drug resistance and membrane alteration in mutants of mammalian cells

DRUG RESISTANCE AND MEMBRANE ALTERATION IN MUTANTS OF MAMMALIAN CELLS l

VICTOR LING The Ontario Cancer Institute, atrd the Department ($Medical Biophysics, C:niversily of Toronto,

Toronto, Ontario, Canada M4X 1K9

Independent colchicine-resistant (CHR) mutants of Chinese hamster ovary cells displaying reduced permeability to colchicine have been isolated. A distinguishing feature of these membrane-altered mutants is their pleiotropic cross-resistance to a variety of unrelated compounds. Genetic characterization of the CHR lines indicates that colchicine resistance and cross-resistance to other drugs are of a dominant nature in somatic cell hybrids. Revertants of CHR have been isolated which display decreased resistance to colchicine and a corresponding decrease in resistance to other drugs. These results strongly suggest that colchicine resistance and the pleiotropic cross-resistance are the result of the same mutation(s). Biochemical studies indicate that although colchicine is transported into our cells by passive diffusion, no major alterations in the membrane lipids could be detected in mutant cells. However, there appears to be an energy-dependent process in these cells which actively maintains a permeability barrier against colchicine and other drugs. The @HR cells might be altered in this process. A new glycoprotein has been identified in mutant cell membranes which is not present in parental cells, and is greatly reduced in revertant cells. A model for colchicine-resistance is proposed which suggests that certain membrane proteins such as the new glycoprotein of 6 H R cells, are modulators of membrane fluidity (mmf proteins) whose molecular conformation regulates membrane permeability to a variety of compounds and that the CHRmutants are altered in their mmf proteins. The possible importance of the CHRcells as models for investigating aspects of chemotherapy related to drug resistance is discussed.

Introduction Currently there is much interest in determining the role of the mammalian cell

membrane in the regulation of cellular proliferation and differentiation, and in the establishment of the neoplastic state. A major part of this effort has been directed toward the elucidation of the molecular organization of mammalian cell membranes (Singer and Nicholson, 1972). However, there is still scant understanding as to how the structural organization of cell membranes mediates the multiplicity of functions attributed to this organelle. The surface membrane, for example, has the unique ability to translate a single signal, such as that represented by the binding of peptide hormones . onto specific cell surface receptors, into a plethora of events inside the cell. Thus, Tomkin's group (Hershko et a l . , 1971) postulated that the cell membrane acts as a "pleiotypic mediator" which coordinates the response of a set of what would seem to be metabolically unrelated biochemical reactions (the pleiotypic response) inside the cell. Moreoever they suggest that the loss of cellular growth regulation or pleiotypic unresponsiveness in transformed cells results from alterations in the cell membrane. Holley (1972) proposed that cellular growth control is regulated by the amount of "low molecular weight nutrients" entering the cell and that the level of these nutrients inside the cell determines the cell's growth rate. Thus, he postulated that in malignant cells, membrane alterations have occurred such that the regulation of nutrient permeability is also altered resulting in the loss of growth control. To date, no definitive evidence in support of these hypotheses has been obtained.

One approach toward the investigation of the many roles played by a complex structure such as the mammalian cell membrane, is to exploit the genetic tools rapidly

'Presented in a Symposium on Cancer and Genetics at the 20th Annual Meeting of the Genetics Society o f Canada, University of Regina, Regina, Saskatchewan, Canada, June 4-6, 1975. Manuscript received August 4. 1975.

Can. J. Genet. Cytol. 17: 503-515,1975.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

504 V. LING

being developed in cultured cell systems (Till et a l . , 1973; Thompson and Baker. 1973). In this approach, the isolation of suitable membrane mutants is a necessary prerequisite for the correlation of changes in membrane function with specific alterations in membrane structure. In this paper, an attempt will be made to outline how the isolation and initial characterization of a class of membrane-altered mutants have contributed to concepts in membrane biology and in medicine.

Isolation of Colchicine-Resistant Cells Starting with independent clones of Chinese Hamster Ovary (CH8) cells, cell lines

resistant to the anti-mitotic drug colchicine were obtained by cloning resistant cells in the presence of the drug (Ling and Thompson, 1974). Independent subclones employed for selections were derived from the parental cell line AUX B 1, an auxotroph requiring glycine, adenosine, thymidine and proline isolated by McBurney and Whitmore (1974). The purpose of employing a parental cell line with auxotrophic markers was to facilitate genetic studies involving the isolation of hybrids from cell-cell fusion experiments described below (c.f. Table 11). Lines able to grow in different concentrations of colchicine were isolated in discrete steps as illustrated in Fig. 1 .

