Am J Physiol Gastrointest Liver Physiol
280: G1187–G1196, 2001.
The intestine expresses pancreatic triacylglycerol
lipase: regulation by dietary lipid
JAMES T. MAHAN, GHANSHYAM D. HEDA, R. HANUMANTHA RAO,
AND CHARLES M. MANSBACH II
Department of Medicine, Division of Gastroenterology, The University of Tennessee, Memphis,
Memphis 38163; and The Veterans Affairs Medical Center, Memphis, Tennessee 38104
Received 17 April 2000; accepted in final form 3 January 2001
lipid absorption
family includes pancreatic
triglyceride (triacylglycerol) lipase (PTL), hepatic
lipase, lipoprotein lipase (13), preduodenal (gastric and
lingual) lipases (41), and two pancreatic lipase-related
proteins (28). Although each of these lipase family
members may play a role in fat metabolism, their
expression varies both developmentally and by tissue.
The preduodenal and pancreatic lipases are known to
hydrolyze dietary fats in the lumen of the gastrointestinal tract of adult mammals. Their combined action
produces monoacylglycerols (MAGs) and fatty acids
that are readily absorbed by enterocytes lining the
small intestine. In addition to lipases in the lumen,
there is also evidence of a lipase in mammalian enterocytes (9, 38). The exact identity and origin of this
intracellular form is unknown.
Our observations indicate that there are two distinct
lipase, or lipase-like, enzymes in the rat small intestine
(31, 37). One is maximally active under acidic conditions and the other under basic conditions. Theoretically, a number of known enzymes could explain this
lipolytic activity. The possibilities include acid lipase,
cholesterol esterase, microsomal triacylglycerol hydrolase, and PTL. For the following reasons, we considered PTL as the most likely candidate enzyme that
could provide lipolytic activity in the intestine. 1) Lipid
hydrolytic activity was expressed in the absence of bile
salts, ruling out significant contribution by cholesterol
esterase (bile salt-stimulated lipase; see Refs. 5 and
21). 2) Because of the pH profile at which the majority
of the lipolytic activity occurred (pH 8.0–8.5; see Ref.
37), acid lipase was ruled out (pH optimum 5.6; see Ref.
30). 3) Microsomal triacylglycerol hydrolase (7) was
improbable because of the cytosolic location of mucosal
lipolytic activity (31). Because many of the characteristics of the observed lipase in the intestine were consistent with that of PTL, we sought to determine if PTL
is synthesized in the intestine and, if so, if it is identical at the molecular level to the PTL made in the
pancreas. Furthermore, because many intestinal enzymes are regulated by their substrates, we also investigated whether the mucosal lipase is responsive to
acute or chronic fat loading.
THE MAMMALIAN LIPASE GENE
Address for reprint requests and other correspondence: C. M. Mansbach II, The Univ. of Tennessee, Memphis, 951 Court Ave., Rm. 555
Dobbs, Memphis, TN 38163 (E-mail: [email protected]).
http://www.ajpgi.org
MATERIALS AND METHODS
Animals and collection of mucosa. Male Sprague-Dawley
rats weighing ⬃230–300 g were maintained on laboratory
chow containing 4% fat (wt/wt) before experiments. To collect
mucosa for RNA isolation or enzyme assay, the small intestine was first divided into quarters. Each quarter was rinsed
two times with cold PBS, split lengthwise, and spread, mucosa up, on a chilled glass plate. The mucosa was collected by
scraping with a glass slide.
In experiments to divert pancreatic fluid from reaching the
intestine, the pancreatic duct was cannulated with a PE-10
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
G1187
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
Mahan, James T., Ghanshyam D. Heda, R. Hanumantha Rao, and Charles M. Mansbach II. The intestine
expresses pancreatic triacylglycerol lipase: regulation by dietary lipid. Am J Physiol Gastrointest Liver Physiol 280:
G1187–G1196, 2001.—We identified the enzyme responsible
for alkaline lipolysis in mucosa of rat small intestine. RTPCR was used to amplify a transcript that, by cloning and
sequencing, is identical to pancreatic triacylglycerol lipase.
In rats fed normal laboratory chow, pancreatic triacylglycerol
lipase mRNA was detected in all four quarters of the small
intestine, with the first quarter expressing about three times
as much of this transcript as was found in the more distal
three-quarters combined. Both acutely and chronically administered dietary fat were shown to regulate pancreatic
triacylglycerol lipase mRNA expression and lipase activity.
The synthesis of pancreatic triacylglycerol lipase protein by
the small intestine was demonstrated by in vivo radiolabeling experiments using [35S]methionine/cysteine followed by
immunoprecipitation with an anti-pancreatic triacylglycerol
lipase antibody. Immunohistochemical studies suggest that
pancreatic triacylglycerol lipase protein expression is restricted to enterocytes throughout the small intestine. To our
knowledge, this is the first report identifying rat small intestinal mucosa as a site of pancreatic triacylglycerol lipase
synthesis and the first demonstration of its modulation in the
mucosa by dietary fat. We propose that pancreatic triacylglycerol lipase is used by the intestine to hydrolyze the
mucosal triacylglycerol that is not transported in chylomicrons.
G1188
INTESTINAL TRIACYLGLYCEROL LIPASE
were determined for all samples on the basis of their ␤-actin
mRNA levels. This was accomplished by producing ␤-actin
cDNA from the sample mRNAs using Promega’s Access RTPCR kit with rat ␤-actin intron-spanning primers (Clontech
Laboratories, Palo Alto, CA). The RT step (AMV-RT) was
performed at 48°C for 45 min. Next, a two-step PCR program
was used to amplify the 764-bp ␤-actin sequence. The reactions were first incubated at 94°C for 2 min followed by 17
amplification cycles by using the following two-step program:
94°C for 30 s and 67°C for 90 s. This was followed by an
additional elongation step of 8 min at 70°C. Finally, the
␤-actin PCR product was electrophoresed and quantified (as
described separately later). These experiments were performed in triplicate, and their mean values were used to
determine equivalent amounts of sample mRNAs.
