Somatic Cell Genetics, Vol. 9, No. 5, 1983, pp. 593~08

Lectin-Resistant CHO Cells: Selection of New Mutant Phenotypes

Pamela Stanley

Department o f Cell Biology, Albert Einstein College o f Medicine, 1300 Morris Park Avenue, Bronx, New York 10461

Received 11 March 1983--Final 12 May 1983

Abstract--Cytotoxic plant lectins select for mutants which exhibit unique structural changes in surface carbohydrates reflecting specific defects in glycosylation reactions. However, lectins are not highly specific selective agents and, as a result, only the most frequently occurring mutants are obtained from single lectin selections. We have previously shown that the specificity of lectin selections may be improved by utilizing a combination of lectins added together or sequentially. This strategy has now been further exploited in the search for novel lectin-resistant mutants of Chinese hamster ovary cells. Five new Lec R phenotypes have been uncovered. One belongs to a new, recessive complementation group, two behave dominantly in somatic cell hybrids, and the remaining two appear to represent new phenotypes which fall into previously described complementation groups.

INTRODUCTION

Glycosylation mutants of animal cells provide an invaluable approach to studying structure-function relationships involving carbohydrates at the cell surface and to defining the pathways of glycoprotein and glycolipid biosynthe- sis in mammalian cells. It is therefore important to isolate the entire family of mutants affected in glycosylation reactions. Such mutants have been obtained to date from a number of different cell lines following selection for resistance to cytotoxic plant lectins (1-4) or to cytotoxic antibodies directed at cell surface antigens (5), or following selection for an inability to metabolize radioactive sugars (4). By far the most common selective agents, however, have been the cytotoxic plant lectins. Lectins from different plants exhibit a wide range of unique carbohydrate-binding specificities (6) and therefore offer the potential of selecting for the largest number of glycosylation

593

0098-0366/83/0900-0593503.00/0 �9 1983 Plenum Publishing Corporation

594 Stanley

mutants. However, two major complications accompany lectin selections: a single lectin usually selects for a variety of different lectin-resistant types (7, 8), and different lectins (which exhibit distinct carbohydrate-binding specifi- cities) often select for mutants which are affected in the same gene (7-9). Thus single lectin selections are far from specific. This lack of specificity is even more characteristic of alternative selection protocols for glycosylation mutants because of the numerous biosynthetic lesions which may lead to alterations in the pathways of complex carbohydrate synthesis in animal cells (10).

Over the last several years, this laboratory has isolated a number of lectin-resistant (Lec R) mutants of Chinese hamster ovary (CHO) cells and classified them into eight recessive complementation groups and one dominant type (7-9). Each class of mutant may be distinguished phenotypically by its unique pattern of cross-resistance and hypersensitivity to a panel of five lectins (7, 8). The latter property of Lec R CHO cells prompted the development of sequential lectin selection protocols in which particular Lec R mutants were selected against by virtue of their hypersensitivity to one or more lectins (8). This approach was successful in dramatically improving the specificity of lectins as selective agents. In addition to isolating previously characterized mutants of known Lec R phenotype, this strategy pointed the way to obtaining new Lec R mutants. In this paper, sequential lectin selections and mixed lectin selections are further exploited in a search for new Lec R glycosylation mutants. A variety of selection protocols have resulted in the isolation of five new Lec R phenotypes from CHO populations. One of the new phenotypes falls into a novel, recessive complementation group, two new phenotypes behave dominantly in somatic cell hybrids, and the remaining two new phenotypes fall into previously defined complementation groups.

MATERIALS AND METHODS

Cell Culture. Cells were cultured in suspension in complete alpha medium containing 10% horse serum and 2% fetal calf serum (Gibco Laboratories, Grand Island, New York). Cells were routinely checked for Mycoplasma by fluorescence microscopy after staining with Hoeschst 33258 (11) and found to be negative. Cell lines were continuously cultured no longer than 4 months before returning to a young frozen stock.

Cell Lines and Nomenclature. Parental cells in these experiments were the auxotrophic mutant clones Pro-5 and Gat -2 which have been previously described (7, 8). All our Lec R cell lines have been derived from these parental auxotrophs. In the past, the Lec R mutants have been named according to the lectin which was used in their selection. However, this practice proved confusing when it became clear that identical mutants were being isolated

Lectin-Resistant CHO Cells 595

using different lectins (7-9). Recently a more simple nomenclature was devised for the Lec R CHO cells isolated by this laboratory (12). Recessive mutants are referred to by the prefix Lec and dominant mutants by the prefix LEC. Different classes of mutants are assigned numbers according to the order of their discovery. For recessive mutants, these numbers correspond to the complementation group to which the mutants were previously assigned (7-9). For example, mutants from complementation group 1 (previously termed Pha R', Wga R', Ric R,, or Lca R') are now referred to simply as Lecl mutants (12). Individual cell lines are denoted by numbers following the general phenotypic designation. Thus Pro-5WgaR'3c becomes Pro-Lecl .3C (see Table 1 and refs 7-9).

Mutagenesis. Pro-5 CHO cells were mutagenized at 34~ and 38.5~ with either ethyl methane sulfonate (EMS; Eastman Chemical Co., Roches- ter, New York) at a concentration of 100 tzg/ml or N-methyl-N-nitrosoguan- idine (MNNG; ICN Co., Plainview, New York) at a concentration of 0.02 #g/ml. Cells in suspension at 2 x 105 cells/ml were incubated for 18 h with EMS or for 2 h with MNNG, washed, and an aliquot removed for the determination of relative plating efficiency. The mutagenized cells were cultured for seven days to allow expression of acquired mutations. Cell survival following mutagen treatment was about 60% for EMS and about 80% for MNNG.

