Báo cáo khoa học: Differential gene expression in periportal and perivenous mouse hepatocytes potx

Many enzymes of intermediary metabolism are not distributed uniformly throughout the liver, but are preferentially expressed in either the periportal or the perivenous hepatocyte subpopu

mouse hepatocytes

Albert Braeuning1, Carina Ittrich2, Christoph Ko¨hle1, Stephan Hailfinger1, Michael Bonin3,

Albrecht Buchmann1and Michael Schwarz1

1 Institute of Pharmacology and Toxicology, Department of Toxicology, University of Tuebingen, Germany

2 Central Unit of Biostatistics, German Cancer Research Center, Heidelberg, Germany

3 Institute for Human Genetics, Microarray Facility, Tuebingen, Germany

Hepatocytes play a pivotal role in both the synthesis

and degradation of numerous endogenous

biomole-cules, thus maintaining metabolic homeostasis, as well

as in the conversion and detoxification of xenobiotic

compounds Based on the location of the blood

ves-sels, the terminal branches of the portal and the

hep-atic (central) veins and on the direction of the blood

flow, hepatocytes of each liver lobule can be divided

into two subpopulations, an upstream ‘periportal’ and

a downstream ‘perivenous’ (pericentral) population

Zonal-specific differences in the metabolic capacities of

many enzymes or other proteins, and – to a lesser

extent) of their corresponding messenger RNAs, have

been subject to extensive studies throughout the last

decades

Many enzymes of intermediary metabolism are not distributed uniformly throughout the liver, but are preferentially expressed in either the periportal or the perivenous hepatocyte subpopulation [1–3] Hence, hepatocytes located in either of the two regions have different, often complementary, functions Whereas, for example, glycolysis is exclusively active in perive-nous hepatocytes, key enzymes of gluconeogenesis, the antagonist pathway, are preferentially expressed in periportal hepatocytes [1] Zonal-specific expression has also been established for enzymes of amino acid and ammonia metabolism, showing, for example, a higher activity of the urea cycle in periportal cells compared to perivenous hepatocytes [3], whereas gluta-mine synthesis is exclusively active in the perivenous

Keywords

metabolic zonation; microarray analysis;

mouse liver; zonal gene expression

Correspondence

M Schwarz, Institute of Pharmacology and

Toxicology, Department of Toxicology,

University of Tuebingen, Wilhelmstr 56,

72074 Tuebingen, Germany

Fax: +49 7071 29 2273

Tel: +49 7071 29 77398

E-mail: michael.schwarz@uni-tuebingen.de

(Received 24 July 2006, revised 15

Septem-ber 2006, accepted 18 SeptemSeptem-ber 2006)

doi:10.1111/j.1742-4658.2006.05503.x

Hepatocytes located in the periportal and perivenous zones of the liver lobule show remarkable differences in the levels and activities of various enzymes and other proteins To analyze global gene expression patterns of periportal and perivenous hepatocytes, enriched populations of the two cell types were isolated by combined collagenase⁄ digitonin perfusion from mouse liver and used for microarray analysis In total, 198 genes and expressed sequences were identified that demonstrated a ‡ 2-fold difference

in expression between hepatocytes from the two different zones of the liver

A subset of 20 genes was additionally analyzed by real-time RT-PCR, val-idating the results obtained by the microarray analysis Several of the differentially expressed genes encoded key enzymes of intermediary meta-bolism, including those involved in glycolysis and gluconeogenesis, fatty acid degradation, cholesterol and bile acid metabolism, amino acid degra-dation and ammonia utilization In addition, several enzymes of phase I and phase II of xenobiotic metabolism were differentially expressed in peri-portal and perivenous hepatocytes Our results confirm previous findings

on metabolic zonation in liver, and extend our knowledge of the regulatory mechanisms at the transcriptional level

Abbreviations

GS, glutamine synthetase.

subpopulation [4] The list of zonally expressed

enzymes can be further extended to metabolic

path-ways such as glycogen synthesis, lipid metabolism and

bile acid formation [1] Many enzymes of xenobiotic

metabolism also exhibit zonal-specific differences in

protein or mRNA levels, with preferential perivenous

expression of the main detoxification enzymes such as

the cytochrome P450 monooxygenase isoforms [5]

Two different types of zonation can be distinguished,

so-called ‘dynamic zonation’ and ‘stable zonation’

Whereas some zonally expressed genes, particularly

those encoding enzymes of carbohydrate metabolism,

undergo dynamic changes in expression as an adaptive

response to changes in hormonal or nutritional status

[6], a second group of genes is more or less stably

expressed within only a few layers of hepatocytes of

the liver lobule, either perivenous or periportal, and

can hardly be affected by external stimuli One of the

best-known examples of proteins with stable zonal

expression is the enzyme glutamine synthetase (GS),

which is expressed at high levels within only one

to two layers of hepatocytes surrounding the central

veins [4]

