[CANCER RESEARCH 52. 3718-3725, July I. 1992]
Differentiation of HT-29 Human Colonie Adenocarcinoma Cells Correlates with
Increased Expression of Mitochondrial RNA: Effects of Trehalose on Cell
Growth and Maturation1
\iang Lu, Teena Walker, John P. MacManus, and Verner L. Seligy2
Institute of Biological Sciences, National Research Council of Canada, Ottawa, Ontario, K1A OK6, Canada
ABSTRACT
The HT-29 human adenocarcinoma cell line has been used extensively
in the study of colonie cell differentiation and colon cancer. We report
here that substitution of glucose with trehalose (a-i>-glucopyranosyl-ai>-glucopyranoside) depresses growth and promotes mucin-producing,
goblet-like maturation of HT-29. An initial characterization of this
process was made by analyzing several cDNA clones whose RNA tem
plates were differentially expressed at elevated levels in cells grown in
trehalose-containing medium. Seven of the 9 clones examined corre
sponded to 6 mitochondria! genes whose expression levels, relative to
those from glucose-grown cells, ranged from approximately 3-fold for
165 r RNA to 8-23-fold for NADH dehydrogenase subunit 4. On the
other hand, levels of mitochondria! DNA copy, measured by using NADH
dehydrogenase subunit 4 cDNA as probe, were shown to be unaffected
by trehalose treatment. Elevation of cellular NADH dehydrogenase
subunit 4 RNA in HT-29 cultures grown in medium containing different
components (sodium butyrate, galactose, no-sugar, glucose, cellobiose)
generally correlated with depressed growth levels and specifically with
increased numbers of mucin-producing cells present. Like butyrate, the
sugar, trehalose, is an effective inducer of HT-29 differentiation, and
may prove useful as a dietary therapeutic, and as a probe for elucidating
mitochondria! involvement in colonie cell differentiation and
transformation.
INTRODUCTION
Among the several adenocarcinoma cell lines derived from
human colonie mucosa, HT-29 is one of the best described (1,
2). As a poorly differentiated, multipotent tumor cell type, HT29 can be induced to mature into both enterocyte-like goblet
and absorptive cell phenotypes by various substances such as
sodium butyrate (3), suramin (4), 12-0-tetradecanoylphorbol13-acetate (5), and galactose (6). Little is known about how
these "inducers" work, and in particular how exogenous glucose
levels (7, 8) directly modulate HT-29 cell growth and
differentiation.
We are interested in the molecular genetic mechanisms that
both trigger and facilitate carbohydrate-dependent differentia
tion and dedifferentiation responses in HT-29 cells for the
purpose of obtaining a better understanding of how normal and
transformed enterocytes evolve. Early investigations have
shown that several sugars can substitute for glucose in support
ing growth of normal and neoplastic cells (9). As a follow-up
to this work, and related studies using insect cells in culture
(10), we were interested in the effects that disaccharides such
as trehalose might have on gene expression and regulation of
colonie cells in general, and in their potential to mediate HT29 differentiation and to attenuate the colonie tumor cell pheReceived 12/26/91 ; accepted 4/20/92.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1This is National Research Council of Canada publication 32005. Work was
supported by IBS intramural research funds.
•¿
To whom requests for reprints should be addressed, at Molecular Cell Biology
Group, Institute of Biological Sciences, Building M-54, Room 2153, NRCC,
Ottawa, Ontario. Canada K1A OR6.
notype. Trehalose («-D-glucopyranosyl-a-o-glucopyranoside) is
a relatively nonlabile glucose derivative that occurs naturally,
often in high abundance, in many lower organisms, plants, and
insects, in which it has been suggested to play various roles in
osmoregularity, regulation of sugar transport and metabolism,
and thermotolerance (10-14). The functions of this sugar and
its hydrolase, trehalase (EC 3.2.1.28) in mammals are not well
understood; they may be involved in active transport of glucose
and carbohydrate metabolism (14). Mammalian trehalase is
associated with brush-border membranes of the small intestine
and kidney (15); it is expressed at low levels in LLC-PK, (pig
kidney) epithelial cells after glucose limitation (16), and in HT29 in the presence of glucose (17).
In this report, we analyzed cell growth, differentiation, and
expression of some of the most abundant transcripts present in
HT-29 cells treated with trehalose. Among cloned cDNAs
detected by differential hybridization using cDNA probes made
from total RNA of cells grown in medium containing either
glucose or trehalose, the majority were found to be encoded by
mt* DNA. A comparison of mt RNA expression in HT-29 cells
exposed to butyrate, galactose, no-sugar, and cellobiose further
indicated that enhanced mt gene expression is linked generally
to cultures exhibiting depressed levels of growth and increased
numbers of differentiated cells. These observations extend the
expression studies of cytochrome c oxidase in HT-29 (18, 19)
using small chain fatty acids, and underscore the importance of
understanding the differences in mitochondria! structure and
expression so far observed for normal and abnormal colonie
tissues (18-23).
MATERIALS
AND METHODS
Cell Culture and Treatment. HT-29 cells (American Type Culture
Collection 8616, cell passage 128) were initially grown in McCoy's 5A
medium, and maintained thereafter in this laboratory on glucose-free
Dulbecco's modified Eagle's medium (Gioco BRL; Grand Island Bio
logical Co., Grand Island, NY) supplemented with 25 HIMglucose, 10%
fetal bovine serum (v/v), and 40 ^g/ml gentamicin at 37°Cand 8.0%
CO2. Cells (passage 130) were subsequently adapted to Dulbecco's
modified Eagle's medium containing one of the following: 25 HIM
cellobiose ((M,4-glucosyl-D-glucoside; Sigma Chemical Co., St. Louis,
MO), 5 HIMgalactose, 25 mivi glucose, 2 mM sodium butyrate, or 25
itiM trehalose (a-o-glucopyranosyl-a-n-glucopyranoside;
Sigma). Since
HT-29 rapidly consumes glucose, glucose-containing medium was re
placed daily to minimize spontaneous differentaition. Medium in all
other culture regimes was replaced every 3 days. For the nutrient shift
experiments, approximately IO7cells grown in glucose-containing me
dium were incubated subsequently for different time intervals in me
dium containing either trehalose or trehalose for 6 days and then shifted
to glucose. Cells were harvested following trypsin-EDTA treatment by
pelleting in a clinical centrifuge and frozen at -75°C until use. A
Neubauer hemocytometer was used to count cells. Monolayers of cells,
fixed in buffered 10% formalin, were stained for acid mucopolysaccha3 The abbreviations used are: mt. mitochondrial; pfu. plaque-forming units;
SSC, standard saline-citrate: SDS, sodium dodecyl sulfate: /VD4, NADH dehy
drogenase subunit 4.