Using the iPTL gene-specific internal primers, iPTL/607
(5⬘-ACGCGGCTGAACCTTACTTCC-3⬘) and iPTL/831 (5⬘-GTCGCGAGTTCCTTCCCAGAT-3⬘), with a Sigma (St. Louis,
MO) avian RT-PCR kit, we performed RT-PCR with equivalent amounts of mRNA as template. In these experiments, a
Sigma enhanced avian myeloblastosis virus RT (AMV-RT)
was used for reverse transcription at 60°C for 35 min. The
reactions were then incubated at 94°C for 2 min followed by
34 amplification cycles using the following PCR program:
94°C for 30 s, 63°C for 30 s, and 69°C for 30 s. The last cycle
was followed by an 8-min additional elongation step at 70°C.
The 245-bp PCR product was electrophoresed and quantified
(as described separately later). These experiments were performed in triplicate, and the mean values were used for
comparison of samples.
A series of preliminary experiments established that, at 17
and 34 PCR cycles for actin and iPTL, respectively, the
amount of RT-PCR product was linearly related to the
amount of starting mRNA (Fig. 1). Control reactions were
routinely conducted in which the RT step was omitted to test
for genomic DNA amplification. Additional RT-PCR controls
included omitting mRNA as another test for contamination
(Fig. 2). To ensure the specificity of the iPTL 607/831 primer
pair, the amplified product (245 bp) was cloned and sequenced. The iPTL 607/831 product was identical to the
nucleotide sequence pPTL (M58369).
Quantification of specific RT-PCR products. RT-PCR products were electrophoresed in 2% agarose gels in 1⫻ Trisborate-EDTA. The gels were soaked in SYBR Green I (Molecular Probes, Eugene OR) at 1:10,000 dilution for 3 h, and
the intensity of the bands was determined by scanning laser
densitometry (STORM 860 and ImageQuant software; Molecular Dynamics, Sunnyvale CA).
Affinity purification of antibody to pancreatic lipase. An
IgG fraction of sheep anti-human pPTL (Biodesign International, Kennebunk, ME) was affinity purified on a column
consisting of pig pPTL (Sigma) coupled to AminoLink Plus
agarose beads (Pierce Chemical, Rockford, IL). By Western
blotting, this anti-PTL antibody preparation identifies a single band [relative molecular mass (Mr) ⫽ 49 kDa] in extracts
of rat small intestine or pancreas (data not shown).
Immunodepletion of lipolytic activity from intestinal cytosol. Intestinal cytosol was prepared (19) and treated with
affinity-purified anti-pPTL antibody or an equivalent amount
of sheep IgG. Cytosol (39 ␮g protein) was incubated with
either 10 or 500 ␮g of lipase antibody (or control IgG) overnight at 4°C and centrifuged at 12,000 g for 15 min. The
supernatant was used for the lipase assay, as previously
described (31).
Synthesis of iPTL by the small intestinal mucosa. The first
quarter of the small intestine in an anesthetized rat was
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
cannula (Clay Adams, Parsippany, NJ). The cannula was
introduced in the duodenal lumen through a stab wound in
the duodenal wall made by an 18-gauge needle, threaded into
the Ampulla of Vater for 2 mm, and tied in place using silk
sutures. The common bile duct was cannulated with another
PE-10 cannula and secured with silk ties. Both cannulas
exited the rat through the right subcostal incision used to
expose the duodenum and common bile duct. The pancreatic
duct cannula drained clear fluid, and the bile duct cannula
drained yellow fluid. A duodenal feeding cannula was also
placed (PE-50) through which the rat received Vivonex as a
protein and calorie source for 48 h. At the conclusion of the
infusion period, the rats had only 6% of the tryptic activity in
their intestinal lumens compared with noncannulated control rats.
RNA isolation. Mucosa from each quarter of the small
intestine was homogenized in guanidine thiocyanate buffer,
and total RNA was isolated by CsCl density gradient ultracentrifugation followed by phenol-chloroform-isoamyl alcohol
(25:24:1) and chloroform-isoamyl alcohol (24:1) extractions
(39). RNA quality and concentration were determined by
electrophoresis on agarose gels with ethidium bromide staining. Only samples with intact 28S and 18S RNA were used.
mRNA was isolated from each total RNA using the Promega
(Madison, WI) PolyATtract system. All mRNA samples were
suspended in 100 ␮l of sterile nuclease-free water containing
75 units recombinant RNasin RNase inhibitor (Promega) and
stored at ⫺80°C until use.
Identification of PTL mRNA in small intestinal mucosa.
First-strand cDNA was synthesized from small intestinal
mRNA using RT (AMV-RT from Promega) and oligo(dT)
primers. Gene-specific PTL primers, 5⬘-ATGCTGATGCTGTGGACATTTGCA-3⬘ and 5⬘-CTAACATGCAGACAGTGTAAGCAGGAC-3⬘ (from GenBank, accession no. M58369), were
used to amplify the cDNA. The amplified DNA was ligated
into the pGEM-Easy vector (Promega), which was then used
to transform DH5␣ Escherichia coli. Plasmid DNA was purified from selected colonies using QIAprep and QIAtip columns (Qiagen, Valencia, CA). The plasmid insert was initially sequenced in both directions by using universal vector
primers [M13(⫺21) forward, M13 reverse, T7, and SP6]. This
generated partial sequences that were extended by using a
series of internal primers to produce contiguous sequences
covering the entire coding region. All sequencing was performed by the dideoxynucleotide dye terminator method. A
consensus cDNA sequence, identified using DNasis software
(DNasis 2.0; Hitachi Software, San Bruno CA), was found to
be 99.6% identical to that reported for pancreatic PTL (pPTL)
cDNA (GenBank accession no. M58369). Alignment of this
consensus sequence [referred to in this report as intestinal
PTL (iPTL)] with the published sequence for PTL from rat
pancreas (referred to in this report as pPTL) was also accomplished using the DNasis 2.0 program.