Previous studies on zonation of metabolic pathways

in the liver were mainly focused on analysis of protein

expression or measurements of enzyme activities in rat

liver, but comparatively few data are available on

mRNA expression levels, especially in mice It is

there-fore largely unknown whether zonal gene expression is

regulated primarily at the transcriptional or

post-trans-lational level We have now investigated by microarray

analysis the global gene expression patterns of

peripor-tal and perivenous hepatocytes to identify those

pro-teins that are subject to differential transcriptional

regulation in the two hepatocyte subpopulations

Results

Expression profiles of perivenous and periportal

hepatocytes

Perivenous and periportal mouse hepatocyte fractions

were obtained by combined digitonin⁄ collagenase

per-fusion of liver of male C3H⁄ He mice The efficiency of

hepatocyte separation was monitored by western

ana-lysis of marker proteins with well-known zonal

differ-ences in expression, e.g GS, a perivenous marker, and

E-cadherin, a periportal marker; this demonstrated the

expected differences in levels between periportal and

perivenous hepatocytes (Fig 1) The RNA expression

patterns of hepatocyte fractions were analyzed by use

of the Affymetrix GeneChip MOE-430A, which

con-tains approximately 22 600 probe sets, including more

than 14 000 well-characterized mouse genes Genes were stratified into two groups according to their pref-erential perivenous or prefpref-erential periportal expres-sion, using as discriminators a Œlog2 expression ratioŒ

‡ 1 (corresponding to a ‡ 2-fold difference in expres-sion) and an adjusted P-value£ 0.1

In total, we identified 198 genes and expressed sequences that were differentially regulated in hepato-cytes from the two different zones of the liver; 99 of these were predominantly expressed in perivenous cells, whereas another 99 were mainly expressed in peripor-tal hepatocytes A detailed list of the differentially expressed genes is provided in supplementary Table S1 Note that the number of probe sets with significant differences in expression was somewhat higher than

198, because several genes were represented more than once on the array Figure 2 shows a so-called Volcano plot, where, for each of the  22 600 probe sets, the mean of the adjusted P-values is plotted against its corresponding mean log2 ratio Probe sets meeting the criteria of significance are shown in the gray areas of the plot Genes with preferential perivenous expression

pv GS

E-Cad.

Gpr49

Cyp1A

GAPDH

pp

Fig 1 Western analysis of marker protein levels in protein extracts from perivenous (pv) and periportal (pp) hepatocyte subpopulations enriched by digitonin ⁄ collagenase perfusion The indicated proteins are known to show marked zonal differences in expression in liver and were therefore chosen as ‘markers’ for periportal and perive-nous hepatocytes Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loading control GS, glutamine synthetase; E-cad, E-cadherin; Gpr49, G-protein-coupled receptor 49; Cyp1A, cytochrome P450 1A.

show positive log2 ratios in the plot, whereas genes

with preferential periportal expression show negative

log2 ratios

To validate these data, the expression of 20 (10%)

of the zonated genes found in the microarray

experi-ment was additionally analyzed by real-time RT-PCR

The PCR data closely resembled the findings of the

microarray analysis A comparison of the results

obtained with the two methods is shown in Table 1

Differences in genes encoding enzymes of

intermediary metabolism

The results of our present microarray analysis clearly

demonstrate differences between perivenous and

peri-portal hepatocytes in the expression of genes encoding

key enzymes of zonated pathways of intermediary

meta-bolism This holds particularly true for genes encoding

enzymes playing a role in pathways that are known to

be stably zonated within the liver lobule A schematic

representation of the observed differences in selected

metabolic pathways is given in Figs 3 and 4

Glycolysis and gluconeogenesis

As shown in Fig 3A, several genes encoding enzymes

participating in glycolysis are preferentially expressed in

perivenous cells These include the genes encoding sorbi-tol dehydrogenase (EC 1.1.1.14), aldehyde reductase (EC 1.1.1.21), 6-phosphofructo-2-kinase (EC 2.7.1.105), dihydrolipoamide-S-transferase (EC 2.3.1.2), and isoci-trate dehydrogenase (EC 1.1.1.41) Only one gene of this pathway, that encoding pyruvate kinase (EC 2.7.1.40), was found to be mainly expressed in periportal hepato-cytes The gene encoding phosphoenolpyruvate carb-oxykinase (EC 4.1.1.32), one of the key enzymes in gluconeogenesis, is primarily expressed in periportal hepatocytes The same holds true for the gene encooding ATP citrate lyase (EC 2.3.3.8), an enzyme forming oxaloacetate and acetyl-CoA from citrate for further utilization in gluconeogenesis and cholesterol synthesis, respectively When the discrimination level was lowered

to aŒlog2expression ratioŒ‡ 0.5, additional genes enco-ding enzymes involved in glucose metabolism were found to be zonated, such as the gene encoding the tricarboxylic acid cycle component citrate synthase (EC 2.3.3.1), which is preferentially expressed in peri-venous hepatocytes

Fatty acid degradation and cholesterol metabolism

Zonal-specific differences in the expression of genes encoding enzymes involved in fatty acid degradation