3718
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITÎXÕIONDRIALGENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
rides (mucin) according to the method of Thompson (24), using 1%
alcian blue-3% acetic acid, pH 2.5 (30 min), followed by counterstaining
in 0.1 % nuclear fast red solution ( 10 min) and mounting in Hydramount
(Gurr's, United Kingdom) for photography with a Reichert Zetopan/
Table 1 Characterization ofHT-29 cells grown in different carbon sources
Nikon UFX assembly.
Construction and Screening of cDNA Libraries. cDNA was prepared
according to the method of Gubler and Hoffman (25) with modifica
tions (26). First-strand synthesis was carried out in a reaction volume
of 50 n\ containing 70 pg of RNA, 5 ^g of oligo-dT, and 36 units of
reverse trancriptase. cDNA was methylated with 100 units of EcoRl
methylase and ligated to phosphorylated
EcoRl linker (5'CCCGAATTCGGG-J';
Pharmacia) at a molar ratio of 100:1. After
excessive digestion with EcoRl, the cDNA was passed through a 1-ml
Bio-Gel A-50 m (Bio-Rad, Ontario, Canada) column to remove extra
neous EcoRl linker. The fractions containing the cDNA were ligated
with EcoRl digested X-gflO and packaged into phage according to
commercial protocols (Stratagene, La Jolla, CA; see also Ref. 27). For
differential screening of 2 cDNA libraries (2 x IO6pfu), double stranded
0.3-0.4
Cellobiose
1.7
0.3-0.4
Galactose
5.2
1.0-2.0
Glucose
3.14.2
0.3-0.4
No sugar
0.3-0.4
Trehalose%
6.6Total
°Culture conditions are described in "Materials and Methods."
"Based on 10' cells counted from 8 and 14 days of culture stained with alcian
blue dye (see "Materials and Methods").
' Cells counted after 8 and 14 days of culture.
Carbon"
sourceButyrate
Mucin-containing
cells*'10.0
cell
no.'x
10"0.3-0.4
cDNA (80 to 100 ng) was prepared as described above and labeled
using random hexaprimers in a reaction containing 100 ^Ci of [a--'2P]
dATP to yield approximately 2x10* cpm (28). Use of this method
was necessary since '2P-labeling of first strand cDNA during synthesis
<o
-2-10
1
23456789
10
11 12
13
14
15
Fig. 1. Characterization of HT-29 cells. A, growth curve of cells in glucose (O)- and trehalose (»(-containing medium. Total cell numbers were determined from
duplicate 25-cmFD culture chambers. B and C, phase contrast micrographs of HT-29 cells cultured in glucose (B) and trehalose (C) after 6 days. D and E, micrographs
of mucin-producing cells in the same cultures as in B and C. Arrows, examples of alcian blue-positive, mucin-producing cells.
3719
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIA!.
GENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
usually gave only about 2 x 10" cpm/^g cDNA. Duplicate Hybond
nylon membrane filters (Amersham, Oakville, Canada) were lifted from
85-mm agar plates, each containing about ISOOpfu of the A-#MOcDNA
library. After denaturation in 0.5 M NaOH, 1.5 M NaCl, neutralization
in 0.5 M Tris, 1.5 M NaCl (pH 7.5), and a wash in 2 x SSC, the filters
were baked for 2 h at 80°C.The duplicate filters of each set were
hybridized to 12P-labeled cDNA made from RNA of cells treated for 6
days with either glucose or trehalose in the medium. After hybridization
and washes, the filters were exposed to Kodak X-Omat film. Autora
diographs of corresponding filters were compared visually, and plaques
that hybridized differently to the 2 probes were picked and subjected to
further rounds of differential hybridization until deemed plaque pure.
DNA Analysis. DNA was isolated using the protocol that allowed
the simultaneous extraction of DNA and RNA (29). Alternatively, the
DNA- and protein-containing supernatant recovered after CsCI centrifugation was dialyzed against TE buffer for 3 h, and SDS was added
(1% final) before extraction (once with phenol and once with chloroform:isoamylalcohol, 24:1) and precipitation with ethanol. For South
ern transfers, DNA was digested with £coRIand electrophoresed in
agarose gels (1%) with TBE buffer for 18 h at 2 V/cm before capillary
blotting (Nytran membrane; Schleicher & Schuell, Dassel, Germany)
in 25 mM sodium phosphate buffer, pH 7.2, cross-linking by exposure
to short-wave UV and vacuum-drying at 80°C.Recombinant X-gtlO
DNA was isolated using a miniscale protocol (30). Sizes of inserts from
the \-gt\0 vector were measured following EcoRl digestion and verified
by amplification (polymerase chain reaction) using commercially avail
able flanking primers and Taq polymerase (Gibco BRL) as described
elsewhere (30). Inserts were subcloned into a plasmid vector for se
quencing. DNA sequencing was carried out (30) with Sequenase (USB
Corp.) using the manufacturer's protocol. Nucleotide sequences of
B
individual clones were compared with these in the EMBL/GenBank
data base by the FASTA method in the GCG package (Version 6) from
the Wisconsin University (31).
