic activity of 10 different transfer factor
preparations was comparable to that obtained with C5a and
the chemotactic lymphokine from antigen-stimulated lympho-
cytes. To determine whether the transfer factor preparations
were chemotactic for granulocytes or merely increasing the
spontaneous motility of the leukocytes, experiments were
conducted in which a small amount of transfer factor (5 IA)
was added to the upper compartment of the chemotactic
chamber, thereby diminishing the gradient of the chemotactic
stimulus. Such experiments reduced the chemotactic response
by 35%, which demonstrates the chemotactic property of
the transfer factor preparations.
In contrast to transfer factor from normal donors, transfer
factor prepared from a patient with Sezary's syndrome had
very weak chemotactic activity, even when the concentration
of the transfer factor was doubled (Table 2). The PBS "re-
agent blank" had no chemotactic activity.
500 Immunology: Gallin and Kirkpatrick Proc. Nat. Acad. Sci. USA 71 (1974)
in Fig. 1. A log-linear chemotactic response was obtained
over a dose-range of 10-200 ul of transfer factor added to the
lower compartment of the chemotactic chamber.
Dialyzable transfer factor was also weakly 6hemotactic
for mononuclear cells in the in vitro Boyden chamber assay.
Whereas the chemotactic activity of transfer factor for granu-
locytes was 82% of the chemotactic activity of C5a, the
chemotactic activity of dialyzable transfer factor for mono-
nuclear cells was only 38% of the activity observed with C5a
Ii
0 (498 55 mononuclear cells/hpf for C5a versus 191 ±t 34
mononuclear cells/hpf for transfer factor preparations).
In Vivo Chemotactic Properties of Dialyzable Transfer Factor.
The ability of transfer-factor preparations to attract leuko-
cytes to intradermal sites in vivo was evaluated by histologic
examination of skin biopsies, as summarized in Table 3. Both
doses of transfer factor were strongly chemotactic with dose-
responses similar to the in vitro studies. The effect was maxi-
mal by 2 hr and, with the smaller dose of transfer factor,
was waning by 5 hr. In contrast, infiltration with mononu-
10 50 100 500 clear cells increased with time and by 24 hr was significantly
TRANSFER FACTOR QL liters) greater than the controls. In the same experiments, the trans-
FIG. 1. Chemotactic responses to varying concentrations of fer factor" from a patient with Sezary's syndrome (which had
dialyzable transfer factor. The bars denote the standard error of feeble in vitro chemotactic activity), the "reagent blank,"
the mean at each dose of transfer factor; cor cpm LF, corrected
counts per minute lower filter. (The numbers on the ordinate and PBS did not produce chemotactic responses.
have been multiplied by 10-'.) Chemotactic Activity of Transfer Factor after Gel Filtration.
Preliminary purification of dialyzable transfer factor was
Heating transfer factor preparations at 560 for 30 min had performed by filtration through Sephadex G-25 gel. Column
variable effects on the chemotactic activity: two preparations fractions were screened for absorbance at 255 nm (Fig. 2)
had decreased activity and one preparation was unchanged. and for chemotactic activity. As shown in Fig. 2, a peak of
All preparations studied lost chemotactic activity when granulocyte chemotactic activity was found in tubes 118-
stored at 4° for 2 weeks. Goat antibodies to components C3 156. The monocyte chemotactic activity resided in the same
or C5 of human complement had no effect on the chemotactic
fractions as the granulocyte chemotactic activity. The chemo-
activity of transfer factor, although antibody to C5 markedly tactic activity of these G-25 fractions was labile when stored
inhibited the chemotactic property of C5a. at 40 for 2 weeks or heated at 56° for 30 min, but was un-
A comparison of the time course of chemotaxis for granu- affected by goat antibody to components C3 and C5 of
locytes exposed to C5a or transfer factor revealed that the human complement.
kinetics of cell movement through the micropore filter were Transfer-Factor Activity of Fractions Eluted From the Sepha-
similar for both stimuli. The chemotactic response of granu- dex G-25 Column. The transfer-factor preparation applied
locytes to varying concentrations of transfer factor is shown to the column in these experiments was derived from a donor
TABLE 3. In vivo chemotactic properties of dialyzable transfer factor
Time
biopsies
taken Neutrophil chemotaxist Mononuclear cell chemotaxist
after
Dose and origin of transfer injection Transfer Transfer
factor (hr) factor Diluent P* factor Diluent P*

0.1 ml, fromnormaldonors 2 180.33 ±55.32 4.00± 1.21 <0.01 2.33± 1.45 1.33 ±0.56 NS
4 194.00 ± 51.12 6.83 ± 1.60 <0.001 12.67 ±- 6.49 2.50 ± 1.15 <0.05
24 103.50 ± 34.50 3.00 ±- 1.00 0.10 18.00 ± 5.00 1.50 ± 1.50 <0.05
0.05 ml, from normal donors 2 188.67 ±-76.67 4.00 ± 1.21 <0.01 2.33 ±t 1.45 1.33 ± 0.56 NS
4 95.33 ±-27.23 2.50 ± 1.15 <0.001 1.33 ± 0.67 2.50 ± 1.15 NS
24 71.00 ± 46.00 1. 50 ±1l.50 >0.20 10.00 ± 3.00 1. 50 ± 1.50 0.05
0.1 ml, from patient with
Sezary's syndrome 2 10.00 ± 6. 66 4.00 ± 1.21 >0.20 1.10 ± 1.00 1.33 ± 0.56 NS
4 6.33 ± 1.33 2.50 ±- 1.15 NS 3.67 ± 1.67 2.50 ± 1.15 NS
* Student t-test; N.S. = P > 0.30.