Fig 2 Volcano plot demonstrating

differ-ences in gene expression between periportal

and perivenous hepatocytes Each of the

 22 600 transcripts is represented by a

sin-gle dot Discriminators (P ¼ 0.1 for the

adjus-ted P-value and |log2expression ratio| ‡ 1)

are indicated by horizontal and vertical lines;

these were chosen to identify genes with

significant alterations in expression (areas

indicated by gray) One hundred and

twenty-nine probe sets (corresponding to 99

transcripts) were predominantly expressed in

perivenous cells, showing positive log 2 ratios,

whereas another 114 probe sets

(corres-ponding to another 99 transcripts) were

mainly expressed in periportal hepatocytes,

showing negative log 2 ratios.

and cholesterol metabolism are shown in Fig 3B For

example, mRNAs coding for phosphatide phosphatase

(EC 3.1.3.4) and apoliprotein C-II, an essential

co-factor for the activation of lipoprotein lipase

(EC 3.1.1.34), are preferentially expressed in periportal

hepatocytes As mentioned above, preferential

peripor-tal expression is also observed for the

acetyl-CoA-forming enzyme ATP citrate lyase (EC 2.3.3.8), which

provides acetyl-CoA for cholesterol synthesis

How-ever, the mRNAs for CoA synthase and

HMG-CoA reductase, the enzymes catalyzing the initial steps

in cholesterol formation from acetyl-CoA, failed the

criteria of significance in our microarray analysis Bile

acid synthesis is another pathway known to be active

only in perivenous hepatocytes [2] The key enzyme in

this pathway is cytochrome P450 7A1 (EC 1.14.13.17)

This enzyme, catalyzing the rate-limiting step in bile

acid formation, is regulated at the level of mRNA, which was found to be expressed to a much greater extent in perivenous than in periportal hepatocytes

At a lower cutoff (Œlog2 expression ratioŒ‡ 0.5) cyto-chrome P450 27A1 (EC 1.14.13.15), an enzyme involved in side chain oxidation of sterol intermediates during bile acid formation, appeared to be zonated (preferentially perivenous)

Amino acid degradation Figure 3C demonstrates differences in expression of the genes encoding enzymes of histidine and serine⁄ glycine catabolism Three genes encoding enzymes of histidine catabolism, histidine ammonia lyase (EC 4.3.1.3), uroc-anate hydratase (EC 4.2.1.49) and glutamate formimi-notransferase (EC 2.1.2.5), are exclusively expressed in periportal hepatocytes Additionally, mRNA for hista-mine-N-methyltransferase (EC 2.1.1.8), an enzyme of histamine metabolism, is also mainly expressed in the periportal hepatocyte subpopulation A comparable periportal zonation on the mRNA level can be observed for genes encoding enzymes of serine⁄ glycine meta-bolism, including glycine decarboxylase (EC 1.4.4.2), serine dehydratase (EC 4.3.1.17), and serine dehydra-tase-like (EC 4.3.1.19) Oxaloacetate, the product of the reaction catalyzed by the latter enzymes, can be used for gluconeogenesis, a pathway that is also mainly located

in periportal hepatocytes [1]

Ammonia utilization

As shown in Fig 3D, ammonia is used in perivenous hepatocytes for glutamine synthesis, as GS (EC 6.3.1.2), the key enzyme, is specifically expressed

in this hepatocyte subpopulation Comparable zona-tion is found for transporters participating in ammonia (rhesus blood group-associated B glycoprotein) and glutamate uptake (solute carriers 1A2 and 1A4), thus providing the substrates for GS In contrast, periportal hepatocytes are lacking GS and use ammonia for urea synthesis With less stringent cutoff conditions, the mRNAs of four enzymes of the urea cycle were found

to be preferentially localized in the periportal area, namely ornithine transcarbamylase (EC 2.1.3.3), argin-inosuccinate synthetase (EC 6.3.4.5), arginargin-inosuccinate lyase (EC 4.3.2.1), and arginase (EC 3.5.3.1), showing log2expression ratios between 0.58 and 0.82

Xenobiotic metabolism

As expected, many genes encoding enzymes of xeno-biotic metabolism were mainly expressed in perivenous

Table 1 Validation of microarray analysis data by real-time RT-PCR.

If genes are represented by more than one probe set on the chip,

their individual log2 ratios are shown Genes preferentially

expressed in perivenous (pv) hepatocytes are indicated by positive

log 2 ratios, and periportal (pp) expression is indicated by negative

log2ratios Log2ratios of PCR analysis represent the mean log2

ratios from comparison of the same three periportal and perivenous

hepatocyte isolates that were used for the microarray analysis.