RNA Analysis. Total RNA was extracted from HT-29 cells (approx
imately 10" cells/sample) by addition of 1.5 ml of 4 M guanidium
isothiocyanate containing buffer and purified by CsCI centrifugation
(30). Purified RNA was dissolved in H20 and extracted once with
phenol. RNA concentration for comparative analysis was estimated
initially by absorbance measurement at 260 nm and by the amount of
ethidium bromide staining of 185 and 285 rRNA per gel lane. Relative
RNA concentration was further determined by using DNA probes (see
Table 2) for the abundant RNA species, elongation factor EFla and
285 rRNA. The murine EFla cDNA probe has a homology of 92% at
the nucleic acid level to the human sequence (32). Total RNA (12 ¿ig)
for each sample was dissolved in 20 n\ of 44% formamide, 20 mM 3(TV-morpholino)propanesulfonic acid, 5 mvi sodium acetate, 1 mM
EDTA, 6% formaldehyde, 0.25% bromophenol blue, 0.25% xylene
cyanol FF, pH 7.0, and heated 10 min at 70°C.The 1% agarose gel
containing running buffer [20 mM 3-(/V-morpholino)propanesulfonic
acid, 5 mM sodium acetate, and l mM EDTA, pH 7.0] and 2%
formaldehyde was developed for 18 h at 2 V/cm. RNA was transferred
to Nytran membranes in 40 m.MTris-acetate, 1 mM EDTA at 4"C for
1 h at 15V, followed by 3 h at 45 V using an electrotransfer unit
(Hoeffer Scientific Instrument, San Francisco, CA).
Hybridization. DNA probes were radiolabeled to a specific radioac
tivity of 5 x 10*-5 x IO1*cpm/Vg using random hexaprimers (28).
Prehybridizations (2 h) were done in 6x SSC, 5x Denhardt's solution,
abcdefg
hi
jkl
Fig. 2. Analysis of cloned cDNA inserts by use of polymerase chain reaction
and differential hybridization. A. ethidium bromide-stained agarose gel showing
lanes with 75-200 ng of insert from clones DP 1.2 (Lane a), DP2.2 (Lane b),
DP4.2 (Lane c), DPI.2 (Lane e). DP6.2 (Lanef). and DPI 1.2 (Lane h) amplified
by polymerase chain reaction using Taq polymerase (see "Materials and Meth
ods"). Reference DNA sequences are: EcoRl digested pCD29 (murine EFlo,
Lane i) and crude preparation of HT-29 ml DNA cut with BamHl (Lanej) or
BamH\ and EcoRl (Lane k), and 1 kilobase pair repeat, marker DNA (Lane I).
B and C, autoradiographs of duplicate DNA filters hybridized first with |"P]
cDNA made from total RNA isolated from HT-29 cells cultured for 6 days in
medium containing glucose (B) and ("PjcDNA made from RNA from cells
cultured in trehalose (C).
0.5% SDS, and 100 i/g/ml denatured salmon sperm DNA (29). Hy
bridizations were carried out for 18 h in 6x SSC, 0.5% SDS, and 100
^g/ml denatured salmon sperm DNA containing 5 x 10M07 cpm/ml
probe. Hybridized membranes were washed for 15 min at 23°Cusing
2x SSC, 0.2% SDS, and twice for 30 min at 65°Cin 0.2x SSC, 0.2%
SDS before exposing to Kodak X-Omat films. Hybridized blots were
reused for control experiments by washing each blot for 30 min at 95°C
in 0.2X SSC, 0.2% SDS.
Scanning and Quantitation. Autoradiographs of hybridizations were
analyzed using a computing scanning densitometer (Model 300A; Mo
lecular Dynamics. CA) with software (ImageQuant version 3.0). Inten
sities of hybridization signals were compared using the "volume inte
gration" option of the software program.
3720
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIAL GENE EXPRESSION AND COLON CANCER CELL DIFFERENTIATION
Table 2 Characterization ofcDNA clones and corresponding transcripts from trehalose-grown HT-29 cells
RNA
size
size
designation
(kilobase
pairs)*1.61.23.01.61.6>101.81.61.41.71.75.1Relative
concentration'3.07.03.68-238-233.18-237.07.01.201.101.15Gene
(kilobase
pairs)"1.61.21.4+
(EMBL/GenBank)loSrRNAATPase
cDNA
cloneDP2.2DP4.2DP5.2DP6.2DP7.2DP8.2DP9.2DP10.2DP11.2DP1.2pCD29pCD22cDNA
8.cytochrome
6. ATPase
IINuclear,
oxidase
unknownNADH
4/4LNADH
dehydrogenase
4/4LNuclear,
dehydrogenase
unknownNADH
4/4L165rRNACytochrome
dehydrogenase
1.61.51.22.41.20.21.21.71.71.1RNA
bEFlaEFla
(murine)28S
rRNA (murine)
°Determined by agarose gel electrophoresis after digestion of cloned DNA with £coRI.
'' Determined from hybridization analysis using RNA from HT-29 cells grown in glucose or trehalose for 6 days.
' Quantitation of hybridizations to RNA from trehalose-grown cells (relative to glucose) using computing densitometer (see "Materials and Methods" and Fig. 4).