Selection Protocols. Cells were aliquoted at 106 cells/100-mm tissue culture dish in alpha medium containing 10% fetal calf serum and the appropriate concentration of the primary selective lectin(s). After six days, the majority of the plates were washed twice with alpha medium and the secondary selective lectin(s) added in alpha medium containing 10% fetal calf serum. After approximately four more days of incubation, 20-30 of the largest colonies were picked into alpha medium containing 10% fetal calf serum and the plates stained with 2% methylene blue in 50% methanol. Control plates which contained no lectin or only the primary selective lectin(s), were stained after 8 days and relative plating efficiencies determined.

Determination of Lectin Resistance. As soon as possible after picking, colonies were frozen and tested for lectin resistance by the semiquantitative P test described previously (8). Those which exhibited the appropriate lectin resistances were cloned by limiting dilution in 96-well microtiter dishes. Clones were subsequently tested for lectin-resistance by P test and eventually their Lec R phenotype was determined by quantitative Dl0 analysis using the five lectins L-PHA (from P. vulgaris), WGA (from T. vulgaris), CON A (from C. ensiformis), RIC (the toxin from R. communis), and LCA (from L. culinaris). D10 values, defined as the concentration of lectin (in #g/ml) which reduces relative plating efficiency to 10%, were determined from survival curves and are reproducible within a standard error _+ 30% (13). Knowledge of

596 Stanley

the Dl0 values for each mutant type allows the design of appropriate lectin selection protocols and the classification of mutants into different phenotypic groups.

The lectins used in our studies were obtained from the following sources: Burroughs Wellcome, London, England (L-PHA); Sigma Chemical Co., St. Louis, Missouri (WGA); Pharmacia, Uppsala, Sweden (CON A); Vector Laboratories, Burlingame, California (RIC and LCA); E.Y. Laboratories, San Mateo, California (RIC); and from our own preparations (RIC and LCA) (8). They were dissolved in deionized, distilled water or phosphate- buffered saline, pH 7.4, filtered through 0.22 ~tm filters, and stored at 4 ~ The concentration of each lectin was determined from the optical density at 280 nm and their respective extinction coefficients as described previously (8).

Complementation Analyses. Somatic cell hybrids were formed by treat- ment with 44% polyethylene glycol (PEG 6000; British Drug Houses, Poole, England) by the method described previously (8). Following fusion, cells were plated at 1.5 x 105-3 • 105 cells/100-mm culture dish in alpha medium lacking proline, adenosine, thymidine, and glycine and containing 10% dialyzed fetal calf serum in the presence and absence of a lectin to which both parental cells were resistant. Plating efficiencies were determined by staining the plates after 8-10 days. Appropriate controls for reversion frequency and spontaneous hybridization between parental lines were included. Hybrid colonies from plates containing no lectin were picked before staining, cultured, and tested more extensively for their Lec g phenotype by P test or by D~0 analysis. Representative hybrid cell lines from each cross were shown to be pseudotetraploid by karyotype analysis.

Karyotyping. Modal chromosome numbers were determined from 25 chromosome spreads as previously described (7-9).

RESULTS

Selection of New Mutants. The selective growth inhibition or elimina- tion of unwanted Lec R phenotypes in order to obtain new mutants is possible because all Lec R CHO cells previously characterized by this laboratory are killed by certain concentrations of one or more plant lectins. This can readily be seen from Table 1 which presents the quantitative lectin-resistance properties of those Lec R cell lines previously isolated via single-step selection protocols. These mutants are thought to carry single Lec a genetic markers. They represent the starting group of CHO mutants which we wished to expand by selecting for new Lec R phenotypes. Two other complementation groups (6 and 7) have been previously defined but only in mutants which contain at least one other mutation to lectin-resistance. Thus, the Lec R phenotypes corresponding to Lec6 and Lec7 mutations, when present as single genetic markers, are not known but would be expected to be novel.

Lectin-Resistant CHO Cells 597

Table 1. Lee R Phenotypes of Previously Isolated Lec R CHO Mutant Clones a

Lectin resistance (DI0, #g/ml)

Comp. Lee R Cell *RIC group phenotype line L-PHA WGA CON A (• -3) LCA

Parental Pro-5 3.5 2.4 19 4.5 '19 Gat-2 2.7 1.7 17 5.6 "16

1 Lecl Pro Lecl.3C >3000 70 4.5 500 >3000 Gat-Lecl . lN >3000 49 2.1 500 >3000

R(>1000) R(30) S(6) R(100) R(>200)

2 Lee2 *Pro-Lec2.6A 3 25 16 0.040 9 Gat-Lec2.4C 1.3 25 18 0.084 5

(S) R(11) - - S(100) S(2)

3 Lec3 Pro-Lec3.4B 1.4 11 14 0.56 7 *Gat-Lec3.6F 2 9 14 0.85 10

(S) R(5) (S) S(10) S(2)

4 Lee4 Pro-Lee4.12-2 >2000 3.5 14 4.0 11 *Gat-Lec4.2D >1000 1.8 14 4.6 7

R(> 1000) (R) (S) (S) (S)