t Mean cells per 10 hpf ± standard error (two different transfer-factor preparations from normal donors and one preparation from a
patient with Sezary's syndrome).
 Proc. Nat. Acad. Sci. USA 71 (1974)
I
E x
(%J
in
-80 100 120 140 160 180 200 220 240 260 280 300 320 340 3
Froction | I | 3 [lXI
SKIN TEST
MUMPS + ++ + _
SK -SD - + - -
FIG. 2. Comparison of chemotactic and skin-test-transferring activities of fractions of transfer factor after gel filtration on Sephadex
G-25. Fraction II contained both biological activities. Column dimensions 5.0 X 90 cm; CPM LF, counts per minute lower filter. (The
numbers on the right ordinate have been multiplied by 10-3.)
with strong delayed allergy to mumps and streptococcal pro-
teins; responses to tuberculin and Candida were negative.
The effluents from the preparative column were pooled into
seven fractions. One pool contained the chemotaotic activity
and the other pools were comprised of peaks of material with
absorbance at 255 nm (Fig. 2). The column fraction contain-
ing the chemotactic activity also contained the skin-test-
transferring activity for both antigens, although the adjacent
fractions also transferred reactivity to mumps. There were no
responses to antigens to which the donor was not reactive.
Protein analysis of the peaks disclosed that the fraction con-
taining both the chemotactic and skin-test-converting prop-
erties contained 65% of the total protein in the effluent.
DISCUSSION
The experiments described in this report identify a chemo-
tactic agent in dialyzable transfer factor preparations which
clearly differs from most previously described chemotactic
substances. For example, the insusceptibility of this chemo-
tactic agent to inactivation with antibodies to human C3 and
C5 and its preparation in the absence of serum suggest that it
is not complement-derived. Although the chemotactic cleav-
age product of the third component of complement (C3a), the
chemotactic enzyme kallikrein (12), and the chemotactic
activity of some preparations of dialyzable transfer factor are
destroyed by heating at 560 for 30 min, the molecular weight
data suggest that the chemotactic component of transfer
factor is different from these substances. Though transfer
factor has not been purified to the extent required for precise
molecular weight measurement, the properties of dialyzability
and penetration of beds of Sephadex G-25 suggest that the
molecular weight is 5000 or less (6, 13, 14); whereas C5a, C3a,
and kallikrein have molecular weights of 17,500, 8,700, and
130,000, respectively (15, 12). The chemotactic lympho-
kine produced by stimulation of sensitive lymphocytes with
specific antigens has a molecular weight of about 40,000 and is
resistant to heating at 56° for 30 min (9). The recently de-
Chemotactic Activity in Transfer Factor 501
-
r
x
5
t
-I
TUBE NUMBER
N I l
_
- - -
scribed neutrophil-derived factor which is chemotactic for
neutrophils is nondialyzable and heat-stable (16).
Lawrence and coworkers originally reported that the skin-
test-converting property of transfer factor was dialyzable
and entered the bed of a Sephadex G-25 column (2). Subse-
quently, Baram, Yuan, and Mosko (13) and Arala-Chavez
et al. (14) confirmed these findings. The latter investigators
(14) identified the skin-test-converting activity in an early
fraction, similar to that shown in Fig. 2, in which the chemo-
tactic activity was eluted in fractions containing transfer-
factor activity for mumps and SK-SD (fraction II) but was
not detectable in fractions I and III which had transfer-
factor activity for mumps. The studies conducted thus far
do not exclude the possibility that multiple molecules are
present, and additional purification procedures may demon-
strate that the chemotactic and transfer factor activities are
separable. In our experiments, chemotactic activity of trans-
fer-factor preparations was variably affected by heating at
56° for 30 min. Lawrence has reported that the skin-test-
coverting activity of transfer factor was heat labile (6), while
Spitler et al. (17) have recently reported that their prepara-
tions of transfer factor were unaffected by heating at 56° for
2 hr. Thus heat lability of both the chemotactic and transfer-
factor activities appears to vary with different preparations.