B
DPlX2
20
30
40
50
i
100
110
.ATOACATTAACACTATTCTCACCA
«0
rCACCTCCCATTCCQATIUUlATCACCTTCCÄCCCTTJlCTÄCÄCfcATCXAAOÄCOCCCTCOOCTTÄ'
TCACCTCCCATTCCaATAAAATCACCTTCCACCCTTACTACACAATCAAAOACOCCCTCaaCTTACT
110
130
140
150
OACCTOTA«KOACCCAaACAATT*TACCCTA«CAACCCC
DP6.2
20
10
40
140
B
50
Z
40
Z
TAATaACATTAACACTATTCTCACCA
R
A
V
C
P
B
W
T
70
io
tO
100
110
10
»o
100
110
ACCACATTAAGAACATAAAACCCTCATTCACACaAaAAAACACCCTCATOTTCATACACCTATCCCCCATTCTC^
20
10
40
50
<0 A
ATTTACACCAACCACCCAACTATCTATAAACCTACKICATaGCCATCCCCTTATOAOCGO^
70
TaAOCM^
«TACACCCCTTATCCCCATACTAOTTATTATCOAAACCATCAOCCTACTCATTCAACCAATAGCCCTOaCC
CACCTACÄCCCCTTATCCCCATACTAaTTATTATCOAAACCATCAOCCTACTCATTCAACCAATAOCCCTOaCC
•¿Â»30
ATAAAaTATAaOCOATAGAAATTaAAACCTGaCOCAATAOATATA
TACCAaACAACCTTAOCCWU^CCATTTACCCAAATAÕÕÓT^
.TAATGAATTAACTAGAAATAAerTTGCAAGGAGAACCAAAüC
'AACTAaAAÕTAÕfirmaCÕÕGOAGAGCCÕAAOC
Fig. 3. Sequence identification of cDNA clones and location on mitochondria! DNA. A, partial sequences of DP2.2, DP4.2, DP6.2, and DPI 1.2 were positively
matched to regions within the int sequence MIHSXX of Ik-la cells (33) available from EMBL/GenBank data bases. Numbers in lower lane, nucleotide position in
MIHSXX. B, location of the 4 cDNA clones relative to MIHSXX. Cyto b. cytochrome b; 16s, 165 rRNA; COII, cytochrome oxidase II; A6, ATPase 6; A8, ATPase
8; ND4, NADH dehydrogenase 4/4L; E, £coRIsite.
RESULTS
Effect of Trehalose on Cell Growth. When glucose in Dulbecco's medium was replaced by trehalose, differences in growth
rate and morphology of HT-29 were observed. As shown in
Fig. IA, cell density after 7 days in medium containing trehalose
was approximately 4- to 5-fold lower than that observed for
cells grown in glucose for the same period of time. At the
microscopic level, cells grown in glucose-containing medium
were seen to form densely packed, disorganized multilayers of
undifferentiated cells, while cells in trehalose-containing me
dium formed uniform monolayers that were often polarized,
polygonal in shape, and delimited by large extracellular spaces
(Fig. 1, /i and C). Staining of both types of cell population with
alcian blue to detect mucin-producing cells (Fig. l, D and E)
revealed that approximately 10-fold more goblet-like enterocytes were produced with trehalose than with glucose in the
medium. Cells shifted back from 6 days of treatment with
trehalose to glucose resulted in recovery of the glucose-culture
cell phenotype after several days of incubation. Similar assays
carried out using 8- and 14-day cultures of HT-29 adapted to
alternate carbon sources (Table 1) indicated that in comparison
to cells grown in trehalose, cultures in butyrate produced more
mucin-producing cells, and cultures in galactose, no-sugar, glu
cose, and cellobiose (relative order) produced less. However,
except for glucose, total cell production was about the same
using any of the other treatment regimes.
Selection of Differentially Expressed Genes. Considering the
morphological and histochemical changes noted for HT-29
after shifting from glucose to trehalose-containing medium, we
expected differences in gene expression to occur. As a start in
the characterization of these genetic events, we constructed 2
recombinant \-gt\0 cDNA libraries from total RNA of treha
lose grown cells and screened them for cDNA clones that
encoded RNA sequences whose abundance changed the most
during HT-29 cell adaptation to trehalose (see "Materials and
Methods" for details). From an initial screening of approxi
mately 6 x IO4clones (pfu) with radiolabeled cDNA made from
RNA of cells cultured with either glucose or trehalose, we
3721
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIA!.
GENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
B
Fig. 4. cDNA hybridization to total RNA from HT-29 cells grown in glucose
or trehalose. I-//, autoradiograms of Northern blots of identically fractionated
total RNA (approximately 12 ,,«Line) from HT-29 cells cultured for 6 days in
25 itiM glucose or trehalose (left and right lanes of each panel, respectively)
hybridized with radiolabeled DNA probes. A-D, hybridization with inserts from
clones DP2.2. DP4.2, DP6.2, and DPI 1.2, respectively. E-H, same filters as in
A-D hybridized with EF1«cDNA probe.
obtained 9 candidate clones. The selectivity of this hybridization
screening procedure is illustrated in Fig. 2, A-C (Lanes b, c, e,
f, and h) using the in vitro amplified (polymerase chain reaction)
cDNA inserts from some of these clones. Hybridization of these
cDNA inserts and HT-29 mt DNA (Fig. 2, A-C, Lanes j and k)
was about 10-fold higher with [':P]cDNA made from RNA of
trehalose-grown cells (Fig. 2C) than with [i;P]cDNA made from
RNA of glucose-grown cells (Fig. 2B). This was in contrast to
results obtained for cDNA of clone DP 1.2 (Fig. 2B, Lane a),
which was initially selected on the basis of equivalent hybridi
zation to [12P]cDNA probe made from either source of RNA.
This cDNA contains sequence coding for elongation factor
EFla, a highly expressed gene that is homologous to the mouse
EFla sequence also used as a standard in this experiment (Fig.
2B, Lane i) and in subsequent analyses (see Figs. 4 and 5).