The levels of drug resistance displayed by various selected clones were measured by their relative colony-forming ability (relative plating efficiency) in the presence of drug as illustrated in Fig. 2. It can be seen that CHRA3 and C H V S are considerably more resistant than parental line AUX BI . As a measure of relative drug resistance, the concentration of drug required to reduce the relative plating efficiency of an isolated cell line by 90 percent, is divided by that required to reduce the relative plating efficiency of the parental line by the same amount. Thus, from Fig. 2, the relative colchicine resistance of CHHA3 can be

0.4 pg/d colchicine calculated to be = 7 while that of C H V 5 is greater than 168.

0.04 pg/rnl colchicine Since the isolated colchicine-resistant cells were found to be stable in the absence of colchicine (Ling and Thompson, 1974), and since their frequency increased with mutagen treatment (Till et al., 1973), these cells were classified as mutants (Thompson and Baker, 1973).

Two complementary assays were developed to determine whether the colchicine- resistant (CHR) mutants were altered either in their permeability to colchicine or in the ability of their cytoplasmic target proteins (tubulins) to bind the drug (Ling and Thompson, 1974). Surprisingly, of more than 20 colchicine-resistant lines characterized, all had reduced drug permeability into whole cells while the colchicine-binding ability of the mutant cell extracts was not reduced. Furthermore, the reduced permeability to labelled colchicine exhibited by mutant cells was strongly correlated with the degree of their resistance to the drug (Ling and Thompson, 1974). Thus it would appear that

AUXB1 0.1 R 3 R 10 R CH A3 B3 mutagenized - CH '5

AUXB1-$4 0.1 0.5 C H ~ S Q ~ H A 5 R cHRs4-2H rnutagenized - CH C4

FIG. I . Stepwise selection of CHR lines of Chinese hamster ovary cells from independent parental clones. Details of selection are previously described (Ling and Thompson, 1974). The selecting concentrations of colchicine in pglml are indicated above the a m u s .

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

DRUG RESISTANCE AND MEMBRANE ALTERATION 505

reduced permeability in CHO cells is the main mechanism of resistance to colchicine. Two related points can be made from these results. First, membrane mutations may be relatively common events in mammalian cells and second, in any selection of drug-resistant mutants where the cytotoxic action of the drug requires entry into the cell, the possibility of obtaining permeability mutants must be seriously considered.

Early on in the characterization of CHR lines it was noted that these lines display cross-resistance to a number of apparently unrelated compounds (Bech-Hansen et al. , 1974; Ling and Thompson, 1974; Bech-Hmsen et al. , 1975). This is illustrated in Table I where the relative drug-resistance to a number of compounds is compared in two

Colchicine (pg/rnl) FIG. 2. Dose response curves of the parental line AUX B1, and two CHR lines (see Fig. 1 for

derivation) assayed by colony-forming ability. The calculation of relative resistance of the mutants compared with the parental line is described in the text.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

506 V. LING

independent but highly colchicine-resistant mutants (CHRC4 and C H T S ) . This pleiotropic phenotype provides additional support for the cc~nclusion that colchicine resistance in CHO cells results from an altered membrane since the simplest explanation for the concomitant cross-resistance to a wide range of chugs (Table 1) displayed by these cells is that the CHH mutants have reduced permeability to all these compounds. Indeed, it has been shown that CCHR lines are also less permeable to labelled puromycin (See et a&. , 1974) and Colcesnid (Carken and Ling, unpeablished observation).

Since the membrane a%teratisn(s) resulting in colchicine resistance apparently engenders pleiotropic responses to other drugs, the cross-resistance pattern displayed by a mutant line provides a more detailed description of the dmg-resistant phenotype; moreover, this pattern may be regarded as a "fingerprint9 of the altered membrane structure. Thus, it appears that this aspect of colchicine resistance should be extremely useful in the elucidation sf some facets of structure and function in mammalian cell membranes.

From Table I it can be seen that the patterns of cross-resistance of two CMR lines are fairly similar even though these two lines were obtained independently from multi-step selections and treated with mutagen during some steps (Fig. I ) . Two aspects of this observation seem significant: it suggests that similar molecular alterations have occurred in the membranes of these two arnutant lines and moreover, that these particular alterations may be comparatively common in mammalian cells.

The prospect that similar membrane alterations may occur in human cells and that the colchicine-resistant mutants may be important models for studying features of chemotherapy related to drug-resistance are discussed in a later section. However, it should be noted here that the colchicine-resistant cells are dso resistant to a number of drugs (denoted with asterisks in Table I) currently used for cancer chemotherapy.