5 Lee5 Pro-Lec5.B211 16 4 33 16 59 Gat-Lec5.D11A 23 3 22 2 61

R(7) (R) (R) R(3) R(3)

8 Lee8 b *Pro-Lec8.3D 32 220 16 8.6 1 *Gat-Lec8.2B 34 265 16 9.0 1.5

R(10) R(100) (S) R(2) S(10)

Dom LEC10 Pro-LEC10.3C 1.2 1.6 19 100 *23 Gat LEC10.2A 0.8 0.8 15 200 '15

S(2) (S) - - R(20) - -

aCIones from each of the Lec R CHO mutant classes previously isolated by this laboratory (7-9) are compared for their lectin-resistance by Dl0 analysis (the concentration of lectin in #g/ml required to reduce the relative plating efficiency of each clone to 10%). Some of these values have been reported previously (7, 8) but are repeated here to facilitate comparisons within this paper. However, all Dlo values for cell lines or lectins marked * have not previously been reported. The Dl0 values for RIC were repeated because the toxicity of our more recent RIC preparations and those obtained commercially are approximately 20-fold greater than the preparation used in our original experiments (7, 9). The values given under the Dlo data for each clone refer to the approximate -fold resistance or -fold sensitivity of the mutant type compared to wild type. Differences which are <two-fold are referred to as (R) or (S), respectively.

~These clones were shown to belong to complementation group 8 by complementation analysis against the WGA-resistant mutants from complementation groups 1, 2, and 3 (data not shown).

T h e r a t i o n a l e s b e h i n d t h e p ro toco l s d e s i g n e d to se l ec t new L e c R C H O

cells a r e s u m m a r i z e d in T a b l e 2. T w o m a i n s t r a t e g i e s w e r e app l i ed . O n e

a p p r o a c h was to use a m i x t u r e o f l ec t in s a d d e d t o g e t h e r . H o w e v e r , t h e m o r e

c o m m o n a p p r o a c h was to use a s e q u e n t i a l s e l ec t ion p ro toco l . In t h i s m e t h o d , a

s ing le l ec t in was a d d e d to cel ls a t a c o n c e n t r a t i o n d e s i g n e d to ki l l p a r e n t a l

598 Stanley

Table 2. Selection Strategies for Obtaining New Lec R Phenotypes a

Lec R phenotypes theoretically inhibited

Lectins 1st 2nd 3rd (ag/ml) lectin lectin lectin

W G A ~ CON A + RIC 10,5,4 1 2,3,4,8 (7.5) (7.5) (0.05) W G A ~ CON A + LCA (2)3,4,5,10 1 (2)8 (30) (25) (9) W G A ~ CON A + LCA (1)2,3,4,5,10 1 8 (60) (20) (10) W G A + CON A + LCA 1,2,3,4,5,8,10 N A N A (20) (10) (14) RIC ~ WGA + C ON A 2,3,4,5,8 10 1 (0.1) (3) (7.5) L-PHA ~ LCA 2,3,5,10 8 N A (15) (lO)

aLectins at the concentrations shown were applied to C HO cell populations either sequentially (denoted by 4 ) or together (denoted by +) . In sequential selections, the time for which cells were incubated in the first selective lectin was usually 6 days and, in the lectin(s) added subsequently, about 4 days. The Lec R mutant types theoretically eliminated at the concentration of lectin(s) used at each stage are given. N A = not applicable.

cells as well as certain Lec R mutant phenotypes. This primary (1 ~ selective iectin was allowed to act for approximately six days prior to its removal and immediate replacement with one or two lectins of different sugar specificity. The secondary (2 ~ lectins were added at concentrations designed to inhibit the growth of Lec R phenotypes predicted to have survived the first lectin selection. A number of lectins were used at varying concentrations and in different combinations in the selection protocols described in Table 2. In addition, most selections were performed at 34~ and 38.5~ in order that mutants which were temperature-sensitive for growth or lectin-resistance might be uncovered. Also, most selection protocols were applied to unmuta- genized cell populations as well as to cells previously mutagenized with EMS or M N N G .

The results of some of the selections performed using the strategies described in Table 2 are presented in Table 3. Potential new mutants were identified by testing the Lec R phenotypes of approximately 20 of the largest colonies which survived each selection condition. In three experiments (Nos. 6, 7 and 10), all of the colonies tested exhibited novel Lec R phenotypes. However, the remaining new Lec R phenotypes were uncovered from selections which gave rise predominantly to the Lecl phenotype. This was not surprising in the case of selection 13 which was actually designed to search for the Lecl phenotype. However, the Lecl phenotype was not expected to predominate in selections which included CON A at 7.5 /~g/ml. These results suggest that

Tab

le 3

. S

elec

tion

of

New

Lec

R C

HO

CeU

s ~

Sur

vivo

rs p

er 1

06

Sel

ecti

ve

colo

ny-f

orm

ing

cell

s S

elec

tion

le

ctin

s T

ern

Lec

tin-

resi

stan

t N

o.