Transfection of COS cells with iPTL transcript. The complete coding region of rat iPTL was subcloned downstream of
the cytomegalovirus promoter into the eukaryotic expression
vector pCR 3.1-Uni (Invitrogen, San Diego, CA). COS-7 cells
were transfected using a lipofectamine procedure (Life Technologies, Gaithersburg, MD) with either the vector containing the iPTL insert or the vector alone. Expression was
analyzed by measuring lipase activity in the medium and
cells at 72 h using a sensitive radiometric assay that has been
described in detail previously (31).
Semiquantitative comparison of iPTL mRNA levels by RTPCR. Relative quantitation of iPTL mRNA levels was
achieved by RT-PCR. First, equivalent amounts of mRNA
INTESTINAL TRIACYLGLYCEROL LIPASE
flushed with a bolus of saline to clear the lumen of partially
digested food and secretions. Ligatures were placed around
the small intestine at 2 and 12 cm below the entrance of the
common bile duct to prevent pancreatic secretions from
reaching the intestine. Next, 5.5 ␮Ci of [35S]methionine/
cysteine were introduced into the isolated lumen. One hour
later, the animal was killed, and the mucosa was collected
and homogenized at 4°C in HEPES buffer containing protease inhibitors. The radiolabeled extract was precleared
with protein G beads (Sigma) to reduce the concentration of
proteins that might bind nonspecifically. Next, an aliquot of
affinity-purified sheep anti-human pPTL IgG and fresh protein G beads was added to the precleared extract. After 1 h of
incubation at 4°C the beads were collected by centrifugation
and washed extensively with 1 M NaCl. The beads were
boiled for 5 min in reducing SDS-PAGE sample buffer, and
the resulting supernatant containing the immunoprecipitated radiolabeled protein was separated on an 8.5% SDSPAGE. The gel was dried and placed against Kodak X-Omat
film until an image developed.
iPTL immunohistochemistry. Segments of small intestine
and pancreas were fixed in 10% buffered formalin and routinely processed in paraffin. Sections (8 ␮m) were cut and
collected on poly-L-lysine-coated glass slides. The primary
antibody, affinity-purified sheep anti-human pPTL IgG, was
used at a final concentration of 12 ␮g/ml. IgG from a nonimmunized sheep was used at 33 ␮g/ml as a negative control.
The secondary antibody was a monoclonal anti-sheep IgG
peroxidase conjugate (Sigma) diluted 1:1,000. Diaminobenzidine-urea-hydrogen peroxide (Sigma) was used as the substrate for color development.
Data analysis. Contiguous nucleotide sequences were
aligned and analyzed using DNasis 2.0. All primers were
designed with Oligo 5.0 software (National Biosciences, Plymouth, MN). Quantitative data are expressed as means ⫾
SE. Variation among group means was determined by oneway ANOVA and P values by the Student-Newman-Keuls
test (GraphPad InStat Software, San Diego, CA).
Fig. 2. Evidence that the 245-bp band produced by RT-PCR using iPTL 607/831 primers and mRNA from the
fourth quarter of the small intestine is not due to amplification of genomic DNA or a pancreatic contaminant. The
mRNA samples in this experiment were obtained from rats in which only the lower part of the abdomen was opened
for removal of the distal-most segment of the small intestine. This surgical approach was used to preclude any
chance of contamination with pancreatic tissue. Next, RT-PCR was performed using the iPTL primers both with
and without RT. The cycle number was increased to 45 to maximize the chance of detecting a faint band upon
electrophoresis. Lane on right contains amplified RT-PCR product from pancreatic duct-diverted rats (see
MATERIALS AND METHODS) using the same primers as in gel. The agarose gel was stained with ethidium bromide and
photographed under UV illumination. Lanes 1, 4, and 8, DNA ladder with the indicated fragment sizes; lanes 2 and
3, mRNA from the fourth quarter of rat 1 (lane 2, ⫺RT and lane 3, ⫹RT); lanes 5 and 6, mRNA from the fourth
quarter of rat 2 (lane 5, ⫺RT and lane 6, ⫹RT); lane 7, no RNA added, ⫹RT.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
Fig. 1. Determination of optimal number of PCR cycles for use in
semiquantitative RT-PCR experiments. In these preliminary experiments, a 10-␮l aliquot was removed from each PCR tube at the
completion of the indicated number of cycles. These samples were
electrophoresed on agarose gels, stained with SYBR Green I, and
quantitated as described in MATERIALS AND METHODS. Based on these
curves, we decided to use 17 cycles for ␤-actin quantitation and 34 for
intestinal pancreatic triglyceride lipase (iPTL) quantitation in all
subsequent experiments. A: actin; B: iPTL.
G1189
G1190
INTESTINAL TRIACYLGLYCEROL LIPASE
Fig. 3. Alignment of discrepant region of the rat intestinal cDNA sequence that we derived with the corresponding
region of the published sequences for pancreatic triglyceride lipase (PTL) in rat pancreas and brain. Nonidentical
nucleotides are enclosed in box. NCBI GenBank accession numbers are in parentheses.
RESULTS
Fig. 4. Predicted amino acid sequence of iPTL and comparison with other members of the lipase gene family.
Differences between pancreas PTL (pPTL; M58369) and iPTL (and the residues at the same positions in the other
sequences) are enclosed in boxes. The partial sequences shown here for comparison with rat iPTL are from the
NCBI GenBank databases (protein or nucleotide). Accession numbers are in parentheses. Rat PTL sequences from
3 different tissues are in italics.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
We used RT-PCR to identify and clone a PTL transcript from the intestinal mucosa. On sequencing, the
PTL cDNA from rat intestine (iPTL) was found to be
99.6% identical to pPTL (GenBank accession no.
M58369). The few nucleotide differences that exist are
the following: 1) iPTL has a C at position 473 rather
than the G reported for pPTL and 2) CT at positions
501–502 and 503–504 in iPTL rather than TC at these
two positions in pPTL (Fig. 3). We sequenced through
this region of iPTL cDNA eight times (four in each
direction), and each time a unique internal primer was
used to be certain that the sequence generated was
correct.