It is possible that the chemotactic activity present in trans-
fer factor preparations is essential for inflammation associated
with mediation of the delayed hypersensitivity and that any
manipulation which deletes or denatures the chemotactic
activity will also indirectly effect skin-test conversion. It
is, for example, still unclear from which cell-type transfer
factor is produced. Perhaps it is derived from granulocytes or
other chemotactically responsive cells and, as such, the chemo-
tactic activity might play a critical role in the local cell re-
cruitment necessary for adequate processing and amplifica-
tion of the transfer factor "message." Support for a biological
role for chemotactic agents in cell-mediated immune respon-
ses has been offered by Ramseier, who found that lymphoid
502 Immunology: Gallin and Kirkpatrick
cells from hamsters and mice released a "product of antigenic
recognition" upon stimulation with histocompatibility anti-
gens and that this substance induced accumulations of poly-
morphonuclear cells when injected into the skin of normal
hamsters (18). This response may be functionally analogous
to the antigen-cell interactions that lead to release of antigen-
specific transfer factor and therefore related to our observa-
tions.
The chemotactic activity in dialyzable transfer factor prep-
arations may explain two poorly understood observations
with transfer factor. Healthy recipients of transfer factor may
show positive responses to skin tests planted at the site of
injection of the transfer factor (local transfer) before they
respond to tests applied at remote sites (systemic transfer)
(6). Furthermore, when anergic patients with sarcoidosis were
given transfer factor, conversion of delayed allergic responses
was observed only with tests injected in the vicinity of the
transfer factor (local transfer), while no conversion occurred
at remote sites (19, 20). The chemotactic activity in transfer
factor could participate in these phenomena by "preparing"
local sites either through facilitation of antigen processing or
amplification of subclinical inflammatory stimuli.
The authors are deeply indebted to Mr. Terrill K. Smith for
his expert technical assistance. Dr. Leonard Altman provided
the antisera to complement components C3 and C5. We are
especially grateful to Dr. Douglas Lorenz for providing facilities
for the rhesus monkeys and Dr. Allen Kaplan for critical review
of the manuscript.
1. Lawrence, H. S. (1955) J. Clin. Invest. 34, 219-230.
Proc. Nat. Acad. Sci. USA 71 (1974)
2. Lawrence, H. S., Al-Askari, S., David, J., Franklin, E. C. &
Zweiman, B. (1963) Trans. Ass. Amer. Phys. 76, 84-89.
3. Kirkpatrick, C. H., Rich, R. R. & Smith, T. K. (1972) J.
Clin. Invest. 51, 2948-2958.
4. Buckner, D., Graw, R. G., Eisele, R. J., Henderson, E. S. &
Perry, S. (1969) Blood 33, 353-369.
5. Zucker-Franklin, D., Melton, J. W. & Quagliata, F. (1973)
J. Clin. Invest. 52, 92a-93a.
6. Lawrence, H. S. (1969) Advan. Immunol. 11, 195-266.
7. Gallin, J. I., Clark, R. A. & Kimball, H. R. (1973) J. Im-
munol. 110, 233-240.
8. Gallin, J. I., Clark, R. A. & Frank, M. M. (1973) Clin. Res.
21, 579.
9. Ward, P. A., Remold, H. G. & David, J. R. (1970) Cell.
Immunol. 1, 162-174.
10. Snyderman, R., Altman, L. C., Hausman, M. S. & Mergen-
hagen, S. E. (1972) J. Immunol. 108, 857-860.
11. Boyum, A. (1968) Scand. J. Clin. Lab. Invest. 97, Suppl.
21, 77-89.
12. Kaplan, A. P., Kay, A. B. & Austen, K. F. (1972) J. Exp.
Med. 135, 81-97.
13. Baram, P., Yuan. L. & Mosko, M. M. (1966) J. Immunol.
97, 407-420.
14. Arala-Chaves, M. P., Lebacq, E. G. & Heremans, J. F.
(1967) Int. Arch. Allergy Appl. Immunol. 31, 353-365.
15. Vallota, E. H. & Muller-Eberhard, H. J. (1973) J. Exp. Med.
137, 1109-1123.
16. Zigmond, S. H. & Hirsch, J. G. (1973) J. Exp. Med. 137,
387-410.
17. Spitler, L. E., Webb, D., von Muller, C. & Fudenberg, H. H.
(1973) J. Clin. Invest. 52, 80a.
18. Ramseier, H. (1969) J. Lxp. Med. 130, 1279-1294.
19. Urbach, F., Sones, M. & Israel, H. L. (1952) N. Engl. J.
Med. 247, 794-797.
20. Lawrence, H. S. & Zweiman, B. (1968) Trans. Ass. Amer.
Amer. Phys. 81, 240-247.