BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 248, 297–302 (1998)
ARTICLE NO. RC988953
Diphtheria Toxin:Receptor Interaction: Association,
Dissociation, and Effect of pH
Joanna S. Brooke, Jeong-Heon Cha, and Leon Eidels1
Department of Microbiology, The University of Texas Southwestern Medical Center,
6000 Harry Hines Boulevard, Dallas, Texas 75235-9048
Received June 8, 1998
Diphtheria toxin (DT) binds to a specific heparin-
binding epidermal growth factor-like growth factor
(HB-EGF) precursor that is expressed in DT-sensitive
cells. DT binds to the cell-surface HB-EGF precursor
with an apparent dissociation constant (KD) of É 1 1
1008 0 1009 M at 47C, a temperature at which toxin binds
but is not internalized. The interaction of DT with the
cell-surface receptor, however, may be influenced by
other cell-surface components. We used a biosensor
method to measure the binding of DT to immobilized
recombinant human HB-EGF (hHB-EGF) at 257C with
no other cellular components present. We observed
that at pH 7.4, using this in vitro two component sys-
tem, DT binds to hHB-EGF with an apparent KD of 2.7
1 1008 M. We also observed that the dissociation of DT
from hHB-EGF at pH values that approach those of the
endosome occurs at a faster rate as the pH is de-
creased. These results suggest that the low pH of the
endosome is sufficient to allow DT to dissociate from
the HB-EGF precursor, prior to the translocation of
the enzymatically active fragment of DT into the cyto-
sol. q 1998 Academic Press
Diphtheria toxin (DT) is an exotoxin that is secreted
from lysogenic Corynebacterium diphtheriae as a poly-
peptide of Mr Å 58,340 (1); DT may be proteolytically
digested to yield two disulfide-linked fragments, an A
fragment of Mr Å 21,167 and a B fragment of Mr Å
37,195 (2). DT kills toxin-sensitive eukaryotic cells by
inhibiting protein synthesis. The toxin binds, via the
receptor-binding domain of its B fragment, to specific
cell-surface receptors, and toxin:receptor complexes
then are internalized via clathrin-dependent endocyto-
sis into an endosomal compartment (3-5). When the
toxin is exposed to the low pH environment of the endo-
1
Author to whom correspondence should be addressed. Fax: (214)
648-5906. E-mail: LEIDEL@MEDNET.SWMED.EDU.
297
AID BBRC 8953 / 6958$$$361 07-06-98 09:36:25
some, the B fragment undergoes a conformational
change and inserts into the endosomal membrane bi-
layer (6); this insertion facilitates the translocation of
the A fragment from the endosome into the cytosol and
is followed by the reduction of the interfragment disul-
fide bond (7). In the cytosol, the A fragment catalyzes
the ADP-ribosylation of elongation factor 2 molecules
resulting in the inhibition of protein synthesis (8-10).
Since all eukaryotic elongation factor 2 molecules can
be ADP-ribosylated but not all eukaryotic cells are sen-
sitive to DT, a functional receptor for DT must be pres-
ent in order to initiate the cytotoxicity process (11,12).
As a result of studying the cytotoxicity of DT in
HeLa cells in the presence or absence of CRM197, a
competitor mutant DT, Ittelson and Gill (13) hypoth-
esized that there were specific cell-surface receptors
for DT. The non-toxic mutant toxin CRM197 has an
intact B fragment but an enzymatically inactive A
fragment (14). Dose-response curves measuring pro-
tein synthesis inhibition of HeLa cells indicated that
CRM197 competitively inhibited the binding of DT
to the cells (13). Ittelson and Gill (13) proposed that
the toxin was binding to a surface receptor prior to
passage through the cell membrane and into the cyto-
sol, and that CRM197 was able to inhibit this binding
to cells. An apparent dissociation constant (KD) of É
1008 M was reported for the CRM197:receptor com-
plex from a Schild plot of the inhibition of DT activity
by CRM197 at 377C (13). Since DT and CRM197 have
identical B fragments and the B fragment is neces-
sary for the entry of DT into the cytosol, the investi-
gators reasoned that the apparent KD value for the