At the amino acid level, iPTL and pPTL are 99.4%
identical. The predicted amino acid sequence for iPTL
differs from pPTL (M58369) as follows: 1) iPTL would
have an F rather than L at position 148 and 2) YS
rather than HP at positions 158–159 (Fig. 4). To determine whether the sequence we found coded for an
enzymatically active protein, we transfected COS-7
cells with the full-length iPTL coding sequence and
later assayed the medium and cells for lipase activity.
At 72 h posttransfection, the iPTL-transfected cells
had produced considerable lipase activity (2.19 ␮mol
oleic acid released 䡠 min⫺1 䡠 well⫺1). The amount of
lipase activity secreted in the medium was nearly 17
times as much as was retained in the cells. In the
mock-transfected cells (vector alone), the total amount
of lipase activity (intracellular and secreted) was neg-
ligible (⬍0.03 ␮mol oleic acid released 䡠 min⫺1 䡠 well⫺1).
These data show that the iPTL sequence we found is
enzymatically active even though it differs slightly
from the sequence reported for pPTL (M58369).
We then examined iPTL mRNA expression along the
length of the small intestine by semiquantitative RTPCR using a pair of primers (see MATERIALS AND METHODS) that produce a 245-bp product from approximately
the middle of the full iPTL transcript. Using this approach, we found that iPTL mRNA was present in the
mucosa of all four quarters of the small intestine. In
rats fed normal laboratory chow, the first quarter contained ⬃77% of the total amount of iPTL mRNA
present in the small intestine. The remaining amount
was distributed over the distal three-quarters in decreasing abundance, from ⬃12% for the second quarter
down to 5% for the fourth quarter (see chow-fed values
in Fig. 7). These data demonstrate that transcription of
iPTL occurs throughout the small intestine.
The finding of iPTL transcripts throughout the intestine suggests that we were not amplifying a contaminant from the pancreas. To support this contention,
we obtained mucosa from 48-h pancreatic duct-diverted rats and repeated the RT-PCR for each of the
four quarters of the intestine. On agarose gel electrophoresis, the amplified product migrated at the expected size for iPTL, 245 bp (shown in Fig. 2, right). In
summary, the data strongly suggest that a pancreatic
contaminant was not the origin of the PTL transcript
that we amplified from the intestine.
INTESTINAL TRIACYLGLYCEROL LIPASE
Fig. 5. Evidence for iPTL protein synthesis by the mucosa of the first
quarter of the small intestine. Mucosal proteins were metabolically
labeled in vivo by [35S]methionine/cysteine incorporation in an isolated bowel segment. A mucosal extract containing the newly synthesized radiolabeled proteins was immunoprecipitated with affinity-purified anti-PTL antibodies. The precipitated protein was
separated by SDS-PAGE, and autoradiography was performed. Lane
1 contains the 35S-labeled whole extract of mucosa. Lane 2 contains
the protein immunoprecipitated from the whole extract by our antiPTL antibody. Arrowhead indicates a relative molecular mass (Mr) of
49 kDa, the size expected for PTL.
clear whether this PTL is of pancreatic or intestinal
origin.
Finally, we investigated the possibility that iPTL
might be regulated by the amount of fat entering the
small intestine. In these experiments, we used semiquantitative RT-PCR to compare the iPTL mRNA levels in the following three groups of rats: 1) rats maintained on normal chow with a fat content of 4% (wt/wt),
2) rats acutely loaded with fat by corn oil gavage, and
3) rats chronically (4 days) fed a high-fat diet (20%
wt/wt). We considered the levels present in the chowfed animals to be the baseline values for comparisons.
Our results for acute loading indicate that iPTL
mRNA levels in the first quarter are significantly (P ⬍
0.01) decreased 2.5 h after fat administration by gavage (see gavage values in Fig. 7). Means for the distal
three-quarters were not different (P ⬎ 0.05) in the
fat-gavaged rats compared with chow-fed animals. Recovery of iPTL mRNA levels from the observed decrease that followed the acute 2.5-h fat loading was
studied at 3.5 and 17 h after the remaining fat had
been removed from the stomach. We found significantly (P ⬍ 0.05) elevated levels of iPTL mRNA in all
four quarters after 3.5 h of recovery compared with the
same quarter of either chow-fed or fat-gavaged animals
(Fig. 7). After 17 h of recovery, the iPTL levels in all
quarters had returned to the baseline values exhibited
by the chow-fed animals (Fig. 7).
We have also studied the effect of acute fat loading
on lipolytic activity and its relationship to iPTL mRNA
levels. To more clearly interpret the lipolytic data, it
was first necessary to determine the percentage of total
lipolytic activity in intestinal cytosol that was due to
iPTL. Affinity-purified pPTL antibody (10 ␮g) reduced
cytosol lipolytic activity by 42%, and 500 ␮g of antibody
reduced activity by 86%, suggesting that most of the
lipolytic activity expressed by the intestine was because of iPTL. The data describing the relationship
between iPTL activity and mRNA levels (Fig. 8) show
that at the early recovery point (3.5 h) lipase enzymatic
activity is extremely low, a finding that might be expected since the iPTL mRNA level was found to be
decreased 3.5 h earlier. At 17 h postgavage recovery,
the lipolytic activity had completely recovered to the
baseline values of chow-fed rats (Fig. 8), presumably as
a result of the dramatic increase in iPTL mRNA seen
earlier at the 3.5-h recovery point.
Because we have shown previously that acute and
chronic fat feeding produces different results with respect to complex lipid synthesis (23, 24), we studied
rats fed a high-fat diet for 4 days. We found that iPTL
mRNA transcript abundance was not affected by
chronic fat feeding in the first quarter of the intestine
but that the remaining three more distal quarters
showed marked increases (Fig. 9), in concert with our
previous findings in the hamster (23). It should be
noted that, despite the general increase in PTL transcripts, the proximal-to-distal gradient of PTL mRNA
was maintained. As shown in Fig. 10, not only was the
iPTL mRNA level increased in response to chronic fat
feeding, but lipolytic activity was increased as well.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
To establish whether the iPTL transcript was actually translated in vivo by the intestine, we conducted
metabolic radiolabeling experiments using [35S]methionine/cysteine in an isolated bowel loop. To concentrate
and identify the protein(s) of interest that was radiolabeled, we immunoprecipitated the iPTL from total
mucosal proteins using an affinity-purified anti-PTL
antibody and identified a protein of Mr 49 kDa, which
is appropriate for PTL. The results of these experiments are shown in Fig. 5. We conclude that the intestine synthesizes iPTL in situ.