Genetic Charc~c*ter.ization The genetic characterization of the colchicine-resistant mutants was undertaken

from t\ao directions. In the first instance, an examination of whether the drug resistance locus acts dominantly or recessively with respect to its wild type was examined by

Cmss-resisGance of colchicine-resistant mutants -- - -- - - - -- -- - -- -- -- - - -.

Relative resistarnce

-- - - -- - -- - - -- Colchicine 1 74 184 Colcemid 1 1 1 16 Puromycin I 29 1 05 Ernetine 1 15 29 CytochaBasin B I 6 11 "Adriamy sin I 7 2 5 * Baunorubicin 1 3 2 74 "Actinomycin D I % 0 10 "Vinblastine 1 - 29

- -

Note: Relativc resistance was calculated in a siirnilru manner as that described in the text following Fig. 2 except that the growth rates of the lines in susper~sion cultures rather than plating efficiencies were detennined. Details are described by Beck-Hansen et ul. (1975). CHT4-S4 and CH1T5-S3 are subclones 4 and 3 of C H T 4 and C H T S respectively.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.


characterizing hybrid cells from appropriate cell-cell fusion experiments. In the second case, techniques for the selection of revertant cells were worked out and several revertant cells were characterized. Details of these experiments will be published elsewhere; however, a number of key results will be presented and discussed here.

In collaboration with Dr. R. M. Baker we determined whether the colchicine-resistance phenotype behaved dominantly or recessively in hybrids constructed between resistant and sensitive cells. In order to select for hybrids in the absence of colchicine, a colchicine-sensitive line E29 was employed which carried auxotrophic markers complementing those carried by the CH " lines. Thus hybrid cells are selected in a deficient growth medium which is not able to support the growth of either of the parents of the hybrid. Such hybrids were then tested for their sensitivity or resistance to colchicine to determine the recessiveness or dominance of the CH" marker. As can be seen in Table 11, the colchicine resistant phenotype in CH RA3 and CH RC4 behaves dominantly (incomplete dominance) in somatic cell hybrids. Several other colchicine resistant lines have also been examined in this manner and they all behave similarly.

If the cross-resistance of the CH " cells to the other drugs is due to a single locus one would expect that the dominance expressed in hybrids with respect to colchicine should extend to the other drugs as well. Preliminary evidence indicates that when hybrid cells are tested for resistance to compounds such as actinomycin D and vinblastine, the cross-resistant phenotype for these drugs also behaves in a dominant manner. These findings support the coilcept that the pleiotropic phenotype of multiple drug resistance observed in GWR lines is the result of the same mutation(s) which confers colchicine resistance to these lines.

The fact that the CH R-phenotype is of a dominant nature eliminates the possibility of employing classical genetic complementation studies for this marker. However, it indicates that such mutants can be obtained in any type of cell, including fully diploid cells, differentiated cells and tumor cells of different ploidy. Thus the colchicine resistance mutation can be employed in studies of membrane structure and function in a wide variety of cells.

As mentioned above, some effort has been placed on working out techniques for isolating revertant cells from the CHR lines. One selection scheme has met with some success. The basic design of this scheme is to employ puromycin, a drug to which the CH " lines are resistant and which is also a reversible inhibitor of protein synthesis, to protect drug-sensitive cells (putative revertants) from incorporating highly labelled

TABLE I1

Incomplete dominance of colchicine resistance

Cell Iine Relative resistance to colchicine

parental cells AuxB l E29 CHRA3 C H T 4

hybrid ctdls AuxB 1 X E29 CMRA, x E29 C H T , x E29

Note: Relative resistance was determined as described in text following Fig. 2.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

508 V. LING

radioactive amino acids, while under the same conditions, the drug resistant cells incorporate labelled amino acids and are killed. This is illustrated in Fig. 3 . Employing this technique a number sf revertant lines have been selected from C H V 4 cells. As shown in Table 111, in which the properties of one revertant line 18-31 are compared with those of CH T 4 and AUX B1 cells, the revertant cells have lost most of their resistance to colchicine; moreover, their resistance to other drugs also decreases correspondingly. Since revertant 18-31 cells were recovered in a single step, in the absence of mutagen, and since the loss of colchicine-resistance is paralleled by the loss of resistance to other drugs, thus further persuasive evidence is provided that the multiple drug resistant phenotype is the result of a single mutation.

