normal ATPase activity (e.g. ). The changes in pH
and ⌬⌿ are sufficient to explain the drug resistance
[5, 6]. Along with other work (e.g. ) it is quite clear
at this point that Pgp overexpression causes changes in
and ⌬⌿ that can easily explain drug resistance
caused by Pgp alone. These data argue strongly for al-
ternate “indirect” models. Various aspects of these mod-
els actually have a rich history (e.g. [4, 8]) that is also
neglected by Sharom.
There are other omissions. It is stated “Pgp ATPase
activity can be further stimulated by the addition of...
drugs...” Itisnotstated that this is also true for other
ATPases, none of which have ever been hypothesized to
pump hundreds of structurally divergent drugs (e.g.
[9–11]). When discussing the altered partitioning model
for Pgp-mediated MDR, Sharom implies that previous
work suggesting Pgp is either a Cl
or ATP channel is
somehow intricately linked to the altered partitioning
model. This is simply not true; they are distinct obser-
vations and hypotheses. In fact our best published data
 suggest Pgp perturbs both Cl
characteristics that are not necessarily “channel-like.”
A number of ion transport models are compatible with
the altered partitioning model, and the details are still
being elucidated. Also, although it is true Ruetz & Gros
do not measure an effect of ⌬
on distribution of vin-
blastine , other investigators have in other model
systems , a point Sharom omits. To conclude this
section, Sharom writes “Although cytosolic alkaliniza-
tion and alterations in ⌬⌿ have been measured in some
MDR cells, these observations are not universal....”
In fact they are universal, when one measures them prop-
erly using cells engineered to overexpress Pgp without
selection on chemotherapeutics (e.g. [5, 7, 14,
15]). Since chemotherapeutics perturb pH
via pathways independent of Pgp , this distinction is
actually quite important.
These are only a few of the points raised by Sharom
that really require much more thorough discussion. Via
the drug pump model, Pgp has been proposed to catalyze
dozens of very surprising (if not outright bizarre) trans-
port reactions, and our intuition demands that we view
this skeptically. Although it may appear more complex
to some, the altered partitioning model for Pgp is actually
much simpler, does not violate fundamental laws of bi-
ology and chemistry as does the drug pump model ,
and can conceivably explain the vast majority of pub-
lished observations made with unadulterated systems.
The same cannot be said for various pump models.
Paul D. Roepe, Ph.D.
Department of Chemistry & Lombardi Cancer Center
Washington, DC 20057
1. Wadkins, R.M., Roepe, P.D. 1997. Int. Rev. Cytol. 171:121–165
2. Simon, S.M., Roy, D., Schindler, M. 1994. Proc. Natl. Acad. Sci.
3. Wadkins, R.M., Houghton, P.J. 1993. Biochim. Biophys. Acta
4. Ramu, A., Pollard, H.B., Rosario, L.M. 1989. Int. J. Cancer 44:
5. Hoffman, M.M., Wei, L.-Y., Roepe, P.D. 1996. J. Gen. Physiol.
6. Robinson, L.J., Roepe, P.D., 1996. Biochem. Pharmacol. 52:1081–
7. Hoffman, M.M., Roepe, P.D. 1997. Biochemistry 36:11153–11168
8. Skovsgaard, T. 1990. In: Drug Resistance: Mechanisms and Re-
versal. pp. 209–224. John Libbey, London
9. McKenna, E., Smith, J.S., Coll, K.E., Magack, E.K., Mayer, E.J.,
Antanavage, J., Weidman, R.T., Johnson, R.G. Jr. 1996. J. Biol.
10. Tsakiris, S., Deliconstantinos, G. 1984. Biochemical J. 220:301–
11. Paik, S.R. 1993. J. Bioenergetics & Biomembranes 25:679–684
12. Reutz, S., Gros, P. 1994. J. Biol. Chem. 269:12277–12284
13. Mayer, L.D., Bally, M.B., Hope, M.J., Cullis, P.R. 1985. Biochim.
Biophys. Acta 816:294–302
14. Robinson, L.J., Roberts, W.K., Ling, T.T., Lomming, D., Stern-
berg, S.S., Roepe, P.D. 1997. Biochemistry 36:11169–11178
15. Luz, J.G., Wei, L.Y., Basu, S., Roepe, P.D. 1994. Biochemistry
16. Wei, L.-Y., Hoffman, M.M., Roepe, P.D. 1997. Am. J. Physiol.
Received: 7 July 1998
A more detailed consideration of the altered partitioning
model advanced by Dr. Roepe exists in the literature; I
did refer to a review article he wrote on this topic (Ref.
80 in Sharom, 1997). I chose to discuss the pump model
of P-glycoprotein (Pgp) in more depth because, in my
opinion, the vast bulk of the experimental evidence sup-
ports it. Dr. Roepe cites a list of observations that he
believes are incompatible with the pump model. How-
ever, one could make a much longer list of experimental
observations that cannot be made to fit the altered par-
Clearly, it has not been easy to test the pump model
in intact MDR cells, since many complex interacting
systems are present. But the same is true for the altered
partitioning model, where all of the work appears to have
been carried out in intact cells. This is why the move
towards purified reconstituted systems has been so im-
72 Letters to the Editor