DT:receptor complex must be similar to that observed
with the CRM197:receptor complex (13).
Middlebrook et al. (12) examined the binding of ra-
diolabeled DT to highly DT-sensitive Vero cells and
derived an apparent association constant (KA) of 9 1
108 M01 (an apparent KD of 1.11 1 1009 M) from a
Scatchard plot of the binding data. In this system DT
was allowed to bind to cells at 47C, a temperature at
0006-291X/98 $25.00
Copyright q 1998 by Academic Press
All rights of reproduction in any form reserved.
bbrcgs AP: BBRC
Vol. 248, No. 2, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
which binding occurs but internalization of the toxin
does not. The low temperature, however, appeared to
alter the rate of association of DT with the cells (12),
and may have also affected the rate of dissociation of
DT from the toxin:receptor complex as compared to the
actual rate observed in vivo.
The ability to measure the binding of DT to DT-sensi-
tive cells facilitated the search for the cell surface DT
receptor. Immunoprecipitation and cross-linking ex-
periments using radiolabeled DT revealed a range of
Mr É 10,000 0 20,000 for the cell surface receptor (15).
In our laboratory the identification and eventual isola-
tion of the gene that encodes the DT receptor was ac-
complished by expressing cDNA from highly DT-sensi-
tive monkey kidney Vero cells in the normally DT-resis-
tant mouse L-cells and by screening for DT-sensitive
mouse cells (16-18). The specific binding of radioiodin-
ated DT to these DT-sensitive mouse L-cells was mea-
sured at 47C, and an apparent KD was obtained that
was approximately 10-fold higher than the apparent
KD obtained with Vero cells (18). The receptor for DT
was identified as a cell surface-expressed heparin-bind-
ing epidermal growth factor-like growth factor (HB-
EGF) precursor (18). Analysis of the amino acid se-
quence of the HB-EGF precursor showed that the struc-
ture contains a signal sequence (amino acid residues 1-
23) and is composed of three domains: an extracellular
domain (residues 24-159), a transmembrane domain
(residues 160-184) and a carboxy-terminal cytoplasmic
domain (residues 185-208) (18). Specific proteolytic
cleavage within the extracellular domain of the HB-
EGF precursor results in the release of mature HB-
EGF (residues 63-148) (19).
Using eukaryotic cells expressing mutant DT recep-
tors, amino acids within the EGF-like domain of the
receptor (corresponding to residues 115-148) have been
identified as important for DT binding (20-24). More
recently we have identified three amino acid residues
that lie within the mature HB-EGF region of the DT
receptor as the most important residues for binding to
DT, for toxin sensitivity, and for toxin affinity (24).
Furthermore, recombinant mature human HB-EGF
(hHB-EGF) has been shown to effectively inhibit the
binding of radiolabeled DT to DT receptor-expressing
cells (20). This result indicates that mature HB-EGF
is able to act as a soluble DT-receptor analogue.
In this study, we report the association and dissocia-
tion rates for DT binding to hHB-EGF at 257C using a
biosensor kinetic analysis. We also compared the disso-
ciation rates of DT from the DT:hHB-EGF complex at
different pH values. We employed the BIAcore instru-
ment (Pharmacia) in which protein-protein interac-
tions are observed as changes in refractive index over
time as one protein (analyte) binds to a second protein
(ligand) that is covalently coupled to a sensor chip (25).
Using this technology, we immobilized hHB-EGF on a
298
AID BBRC 8953 / 6958$$$362 07-06-98 09:36:25
biosensor chip and examined the kinetics of DT binding
to hHB-EGF. We found that DT dissociates from hHB-
EGF at neutral pH with an apparent KD É 1008 M that
is similar to values reported for binding of DT to DT
receptor-expressing cells. We found that in an acidic
environment DT dissociates from hHB-EGF at a faster
rate compared to that at neutral pH, supporting a
model in which DT dissociates from the receptor at the