The ability to select for revertants of CH " cells has several implications for future studies in this system. For example, it should permit a more detailed analysis of the nature of the mutation itself as well as the association of an observed phenotype with the primary mutation. Furthermore, it may be possible to obtain revertants in which the revertant phenotype is temperature dependent. Such revertants would be of obvious analytical advantage for relating structure-function relationships.

Biochemical Characterization Since our earlier results indicate that. the major lesion in CH " cells was located in

the cell membrane expressed in the form of reduced drug permeability, experiments were undertaken to elucidate the mechanism of colchicine uptake into whole cells. We reasoned that knowledge of this nature would facilitate the location of the molecular lesion in the mutant cell membranes; moreover, a variety of compounds including the chemotherapeutic drugs to which the mutant cells are cross-resistant (see Table I) might also be transported across mammalian cell membranes by the same mechanism.

The possibility that colchicine is transported into the cell via passive diffusion was investigated. This form of transport is noncarrier mediated and is characterized by nonsaturating kinetics of uptake, lack of competition by related analogues, and a lack of dependence on metabolic energy for uptake. In general, passively diffusing compounds are thought to permeate the cell through the lipid bilayer of the cell membrane. Initial observations are compatible with this mode of transport for colchicine. For example, as documented above, the mutation(s) conferring reduced colchicine permeability also resulted in the reduced permeability of a number of different compounds to which CH cells are cross-resistant. This indicates that the transport of colchicine and other drugs involves a common mechanism, and it is unlikely that the transport of such diverse

no puromyci n uptake, wash

Resistant cells i ncorporation - Ki l l ing 3

'H-leucine of H-leucine incubate

pu rornycin pu rornyci n uptake, wash

Sensitive cells / no incorporation No Ki l l ing (revertants) 3

3 ~ - l e u c i ne of H-leueine incubate

10810

FIG. 3 . Selection of revertants from CH "cells.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.


compounds is mediated by a common carrier protein. To test the hypothesis that colchicine transport is not carrier-mediated, Carlsen (unpublished results) measured the initial rates of colchicine uptake as a function of colchicine concentration and found that the kinetics of uptake was nonsaturating, even up to molar of colchicine; moreover, no competition of colchicine uptake by a number of drugs was observed. These observations provide some direct support for the hypothesis that colchicine is transported into the cell by passive diffusion.

Recently Bech-Hansen et al . (1975) have discovered that the drugs to which CH" cells are resistant (c.f. Table I) are hydrophobic as measured by partition in an octano1:phosphate buffered saline system. These compounds partition into octanol in 4 fold or greater concentrations than in aqueous buffer. It appears also that there may be a positive correlation between the amount of resistance and the hydrophobicity of the drug as measured by this system. This suggests that colchicine and other drugs diffuse into the cell through the hydrophobic portion of the surface membrane, possibly the membrane lipids.

In an attempt to discover the molecular alteration in the mutant cell membranes which give rise to reduced drug permeability, we next determined the lipid composition of the parental and mutant membranes. In a number of model systems it has been shown that membrane permeability to passively diffusing compounds is dependent on the amount of cholesterol/phospholipid in the cell membrane, and/or on the relative amount of unsaturated fatty acids (McElhaney et al. , 1973; DeKruyff et a l . , 1973; Davis and Silbert, 1974; Grunze and Beuticke, 1974). In each case, rather major changes in these parameters led to changes in membrane permeability. In analyzing the lipids in the membranes of parental and mutant cells, relatively little differences were observed; for example, the cholestero1:phospholipid ratios were found to be 0.75 and 0.72 respectively, and the percent of unsaturated fatty acids were 40 and 41% respectively (Y. P. See, unpublished results). Thus, it appears that the drug permeability in the mutant cells is not the result of major changes in their lipid composition and suggests that factors modulating colchicine permeability may be different from those thought to be important for other passively diffusing compounds mentioned above.

In performing experiments to determine whether or not colchicine transport was dependent on metabolic energy, we were surprised to discover that in the presence of metabolic inhibitors such as cyanide, and in the absence of glucose, the rates of cholchicine uptake were not reduced but instead increased markedly in both parental and mutant cells so that both lines eventually attained similar rates (See et al . , 1974). The effect of cyanide was rapid and could be reversed with a metabolizable sugar such as glucose while nonmetabolizable sugars had no effect. Subsequently it has been

Relative drug resistance

Cell line

Bmg AUXB 1 6HT4 18-31

Colchicine B 1 50 6 Colcemid 1 40 4 Vinblastine 1 2Q 3 Baunombicine 1 50 3 hromycin 1 45 2

Note: Relative drug resistance was determined as in Table 11.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

518 V . LING

discovered that the degree of membrane drug permeability in CHO cells is correlated with the amount of ATP inside the cell (Carlsen, unpublished results). The mutant cells behaved differently from parental cells, being more refractile to the cyanide-induced stimulation of drug permeability but more sensitive to the glucose prevention of cyanide induction. We have also determined that the uptake of other drugs such as puromycin and actinomycin D is also stimulated by cyanide in both mutant and parental cells (See e t a / . , 1974).