To reveal the identity of the 9 cDNA clones, cross-hybridi
zation of the cDNA inserts and partial nucleotide sequencing
were carried out. These results are summarized in Table 2 and
Fig. 3. The cDNA of DPS.2 and DPS.2, which are unrelated to
any of the other clones, also did not show any significant
sequence analogues in the EMBL/GenBank database. How
ever, sequences of the remaining 7 clones indicated them to be
derived from mt DNA-encoded transcripts. The homologies
shown in Fig. 3 between HeLa mt DNA sequence (33) and the
cDNA sequences were: 99.6% in 227 base pairs of DP2.2 with
165 rRNA; 99.5% in 199 base pairs of DP4.2 with DNA
spanning the coding region for ATPase 6, ATPase 8, and
cytochrome oxidase II (referred to here as ATP6/ATP8/COU);
100% in 113 base pairs of DP6.2 with DNA of NADH dehydrogenase subunit 4/4L (referred to as ND4); 95.2% in 188
base pairs of DPI 1.2 with cytochrome b (referred to here as
Cyto b).
Inserts from all 4 clones were sequenced from both ends, and
sequence integrity of these clones was further checked by re
striction site mapping using enzymes Hindi, Kpnl, and .Vu«
II
for DP2.2 (165rRNA); Acci, Apal, and Xbal for DP4.2 (ATP6/
ATP%/CO\\); Hindlll and Spel for DP6.2 (ND4); and Acci for
DPI 1.2 (Cyto b). All clones gave the expected diagnostic re
striction cuts (data not shown). DP4.2 (ATP6/ATPS/COII) is
unusual and appears to have been generated from an unpro
cessed, polycistronic RNA. The one nucleotide mismatch noted
in ATP6/ATPS/COII (A to G at nucleotide position 8, 860)
would result in an amino acid change from threonine (polar
uncharged) to an alanine (nonpolar). A significant sequence
mismatch of 8 nucleotides was found in Cyto h at nucleotide
positions 15,545 to 15,567, resulting in a change in sequence
from HURA WC to HIKPEWY (single letter amino acid code).
Since the Cyto b was sequenced from the noncoding strand, the
mismatches were at the beginning of the analyzed sequence and
therefore not a result of incorrect reading.
Quantitation of Elevated Transcript Levels. Data in Fig. 4, AD, show that cDNA of clones DP2.2, DP4.2, DP6.2, and
DP 11.2 all hybridize to RNA, whose expression levels were
highest in cells grown in trehalose rather than in glucose. This
confirms the results obtained in Fig. 2 and indicates that specific
cDNA synthesis and abundance generally reflected template
RNA copy levels and not RNA template preference. Equality
in concentration of total RNA in samples (see "Materials and
Methods") was confirmed on the Northern blots using cDNA
probes for 285 rRNA (data not shown) and EFla (Fig. 4, EH). This is also supported by data presented in Fig. 2 using
both murine and HT-29 EFla cDNA as standards.
Since the cDNA sequences were identified, most of the RNA
sizes revealed by Northern analysis in Fig. 4 could be inter
preted. For example, the 2 major bands in Fig. 4B are consistent
with their being ATPase 6/ATPase 8 and cytochrome c oxidase
II mRNA (33, 34), and the single bands in Fig. 4, C and D, are
consistent with their being NADH dehydrogenase subunit 4
and cytochrome b, respectively. However, the faint 2.4-kilobase
band near the 185 rRNA in Fig. 4B could not be identified.
The intensities of the hybridizations to the different RNA
species were further quantified using a computing densitometer.
While all mt RNAs showed an elevated expression, the magni
tude of their increase varied from 3-fold for 165 rRNA to 8- to
23-fold for ND4 (Table 2). The elevation of transcript for each
mt gene is considered real since the same Northern blots when
stripped and hybridized to probes for commonly expressed
RNA species (285 rRNA and EFla) gave comparable results
(<1.20).
Mitochondria! DNA and ND4 RNA Concentration. To distin
guish whether the observed increases in mt RNA were due to
an increase in stable transcription or an increase in mt DNA
copy during induced cell differentiation, RNA and DNA were
isolated from the HT-29 cells, 1 and 3 h, and 1, 2, 4, and 6
3722
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIA!.
GENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
CO
Fig. 5. Expression of ND4 and mitochon
dria! DNA copy. Top insert, autoradiogram of
Northern blot of total RNA hybridized with
ND4 cDNA. RNA (approximately 12 jig/lane)
was isolated from HT-29 cells cultured for 0,
1, and 3 h. and 1. 2. 4. and 6 days in 25 HIM
trehalose. Bottom insert, corresponding South
ern blot of DNA hybridized to /VD4 cDNA.
The DNA (10 ng/lane) was isolated from (he
same cells as the RNA described above, and
cut with EcoR\. Dotted bars, quantitation of
hybridization signals of HT-29 RNA to MM
cDNA; solid bars, quantitation of HT-29 DNA
hybridized to A/D4 cDNA; hatched bars, quan
titation of HT-29 DNA hybridized to EFla
cDNA. Numerical values correspond to the
integrated
volume (see "Materials
and
Methods").
6d
exposed to trehalose was established by comparing specific
RNA and DNA levels of cultured cells maintained in medium
containing either glucose or trehalose with levels in cells trans
ferred from trehalose back to glucose. As shown in Fig. 6, the
expression of the ND4 RNA remained high as long as HT-29
cells were maintained in trehalose medium.
ND4 Expression in Cells Exposed to Other Sugars or Butyrate.
We used the ND4 cDNA as a probe to measure ND4 RNA
levels in RNA isolated from HT-29 cells cultured in medium
containing different carbon regimes as described in Table 1.
The hybridization analysis summarized in Fig. 7 indicates that
except for cellobiose, expression levels of cellular ND4 RNA
correlated approximately with the amount of cell differentiation
that had taken place as estimated by measurement of mucinproducing cells (Table 1).