low pH of the endosomal compartment prior to inser-
tion and translocation.
MATERIALS AND METHODS
Materials. DT was purchased from Connaught Laboratories and
purified as described (15). Recombinant hHB-EGF was obtained from
R & D Systems. We used hHB-EGF as a DT-receptor analogue; the
amino acid sequences for the mature HB-EGF of the human and the
monkey proteins differ in only one amino acid that is not important
for binding to DT, residue 87 which is Asn in the human and Ser in
the monkey (18). The dehydrated tissue culture medium 199 was
purchased from Gibco BRL. All other chemicals were purchased from
Sigma.
Instrumentation and reagents. The BIAcore 2000 system and
CM5 sensor chips were obtained from Pharmacia. The buffers used
with the BIAcore 2000 system were: HEPES-buffered saline (HBS)
(10 mM HEPES, pH 7.4, 150 mM NaCl and 3.4 mM EDTA); HBS-
P20 (HBS containing 0.05% (v/v) BIAcore Surfactant P20); running
buffer (tissue culture medium 199 with Hanks salts, 50 mg/ml BSA,
100 mg/ml gelatin, 20 mM HEPES, pH 7.4), and three other running
buffers with the same components as the previous buffer but with
20 mM sodium maleate buffer substituted for the HEPES buffer, and
the pH of these buffers was adjusted to 6.9, 6.4, and 5.8, respectively;
regeneration buffer (10 mM glycine/HCl, pH 2.0, with 0.005% (v/v)
Tween 20, 150 mM NaCl); carboxymethylated dextran activation
solutions (0.1 M N-hydroxysuccinimide and 0.1 M N-ethyl-N*-(3-di-
ethylaminopropyl carbodiimide); coupling buffer (10 mM sodium ace-
tate, pH 4.0), and deactivation solution (1 M ethanolamine, pH 8.5).
Protein immobilizations were performed using the standard BIAcore
amine coupling protocol provided by Pharmacia.
Sensor chip configuration. Recombinant hHB-EGF was immobi-
lized by standard amine coupling using 10 mM sodium acetate, pH
4.0 as the coupling buffer and using a flow rate of 10 ml/min. A
running buffer flow rate of 30 ml/min was used in the kinetic experi-
ments.
Kinetic analysis. The coupling of hHB-EGF onto the chips re-
sulted in a range of 125-289 resonance units (RU) immobilized on the
sensor chips. DT, resuspended in running buffer at concentrations of
25 nM-1 mM, was injected using the kinject method of injection and
allowed to flow at a temperature of 257C over the CM5 chip with
immobilized hHB-EGF. DT at the concentrations of 25-200 nM when
injected over the immobilized hHB-EGF did not produce enough data
points to use in the 1:1 analysis of the DT:hHB-EGF complex, and
thus were not employed. A negative control used to monitor non-
specific binding to the sensor chip surface was achieved by activating
and inactivating the carboxymethylated dextran in the first flow cell
without coupling any ligand to this flow cell. To eliminate possible
bulk refractive index effects, the surface plasmon response observed
for running buffer passed over this first flow cell was subtracted from
the binding curves observed for the interaction of DT with hHB-
EGF. For the kinetic analysis of DT dissociation from hHB-EGF
at different pH conditions, the surface plasmon resonance response
observed for running buffer passed over the immobilized hHB-EGF
was subtracted from the binding curves observed for the dissociation
bbrcgs AP: BBRC
Vol. 248, No. 2, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Binding of DT to immobilized hHB-EGF. The first arrow represents the start of the injection of DT over the immobilized hHB-
EGF. The arrow with a ball represents the end of the DT injection and the beginning of the flow of the running buffer. From the bottom
curve to the top curve, the respective binding curves for DT of 400 nM, 500 nM, 600 nM, 700 nM, 800 nM, and 1 mM concentrations are
shown. The numbers ①, ②, and ③ represent the baseline of running buffer, the association phase of the binding of DT to hHB-EGF, and
the dissociation phase of the release of DT from hHB-EGF, respectively. RU, resonance units. This figure shows a representative experiment
(see Materials and Methods).