In other experiments, we performed immunohistochemistry to examine the distribution of iPTL protein
along the length of the small intestine. These results
(Fig. 6) suggest that iPTL immunoreactivity is restricted to enterocytes. The proportion of enterocytes
that were iPTL positive was highest in the first quarter
of the intestine. The distal three-quarters had considerably fewer enterocytes that stained for iPTL. Thus
the distribution of iPTL protein as detected by immunohistochemistry is consistent with the distribution of
iPTL mRNA as determined by RT-PCR. No iPTL staining was noted in the lamina propria or smooth muscle
of the intestine. Also, the absence of iPTL reactivity in
the intercellular spaces between enterocytes and in the
basal lamina that underlies the enterocytes suggests
that iPTL is not secreted basolaterally. Occasionally, a
small amount of additional staining was present at the
apical membrane of villus tip enterocytes, but it is not
G1191
G1192
INTESTINAL TRIACYLGLYCEROL LIPASE
This would be the expected result if the mRNA iPTL
transcripts are translated into protein in the small
intestine.
DISCUSSION
These experiments were undertaken to more clearly
define the origin of intestinal lipolytic activity originally found by DiNella et al. (9) and confirmed in our
laboratory (29–31, 37). Although we discovered both
acid and alkaline lipolytic activities to be present in the
intestine (29, 31), it was the alkaline active one that
appeared to be physiologically important (37). The
finding of no diminution in alkaline lipase activity in
mucosal cytosol after tying off the pancreatic duct for
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
Fig. 6. Immunohistochemical localization of PTL in the
following 4 quarters of the small intestine: first quarter
(a), second quarter (b), third quarter (c), and fourth
quarter (d). Intense PTL immunoreactivity (brown reaction product) is present in villus enterocytes and
cryptal cells. The percentage of villus enterocytes that
are PTL positive is highest in the first quarter. The
reaction product visible at the villus tips in the third
and fourth quarters appears to be on the luminal surface of the cells and is therefore presumed to be of
pancreatic origin. a-d: Magnification ⫻375. Adjacent
sections treated with control IgG were completely negative in all four quarters. e: higher magnification
(⫻4,000) of midvillus enterocytes from the first quarter;
f: section from the first quarter treated with control IgG
instead of anti-PTL IgG (⫻1,000); g: section from the
first quarter of the intestine using anti-PTL IgG preabsorbed with porcine pPTL (see MATERIALS AND METHODS;
magnification ⫻2,500).
48 h suggested that the activity originated in the intestine (37). The ability to clearly differentiate the
lipase described by our laboratory from lipase of pancreatic origin (pPTL) is not a trivial point, since the
activities of cholesterol esterase (12), amylase (2), and
RNase (1) have all been shown to be present in the
intestinal mucosa, but the majority of each protein was
found to originate in the pancreas. In this report, using
a molecular approach, we show that the intestine does
indeed synthesize lipase locally, that this lipase is
almost identical to pPTL, and that it is responsive to
lipid feeding.
RT-PCR studies clearly identified a transcript in the
mucosa that is identical to that reported for PTL cDNA
INTESTINAL TRIACYLGLYCEROL LIPASE
Fig. 9. Effect of chronic fat feeding on iPTL mRNA levels in the small
intestine. The relative amount of iPTL mRNA was determined by
RT-PCR for each quarter of the small intestine. The fat-fed group
was maintained on a mixture of their normal chow and 20% corn oil
(wt/wt) for 4 days. Each bar represents the mean and SE (n ⫽ 3).
*Significant difference (P ⬍ 0.05) between chow-fed and fat-fed
groups for the distal three quarters; the first quarters are not
different in the 2 treatment groups.
from the pancreas. Because the pancreas expresses far
more PTL than does the intestine, we were concerned
that we might have amplified a pancreatic contaminant. To lessen this potential, we amplified a transcript of appropriate size for PTL from the distal quarter of the intestine, which was removed without
disturbing the pancreas (described in the legend to Fig.
2). From the same animals, we also isolated skeletal
muscle from the incision used to obtain the segment of
intestine. In each instance with skeletal muscle, no
PTL message was found. In a second approach to this
point, we were unable to find in the intestine the
pancreas-specific transcription factor Nkx6.1 (17, 22,
34). Third, we were able to show directly that the
intestine can synthesize a protein that is immunoidentifiable as PTL and that is of the appropriate Mr. This
establishes that the PTL transcript is actively trans-
lated by the intestine and provides a rationale for the
finding of lipolytic activity expressed by the intestine.
Indeed, our immunodepletion experiments using purified anti-PTL antibody established that most of the
intestinal lipolytic activity was a result of iPTL. Last,
we diverted pancreatic juice from the intestine for 48 h
and found no diminution in PTL activity in the intestine; we continued to be able to isolate iPTL from the
intestine. In summary, the data show that our amplified lipase transcript was most likely of intestinal and
not pancreatic origin. Our data confirm and extend
those of others who have found PTL outside the pancreas, including extrahepatic peribiliary gland epithelium (36), rat brain (38), ground squirrel heart (3), and
Caco-2 cells, a colonocyte cell line (35).