Apparent molecular weight x 8230

FHC~. 4. Fractionation of externally labelled membrane glycoproteins by polyacrylamide gel electrophoresis. Details are described by Juliancd et ul. (1975). Labelling was performed by the galactose oxidase-%-borohydride technique. Separated proteins were visualized by an autoradiogram and the autoradiographis image scanned with a densitometer to quantitate the amount of label. (Figure adapted from Juliano et al. , 1975, with permission).

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.


These findings strongly suggest that metabolic energy in the form of APT is required for the maintenance of a functional membrane barrier against a variety of drugs and also suggest that these cells are able to regulate their membrane permeability in some way not yet understood, to certain passively diffusing compounds. Moreover, since the responses of the CHR lines to cyanide and sugars were somewhat different from the parental cells, the possibility was raised that the CHR lines were altered in this energy dependent modulation of drug permeability (See et a l . , 1974).

An approach to the analysis of the structure of mammalian cell membranes is to examine the content or disposition of their glycoproteins. It is possible for example, to compare glycoproteins expressed at the surface by specific labelling techniques (Sahmberg and Hdcomori, 1973; Juliano and Behar-Bannelier, 1975). In collaboration with Dr. R. Juliano we have examined our cells for changes in the surface labelling of their membrane glycoproteins. Preliminary indications, employing the galactose oxidase procedure to label cell surface carbohydrates with radioactive sodium borohydride (JuIiano et al . , 1975), are that the cell surfaces of mutant and parental lines are different. It can be seen (Fig. 4) that when externally labelled glycoproteins of CH %24 membranes are fractionated by polyacrylamide gel electrophoresis, a new labelled glycoprotein peak 111 with an apparent molecular weight of 165,000 is present which is not observed in the parental line. In addition, when the membranes of the revertant 18-31 (see Table 111) are examined by this same technique (Fig. 4 lower panel), this glycoprotein is seen to be significantly decreased. Thus there seems to be a positive correlation between the amount of the 165,000 molecular weight glycoprotein on the cell surface labelled by this technique and the degree of drug resistance. It should also be mentioned that the appearance of a different glycoprotein on the surface of the mutant cells is completely compatible with the dominance of the CHR phenotype documented above. Experiments are presently underway to discover whether or not this glycoprotein is also expressed in hybrid cells.

Model for Mechanism of Drug Resistane~ The genetic and biochemical characterization of the CW mutants has yielded

substantial information which allows us to postulate a general model for the mechanism of resistance to colchicine. It is proposed that certain glycoproteins in the mammalian cell membranes, e.g. peak I11 of Fig. 4 are modulators of membrane fluidity (mmf

Cross-resistance in various mammalian drug-resistant cells

Cell type/ drug resistance Cross-resistance Reference

Chinese hamster Act R, VLB R, PuroR, This paper ovary/CHR CMR, DaunoR, +others

Chinese hamster VLB R, h r o R. CM R, Biedler and lung/Act BaunoR+others Riehrn (1 970)

Mouse mast BaunoR e l l leukemia P8 15/VLB a

Kessel et al. ( 1 968)

Syrian hamster PuroR Langelier et al. SV40-transformed/ (1 974) ActR

CH = colchicine; Act = actinomyein B; VLB = vinblastine; Puro = puromycin; CM = Colcemid; Dauno = daunorubicine

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

512 V . LING

proteins) which regulate lipid fluidity by changes in their conformation and that these changes consequently affect dmg permeability. The conformation of mrnf proteins in turn is assumed to be modulated by an ATP-dependent process, possibly by a phosphorylation-dephosphorylation process mediated by specific protein kinases and phosphatases. Phosphorylation and dephosphorylation of proteins have been proposed in a number of systems as a mechanism of modulating enzyme activity and protein structure (Walsh and Ashby, 1973; Louie and Bixon, 1973; Bradbury et a!. , 1974; Stein e f al . , 1974)- The mechanism by which the putative m n ~ f proteins affects lipid fluidity is not clear, but there are at least two possibilities.'First, conformation changes in an mrnf protein could directly affect the packing of lipid molecules within its domain, thus affecting lipid fluidity. A second possibility is that a conformation change in mrnf proteins triggers other events which subsequently lead to alterations in lipid fluidity possibly by redistribution of certain lipid components. It is postulated that CH mutants are altered in their rnmf proteins such that their membrane fluidity is affected resulting in reduced permeability to colchicine as well as to a number of structurally unrelated drugs.