(#g/

ml)

C

ells

(~

1 o

2 ~

ph

enot

ypes

., ,=

1 W

GA

~

CO

N A

+

RIC

P

ro-5

34

50

3

Par

enta

l (1

1/13

) (7

.5)

(7.5

) (0

.05)

L

ec 5

(2

/13)

2

EM

SP

ro

5 34

~

800

8 L

ecl

(6/6

) 3

MN

NG

Pro

5

34

80

3 L

ecl

(2/3

) N

EW

(1

/3)

4 E

MS

Pro

-5

38.5

~

1000

~

100

Lec

l (1

8/18

) 5

MN

NG

Pro

- 5

38.5

~

100

- 10

0 L

ee 1

(1

0/10

)

6 W

GA

~

CO

N A

+

LC

A

Pro

-5

38.5

13

0 4

NE

W

(18/

18)

(30)

(2

5)

(9)

7 W

GA

+

CO

N A

+

LC

A

EM

SP

ro

5 38

.5

<1

NA

N

EW

(1

/1)

(20)

(1

0)

(14)

8

MN

NG

Pro

5

38.5

<

1 N

A

-

9 W

GA

~

CO

N A

+

LC

A

EM

SP

ro-5

38

.5

40

<1

- (6

0)

(10)

(1

0)

10

MN

NG

Pro

5

38.5

3

<1

NE

W

(3/3

)

11

RIC

~

WG

A +

C

ON

A

EM

SP

ro-5

38

.5

14

10

Lec

l (9

/9)

(0.1

) (3

) (7

.5)

12

MN

NG

Pro

5

38.5

4

5 L

ecl

(10/

10)

13

L-P

HA

~

LC

A

EM

SP

RO

-5

37

14

~50

L

ecl

(1/8

) (1

5)

(10)

N

EW

(5

/8)

=.

r

aUnm

utag

eniz

ed P

ro

5 ce

lls

or P

ro-5

pop

ulat

ions

mut

agen

ized

wit

h E

MS

or

MN

NG

wer

e su

bjec

ted

to s

elec

tion

con

diti

ons

1-13

. T

he a

ppro

xim

ate

num

ber

of s

urvi

vors

at

each

ste

p of

the

sel

ecti

on (

1 ~

and

2 ~

was

det

erm

ined

fro

m s

tain

ed p

late

s. A

fter

the

2 ~

phas

e of

the

sel

ecti

on,

the

larg

est

surv

ivin

g co

loni

es w

ere

pick

ed i

nto

nons

elec

tive

med

ium

, cul

ture

d, a

nd t

este

d fo

r th

eir

lect

in-r

esis

tanc

e pr

oper

ties

by

the

sem

iqua

ntit

ativ

e P

test

. Bas

ed

on p

atte

rns

of r

esis

tanc

e to

the

lec

tins

L-P

HA

, W

GA

, C

ON

A

, R

IC,

and

LC

A,

they

wer

e cl

assi

fied

as

new

Lec

R p

heno

type

s or

as

prev

ious

ly

char

acte

rize

d L

ec R

mu

tan

t ty

pes.

The

ass

ignm

ent

of t

he t

wo

Lec

5 co

loni

es i

n se

lect

ion

1 w

as c

onfi

rmed

by

D~0

ana

lysi

s an

d by

com

plem

enta

tion

an

alys

is.

The

nov

elty

of

each

Lec

R p

heno

type

des

igna

ted

NE

W

was

als

o su

bseq

uent

ly i

nves

tiga

ted

by D

10 a

naly

sis

and

com

plem

enta

tion

ana

lysi

s (T

able

s 4-

9).

NA

= n

ot a

ppli

cabl

e; -

-ind

icat

es n

o su

rviv

ors

from

up

to 2

x 1

07 c

ells

pla

ted;

num

bers

in p

aren

thes

es r

efer

to

thos

e co

loni

es e

xhib

itin

g a

part

icul

ar L

ec R

phe

noty

pe o

f th

e to

tal

colo

nies

tes

ted.

r

600 Stanley

higher concentrations of CON A should be used to effectively select against this phenotype. In a similar selection series using WGA (7.5/~g/ml) followed by CON A (7.5 ~tg/ml) and LCA (10 ~tg/ml), the counterselective lectin concentrations also appeared too low since, out of 80 colonies tested, all appeared to belong to complementation groups 1, 2, 3, or 8 (data not shown). By contrast, this mixture of lectins added together is extremely toxic. In six selections using this combination, the survival frequency varied from approxi- mately 2 x 10 -6 to less than 4 • 10 -7. Only in one selection with this lectin combination were cells with a typical Lecl phenotype isolated and those cells exhibited an extremely fibroblastic morphology, quite uncharacteristic of Pro-5 or Lecl CHO cells.

Lec R Phenotypes of New Isolates. The lectin-resistance properties of cloned cell lines from each of the new Lec R mutants were determined by quantitative D10 analyses (Table 4). All of the isolates exhibited novel Lec R phenotypes. The nomenclature of the new mutants, although derived on the basis of the complementation results described below, is included in Table 4 to facilitate discussion in the text.

The clones from selection 3 (designated Lec9 cells) were found to be moderately RIC-resistant and slightly resistant to CON A and LCA. They were not hypersensitive to any of the five lectins tested. However, they were found to be quite temperature-sensitive for growth. Their plating efficiency at 38.5~ was reduced 10- to 100-fold compared with their plating efficiency at 34~

The clones from selection 6 (Lec2B cells) gave Dl0 values very similar to cells belonging to complementation group 2 (Table 1) except for their increased resistance to WGA and decreased sensitivity to RIC. By contrast, the clones from selection 7 (LEC11 cells) expressed a completely novel Lec R phenotype. They were highly resistant to WGA, moderately resistant to L-PHA and LCA, and extremely sensitive to the toxicity of RIC. This Lec R phenotype is qualitatively similar to that expressed by the clones obtained from selection 10 (LEC12 cells). However, the LEC11 and LEC12 pheno- types differ significantly in degree. LEC11 cells are 8-fold resistant to WGA and 25-fold hypersensitive to RIC, whereas LEC12 cells are approximately 50-fold resistant to WGA and only 4-fold hypersensitive to RIC. Hence it seems likely that LEC11 and LEC12 mutants result from different muta- tions.