of DT from hHB-EGF. Version 2.1 of the BIAcore software package
was used to analyze the curves. The values for the dissociation rate
constant (kd) were obtained using the dissociation part of the curve
and plotting ln(RUi/Rt) vs. time, where RUi represents the initial
resonance units response and Rt represents the resonance units at
a given time. The slope of this plot was used to derive the kd value.
To obtain the values for the association rate constant (ka), first the
plots of dRU/dt vs. R for each concentration of DT were derived, the
slopes of these curves were then plotted vs. DT concentration, and
the slope of this second plot yielded the ka value. The values for the
apparent dissociation constant, KD , were determined by using the
formula KD Å kd/ka . The kinetic experiment was performed twice,
each time using a different chip. The pH experiment was performed
twice (once at 600 nM DT and once at 800 nM DT), each time using
a different chip. The values obtained for the binding of DT to hHB-
EGF cannot be compared between the kinetics experiment (Fig. 1)
and the effect of pH experiment (Fig. 2) because a HEPES-based
running buffer was used in the former and maleate-based running
buffers were used in the latter.
RESULTS AND DISCUSSION
DT binds to recombinant human HB-EGF with an
apparent KD of 2.7 1 1008 M. To examine the bind-
ing of DT to hHB-EGF, increasing concentrations of
DT were injected over the hHB-EGF immobilized on
the CM5 sensor chip. No binding was observed when
running buffer was passed over the immobilized
hHB-EGF (Fig. 1, ①), indicating that none of the com-
ponents in the running buffer bind to hHB-EGF. The
binding of DT (400 nM-1 mM) to hHB-EGF produced
the binding curves shown in Fig. 1; the interaction
of DT with hHB-EGF displayed first order kinetics
and permitted us to use a 1:1 analysis for the study
of DT association and dissociation in the DT:hHB-
EGF complex. The calculated average values for ka
and kd were 7.9 1 104 M01s01 and 2.0 1 1003 s01,
respectively, and the calculated average value for the
apparent KD was 2.7 1 1008 M (Table 1). This KD
value is within the range of biological affinities (1008
0 1009 M) described using biosensor analysis for such
ligand-receptor interactions as EGF with its recep-
tor, ligands for receptor tyrosine kinases, and human
interleukin 5 with its receptor (26-29). The apparent
KD value of 2.7 1 1008 M obtained here is in agree-
ment with the reported value of 1.5 1 1008 M ob-
tained from an in vivo binding assay using 125I-la-
belled DT and DT receptor-expressing L-M(TK0) cells
(30). In contrast, the binding of radioiodinated DT to
Vero cells resulted in an apparent KD of 1.1 0 2.0 1
1009 M (12, 30); this KD is approximately 10-fold
lower than the KD we obtained in this study. There-
fore, since the value of the apparent KD determined
in the present study is derived from the kinetics of
the interaction of only two components-DT and hHB-
EGF- it is reasonable to expect that in Vero cells,
which display a higher toxin affinity, there may be
additional factor(s) that also participate in the inter-
action of DT and the cell-surface HB-EGF precursor.