This model for colchicine-resistance takes into account several aspects of the CH cells characterized above: (a) transport of colchicine and other drugs to which the CH cells are cross-resistant, is mediated by passive diffusion, through the hydrophobic lipid portion of the plasma membrane; (b) no magor compositional changes in the membrane lipids of mutant cells are observed; (c) the maintenance of the drug permeability barrier requires ATP and the response of the permeability barrier of the parental and mutant cells to various agents (e.g. cyanide and metabolizable sugars) are different; (d) the presence of a 165,000 molecular weight surface glycoprotein (peak 111 of Fig. 4) is related to the degree of drug resistance. Certain facets of this model are testable. It predicts that suitably labelled compounds able to monitor the fluldity of the diffusion path of colchlcine should indicate that mutant cell membranes are less fluid compared with parental cells. Furthermore, this difference should be eliminated or greatly reduced in metabolically starved mutant and parental cells where presumably the mmf proteins are not able to function. Measurements of membrane fluidity could be performed with fluorescently labelled compounds by fluorescence polarization (Radda and Vanderkooi, 1972; lnbar et a l . , 1974) or with compounds labelled with a reporter group for electron-spin-resonance measurements (McFarland, 197%; Keith et al . , 1973). Experi- ments to test this model by these means are presently underway.

It is likely that this model can be tested more rigorously when the 165,000 molecular weight glycoprotein (peak III of Fig. IV) of CHR cells is isolated and characterized. Questions such as whether or not this presumptive rnmf protein is

TABLE V

Collateral sensitivity of colchicine-resistant mutants -- - -

Relative resistance

Compound AUXBI CHHC9S4 CHR&IrS3

Cslchicine 1 74 184 Tetracaine 1 0.3 0.2 Xylocaine 1 0.3 6) . 1 Progesterone 1 0.6 0.1 1 -Dehydrc~testosterone 1 0.8 0.1

Note: Relative resistance was determined as in Table I .

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

DRUG RESISTANCE AND MEMBRANE ALTERATION 5 13

modified by phosphorylation can be asked. Another question sf interest is whether the peak 111 glycoprotein is comprised of a single homogeneous macromolecule or whether a number of closely related but different glycoproteins are involved. It might be expected that different mmf proteins regulate membrane fluidity in different domains of the cell membrane. In this context it seems important to isolate other drug resistant mutants with altered membrane permeability, with cross-resistance patterns different from the CH " lines, for comparison and study.

This proposed model has certain wider implications. A variety of experiments have shown that the activity of a number of membrane enzymes are profoundly affected by the fluidity of the environment in which they are located (Triggle, 1970; Bloj et al . , 1973; Kimelberg and Papahadjopoulos, 1974; Electr et a l . , 1974, Warren et a / . , 1974). Tt is entirely possible that presumptive mmf proteins could regulate a whole host of membrane enzyme activities by modulating membrane fluidity. Thus, the "'pleiotypic response" proposed by Hershko et al. (1971) could be mediated by mmf proteins. Certain hormone receptors on cell membranes for example may be a class of mmf protein.

Possible Irnportancx? cpfCHRCells to Chcrnatherupy The study of CHR cells has also pointed to areas of investigation which may have

important applications in medicine. As mentioned earlier, the cross-resistance of CHR lines to other compounds extends to drugs commonly used in cancer chemotherapy (denoted with asterisks in Table I). It is important therefore to know whether mutational events which result in drug resistance observed in our system are common in humans, and if these events do occur, whether they result in therapy-resistant tumor cells. While the answers to these questions are not presently available, it is known that drug- resistance with concomitant cross-resistance to another unrelated drug is observed in different cell types (Table IV). Furthermore, in each case, an altered membrane permeability has been implicated as the basis of resistance. Indeed, drug resistance of this type has also been observed in yeast cells (Rank et al., 1975) and Chlaymdamonas (Wan and Gibbons, 1974). Thus it would be surprising if human tumor cells did not possess the potential for this phenotype. The fact that the CHR mutation is dominant adds further weight to the notion that such mutants would be possible irrespective of the ploidy of the tumor cells.