The clones from selection 13 (LeclA cells) possess a Lec R phenotype qualitatively similar to Lecl CHO cells. However, their lectin-resistance pattern differs significantly in degree. Both LeclA clones are approximately 3-fold less resistant to WGA and LCA and at least 10-fold less resistant to RIC than Lecl cells (see Table 1).

Karyotype Analyses. Modal chromosome numbers were determined from mitotic spreads of each of the new Lec R mutant clones. All but one of the

m~

u

Sel

ecti

on

Sel

ecti

on

num

ber

prot

ocol

Tab

le 4

. L

ecti

n R

esis

tanc

e P

rope

rtie

s of

New

Lec

R Is

olat

es a

Lec

R p

heno

type

(D10

, t~

g/m

l)

New

C

lone

s R

IC

mu

tan

t te

sted

L

-PH

A

WG

A

CO

N A

(x

10 -

3)

LC

A

type

r

3 W

GA

~

CO

N A

+

RIC

4A

4.

5 4

17

52

25

(7.5

) (7

.5)

(0.0

5)

4C

4.0

3 19

62

27

L

ec9

(R)

(R)

--

R(1

0)

(R)

6 W

GA

~

CO

N A

+

LC

A

3A

1.5

54

13

1.5

11

(30)

(2

5)

(9)

2B

1.5

52

16

1.6

12

Lec

2B

(S)

R(2

5)

(S)

S(3

) (S

)

7 W

GA

+

CO

N A

+

LC

A

E5

13

14

17

0.2

57

(20)

(1

0)

(14)

E

7 t4

19

18

0.

2 80

L

EC

ll

R(4

) R

(8)

--

S(2

5)

R(3

)

l0

WG

A -

-, C

ON

A +

L

CA

1B

10

10

2 16

1.

4 46

(6

0)

(10)

(1

0)

2E

9 10

6 16

1.

2 42

L

EC

12

R(3

) R

(50)

--

S

(4)

R(2

)

13

L-P

HA

~

LC

A

3B

> 10

00

18

3 45

70

0 (1

5)

(10)

3E

>

1000

19

4

40

650

Lec

lA

R(>

3000

) R

(9)

S(5

) R

(10)

R

(35)

aSub

elon

es d

eriv

ed f

rom

the

col

onie

s ex

pres

sing

new

Lec

R p

heno

type

s (T

able

3)

wer

e su

bjec

ted

to D

10 a

naly

sis

as d

escr

ibed

in

Mat

eria

ls a

nd

Met

hods

.

602 Stanley

mutan t types were pseudodiploid. LEC 11 clones were hyperdiploid exhibiting modal chromosome numbers of 32 for both LEC11.E5 and LEC11.E7.

Complementation Analysis. To determine which of the new phenotypes

represent new genetic classes of Lec R C H O mutants , each clone was hybrid- ized with parental cells and a number of the previously characterized recessive Lec R mutants which were most similar in lectin-resistance properties to the putative new mutants . Three of the new mutan t types (Lec9, L E C l l , and LEC12) appear to represent novel classes of Lec R C H O mutants . However, two of the new Lec R phenotypes ( L e c l A and Lec2B) were found to belong to previously defined complementat ion groups (Tables 5 to 9).

Hybrids formed between parental and Lec9 C H O cells exhibited paren- tal sensitivity to the toxicity of R IC whether tested immediately after fusion or following a period of culture in the absence of the lectin (Table 5). Thus the Lec9 mutat ion behaves recessively in somatic cell hybrids. The Lec9 mutat ion also behaved recessively when Lec9 cells were crossed with all the previously described recessive Lec R mutants , Since complementat ion of each Lec R

marker was observed, Lec9 cells represent a new, recessive complementat ion group (number 9) of Lec R C H O cells.

Another new class of C H O mutan t was obtained from selection 7. LEC11 cells crossed with parental cells and with four WGA-res i s tan t Lec R

mutants gave rise to a hybrid population which was markedly WGA-res i s tan t

Table 5. Complementation Analysis of Lec9 CHO Cells"

Colonies per 3 x 105 cells plated in

Cell lines crossed RIC (ug/ml x 10 -3) with Pro Lec9.4C 0 10 Complementation

Parental 454 6 + Lec 1 500 13 + Lec2 200 1 + Lec3 400 10 + Lec4 400 15 + Lec5 550 12 + Lecl .Lec6 500 30 + Lec2.Lec7 450 3 + Lec8 220 15 +

"Pro-Lec9.4C cells were fused with parental and L e c R cell lines carrying the Gat- auxotrophic marker as described in Materials and Methods. The number of colonies which arose in the absence of RIC and in the presence of 1, 5, and 10 ng/ml RIC was determined. The relative plating efficiency of Lee9 cells in 10 ng/ml RIC was >_90%. The frequency of auxotrophic revertants for each cell line was 10 -5. Three colonies from the cross with parental CHO cells were picked, cultured, and shown by P test to exhibit wild-type sensitivity to RIC. They were also shown to be pseudotetraploid. The result presented for Lec5 cells was performed with Pro-Lec9.4A cells and the sensitivity of the hybrids to RIC was verified by P test. Parental and Lecl CHO cells were also crossed with Pro-Lec9.4A cells with results similar to those obtained for Pro-Lec9.4C cells.