Gene

Log 2 ratio(s)

pv versus pp (microarray)

Log 2 ratio

pv versus

pp (PCR) Glutamine synthetase (glul) 6.23 ⁄ 3.72 4.21

6-Phosphofructo-2-kinase ⁄

fructose-2,6-bisphosphatase (Pfkfb1)

Phosphoenolpyruvate carboxykinase 1,

cytosolic (Pck1)

Sulfotransferase 5a1 (Sult5a1) ) 3.03 ) 3.39

Aldehyde dehydrogenase 1B1

(Aldh1b1)

Cytochrome P450 7a1 (Cyp7a1) 3.42 ⁄ 2.55 2.69

ATP citrate lyase (Acly) ) 1.36 ⁄ ) 1.44 ⁄

G protein-coupled receptor 49 (Gpr49) 2.63 8.10

Constitutive androstane receptor

(Nr1i3)

Hairy and enhancer of split 1 (Hes1) ) 1.90 ) 1.49

Catenin beta interacting

protein 1 (Ctnnbip1)

Rhesus blood group-associated

B glycoprotein (Rhbg)

hepatocytes A list of these genes is given in

Table 2 This holds true both for enzymes of phase I

xenobiotic metabolism, e.g various cytochrome P450

monooxygenases, and for phase II enzymes, e.g several isoforms of glutathione-S-transferases and sulfotransf-erases Other genes involved in xenobiotic metabolism,

periportal perivenous

perivenous (only

on protein level)

no zonation RNA periportal, protein

equally distributed

fructose-2,6-bis-P

glucose

fructose-6-P

pyruvate

acetyl-CoA

Phosphoenolpyruvate

1.1.1.21

2.7.1.105/

3.1.3.46

2.3.1.12

sugar alcohols

1.1.1.14

sugars

4.1.1.32

allosteric activation +

2.7.1.2

fructose-1,6-bis-P

2.7.1.11

oxaloacetate

glucose-6-P

2.7.1.40

citrate

2.3.3.8

oxaloacetate citrate

A

1.1.1.41

glutamine

ammonia glutamate

amino acid degradation

Rhbg Slc1A4 Slc1A2

6.3.1.2

urea cycle

2.6.1.13

ornithine ammonia

D

urocanate

N-formimino-glutamate

histidine

N-methyl-histamine histamine

glutamate

2.1.2.5

4.3.1.3

4.2.1.49

2.1.1.8

serine glycine

pyruvate

C

excretion

oxaloacetate gluconeogenesis

4.3.1.17 1.4.4.2

4.1.1.32 4.3.1.19

cholesterol

citrate acetyl-CoA

1.14.13.17

bile acids

2.3.3.8

oxaloacetate

4.1.1.32

gluconeo-genesis

diacylglycerol

diacylglycerol-P

3.1.3.4

APOC2

activation

+

3.1.1.34

B

Lipo-proteins

Fig 3 Zonal differences in expression of genes encoding enzymes and other proteins involved in intermediary metabolism Perivenous expression is indicated by green, and genes with preferential periportal expression are indicated by red (A) Perivenous zonation of glycolysis and periportal zonation of gluconeogenesis (B) Fatty acid degradation and cholesterol metabolism in periportal hepatocytes (C) Elevated amino acid degradation in periportal hepatocytes (D) Ammonia utilization for glutamine synthesis in perivenous hepatocytes EC 1.1.1.14,

L -iditol-2-dehydrogenase (sorbitol dehydrogenase) (gene name: Sdh); EC 1.1.1.21, aldehyde reductase (Akr1b3); EC 2.7.1.2, glucokinase;

EC 2.7.1.105⁄ EC 3.1.3.46, 6-phosphofructo-2-kinase ⁄ fructose-2,6-bisphosphatase (bifunctional enzyme) (PfkFB1); EC 2.7.1.11, 6-phospho-fructokinase; EC 2.7.1.40, pyruvate kinase liver and red blood cell (Pklr); EC 2.3.1.12, dihydrolipoamide-S-transferase (E2 component of pyru-vate dehydrogenase complex) (Dlat); EC 1.1.1.41, isocitrate dehydrogenase NAD + (Idh3a); EC 4.1.1.32, phosphoenolpyruvate carboxykinase

1, cytosolic (pck1); EC 2.3.3.8, ATP citrate lyase (Acly); APOC2, apolipoprotein C-II (essential cofactor for the activation of lipoprotein lipase);

EC 3.1.1.34, lipoprotein lipase; EC 3.1.3.4, phosphatide phosphatase type 2c (Ppap2c); EC 1.14.13.17, cytochrome P450 7A1 (cholesterol-7-a-monooxygenase) (Cyp7a1); EC 4.3.1.3, histidine ammonia lyase (Hal); EC 4.2.1.49, urocanase domain containing 1 (urocanate hydratase) (Uroc1); EC 2.1.2.5, glutamate formiminotransferase (Ftcd); EC 2.1.1.8, histamine-N-methyltransferase (Hnmt); EC 1.4.4.2, glycine decarboxy-lase (part of glycine dehydrogenase complex) (Gldc); EC 4.3.1.17, serine dehydratase (Sds); EC 4.3.1.19, serine dehydratase-like (Sdsl);