CO
CO
2 15-
0)
3
10-
eO
TRE
TRE->GLU
DISCUSSION
GLU
Fig. 6. Changes in level of ND4 mRNA with cell culture conditions. Left to
right, HT-29 cells cultured for 6 days in trehalose. 6 days after transfer from
trehalose to glucose, and 6 days in glucose. Insert, autoradiogram of corresponding
Northern blot of RNA (12 ng/lane) from cultured cells. Hybridization to /VD4
cDNA probe was quantiated as described in Fig. S and "Materials and Methods."
days after transferring from glucose to trehalose. To monitor
mt DNA copy, total DNA was digested with EcoRl to generate
the 6.5-kilobase mt DNA fragment containing ND4 (see Fig. 3,
map). For quantitation, both Southern and Northern blots were
hybridized to the ND4 cDNA probe. Fig. 5 shows that while
the ATM-related mt DNA content remained constant, an in
crease in the amount of/VD4 mRNA could be observed as early
as a few hours after shift from glucose to trehalose. mt DNA
content in total DNA isolated from cells at 0 and 6 days after
transfer to trehalose were also the same when probed with total
mt DNA.
The quantitation presented in Fig. 5 further shows that
maximal relative increases in ND4 mRNA took place during
the first 2 to 4 days, after which time levels remained constant.
Confirmation of the elevated expression level of ND4 in cells
When glucose was replaced by trehalose in the culture me
dium, HT-29 cells were found to undergo phenotypic changes
that are more similar to those observed using either sodium
butyrate or galactose than to those using no-sugar or glucose.
Among the genes identified so far whose expression changed
the most, 7 were mt DNA encoded and 2 others were nuclear
DNA encoded but unrelated to any gene sequence presently in
the EMBL/GenBank databases. The corresponding transcript
of one of these clones (DPS.2) is unusually large and is currently
being analyzed. Use of internal standards such as EFla and
285 rRNA for quantitation of mt RNA in total RNA extracted
from cells further revealed probable differences in relative
amounts of mt RNA ranging in magnitude from approximately
3-fold for 165 RNA to 8-23-fold for ND4. The level of ND4
RNA expression in total RNA from cells exposed to sodium
butyrate, galactose, no-sugar, and cellobiose (order of dimin
ished expression) correlates favorably with decreased cell
growth and particularly with increased numbers of mucinproducing cells present. With refinement, levels of mt RNA
such as ND4 may serve as a generic indicator of the state of
3723
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIAL
V)
GENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
relative to COIII RNA have not been measured in HT-29 to
determine which is the highest. However, the increased levels
of expression of these RNAs in trehalose-induced (this study)
and butyrate-induced (18) HT-29 cells are not accounted for by
any demonstrable amplification or deletion in mt DNA that
may have occurred during cell adaptation -to-these carbon
sources, mt DNA levels from colonie tumors and polyps are
also similar to levels in normal colonie tissue (22, 23, 37). The
finding of elevated levels of RNA encoded by 4 mt genes,
including subunits COI and COI I, in biopsies from familial
polyposis coli patients (22), would appear to be inconsistent
with a similar analysis using COIII cDNA as probe (18). Age
and origin of cell mass in such samples are likely important
factors that will have to be qualified further (38). We have noted
that trehalose does not promote elevated expression of mt genes
in the monocyte line HL-60.5 Recent studies on noncolonic
8
CO
CO
0
4
(O
o>
0
2
G
B
C
Ga
NS
T
Fig. 7. yVD4 mRNA levels in HT-29 cells after exposure to various cell
differentiation regimes. Insert, autoradiogram of a Northern blot of RNA (12 un
lane) from cells cultured for 6 days in medium containing one of the following:
25 m\i cellobiose (C), 5 m\i Na butyrate (B). 25 m\i glucose (G). no-sugar (NS),
25 m\i galactose (Ga). 25 m\i trehalose (T). RNA obtained from these cells was
hybridized to MM cDNA probe and quantiated as described in Fig. 5 and
"Materials and Methods."
HT-29 differentiation. Because of the relatively large changes
in ND4 (to COll for example), the ND4 gene might also be
useful as a probe in the assessment of colon biopsies.
The effects of trehalose on HT-29 production of mt RNA
and goblet-like cells are not equivalent to those obtained by
removal of glucose (6-8) or by substitution with other carbo
hydrates such as galactose or cellobiose, a ß-l,4-linked dimer
of glucose. Also, unlike butyrate (19), it appears that trehalose
is not metabolized to any major extent by HT-29 cells, and
unlike insect cells in culture that exhibit 4- to 8-fold increases
in trehalase activity when exposed to trehalose (10), HT-29
cells in the presence of trehalose4 or glucose (17) express only
low levels of trehalase activity. HT-29 cells are also in apparent
contrast to LLC-PK,, pig kidney epithelial cells that exhibit
glucose-dependent changes in trehalase expression (16). It is
possible that removing or substituting glucose with another
sugar stimulates the cells to respire. However, in view of trehalose's purported role in osmoregularity, thermotolerance, and
carbohydrate metabolism in various organisms (10-14), treha
lose may also affect HT-29 membrane permeability and general
uptake of nutrients. It will be interesting to determine whether
the changes in ND4 RNA levels after treatment with HT-29
cells with the various carbon regimes described here are due to
a general stress response involving modulated expression of
glucose/oxygen/heat shock-regulated proteins (35, 36).
The finding that expression levels of several mt genes, includ
ing COll and cytochrome c oxidase subunit II, are elevated
when HT-29 cells are induced to differentiate by culturing with
sugars other than glucose or no-sugar extends observations
showing elevation of COI and COIII RNA in HT-29 cells
cultured with small chain fatty acids (19), and the depression
of COIII RNA in tissue biopsies of colon adenomas and carci
nomas (18). At this time, levels of expression of ND4 RNA
* A. P. Jahagirdar. I. Rasquinha. J. P. MacManus. and V. L. Seligy, Carbon
source induced variation in expression of disaccharidase and peptidase activities
of HT-29 cells, manuscript in preparation.
cells clearly demonstrate that neoplastic transformation of
either rat hepatoma cells (39) or human fibroblasts (40) is
associated with increased levels of mt RNA.