L e c t i n - R e s i s t a n t CHO Cells 603

Table 6. Complementation Analysis of LEC11 CHO Cells a

Colonies per 3 • 105 cells Cell lines plated in W G A (#g/ml) D~o of

crossed with hybrids for Pro- LEC I I.E7 0 3 W GA (/zg/ml) Complementation

Parental 300 1 2 2 ~ 10

Lecl 315 97 >10 Lec2 330 77 - 10 Lec3 160 68 > 10 Lec8 385 90 > 10

aPro-LEC11.E7 cells were fused with parental and WGA-resistant Lec ~ mutants carrying the Gat auxotrophic marker as described in Materials and Methods. The number of colonies arising in 0, 1, 2, and 3 #g /ml W GA was determined for each cross. The frequencies of auxotrophic revertants for each cell line was _<2 x 10 -5. Two colonies from each cross were cultured and tested for resistance to W G A by D~o analysis. They all exhibited relative plating efficiencies of between 10 and 30% at 10 #g /ml W G A and were pseudotetraploid by karyotype analysis. Three of these colonies were tested for their resistance to L-PHA, WGA, C O N A, RIC, and LCA and found to exhibit a complete Lec R phenotype typical of Pro LEC11 cells.

(Table 6). Although this suggested a dominant Lec R phenotype, it did not appear to be complete since dominant expression of the WGA resistance of LEC11 cells should give approximately 100% relative plating efficiency at 3 tzg/ml WGA (see Table 4). In fact, it was found that hybrid colonies selected in the absence of WGA and subsequently P-tested for lectin resistance did exhibit Lec R properties typical of clones E5 and E7 (Table 6). Thus the mutation expressed by L E C l l cells behaves dominantly in somatic cell hybrids provided the test for lectin resistance is not performed immediately following hybrid formation. This suggests that a period of culture is required before expression of the dominant phenotype is complete.

The mutants isolated in selection 10 (LEC 12 cells) behaved similarly to LEC 11 mutants in somatic cell hybrids. The resistance of hybrids formed with LEC 12 cells to WGA was also partially dominant if the hybrid population was tested immediately following fusion (Table 7). However, hybrid colonies picked from plates containing no WGA and subsequently tested for lectin resistance exhibited a Lec R phenotype typical of the LEC12 clones (Table 7). Therefore the Lec R phenotype of LEC12 ceils is also completely dominant, provided time for expression of the phenotype is adequate. Despite the apparent similarlity in the behavior of hybrids involving LEC 11 and LEC 12 CHO cells, there are highly significant differences between these two mutant types in the degree of their lectin resistances (Table 4). As mentioned previously, LEC11 cells are approximately sixfold less resistant to WGA but approximately sixfold more sensitive to RIC than LEC12 cells. For this, and other reasons outlined in the Discussion, it seems likely that LEC 12 cells are representative of a new mutant class.

The new Lec R phenotypes designated LeclA and Lec2B (Table 4) pose somewhat of a classification dilemma. The Lec2B clones were found to behave

604 Stanley

Table 7. Complementation Analysis of LEC12 CHO Cells"

Colonies per 3 x 105 cells Cell lines plated in WGA (~tg/ml) P test of

crossed with hybrids for Pro- LEC 12.1B 0 3 WGA (/~g/ml) Complementation

Parental 97 47 -75 Lecl 235 174 75-100 Lec2 105 57 50-75 Lec3 131 65 50-100 Lec8 69 66 ~75

a Pro-LEC 12.1B cells were fused with parental and WGA-resistant CHO cells carrying the Gat- auxotrophic marker as described in Materials and Methods. The number of colonies which arose in 0, 1, 2, and 3/zg/ml WGA was determined. Similar results were obtained from an identical set of crosses performed with Pro LEC12.2E cells. The frequency of auxotrophic revertants for each cell line varied from 1 to 3 per 105 cells plated. Three colonies from each cross were picked from plates containing no WGA, cultured, and P tested for resistance to WGA. Hybrids from the parental x LEC12 cross were also tested against RIC and LCA and found to exhibit a phenotype similar to that of LEC 12 cells. Karyotype analysis of hybrids from each cross revealed that the cells were pseudotetraploid.

r eces s ive ly in h y b r i d s a n d to c o m p l e m e n t al l W G A - r e s i s t a n t C H O cell

m u t a n t s e x c e p t t hose b e l o n g i n g to c o m p l e m e n t a t i o n g r o u p 2 ( T a b l e 8). T h e y

a p p e a r , t h e r e f o r e , to be m e m b e r s of c o m p l e m e n t a t i o n g r o u p 2, a l t h o u g h t h e y

e x p r e s s a q u a l i t a t i v e l y d i f f e r e n t L ec R p h e n o t y p e f r o m l ines p rev ious ly s h o w n

to fa l l i n to th i s g roup . S i n c e a n u m b e r o f i so la tes w i t h t h i s p h e n o t y p e we re

o b t a i n e d f r o m a v a r i e t y of s e l ec t i on p r o c e d u r e s a n d w e r e a l w a y s f o u n d to

e x h i b i t n o n c o m p l e m e n t a t i o n w i t h L e c 2 m u t a n t s , i t s e e m s l ike ly t h a t t h e y

r e p r e s e n t a subc l a s s o f t h e m u t a n t s w h i c h fa l l in to t h i s c o m p l e m e n t a t i o n