With the viewpoint that tumors resistant to multiple drugs could arise in humans, another aspect of the pleiotropy observed in CH lines is of interest.From Table V. it is seen that the acquisition of drug-resistance in the CHR lines also confers increased sensitivity of these cells to a limited number of compounds such as local anesthetics and steroid hormones. At present the basis of this collateral sensitivity is not understood. Further work is elucidating the nature and mechanism of this collateral-sensitivity could provide important concepts for the chemotherapeutic treatment of drug resistant tumors.

Acknowledgments I wish to thank the various scientists with whom I had the opportunity of

collaborating in the experiments described in this paper. These include, R. M. Baker (M.I.T., Massachusetts), N. T. Bech-Hansen, S. A. Carlsen, R. L. Juliano (Hospital for Sick Children, Toronto), L. H. Thompson (Lawrence Livermore, California) and Y. P. See (University of Ottawa, Ontario). I also thank L. Siminovitch, J. E. Till and J . E. Aubin for helpful comments concerning the preparation of this manuscript. The work described in this paper was supported by grants from the Medical Research Council of Canada and the National Cancer Institute of Canada, and by a contract from the National Institutes of Health of the United States.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

References Bloj, B., Morero, W. B. and Farias, R. N. 1973. Membrane fluidity, cholesterol and allosteric transitions

of membrane-bound Mg2'ATPase, (Na+ & K')-ATPase and acetylcholinesterase from rat erythrocytes. FEBSLett.38: 101-105.

Bech-Wansen, N.T., Ling, V.. Till, J. E. and Aubin. J . 1974. Pleiotropic phenotype of mutants of Chinese hamster ovary (CHO) cells selected for resistance to colchicine. Can. J . Genet. Cytol. 16: 702 (Abstr.).

Bech-Plansen, N. T., Till, J . E, and Ling, V. 1975. Pleiotropic phenotype of colchicine-resistant CHO cells: cross-resistance and collateral sensitivity. Submitted for publication.

Biedler, 3. L. and Riehm, H. 1970. Cellular resistance to Actinomycin D in Chinese hamster cells in vitro: cross-resistance, radioautographic and cytogenetic studies. Cancer Res. 38: 1 174-1 184.

Bradbury, E. M., Inglis, R. 3. and Matthews, H. R. 1974. Control of cell division by very lysine rich histone (KH) phosphorylation. Nature (London) 247: 257-260.

Davis, M-T.B. and Silbert, D. K. 1974. Changes in cell permeability following a marked reduction of saturated fatty acid content of Escherdchica coli K- 12. Biochim. Biophys. Acta 373: 224-241.

DeKruyff, B., BeGreef, W. J., VanEyk, R, V. W., Bemel, R. A. and VanDeenen, L. L. M. 1973. The effect of different fatty acid and sterol composition on the erythritol flux through the cell membrane of Achobegbasma kaidlawii. Biochim. Biophys. Acta 298: 479-499.

Electr, S., Williams, M.A., Watkins, T. and Keith, A.D. 1974. Perturbations of the dynamics of lipid alkyl chains in membrane systems. Effect on the activity of membrane bound enzymes. Biochim. Biophys. Acta 339: 190-201.

Gahrnberg. C. G. and Hakomori. S.-I. 1973. Altered growth behavior of malignant cells associated with changes in externally labeled glycoprotein and glycolipid. Proc. Natl. Acad. Sci. U.S.A. 78: 3329-3333.

Grunze, M. and Beuticke, B . 1974. Changes of membrane permeability due to extensive cholesterol depletion in mammalian erythrocytes. Biochim. Biophys. Acta 356: 125-130.

Hershko, A., Mamont, B, , Shields, R. and Tomkins, G. M. 1971. "Pleiotypic response". Nature (London), New Biology 232: 286-2 1 1.

Holley, 8. W. 1972. A unifying hypothesis concerning the nature of malignant growth. Proc. Natl. Acad. Sci. U.S.A. 69: 28%Q-2841.

Inbar, M., Shinitzky, M. and Sachs, L. 1974. Microviscosity in the surface membrane lipid layer of intact normal lymphocytes and leukemia cells. FEBS Lett. 38: 268-270.

Juliano, R. L. and Behar-Bmnelier, M. 1975. An evaluation of techniques for labelling the surface proteins of culture mammalian cells. Biochim. Biophys. Acta 375: 249-267.

Juliano, R., Ling, V. and Graves, J. 1975. Drug resistant mutants of CHO cells possess an altered surface carbohydrate component. 5 . Supramol. Struct. In press.