299

AID BBRC 8953 / 6958$$$362 07-06-98 09:36:25 bbrcgs AP: BBRC

Vol. 248, No. 2, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABLE 2
Dissociation of DT from hHB-EGF at Acidic pH Conditions

DT pH of kd 1 103
(nM) running buffer (1/s)

600 6.93 1.3 (0.1)

600 6.42 1.9 (0.1)
600 5.84 3.2 (0.1)

Note. The dissociation of DT from hHB-EGF was measured at 257C

using the maleate-based running buffers described in Materials and
Methods. The running buffers were used at a flow rate of 30 ml/
min. The standard error values are shown in parentheses. This table
shows values for a representative experiment (see Materials and
Methods). The kd values obtained in this experiment cannot be di-
rectly compared to those obtained in the kinetics experiment (Table
1) where a HEPES-based buffer system was employed.

receptor participates in the insertion and translocation

process. Therefore, to test whether DT dissociates from
hHB-EGF at acidic pH values, the following experi-
FIG. 2. Dissociation of DT from immobilized hHB-EGF in run- ment was performed. hHB-EGF was immobilized on a
ning buffers of decreasing pH. The results for DT of 600 nM concen-
tration in running buffers of pH 6.9, 6.4, and 5.8 are shown. The
CM5 chip and DT at a concentration of 600 nM was
association phase (at pH 7.4) of these curves is not shown. The origin injected in running buffer (pH 7.4) over the hHB-EGF.
represents the end of the DT injection and is the time at which the After the injection had finished, the dissociation of DT
running buffer of specific pH has started flowing over the sensor from hHB-EGF was allowed to occur in one of the run-
chip. RU, resonance units. This figure shows a representative experi- ning buffers of pH 6.9, 6.4, or 5.8 (see Materials and
ment (see Materials and Methods).
Methods); as the pH of the running buffer decreased,
DT dissociated at faster rates from hHB-EGF (Fig. 2
DT dissociates from hHB-EGF at faster rates under and Table 2). Interestingly, elution at pH 5.3 (and be-
conditions of decreasing pH. The exact nature of the low) did not result in a decrease in response (data not
events occurring in the endosomal compartment prior shown), suggesting that the toxin had become dena-
to insertion and translocation have not been fully eluci- tured and adhered to the chip. The denatured and pos-
dated. It is possible that the toxin dissociates from its sibly aggregated toxin bound to hHB-EGF on the BIA-
core chip could not be dissociated using a variety of
receptor at the low pH of the endosome or that the
such regeneration conditions as high pH, salts, and
nonionic detergents (data not shown). Our result of DT
TABLE 1 denaturation at pH 5.3 is interesting because at this
Binding of DT to hHB-EGF
pH at 377C, DT is in an environment that approaches
the pH found in endosomes (31, 32), and the toxin is
DT ka 1 1004 kd 1 103 KD 1 108 in a more hydrophobic state than at neutral pH (33-
(nM) (1/Mr1/s) (1/s) (M) 35). Furthermore, at a low pH, DT binds nonionic deter-
gents (33, 36, 37), binds lipids tightly and rapidly, and
400 12.2 (4.4) 1.9 (0.1) 1.6 penetrates them due to hydrophobic and electrostatic
500 9.0 (1.3) 1.9 (0.1) 2.1
600 5.3 (1.5) 2.1 (0.1) 4.0 interactions (38). It has been suggested that at this low
700 7.5 (0.8) 2.0 (0.1) 2.7 pH DT is in a transition state where it is capable of
800 6.6 (1.4) 1.9 (0.1) 2.9 inserting into the endosomal membrane prior to the
1000 6.9 (4.2) 1.9 (0.1) 2.8 translocation of the A fragment into the cytosol (6).
Average 7.9 (2.3) 2.0 (0.1) 2.7 (0.6) Blewitt et al. (37) reported that the toxin undergoes
Note. The binding of DT to hHB-EGF was measured at 257C using conformational changes when at low pH (° pH 5 at
running buffer (tissue culture medium 199 with Hanks salts, 50 mg/ 237C, ° pH 5.3 at 377C); tryptophan residues concealed
ml BSA, 100 mg/ml gelatin, 20 mM HEPES, pH 7.4) at a flow rate within the toxin structure are exposed and the toxin
of 30 ml/min. The individual values for ka and kd are shown with becomes hydrophobic and irreversibly aggregates. In
their standard errors in parentheses. The calculated averages for ka the cell at the low pH of the endosome (É pH 5), instead
and kd are shown with their average standard errors in parentheses.
The calculated average Kd is shown with its average deviation in of aggregating, the hydrophobic toxin penetrates the
parenthesis. This table shows values for a representative experiment endosomal membrane and the translocation process is
(see Materials and Methods). initiated.