Table 8. Complementation Analysis of Lec2B CHO Cells a

Colonies per 4 x 105 cells plated in WGA (#g/ml)

Cell lines crossed with Pro-Lec2B.3A 0 3 Complementation

Parental 400 33 + Lecl 230 40 + Lec2 159 143 - Lec3 114 41 + Lec8 256 28 +

aPro-Lec2B.3A cells were fused with WGA-resistant CHO cells carrying the Gat- auxotrophic marker as described in Materials and Methods. The number of colonies which arose in 0, 1, 2, and 3 #g/ml WGA was determined. Although the frequency of revertants of each Gat-Lec g line was <2 x 10 -5, Pro-Lec2B.3A cells reverted to Pro + at a frequency of approximately 40 colonies per 4 x 105 cells plated. The colonies appearing on plates containing 3 #g/ml WGA in the crosses with parental, Lecl, Lec3, and Lec8 cells exhibited a morphology typical of Lec2B.3A revertants and therefore probably do not represent hybrid colonies resistant to WGA. A similar series of crosses with Pro-Lec2B.2B cells gave the same complementation results.

Lectin-Resistant CHO Cells 605

Table 9. Complementation Analysis of LeclA CHO Cells a

Colonies per 3 • 10 5 cells plated in WGA (~tg/ml)

Cell lines crossed with Pro-Lec 1A.3B 0 3 Complementation

Lecl 358 356 - Lec2 147 1 + Lec3 236 1 + Lec4 227 16 + Lec8 354 6 +

~ 1A.3B cells were fused with representatives of complementation groups 1, 2, 3, 4, and 8 which carry the Gat- auxotrophic marker as described in Materials and Methods. The number of colonies which arose in the presence of 0, 1, 2, and 3 ug/ml WGA was determined. Essentially identical results were obtained in a similar series of crosses with Pro-LeclA.3E cells. The hybrids formed between LeclA.3B or LeclA.3E and Lec4 cells were also shown to exhibit parental sensitivity to the toxicity of L-PHA. The Gat- lines reverted at a frequency of <2 • 10 -s while Pro-Lecl A.3B cells reverted at approximately 2 x 10 -4.

group. W e have previously observed another subclass of Lec2 mutan t s which differ from other Lec2 mutan ts in their ab i l i ty to bind [125I]WGA (8). These mutan t s have been t e rmed Lec2A (12). The class of m u t a n t represented by clones 3A and 2B will therefore be referred to as Lec2B.

The mutan t s des ignated L e c l A also fall into a previously defined complementa t ion group, despi te their novel Lec R phenotype (Table 9). Lec 1A cells behave recessively in somat ic cell hybrids but exhibi t noncomplementa- t ion with Lec l C H O cells. They apparen t ly belong to complementa t ion group 1, a l though they are signif icantly less res is tant to W G A , RIC , and L C A than L e c l C H O cells (Tab le 4).

D I S C U S S I O N

Cytotoxic p lant lectins provide the most direct appproach to selecting a b road range of an imal cell mutan t s which express a l te red ca rbohydra te s at the cell surface. Al though other selection protocols have given rise to glycosyla- t ion mutants , their specificity is less c lear ly d i rec ted at obta in ing mutan t s of this phenotype. The specificity of lectins in selecting for glycosylat ion mutan t s p re sumab ly derives from the fact tha t they must bind to surface carbohy- dra tes before exer t ing their toxic effects. Therefore , most of the lectin- res is tant mutan t s isolated to da te are " lec t in receptor" mu tan t s which exhibi t reduced binding of the selective lectin at the cell surface (1-4 , 13). The reason tha t lectins are not specific as selective agents, a l though they bind only to pa r t i cu la r ca rbohydra t e configurations, is due to the s t ruc tura l na ture and dis t r ibut ion o f cell surface carbohydra tes . Thus lectins of different sugar specificity have been shown to se lec t for ident ical g lycosyla t ion muta t ions (7,

606 Stanley

8). Secondly, all cytotoxic lectins examined so far select for more than one type of glycosylation mutation. In addition the different mutants which arise may be represented at very different frequencies among surviving colonies. For example, Lecl mutants are comparatively common in CHO populations and, since they are highly resistant to L-PHA, WGA, RIC, and LCA (and probably other lectins), they often predominate among survivors of single lectin selections. They have also been isolated from CHO cells using a 3H-fucose suicide selection protocol (14).

Since a great amount of time is involved in characterizing survivors to determine which, if any, represent novel mutations, this laboratory has improved the specificity of lectin selections by designing protocols to select against known Lec R phenotypes. This strategy was previously successful in selecting for WGA-resistant cell lines belonging to complementation groups 1, 2, 3, and 8 (8). In this paper, a similar approach was used to search for new glycosylation mutants of CHO cells.