EC 2.6.1.13, ornithine aminotransferase (oat); EC 6.3.1.2, glutamate ammonia ligase (glutamine synthetase) (glul); Slc1A2, solute carrier 1A2; Slc1A4, solute carrier 1A4; Rhbg, rhesus blood group-associated B glycoprotein.

such as cytochrome P450 oxidoreductase and receptors

activating enzymes of xenobiotic metabolism, namely

the constitutive androstane receptor and the aryl

hydrocarbon receptor, are also preferentially expressed

in hepatocytes near the central veins Some exceptions

with preferential expression in periportal hepatocytes,

however, were also observed, such as sulfotransferase

5a1 and glutathione-S-transferase alpha 2

Discussion

Upon separation of hepatocytes from the periportal

and perivenous zones of the liver lobule,  200 genes

or expressed sequence tags were identified, which were

differentially expressed between the two cell

subpopu-lations These included several genes encoding enzymes

that are rate-limiting in distinct metabolic pathways of

intermediary metabolism and show well-established

zonal heterogeneity in liver [1–3], demonstrating

that they are, at least in part, regulated by zonal

differ-ences at the transcriptional level or by zonal-specific

post-transcriptional mechanisms affecting the stability

of their mRNAs

The zonally expressed pathways showing striking differences in mRNA levels of several genes are shown

in Figs 3 and 4 Among them are several genes enco-ding enzymes participating in glycolysis that are prefer-entially expressed in perivenous cells, whereas mRNAs for key enzymes in gluconeogenesis, the antagonist pathway to glycolysis, are primarily expressed in peri-portal hepatocytes These results confirm previous observations on the zonation of glycolysis and glucone-ogenesis in the liver, but also reveal zonal expression of genes involved in glucose metabolism that have not previously been reported as zonated For example, peri-venous localization was demonstrated for the mRNAs

of sorbitol dehydrogenase and aldehyde reductase, which are involved in carbohydrate conversion processes that provide glucose for further metabolism

in glycolysis Neither of these enzymes has been pre-viously reported to be differentially expressed between perivenous and periportal hepatocytes However, the

oxygen tension

glycolysis gluconeogenesis

cholesterol

amino acid degradation

glutamine synthesis metabolism of xenobiotics

blood flow

fatty acid degradation

hormones,

growth factors

-catenin signaling

Fig 4 Schematic representation of metabolic processes taking place in different hepatocyte subpopulations along the portocentral axis The figure summarizes the activities of zonated pathways shown in detail in Fig 3 and additionally shows the gradients in oxygen tension, hor-mones ⁄ growth factors, and b-catenin signaling that have been suspected to influence zonal gene expression in the liver [1,5,22,25].

mRNA for glucokinase, the key enzyme in initiation of

glucose degradation, was not found to be zonated in

our microarray analysis, which is in line with previous

observations that the preferential perivenous

locali-zation of glucokinase activity is regulated on a

post-translational level [7] Other mRNAs, mainly expressed

in perivenous cells, code for 6-phosphofructo-2-kinase,

which produces fructose-2,6-bisphosphate, an allosteric

activator of the glycolytic enzyme

6-phosphofructo-1-kinase and an inhibitor of the gluconeogenic enzyme

fructose-1,6-bisphosphatase [8] The mRNA for the

latter enzyme was not zonated in our analysis,

confirm-ing a previous study describconfirm-ing homogeneous

distri-bution of 6-phosphofructo-1-kinase in the liver [9]

Dihydrolipoamide-S-transferase, the core component

of the pyruvate dehydrogenase complex, and isocitrate

dehydrogenase, an enzyme of the tricarboxylic acid

cycle, were also preferentially expressed in perivenous

hepatocytes Perivenous zonation of citrate synthase is

in line with previous observations describing higher

perivenous activity of the enzyme in rat liver [10]

Although there are no previous reports describing zonation of dihydrolipoamide-S-transferase, the prefer-ential mRNA expression of isocitrate dehydrogenase

in perivenous hepatocytes is in accordance with the known perivenous activity of this enzyme [11,12] The only gene of the glycolytic pathway that is mainly expressed in periportal hepatocytes is that encoding pyruvate kinase The metabolic capacity of the res-pective protein, however, was found to be equally distributed throughout the liver lobuli [13], or to be even higher in the perivenous zone [14], suggesting a post-transcriptional regulation mechanism for this enzyme On the other hand, periportal zonation

of gluconeogenesis and particularly of phosphoenol-pyruvate carboxykinase, has been reported before [1] Fatty acid degradation is another metabolic pathway underlying zonal expression in liver, as two genes, those encoding phosphatide phosphatase and apolipo-protein C2, a cofactor for activation of lipoapolipo-protein lipase, were found in our study to be preferentially expressed in the periportal hepatocyte subpopulation Zonal-specific expression of these genes has not been reported so far, but our findings are in line with previ-ous observations of the periportal localization of fatty acid degradation [15] Bile acid synthesis from choles-terol takes place in perivenous hepatocytes, as the key enzyme of this metabolic pathway, cholesterol-7-a-monooxygenase, is preferentially expressed in hepato-cytes surrounding the central veins [16] Our analysis confirms previous findings on the zonation of choles-terol-7-a-monooxygenase protein [17] and mRNA [16] The mRNA levels for ATP citrate lyase, an enzyme forming acetyl-CoA from citrate, thus providing sub-strate molecules for cholesterol synthesis, were higher

in periportal hepatocytes, which is in line with a report

by Evans et al [18] describing periportal localization

of the protein However, another study found ATP cit-rate lyase protein to be more active in the perivenous zone [19] Our microarray data, confirmed by real-time PCR experiments, clearly demonstrate periportal local-ization of the mRNA for this enzyme