Differences in morphology, rhodamine-123 staining, and en
zyme properties of mitochondria from human colon carcinoma
cell lines have been noted previously (20, 21). At this time, we
do not know whether the increases in the levels of RNA encoded
by the different mt genes corresponds to similar increases in
activity of the enzyme products. Treatment of HT-29 cells with
small chain fatty acids enhances cytochrome c oxidase activity
(19). It is uncertain whether the nucleotide sequence mis
matches we detected between Hela mt genes (33) and the mt
cDNA clones have any functional significance in HT-29. The
cDNA clones we identified code for 6 mt genes that are distrib
uted rather uniformly over the mt genome and are themselves
the products of polycistronic RNA corresponding to almost the
total length of the mi genome (33, 34). The differences in
specific mt RNA levels, which ranges from 3-fold to 23-fold,
may be a result of different rates of turnover or processing. The
unusually high level of ND4 transcript expression has no ap
parent relationship to location of ND4 coding region relative
to the 2 major mt DNA promoters. Interestingly, Corral et al.
(39) noted that during rat hepatocarcinogenesis
165 rRNA,
ND\, ND6, COll, and Cyto b mt RNA levels were moderately
increased by 2- to 3-fold over control levels and that ND5 mt
RNA was increased by about 10-fold. In this case, mt DNA
heteroplasmy and/or mutations may explain the differences in
RNA expression (39). Further elucidation of mt genome struc
ture, and the molecular events involved in coordinating expres
sion of nuclear and mt genes in HT-29 cells, would appear to
be important for understanding the metabolic relationships
between growth and differentiation phases of these cells.
ACKNOWLEDGMENTS
We wish to thank M. J. Dove for general technical assistance, I.
Rasquinha for tissue culture assistance, and T. Devercseri for photog
raphy. Critical reading of the manuscript by Drs. A. P. Jahagirdar and
N. Seto is appreciated.
REFERENCES
1. Fogh, J., and Tremp, G. New tumour cell lines. In: J. Fogh (ed.). Human
Tumour Cells "in Fi/ro," pp. 115-141. New York: Plenum Publishing Corp.,
1975.
*X. Lu and V. L. Seligy. HspoO/chaperonin gene expression and differentiation of human colon adenocarcinoma and multipotent leukemic cells, manuscript
in preparation.
3724
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
MITOCHONDRIA!.
GENE EXPRESSION
AND COLON CANCER CELL DIFFERENTIATION
2. Zweibaum, A., Laburthe, E., Grasset, E., and Louvard, D. Use of cultured
cell lines in studies of intestinal cell differentiation and function. In: M. Field
and R. A. Frizzel (eds.). Handbook of Physiology. The Gastrointestinal
System. Vol. IV, Intestinal Absorption and Secretion, pp. 223-255. Wash
ington. DC: American Physiological Society. 1991.
). Augeron, C., and Laboisse, C. L. Emergence of permanently differentiated
cell clone using a human colonie cancer cell line in culture after treatment
with sodium butyrate. Cancer Res., 44: 3961-3969, 1984.
4. Fantini, J., Rognoni, J. B., Roccabianca, M., Pommier, G., and Marvaldi, J.
Suramin inhibits cell growth, glycolytic activity and triggers differentiation
of human colic adenocarcinoma cell clone (HT29-4). J. Biol. Chem., 264:
10282-10286. 1989.
5. Rochette-Egly, C., Kedinger, M.. and Haffen, K. Modulation of HT-29
human colonie cancer cell differentiation with calmidazolium and 12-0tetradecanoylphorbol-13-acetate.
Cancer Res.. 48: 6173-6182, 1988.
6. Pinto, M., Appay, M. D., Simon-Assmann, P., Chevalier, G.. Dracopoli. N.,
Fogh, J., and Zweibaum, A. Emerocytic differentiation of cultured human
cancer cells by replacement of glucose by galactose in the medium. Biol.
Cell., 44: 193-196, 1982.
7. Zweibaum, A., Pinto, M., Chevalier, G., Dussaulx, E., Triadou, N., Lacroix,
B., Haffen. K., Brun, J. L., and Roussel. M. Enterocytic differentiation of a
subpopulation of the human colon tumor cell line HT29 selected for growth
in sugar free medium and its inhibition by glucose. J. Cell. Physiol., 722: 2129, 1985.
8. Biais, A. Expression of Na*-coupled sugar transport in HT29 cells: modula
tion by glucose. Am. J. Physiol.. 260.-C1245-C1252, 1991.
9. Eagle, H., Barban, S., Levy, M., and Schulze, H. O. The utilization of
carbohydrates by human cell cultures. J. Bio!. Chem.. 233: 551-558, 1958.
10. Jahagirdar, A. P., and Seligy, V. L. Modulation of trehalase activity in two
insect cell lines by virus infection and trehalose. Biochem. Int., 23: 10491054, 1991.
11. Crowe, J. H., Crowe, L. M., and Champman, D. Preservation of membranes
in anhydrobiotic organisms: role of trehalose. Science (Washington DC),
223: 701-703. 1984.
12. Thevelin, J. Regulation of trehalose mobilization in fungi. Microbiol. Rev.,
«.•42-59,1984.
13. Neves, M. J., Jorge, J. A., Francois, J. M., and Terenxzi, H. F. Effects of
heat shock on the level of trehalose and glycogen. and on the induction of
thermotolerance in Neurospora crossa. FEBS Lett.. 283: 19-22. 1991.
14. Eze, L. C. Plasma trehalase activity and diabetes mellitus. Biochem. Genet.,
27:487-495, 1989.
15. Wacker. J. R., James, P., Maffia. M., Seiler, P., Galand. G., von Kieeckebusch, A. Semenza, G., and Mantei, N. Rabbit small intestine trehalase. J.
Biol. Chem., 265; 15034-15039, 1990.
16. Yoneyma, Y., and Lever, J. Increased trehalase expression after glucose
limitation of LLC-PK, clones. Am. J. Physiol., 255: C816-C821, 1988.