Keith, A. D., Sharnaff, M. and Cohn, @. E. 1973. A summary and evaluation of spin labels used as probes for biological membrane structure. Biochim. Biophys. Acta 3QO: 379-419.

Kessel, B., Botterill, V. and Wodinsky, 1. 1958. Uptake and retention of daunomycin by mouse leukemic cells as factors in drug response. Cancer Wes. 28: 938-941.

Kimelberg, H. K. and Bapahadjopoulos, B. 1974. Effects of phospholipid acyl chain fluidity, phase transitions, and cholesterol on (NA' &"K+j-stimulated adenosine triphosphatase. J. Biol. Chem. 249: 1071-1080.

Langelier, Y., Simard, R. and Brailovsky, C. 1974. Mechanism of actinomycin resistance in SV40- transformed hamster cells. Differentiation 2: 26 1-267.

Ling, V. and Thompson, L. PI. 1974. Reduced permeability in CHB cells as a mechanism of resistance to colchicine. J. Cell. PhysioI,83: 103- 1 16.

Louie, A. 3. and Bixon, G o H. 1973. Kinetics of phosphorylation and dephosphorylatian of testis histones and their possible role in determining chromosomal structure. Nature (London), New Biology 243: 164-158.

McBurney, M. W. and Whitmore, G . F. 1974. Isolation and biochemical characterization of folate deficient mutants of Chinese hamster cells. Cell 2: 173- 182.

McElhaney, R. N., BeGier, J . , VanDer Neut-Kok, E. C. M. 1973. The effect of alterations in fatty acid composition and cholesterol content on the nonelectrolyte permeability of Ackobeplasma laidlawii B cells and derived liposomes. Biochim. Biophys. Acta 298: 500-5 12.

McFarland, B. G. 1972. The molecular basis of fluidity in membranes. Chem. Phys. Lipids 8: 303-3 13.

Radda, G. K . and Vanderkooi, 9. 1972. Can fluorescent probes te14 us anything about membranes? Biochim. Biophys. Acta 265: 509-549.

Rank, G. PI., Robertson, A. and Phillips, K. 1975. Reduced plasma membrane permeability in a multiple cross resistant strain of Sacclaaromyces cerevisiae . J. Bacterial. In press.

See, Y. P., Czlsen, S. A., Till, J. E. and Ling, V. 1974. Increased drug permeability in Chinese hamster ovary cells in the presence of cyanide. Biochim. Biophys. Aeta 373: 242-252.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.


Singer. S. J. and Niccslson, 6 . L. 1972. The fluid mosaic model of the structure c~f cell membranes. Science 175: 720-73 1 .

Stein, G. S., Spelsberg, T. C. and Kleinsmith, L. J . 1974. Nonhistone chromosomal proteins and gene regulation. Science 183: 8 l 7-824.

Thompson, L. H. and Baker, R. M. 1973. Isolation of mutants of cultured mammalian celrs. In Methods in cell bicalogy. Edited by D. M . Prescott. Academic Press, New Yorko 6: 209-281.

Till, J. E., Baker, R. M., Brunette, D. M.. Ling, V . , Thompson, E. H. and Wright, 9. A. 1973. Genetic regulation of membrane function in mammalian cells in culture. Fed. Proc., Fed. Am. Soc. Exp. Biol. 32: 29-33.

Triggle, D. 9. 1970. Some aspects of the role lipids in lipid-protein interactions and cell membrane structure and function. In Recent progress in surface science. Edited by J. F. Danielli et u l . , Academic Press, New York and London. 3: 273-290.

Walsh, D. A. and Ashby, C. D. 1973. Protein kinases: Aspects of their regulation and dikersity. In Recent progress in horomone research. Edited by R . 0. Geep . , Academic Press, New York and London. 29: 329-359.

N'arr. 9. R. and Gibbons, D. 1974. Further studies on colchicine resistant mutants d Chalmyslomonas reinlzardi . Exp. Cell Res. 85: 1 17- 122.

Warren, G . B.. Toon. P. A.. Birdsall, N. 9. M., Lee. A. 6. Metcaife, J. C. 1974. Complete ccmtrol of the lipid environment of membrane-bound proteins: application to a colchicine transport system. FEBS Lett. 41: 122-124.

Can

. J. G

enet

. Cyt

ol. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y U

NIV

CH

ICA

GO

on

08/0

3/13

For

pers

onal

use

onl

y.

drug resistance and membrane alteration in mutants of mammalian cells

Documents