The metabolism of several amino acids has also been reported to be differentially regulated in the two zones

of the liver [1–3] Our present results indicate that his-tidine degradation seems to take place mainly in peri-portal hepatocytes, as mRNAs for several enzymes of histidine catabolism, namely histidine ammonia lyase, urocanate hydratase, glutamate formiminotransferase and histamine-N-methyltransferase, are preferentially expressed in periportal hepatocytes Zonal-specific expression of these genes in mouse liver has not been reported in the literature so far Comparable periportal zonation on the mRNA level can be observed for the

Table 2 Zonated genes involved in xenobiotic metabolism If

genes are represented by more than one probe set on the chip,

their individual log2 ratios are shown Genes preferentially

expressed in perivenous (pv) hepatocytes are indicated by positive

log 2 ratios, and periportal (pp) expression is indicated by negative

log2ratios.

Gene

Log2ratio(s)

pv versus pp

enzymes of serine⁄ glycine metabolism: Glycine

decarb-oxylase, a component of the glycine dehydrogenase

complex that converts glycine to serine, is

predomin-antly expressed in periportal cells; the same holds true

for the enzymes of serine catabolism, namely serine

dehydratase and serine dehydratase-like This is in

agreement with results of previous studies describing

periportal localization of serine dehydratase at the

mRNA level [20] Enhanced serine metabolism in

peri-portal hepatocytes may contribute to the availability

of substrates for gluconeogenesis, a pathway that is

also located mainly in periportal hepatocytes [1] By

contrast, glutamine synthesis occurs mainly in

peri-venous hepatocytes, as mRNA for GS, the key

enzyme, is specifically expressed in this hepatocyte

sub-population, which is in accordance with previous

observations [4] Notably, mRNAs for transporters

participating in ammonia and glutamate uptake also

show preferential perivenous localization, thus

provi-ding the substrates for GS The ammonium transporter

rhesus blood group-associated B glycoprotein has been

previously reported to be expressed only in perivenous

hepatocytes at both the protein [21] and mRNA [22]

levels Zonation of the glutamate transporter solute

carrier 1A2 has already been established [23], whereas

the preferential perivenous localization of solute carrier

1A4 has not been described so far The urea cycle has

been reported to be mainly localized in periportal

hepatocytes [1,2] Our data now suggest that this

localization is based on differences in mRNA levels of

four enzymes of the urea cycle

Whereas the zonal-specific expression of the main

enzymes of drug and xenobiotic metabolism has been

the subject of extensive research (e.g cytochrome P450

zonation [5]), the zonal expression profiles of the more

uncommon cytochrome P450 isoforms have mostly not

been described in the literature For example, up to

now cytochrome P450 2G1 (Cyp2g1) has been

consid-ered to be exclusively expressed in the olfactory

mucosa in mammals [24] Whereas the mRNAs for

most xenobiotic-metabolizing enzymes are mainly

expressed in hepatocytes near the central veins, a small

number of these enzymes exhibit preferential periportal

expression The periportal localization of these

mRNAs has not been reported in the literature so far

The mechanisms underlying zonal gene expression in

the liver are not yet fully understood Based on

compari-sons of mRNA⁄ protein expression patterns of

perive-nous and periportal hepatocytes with those of liver

tumors containing activating mutations in either the

Ha-rasor ctnnb1 (catnb; b-catenin) gene, we developed

the hypothesis that two opposing signaling pathways

triggered by Ha-ras- and b-catenin-dependent factors

may determine zonal differences in gene expression in murine liver [22] In addition, the adenomatous poly-posis coli (APC) tumor suppressor gene, an important regulator of b-catenin signaling, has also been estab-lished as a ‘zonation-keeper’ in mouse liver [25] Further studies, however, are required to unravel the molecular details and interplay of the various players involved

In summary, our findings show that several of the well-documented zonal differences in the levels and activities of key enzymes of various pathways of inter-mediary metabolism can be explained, at least in part,

by corresponding differences at the mRNA level in periportal and perivenous hepatocytes, indicating that regulation at the transcriptional level or by mecha-nisms controlling mRNA stability are important fac-tors determining their zonal expression in liver In addition, we found that several other genes with unknown localization in the liver show distinct expres-sion differences between periportal and perivenous hepatocytes Among these are genes coding for proteins involved in well-established zonated but also other pathways that have not been described so far as being differentially expressed in murine liver