Because the mucosa is comprised of multiple cell
types, we employed immunohistochemistry to show
that immunoidentifiable lipase was present specifically
in the enterocytes throughout the small intestine. As
shown by this technique, more PTL was present in the
proximal quarter of the intestine compared with the
more distal gut. These data are consistent with the
lipolytic activity and the quantity of mRNA iPTL tran-
Fig. 8. Effects of acute fat loading on the levels of iPTL mRNA and
lipolytic activity in the small intestine. Small intestine is considered
as a whole rather than by quarters for the sake of simplicity. The
mRNA values are derived from the data presented in Fig. 7. Lipase
activity was determined by a sensitive radiometric assay (see MATERIALS AND METHODS). The treatment groups are the same as described
in Fig. 7. **P ⬍ 0.01, 3.5-h recovery value is significantly different
from the other treatment groups for both mRNA and lipase activity.
Fig. 10. Effects of chronic fat feeding on the levels of iPTL mRNA
and lipolytic activity in the small intestine. Small intestine is considered as a whole rather than by quarters for the sake of simplicity.
The mRNA values are derived from the data presented in Fig. 9.
Lipase activity was determined by a sensitive radiometric assay (see
MATERIALS AND METHODS). The treatment groups are the same as
described in Fig. 9. *Significant difference (P ⬍ 0.05) between chowfed and fat-fed groups for both mRNA and lipase activity. FFA, free
fatty acid.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
Fig. 7. Effect of acute fat loading on the amount of iPTL mRNA in
the small intestine. The relative amount of iPTL mRNA was determined by RT-PCR for each quarter of the small intestine. Chow-fed
rats were used to establish baseline control values for comparison
purposes. The remaining 3 groups all received identical 2-ml corn oil
gavages followed by a 2.5-h absorption period. iPTL mRNA levels
were determined either immediately after the absorption period
(gavage group) or after a recovery phase of 3.5 h (Gav ⫹ 3.5 h) or 17 h
(Gav ⫹ 17 h). Each point represents the mean and SE (n ⫽ 3). **First
quarter in the gavage group is significantly (P ⬍ 0.05) different from
the first quarter in each of the other 3 treatment groups. *Second,
third, and fourth quarters in the gavage group are all different (P ⬍
0.05) from the comparable quarters in the group that recovered for
3.5 h but are not different from the chow-fed controls or from the
animals recovered for 17 h.
G1193
G1194
INTESTINAL TRIACYLGLYCEROL LIPASE
of fat quickly and efficiently (27), and the rats quickly
recovered to control levels of iPTL mRNA and lipolytic
activity, making this speculate unlikely.
Chronically (4 days) feeding a high-lipid load increased both steady-state iPTL mRNA levels and lipolytic activity. These experiments parallel our previous
data, which show that the chronic feeding of a high-fat
diet increases the activity of diacylglycerol acyltransferase and lysophosphatidylcholine acyltransferase in
hamsters (23). Furthermore, other studies have shown
an effect of chronic fat feeding on the pancreas in which
secretin was the proposed hormone associated with an
increase in pPTL secretion (32). We interpret the coordinate upward regulation of iPTL mRNA and lipase
activity along the whole length of the bowel to suggest
that iPTL plays an important physiological role in
intestinal lipid metabolism.
Two possibilities exist to explain our findings with
regard to chronic fat feeding. The first involves the
peptide tyrosine tyrosine (PYY), which is released from
the distal small bowel when lipid reaches the ileal
lumen (4, 20). An elevated PYY level in the circulation
may increase iPTL expression. PYY has already been
shown to increase both intestinal fatty acid binding
protein mRNA (14) and apolipoprotein A-IV (18). Alternatively, free fatty acids may enter the nucleus,
activate transcription factors, and thus modulate iPTL
gene expression (40).
Two actions have been suggested for iPTL in the
intestinal mucosa. The first was described by Tsujita et
al. (38), who proposed that the reverse, i.e., acylation,
activity of PTL is expressed in the intestine. In these
studies, MAG was shown to be acylated by an enzyme
that they called MAG acyltransferase, even though it
reacted with antibodies directed against PTL. However, we predict that, under physiological conditions,
hydrolytic activity would predominate, since this is the
expected reaction of PTL at intracellular pH (6). In this
regard, a potential physiological role for the hydrolytic
activity of iPTL was recently reported from our laboratory (25). In that study, we showed that endogenous
esterified acyl groups like TAG do not readily cross the
endoplasmic reticulum (ER) membrane but remain on
its cytosolic face. Although we showed that this triacylglycerol (TAG) could be rapidly hydrolyzed, the hydrolytic activity was provided by exogenous lipase
added in vitro. However, we have previously shown
that active lipolytic activity exists in mucosal homogenates (37), suggesting that lipase is active under in
vivo conditions. As shown in the present study, most of
the lipolytic activity in the cytosol is iPTL. We therefore propose that a function of iPTL is to release fatty
acids from endogenous acyl groups that do not translocate across the ER membrane. In this manner, the
free fatty acids may exit the enterocyte by the basolateral surface to be transported to the liver in the portal
vein (26).
There is no evidence that intestinally expressed PTL
is secreted in the intestinal lumen by the enterocytes in
amounts that are physiologically meaningful. In pancreatic duct-ligated rats, [3H]glyceryltrioleate, infused
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
scripts, which also demonstrated a similar proximalto-distal gradient. These parallel data all support the
synthesis of iPTL by the enterocytes of the small intestine.
The DNA sequence that we derived for iPTL differed
in only minor respects in two areas from that published
as M58369 in the GenBank database. The first deviation is our finding of a nucleotide sequence coding for
phenylalanine (F) at amino acid residue 148 instead of
leucine (L). Phenylalanine is present at this position in
PTL from pig, human, nutria, and ground squirrel,
whereas PTL from rat pancreas and brain and pancreas from rabbit, guinea pig, and horse have leucine at
the same position (see Fig. 4). The second difference is
at positions 158 and 159, where we predict tyrosineserine (YS) rather than histidine-proline (HP). A comparison of all of the known PTL sequences (14 are
known; see Fig. 4) reveals that the YS motif is highly
conserved across species. Even the two PTL-related
proteins (PLRP-1 and PLRP-2) in rat have the YS motif
at this position. Only two do not have the YS motif: rat
pancreas which has HP and guinea pig pancreas which
has YP.