300

AID BBRC 8953 / 6958$$$362 07-06-98 09:36:25 bbrcgs AP: BBRC

Vol. 248, No. 2, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
A recent study (39) has reported, by following the
fate of DT through the classical endocytic pathway,
that most of the DT passes through the early endo-
somes, endocytic carrier vesicles, late endosomes, and
finally is degraded within lysosomes. DT fragment A
is translocated from the early endosomes into the cyto-
sol in an acidification-dependent, ATP-dependent, and
cytosolic factor-dependent manner (39). The rate of re-
lease of DT fragment A from the endosome can be con-
sidered as a rate-limiting step in the cytotoxic action
of DT. In the case of DT and its receptor, since the
translocation of fragment A of DT from the endosome
to the cytosol can take minutes to occur (40), it is feasi-
ble to suggest that an increased rate of dissociation
of DT from the HB-EGF precursor/receptor within the
endosome would facilitate this translocation.
We have used a biosensor technology to look at the
continuous binding of DT to hHB-EGF with time. The
main advantage of the biosensor method is that only
two components are necessary, soluble analyte and im-
mobilized ligand. The system allows us to study the
effect on the DT:hHB-EGF interaction of such environ-
mental factors as temperature, osmolarity, and pH
without being limited by the effect of these factors on
the cells per se. In this study, we have demonstrated
that the strong interaction between DT and its receptor
measured on the surface of cells can be simulated em-
ploying purified toxin and immobilized hHB-EGF. We
have further demonstrated that DT dissociates faster
from hHB-EGF as the pH is decreased, an observation
consistent with a model in which the toxin dissociates
from its receptor prior to insertion into the endosomal
membrane.
ACKNOWLEDGMENTS
We thank Robert S. Munford, Sergei Popov, and Christine Ward
for critical review of the manuscript. We also thank Kathy Potter
and Sally Ward for helpful discussions in the early part of these
studies. The editorial assistance of Eleanor R. Eidels is greatly ap-
preciated. This research was supported by United States Public
Health Service Grant AI-16805.
REFERENCES
1. Greenfield, L., Bjorn, M. J., Horn, G., Fong, D., Buck, G. A., Col-
lier, R. J., and Kaplan, D. A. (1983) Proc. Natl. Acad. Sci. USA
80, 6853–6857.
2. Pappenheimer, A. M., Jr., (1977) Annu. Rev. Biochem. 46, 69–
94.
3. Keen, J. H., Maxfield, F. R., Hardegree, M. C., and Habig, W. H.
(1982) Proc. Natl. Acad. Sci. USA 79, 2912–2916.
4. Morris, R. E., Gerstein, A. S., Bonventre, P. F., and Saelinger,
C. B. (1985) Infect. Immun. 50, 721–727.
5. Moya, M., Dautry-Varsat, A., Goud, B., Louvard, D., and Boquet,
P. (1985) J. Cell Biol. 101, 548–559.
6. London, E. (1992) Mol. Microbiol. 6, 3277–3282.
301
AID BBRC 8953 / 6958$$$363 07-06-98 09:36:25
7. Papini, E., Rappuoli, R., Murgia, M., and Montecucco, C. (1993)
J. Biol. Chem. 268, 1567–1574.
8. Collier, R. J. (1975) Bacteriol. Rev. 39, 54–85.
9. Eidels, L., Proia, R. L., and Hart, D. A. (1983) Microbiol. Rev.
47, 596–620.
10. Middlebrook, J. L., and Dorland, R. B. (1984) Microbiol. Rev. 48,
199–221.
11. Middlebrook, J. L., and Dorland, R. B. (1977) Can. J. Microbiol.
23, 183–189.