Experimental procedures Animal experiments

For microarray analysis of mRNA expression patterns in periportal and perivenous hepatocyte subpopulations, male C3H⁄ He mice were killed at 10 weeks of age and hepatocyte fractions were isolated as described below Mice were kept on

a 12 h dark⁄ light cycle and were killed between 9 and 11 a.m

to avoid circadian influences Animals received humane care, and protocols complied with institutional guidelines

Isolation of hepatocytes Periportal and perivenous subpopulations of hepatocytes were isolated and enriched by combined digitonin⁄ collage-nase perfusion of the liver according to Taniai et al [26], with minor modifications as described previously [22] First, the liver was perfused for 10 min with Krebs⁄ Henseleit buf-fer at 37C To obtain periportal hepatocyte subpopula-tions, a 5 mm digitonin solution was infused for 10 s through the vena cava and then immediately flushed out from the opposite direction To obtain perivenous hepato-cytes, the digitonin solution was infused through the portal vein After digitonin treatment, the liver was perfused with collagenase solution Subsequently, viable hepatocytes were separated by density gradient centrifugation Viability of the resulting hepatocyte fractions was always  80–90% as determined by trypan blue staining The efficiency of

separation of hepatocytes into periportal and perivenous

subfractions was determined by real-time RT-PCR analysis

of GS expression and western blotting for marker proteins

as described below

Microarray analysis and statistical evaluation

of data

The Affymetrix GeneChip MOE-430A (Affymetrix, Santa

Clara, CA, USA) was used for mRNA expression profiling

Six chips were hybridized with cRNA from three periportal

and three perivenous hepatocyte isolates, obtained from

independent liver perfusions RNA quality was controlled

with the Laboratory-on-Chip-System Bioanalyzer 2100

(Agilent, Palo Alto, CA, USA) Data normalization and

statistical analysis was carried out essentially as previously

described [27] To analyze expression differences in the two

cell populations, we used a threshold of 0.1 for the false

discovery rate adjusted P-values and selected only those

probe sets that showedŒlog2expression ratiosŒ‡ 1

(equival-ent to a‡ 2-fold change) The former cutoff was chosen to

keep the expected proportion of false-positives below 10%

The latter was chosen because expression differences

smal-ler than two-fold are very difficult to detect by quantitative

RT-PCR, which was used in this study for verification of

the microarray results We additionally investigated our

dataset, however, using less stringent conditions (Œlog2

expression ratioŒ ‡ 0.5), but the results are only mentioned

in the text in those instances when genes are affected

encoding enzymes within pathways that were found to

show significant differences in mRNA levels when analyzed

at the more stringent cutoff

Western analysis Cells were homogenized in WCE [50 mm Hepes, 150 mm NaCl, 10% (v⁄ v) glycerol, 1% (v ⁄ v) Triton X-100, 1.5 mm MgCl2, 1 mm EGTA, 100 mm NaF, 10 mm NaP2O7,

200 lm Na3VO4] buffer plus protease inhibitor cocktail (Complete Mini, Roche, Mannheim, Germany) Protein concentrations were estimated using the Bradford assay Western blotting was carried out as recently described [22] using antibodies against GS (1 : 5000 dilution; Sigma, Tauf-kirchen, Germany), E-cadherin (1 : 1000; Transduction Laboratories, Lexington, KY, USA), G-protein-coupled receptor 49 (1 : 1000; Affinity BioReagents, Golden, CO, USA), glyceraldehyde-3-phosphate dehydrogenase (1 : 1,000; Chemicon, Hampshire, Chandler’s Ford, UK) and cyto-chrome P450 1A (1 : 1000; gift of R Wolf, Biomedical Research Centre, University of Dundee, UK) Antibody binding was visualized using appropriate alkaline phospha-tase-conjugated secondary antibodies (1 : 10 000; Tropix, Applied Biosystems, Weiterstadt, Germany) and CDP-Star

as a substrate Chemoluminescence signals were monitored

by use of a CCD camera system

Quantitative determination of mRNAs by RT-PCR Total RNA was isolated with Trizol reagent (Invitrogen, Karlsruhe, Germany) RNA was purified using the RNeasy Table 3 PCR primers.

Gene

Forward (5¢- to 3¢)

Reverse (5¢- to 3¢)

Mini Kit (Qiagen, Hilden, Germany) Five hundred

nano-grams of total RNA were reverse transcribed into cDNA

by avian myeloblastosis virus-RT (Promega, Mannheim,

Germany) using standard methods and oligo(dT)18 and

random(dN)6 primers Expression analysis was performed

using the LightCycler real-time PCR system (Roche,

Mannheim, Germany) Expression of 18S rRNA was used

for normalization The primer pairs used for PCR

amplifi-cation are given in Table 3

Acknowledgements

We gratefully acknowledge the excellent technical

assistance of Elke Zabinsky and Silvia Vetter We also

thank Dr R Wolf for the gift of Cyp1A antibody

This study was supported by the Deutsche Krebshilfe

(grant 106356)