The differences between iPTL as reported here and
rat pancreas and brain could simply be due to sequencing errors. For example, a misreading of the nucleotides CT as TC at two places (nucleotides 500–501 and
503–504) would result in the prediction of amino acids
HP rather than YS (amino acids 158 and 159 in Fig. 4).
On the other hand, the three distinct nucleotide
sequences may reflect PTL transcript isoforms found in
vivo. This possibility could occur if only one of the
transcripts is encoded by genomic DNA and the other
two by posttranscriptional events such as RNA editing.
Further characterization of the rat PTL gene is needed
to distinguish between all of the possibilities raised. In
any case, these minor changes are not likely to be of
physiological importance, since they are not part of the
putative lipid recognition site (G-X-S-X-G) centered
around the active-site serine 169 or the lid domain at
amino acid residues 238–262 (42). Our transfection
results indicate that the minor differences in iPTL
mRNA sequence discussed above are not sufficient to
block the enzymatic activity of this lipase.
Another interesting aspect of this study is the regulation of lipase activity and mRNA by dietary fat. The
iPTL response to acute lipid loading was dramatically
different from chronic feeding. When oil was given by
gavage, lipase activity level fell acutely and then returned to pregavage levels 17 h after oil remaining in
the stomach had been removed. mRNA levels followed
a similar pattern, with mRNA transcripts increasing
before the increase in lipase activity, as would be
expected. It is not clear why iPTL levels decreased in
response to the acute lipid load. However, these data
are consistent with our previous finding of a reduced
lipolytic activity in the proximal intestine in response
to lipid infusion (31). One potential explanation for our
observation is toxic injury from the fatty acids released
on hydrolysis by pPTL in the lumen of the intestine.
However, the rat intestine can process large amounts
INTESTINAL TRIACYLGLYCEROL LIPASE
NOTE ADDED IN PROOF
We have sequenced rat pPTL cDNA and find that positions
500 –504 are CTACT, the same as we report for intestinal
iPTL and as has been reported for rat brain PTL (Gen Bank
no. D88534).
This study was supported by The Office of Research and Development Medical Research Service, the Department of Veterans Affairs,
and National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-38760.
REFERENCES
1. Alpers DH and Isselbacher KJ. Protein synthesis by rat
intestinal mucosa. J Biol Chem 242: 5617–5622, 1967.
2. Alpers DH and Solin M. The characterization of rat intestinal
amylase. Gastroenterology 58: 833–842, 1970.
3. Andrews M, Squire T, Bowen C, and Rollins M. Low-temperature carbon utilization is regulated by novel gene activity in
the heart of a hibernating mammal. Proc Natl Acad Sci USA 95:
8392–8397, 1998.
4. Aponte GW, Fink AS, Meyer JH, Tatemoto K, and Taylor
IL. Regional distribution and release of peptide YY with fatty
acids of different chain lengths. Am J Physiol Gastrointest Liver
Physiol 249: G745–G750, 1985.
5. Barrowman JA and Borgstrom B. Specificity of certain methods for the determination of pancreatic lipase. Gastroenterology
55: 601–607, 1968.
6. Borgström B. Influence of bile salt, pH, and time on the action
of pancreatic lipase: physiological implications. J Lipid Res 5:
522–531, 1964.
7. Coleman RA and Haynes EB. Differentiation of microsomal
from lysosomal triacylglycerol lipase activities in rat liver. Biochim Biophys Acta 751: 230–240, 1983.
8. Dartsch H, Kleene R, and Kern HF. In vitro condensationsorting of enzyme proteins isolated from rat pancreatic acinar
cells. Eur J Cell Biol 75: 211–222, 1998.
9. DiNella RR, Meng HC, and Park CR. Properties of intestinal
lipase. J Biol Chem 235: 3076–3081, 1960.
10. Fredrikzon B and Blackberg L. Lingual lipase: an important
lipase in the digestion of dietary lipids in cystic fibrosis. Pediatr
Res 14: 1387–1390, 1980.
11. Freedman SD and Scheele GA. Regulated secretory proteins
in the exocrine pancreas aggregate under conditions that mimic
the trans-Golgi network. Biochem Biophys Res Commun 197:
992–999, 1993.
12. Gallo L, Chiang Y, Vahouny G, and Treadwell C. Localization and origin or rat intestinal cholesterol esterase determined
by immunocytochemistry. J Lipid Res 21: 537–545, 1980.
13. Giller T, Buchwald P, Blum-Kaelin D, and Hunziker W.
Two novel human pancreatic lipase related proteins, hPLRP1
and hPLRP2: differences in colipase dependence and in lipase
activity. J Biol Chem 267: 16509–16516, 1992.
14. Hallden G and Aponte G. Evidence for a role of the gut
hormone PYY in the regulation of intestinal fatty acid-binding
protein transcripts in differentiated subpopulations of intestinal
epithelial cell hybrids. J Biol Chem 272: 12591–12600, 1997.
15. Hamosh M and Scow RO. Lingual lipase and its role in the
digestion of dietary lipid. J Clin Invest 52: 88–95, 1973.
16. Henkel JR, Popovich JL, Gibson GA, Watkins SC, and
Weisz OA. Selective perturbation of early endosome and/or
trans-Golgi network pH but not lysosome pH by dose-dependent
expression of influenza M2 protein. J Biol Chem 274: 9854–
9860, 1999.
17. Jensen J, Serup P, Karlsen C, Nielsen T, and Madsen O.
mRNA profiling of rat islet tumors reveals Nkx 6.1 as a ␤-cellspecific homeodomain transcription factor. J Biol Chem 271:
18749–18758, 1996.
18. Kalogeris TJ, Qin X, Chey WY, and Tso P. PYY stimulates
synthesis and secretion of intestinal apolipoprotein A-IV without
affecting mRNA expression. Am J Physiol Gastrointest Liver
Physiol 275: G668–G674, 1998.
19. Kumar NS and Mansbach CM II. Determinants of triacylglycerol transport from the endoplasmic reticulum to the Golgi in
intestine. Am J Physiol Gastrointest Liver Physiol 273: G18–
G30, 1997.