Five new Lec R phenotypes were uncovered by the different selection protocols described in Table 2. Each one of them probably expresses unique carbohydrate structural changes at the cell surface giving rise to their novel lectin-resistance properties. The mutants referred to as Lec9, L E C l l , and LECI2 appear to represent mutations in previously unidentified CHO genes. Lec9 cells belong to a new recessive complementation group. LEC11 and LEC12 cells are both dominant but express significantly different Lec R phenotypes and are therefore likely to be the result of mutations in different genes. Additional evidence in support of this probability has been obtained from carbohydrate structural studies of glycopeptides derived from Vesicular Stomatitis virus grown in L E C l l and LEC12 cells and their revertants (Campbell and Stanley, manuscript in preparation). Both LEC11 and LEC12 cell-free extracts exhibit increased N-acetylglucosaminide a(1,3)fucosyl- transferase activity. However there is evidence that the substrate specificities of the altered fucosyltransferase(s) from the two cell lines differ, providing a further indication that LEC11 and LEC12 cells are the products of different mutations (Campbell and Stanley, manuscript in preparation).

By contrast, the new mutants which belong to complementation groups 1 and 2 (LeclA and Lec2B, respectively) do not appear to result solely from mutations in new genes. They may reflect new mutations in the genes previously identified with the Lecl and Lec2 phenotypes. Alternatively, they may reflect the expression of Lecl and Lec2 mutations on new genetic backgrounds (i.e., they may be the product of double mutations). Whatever the genetics of their origins, they are definitely of interest in structure- function studies of cell surface carbohydrate because of the potentially subtle differences in their cell surface carbohydrates compared with other mutants in the same complementation group. For example, Lec 1A cells are markedly less

Lectin-Resistant CHO Cells 607

tumorigenic than Lecl cells in nude mice (Ripka, Stanley and Shin; manu- script in preparation). Future experiments are aimed at determining whether this difference is a direct consequence of alterations in carbohydrate struc- tures at the cell surface.

Although the use of a combination of lectins is clearly one approach to improving the specificity of lectin selections, it has certain drawbacks. The most important one is the possibility of eliminating novel mutants which exhibit lectin sensitivities identical to the sensitivities of previously character- ized mutants which the selection is designed to eliminate. Another difficulty is the prevalence of mutants belonging to complementation groups 1 and 2 in CHO populations. To reduce the likelihood of isolating these phenotypes, it is clearly important to increase the counterselective pressures against them. It will also be important to explore the use of mutagens besides EMS and MNNG, since different mutagens appear to give rise preferentially to different types of mutations.

Mutations affecting the biosynthesis of the carbohydrate moieties of glycoproteins and glycolipids have already provided a wealth of information on the biochemical pathways of glycosylation in animal cells, on the biological roles of carbohydrate associated with certain membrane molecules, and on the effects of expressing altered carbohydrates at the cell surface (1-4, 15). Clearly it is important to isolate the entire family of these mutants. Although other selective agents give rise to glycosylation mutants, the potential for designing specific selections aimed at obtaining particular types of cell surface carbohydrate alteration or glycosylation mutation would appear to lie with the full exploitation of lectins and lectin-toxin conjugates as selective agents.

ACKNOWLEDGMENTS

The author gratefully acknowledges the excellent technical assistance of Lisa Youkeles and Barbara Dunn and the critical comments of Christine Campbell on the manuscript. This work was supported by a grant from the National Science Foundation (PCM80-23672) until August 1981 and there- after by grant CA 30645 from the National Institutes of Health. Additional support was provided by NCI Core Cancer Grant (1P01 CA13330). P.S. is the recipient of an American Cancer Society Faculty Award.

LITERATURE CITED

1. Stanley, P. (1980). In Biochemistry of Glycoproteins and Proteoglycans, (ed.) Lennarz, W.J. (Plenum Publishing, New York), pp. 161-189.

2. Wright, J.A., Lewis, W.H., and Parfett, C.L.J. (1980). Can. J. Genet. Cytol. 22:443-496. 3. Briles, E.B. (1982). Int. Rev. Cytol. 75:101-165.

608 Stanley

4. Baker, R.M., Hirschberg, C.B., O'Brien, W.A., Awerbuch, T.E., and Watson, D. (1982). Methods Enzymol. 83:444-458.

5. Hyman, R., and Trowbridge, I. (1978). Cold Spring Harbor Conf. Cell Proliferation 5:741-754.

6. Goldstein, I.J., and Hayes, C.E. (1978). Adv. Carbohydr. Chem. Biochem. 35:127-340. 7. Stanley, P., Caillibot, V., and Siminovitch, L. (1975). Cell 6:121-128. 8. Stanley, P. (1981), Mol. CellBiol. 1:687-696. 9. Stanley, P., and Siminovitch, L. (1977), Somat. Cell Genet. 3:391-405.

10. Schachter, H., and Roseman, S. (1980). In The Biochemitry of Glycoproteins and Proteoglycans, (ed.) Lennarz, W.J. Plenum Publishing, New York), pp. 85-160.

11. Chen, T.R. (1977). Exp. Cell Res. 104:255-262. 12. Stanley, P. (1983). Methods Enzymol. 96:157-184. 13. Stanley, P., and Carver, J.P. (1977). Adv. Exp. Med. Biol. 84:265-282. 14. Hirschberg, C.B., Baker, R.M., Perez, M., Spencer, L.A., and Watson, D. (1981). Mol. Cell

Biol. 1:902-909. 15. Kerbel, R.S., Dennis, J.W., Lagarde, A.E., and Frost, P. (1982). Cancer Metastasis Rev.

1:99-140.