References

1 Jungermann K & Katz N (1989) Functional

speciali-zation of different hepatocyte populations Physiol Rev

69, 708–764

2 Gebhardt R (1992) Metabolic zonation of the liver:

regulation and implications for liver function

Pharma-col Ther 53, 275–354

3 Jungermann K & Kietzmann T (1996) Zonation of

par-enchymal and nonparpar-enchymal metabolism in liver

Annu Rev Nutr 16, 179–203

4 Gebhardt R & Mecke D (1983) Heterogeneous

distribu-tion of glutamine synthetase among rat liver parenchymal

cells in situ and in primary culture EMBO J 2, 567–570

5 Oinonen T & Lindros KO (1998) Zonation of hepatic

cytochrome P-450 expression and regulation Biochem J

329, 17–35

6 Lemaigre FP & Rousseau GG (1994) Transcriptional

control of genes that regulate glycolysis and

gluconeo-genesis in adult liver Biochem J 303, 1–14

7 Eilers F, Bartels H & Jungermann K (1993) Zonal

expression of the glucokinase gene in rat liver

Dynamics during the daily feeding rhythm and

starva-tion–refeeding cycle demonstrated by in situ

hybridiza-tion Histochemistry 99, 133–140

8 Rider MH, Bertrand L, Vertommen D, Michels PA,

Rousseau GG & Hue L (2004) 6-Phosphofructo-2-kinase

⁄ fructose-2,6-bisphosphatase: head-to-head with a

bifunctional enzyme that controls glycolysis Biochem J

381, 561–579

9 Frederiks WM, Marx F & van Noorden CJ (1991)

Homogeneous distribution of phosphofructokinase in

the rat liver acinus: a quantitative histochemical study

Hepatology 14, 634–639

10 Tosh D, Alberti GM & Agius L (1988) Glucagon

regu-lation of gluconeogenesis and ketogenesis in periportal

and perivenous rat hepatocytes Heterogeneity of hor-mone action and of the mitochondrial redox state Biochem J 256, 197–204

11 Wimmer M & Pette D (1979) Microphotometric studies

on intraacinar enzyme distribution in rat liver Histo-chemistry 64, 23–33

12 Agius L & Tosh D (1990) Acinar zonation of cytosolic but not organelle-bound activities of phosphoenolpyru-vate carboxykinase and aspartate aminotransferase in guinea-pig liver Biochem J 271, 387–391

13 Jones CG & Titheradge MA (1996) Measurement of metabolic fluxes through pyruvate kinase, phosphoenol-pyruvate carboxykinase, phosphoenol-pyruvate dehydrogenase, and pyruvate carboxylate in hepatocytes of different acinar origin Arch Biochem Biophys 326, 202–206

14 Zierz S, Katz N & Jungermann K (1983) Distribution

of pyruvate kinase type L and M2 in microdissected periportal and perivenous rat liver tissue with different dietary states Hoppe Seylers Z Physiol Chem 364, 1447–1453

15 Guzman M & Castro J (1989) Zonation of fatty acid metabolism in rat liver Biochem J 264, 107–113

16 Twisk J, Hoekman MF, Mager WH, Moorman AF, de Boer PA, Scheja L, Princen HM & Gebhardt R (1995) Heterogeneous expression of cholesterol 7 alpha-hydro-xylase and sterol 27-hydroalpha-hydro-xylase genes in the rat liver lobulus J Clin Invest 95, 1235–1243

17 Ugele B, Kempen HJ, Kempen JM, Gebhardt R, Meijer

P, Burger HJ & Princen HM (1991) Heterogeneity of rat liver parenchyma in cholesterol 7 alpha-hydroxylase and bile acid synthesis Biochem J 276, 73–77

18 Evans JL, Quistorff B & Witters LA (1989) Zonation of hepatic lipogenic enzymes identified by dual-digitonin-pulse perfusion Biochem J 259, 821–829

19 Katz NR, Fischer W & Ick M (1983) Heterogeneous distribution of ATP citrate lyase in rat-liver parench-yma Microradiochemical determination in microdis-sected periportal and perivenous liver tissue Eur J Biochem 130, 297–301

20 Ogawa H & Kawamata S (1995) Periportal expression

of the serine dehydratase gene in rat liver Histochem J

27, 380–387

21 Weiner ID, Miller RT & Verlander JW (2003) Localiza-tion of the ammonium transporters, Rh B glycoprotein and Rh C glycoprotein, in the mouse liver Gastro-enterology 124, 1432–1440

22 Hailfinger S, Jaworski M, Braeuning A, Buchmann A & Schwarz M (2006) Zonal gene expression in murine liver: lessons from tumors Hepatology 43, 407–414

23 Cadoret A, Ovejero C, Terris B, Souil E, Levy L, Lamers WH, Kitajewski J, Kahn A & Perret C (2002) New targets of beta-catenin signaling in the liver are involved in the glutamine metabolism Oncogene 21, 8293–8301