20. Lin HC, Zhao X-T, and Wong H. Fat-induced ileal brake in the
dog depends on peptide YY. Gastroenterology 110: 1491–1495,
1996.
21. Lindstrom MB, Sternby B, and Borgstrom B. Concerted
action of carboxyl ester lipase and pancreatic lipase during lipid
digestion in vitro: importance of the physicochemical state of the
substrate. Biochim Biophys Acta 959: 178–184, 1988.
22. Madsen OD, Jensen J, Peterson HV, Pederson EE, Oster
A, Anderson FG, Jorgensen MC, Jensen PB, Larsson LI,
and Serup P. Transcription factors contributing to the pancreatic beta-cell phenotype. Horm Metab Res 29: 265–270,
1977.
23. Mansbach CM II. Effect of fat feeding on complex lipid
synthesis in hamster intestine. Gastroenterology 68: 708–714,
1975.
24. Mansbach CM II. The origin of chylomicron phosphatidylcholine in the rat. J Clin Invest 60: 411–420, 1977.
25. Mansbach CM II and Dowell R. The effect of increasing lipid
loads on the ability of the endoplasmic reticulum to transport
lipid to the Golgi. J Lipid Res 41: 605–612, 2000.
26. Mansbach CM II, Dowell RF, and Pritchett D. Portal transport of absorbed lipids in the rat. Am J Physiol Gastrointest Liver
Physiol 261: G530–G538, 1991.
27. Mansbach CM II and Nevin P. Intracellular movement of
triacylglycerols in the intestine. J Lipid Res 39: 963–968, 1998.
28. Payne R, Sims H, Jennens M, and Lowe M. Rat pancreatic
lipase and two related proteins: enzymatic properties and mRNA
expression during development. Am J Physiol Gastrointest Liver
Physiol 266: G914–G921, 1994.
29. Rao RH and Mansbach CM II. Acid lipase in rat intestinal
mucosa: physiological parameters. Biochim Biophys Acta 1043:
273–280, 1990.
30. Rao RH and Mansbach CM II. Purification and partial characterization of intestinal acid lipase. Biochim Biophys Acta 1046:
19–26, 1990.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017
into the duodenum, was poorly hydrolyzed and absorbed (R. F. Dowell and C. M. Mansbach, unpublished
observations). The amount of lipid that can be absorbed under conditions of pancreatic exocrine insufficiency (30%; see Ref. 10) is consistent with the amount
predicted from the action of preduodenal lipases, which
are known to primarily hydrolyze acyl groups esterified
at the sn-3 position of TAG (33). These data are also
consistent with the steady-state percentage of TAG
remaining (70%) after prolonged hydrolysis by preduodenal lipases in the rat (15). Because the pancreatic
secretion of digestive enzymes, including pPTL, is almost exclusively from zymogen granules, it may be
that enterocytes are unable to secrete iPTL because
they lack the prerequisite zymogen granule apparatus.
The intense PTL immunoreactivity (see Fig. 6) that is
located near the apical aspect of the nucleus may
represent aggregated iPTL that has precipitated in the
slightly acidic trans-Golgi network (8, 16) and is therefore unable to be secreted. Prior work has shown that
pancreatic secretory proteins, including pPTL, precipitate under these conditions in the absence of zymogen
granule membranes (11). We therefore speculate that
enterocytes retain the iPTL that they synthesize and
that this retention is necessary if iPTL is to play an
intracellular role in fat processing, as we propose.
G1195
G1196
INTESTINAL TRIACYLGLYCEROL LIPASE
31. Rao RH and Mansbach CM II. Alkaline lipase in rat intestinal
mucosa: physiological parameters. Arch Biochem Biophys 304:
483–489, 1993.
32. Rausch U, Rodgers K, Vasiloudes P, Kern H, and Scheele
G. Lipase synthesis in the rat pancreas is regulated by secretin.
Pancreas 1: 522–528, 1986.
33. Regalska E, Ransac S, and Verger R. Stereoselectivity of
lipases II: stereoselective hydrolysis of triglycerides by gastric
and pancreatic lipases. J Biol Chem 265: 20271–20276, 1990.
34. Sander M and German MS. The beta cell transcription factors
and development of the pancreas. J Mol Med 75: 327–340, 1997.
35. Spalinger J, Seidman E, Menard D, and Levy E. Endogenous lipase activity in Caco-2 cells. Biochim Biophys Acta 1393:
119–127, 1998.
36. Terada T, Kitamura Y, Ashida K, Matsunaga Y, Kato M,
Harada K, Morita T, Ohta T, and Nakanuma Y. Expression
of pancreatic digestive enzymes in normal and pathologic epithelial cells of the human gastrointestinal system. Virchows
Arch 431: 195–203, 1997.
37. Tipton AD, Frase S, and Mansbach CM II. The isolation and
characterization of a mucosal triacylglycerol pool undergoing
hydrolysis. Am J Physiol Gastrointest Liver Physiol 257: G871–
G878, 1989.
38. Tsujita T, Miyazaka T, Tabei R, and Okuda H. Coenzyme
A-independent monoacylglycerol acyltransferase from rat intestinal mucosa. J Biol Chem 271: 2156–2161, 1996.
39. Turpen T and Griffith O. Rapid isolation of RNA by a guanidinium thiocyanate/cesium chloride gradient method. Biotechniques 4: 11–15, 1986.
40. Van der Lee K, Vork M, De Vries J, Willemsen P, Glatz J,
Reneman R, Van der Vusse G, and Van Bilsen M. Longchain fatty acid-induced changes in gene expression in neonatal
cardiac myocytes. J Lipid Res 41: 41–47, 2000.
41. Winzell MS, Lowe ME, and Erlanson-Albertson C. Rat gastric procolipase: sequence, expression, and secretion during
high-fat feeding. Gastroenterology 115: 1179–1185, 1998.
42. Yang Y and Lowe ME. The open lid mediates pancreatic lipase
function. J Lipid Res 41: 48–57, 2000.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.5 on March 30, 2017