12. Middlebrook, J. L., Dorland, R. B., and Leppla, S. H. (1978) J.
Biol. Chem. 253, 7325–7330.
13. Ittleson, T. R., and Gill, D. M. (1973) Nature (London) 242, 330–
332.
14. Uchida, T., Pappenheimer, A. M. Jr., and Harper, A. A. (1972)
Science 175, 901–903.
15. Cieplak, W., Gaudin, H. M., and Eidels, L. (1987) J. Biol. Chem.
262, 13246–13253.
16. Naglich, J. G., and Eidels, L. (1990) Proc. Natl. Acad. Sci. USA
87, 7250–7254.
17. Naglich, J. G., Rolf, J. M., and Eidels, L. (1992) Proc. Natl. Acad.
Sci. USA 89, 2170–2174.
18. Naglich, J. G., Metherall, J. E., Russell, D. W., and Eidels, L.
(1992) Cell 69, 1051–1061.
19. Higashiyama, S., Lau, K., Besner, G. E., Abraham, J. A., and
Klagsbrun, M. (1992) J. Biol. Chem. 267, 6205–6212.
20. Hooper, K. P., and Eidels, L. (1995) Biochem. Biophys. Res. Com-
mun. 206, 710–717.
21. Mitamura, T., Higashiyama, S., Taniguchi, N., Klagsbrun, M.,
and Mekada, E. (1995) J. Biol. Chem. 270, 1015–1019.
22. Hooper, K. P., and Eidels, L. (1996) Biochem. Biophys. Res. Com-
mun. 220, 675–680.
23. Mitamura, T., Umata, T., Nakano, F., Shishido, Y., Toyoda, T.,
Itai, A., Kimura, H., and Mekada, E. (1997) J. Biol. Chem. 272,
27084–27090.
24. Cha, J.-H., Brooke, J. S., and Eidels, L. (1998) Submitted.
25. Karlsson, R., and Stahlberg, R. (1995) Anal. Biochem. 228, 274–
280.
26. Zhou, M., Felder, S., Rubinstein, M., Hurwitz, D. R., Ullrich, A.,
Lax, I., and Schlessinger, J. (1993) Biochemistry 32, 8193–8198.
27. Nagata, K., Ohashi, K., Nakano, T., Arita, H., Zong, C., Hana-
fusa, H., and Mizuno, K. (1996) J. Biol. Chem. 271, 30022–
30027.
28. Laminet, A. A., Apell, G., Conroy, L., and Kavanaugh, W. M.
(1996) J. Biol. Chem. 271, 264–269.
29. Johanson, K., Appelbaum, E., Doyle, M., Hensley, P., Zhao, B.,
Abdel-Meguid, S. S., Young, P., Cook, R., Carr, S., Matico, R.,
Cusimano, D., Dul, E., Angelichio, M., Brooks, I., Winborne, E.,
McDonnell, P., Morton, T., Bennett, D., Sokoloski, T., McNulty,
D., Rosenberg, M., and Chaiken, I. (1995) J. Biol. Chem. 270,
9459–9471.
30. Brown, J. G., Almond, B. D., Naglich, J. G., and Eidels, L. (1993)
Proc. Natl. Acad. Sci. USA 90, 8184–8188.
31. Geisow, M. J., and Evans, W. H. (1984) Exp. Cell Res. 150, 36–
46.
32. Tycko, B., and Maxfield, F. R. (1982) Cell 28, 643–651.
33. Blewitt, M. G., Zhao, J.-M., McKeever, B., Sarma, R., and Lon-
don, E. (1984) Biochem. Biophys. Res. Commun. 120, 286–290.
34. London, E., Zhao, J.-M., Chattopadhyay, A., Blewitt, M. G.,
McKeever, B., and Sarma, R. (1984) Ann. N.Y. Acad. Sci. 435,
558–560.
35. Collins, C. M., and Collier, R. J. (1987) in Membrane-Mediated
bbrcgs AP: BBRC
Vol. 248, No. 2, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Cytotoxicity, (Bonavida, B., and Collier, R. J., Eds.), pp. 41–52,
Liss, New York.
36. Sandvig, K., and Olsnes, S. (1981) J. Biol. Chem. 256, 9068–
9076.
37. Blewitt, M. G., Chung, L. A., and London, E. (1985) Biochemistry
24, 5458–5464.
38. Chung, L. A., and London, E. (1988) Biochemistry 27, 1245–
1253.
39. Lemichez, E., Bomsel, M., Devilliers, G., vanderSpek, J., Mur-
phy, J. R., Lukianov, E. V., Olsnes, S., and Boquet, P. (1997) Mol.
Microbiol. 23, 445–457.
40. Marnell, M. H., Shia, S.-P., Stookey, M., and Draper, R. K. (1984)
Infect. Immun. 44, 145–150.

302

AID BBRC 8953 / 6958$$$363 07-06-98 09:36:25 bbrcgs AP: BBRC