17. Chung, Y. S., Song, I. S.. Erickson. R. H., Sleisenger, M. H., and Kim. Y.
S. Effect of growth and sodium butyrate on brush-border membrane-associ
ated hydrolases in human colorectal cancer cell lines. Cancer Res., 45: 29762982, 1985.
18. Heerdt, B. G., Halsey, H. K., Lipkin, M., and Augenlicht. L. H. Expression
of mitochondrial cytochrome i oxidase in human colonie cell differentiation,
transformation, and risk for colonie cancer. Cancer Res., 50: 1596-1660,
1990.
19. Heerdt, B. G., and Augenlicht, L. H. Effects of fatty acids on expression of
genes encoding subunits of cytochrome c oxidase and cytochrome c oxidase
activity in HT-29 human colonie adenocarcinoma cells. J. Biol. Chem.. 266:
19120-19126.1991.
20. Denis-Pouxviel. C.. Riesinger, I. Biilher. C.. Brdiczka, D.. and Mural, J. C.
Regulation of mitochondrial hexokinase in cultured HT29 human cancer
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
cells: an ultrastructural and biochemical study. Him him. Biophys. Acta, 972:
335-348, 1987.
Medica-Napolitano. J. S., Steele. G. D., Jr.. and Chen, L. B. Abberant
mitochondria in two human colon carcinoma cell lines. Cancer Res., 49:
3369-3373, 1989.
Yamamoto, A., Horari, S.. and Yasuhito. Y. Increased level of mitochondrial
gene expression in polyps of familial polyposis coli patients. Biochem.
Biophys. Res. Commun., 159: 1100-1106, 1989.
Cui, W., Talbot, I. C., and Northover, J. M. Restriction enzyme analysis of
mitochondrial DNA in colorectal tumors. Biochem. Biophys. Res. Commun.,
180: 85-90, 1991.
Thompson, S. W. Selected Histochemical and Histopathological Methods,
pp. 462-267. Springfield. IL: Charles C Thomas Publisher, 1966.
Gubler, I '.. and Hoffman. B. J. A simple and very efficient method for
generating cDNA libraries. Gene, 25: 263-269, 1983.
In. X., and Werner, D. Construction and quality of cDNA libraries started
from cytoplasmic RNA not enriched in poly(A)* RNA. Gene, 71: 157-164,
1988.
Huynh, T. V., Young, R. A., and Davis, R. W. Construction and screening
cDNA libraries in lambda gtlO and gtll. In: D. M. Glover (ed.). DNA
Cloning, A Practical Approach, Vol. 1. pp. 49-78. Oxford: IRL Press, 1985.
Feinberg, A. P., and Vogelstein, B. A technique for radiolabelling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem.,
/J2.-6-13, 1983.
Krieg, P., Amtmann, E., and Sauer, G. The simultaneous extraction of highweight DNA and of RNA from solid tumors. Anal. Biochem., 134: 288-294,
1983.
Sambrook, J., Fritsch, E. F., and Maniatis, T. Molecular Cloning, A Labo
ratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory
Press, 1989.
Devereux, J., Haeberli, P., and Smithies, O. A comprehensive set of sequence
analysis programs for the VAX. Nucleic Acids Res., 12: 387-395, 1984.
Lu, X., and Werner, D. The complete cDNA sequence of mouse elongation
factor 1 alpha (EF 1 alpha) mRNA. Nucleic Acids Res., 17:442, 1989.
Anderson, S., Bankier, A. T., Barrel!. B. G., de Bruijn, M. H. L.. Coulson,
A. R., Drouin, J., Eperon, I. C., Nierlich, D. P., Roe, B. A., Sanger, F.,
Schreier, P. H., Smith. A. J. H., Staden, R., and Young, 1. G. Sequence and
organization of the human mitochondrial genome. Nature (Lond.). 290:457465. 1981.
Chomyn, A., and Attardi. G. Mitochondrial gene products. Curr. Topics
Bioenerg.. 15: 295-329, 1987.
Butler, A. J.. Eaglton. M. J.. Wang. D.. Howell, R. L., Srauch, A. R.,
Khasgiwala, V.. and Smith, H. C. Induction of the proliferative phenotype
in differentiated myogenic cells by hypoxia. J. Biol. Chem., 266: 1825018258, 1991.
Kozutsumi, Y., Segal, M., Normington, K., Gething, M-J., and Sambrook,
J. The presence of malfolded proteins in the endoplasmic reticulum signals
the induction of glucose-regulated proteins. Nature (Lond.), 332: 462-464,
1988.
Heerdt, B., and Augenlicht. L. H. Absence of detectable deletions in the
mitochondrial genome of human colon tumors. Cancer Commun., 2: 109111, 1990.
Bandy, B., and Davison. A. J. Mitochondrial mutations may increase oxidative stress: implications for carcinogenesis and aging? Free Radical Biol.
Med., S: 523-539, 1990.
Corral, M., Paris. B.. Baffet. G., Tichonicky, L., Gugen-Guillouzo, C., Kruh,
J., and Defer, N. Increased level of the mitochondrial ND5 transcript in
chemically induced rat hepatomas. Exp. Cell Res., ¡84:158-166, 1989.
Torroni, A., Stepien, G., Hodge. J. A., and Wallace, D. C. Neoplaslic
transformation is associated with coordinate induction of nuclear and cyto
plasmic oxidative phosphory lation genes. J. Biol. Chem.. 265:20589-20593,
1990.
3725
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.
Differentiation of HT-29 Human Colonic Adenocarcinoma Cells
Correlates with Increased Expression of Mitochondrial RNA:
Effects of Trehalose on Cell Growth and Maturation
Xiang Lu, Teena Walker, John P. MacManus, et al.
Cancer Res 1992;52:3718-3725.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/52/13/3718
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1992 American Association for Cancer Research.