| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 34, 31950-31957, August 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶ || || 
From the
Endocrine Institute, Sheba Medical
Center, Tel-Hashomer 52621, **Internal Medicine B,
Meir Medical Center, Kfar-Saba 44281, Human
Genetics, Sackler School of Medicine, Tel-Aviv University, Tel-Aviv 67978,
||Life Science, Bar-Ilan University, Ramat-Gan
52900, and Pathology Institute, Sheba
Medical Center, Tel-Hashomer 52621, Israel
Received for publication, March 26, 2003 , and in revised form, May 29, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-cell
function. Here we analyze whether PDX-1 functions as a pancreatic
differentiation factor and a bona fide master regulator when
ectopically expressed in mature fully differentiated liver in vivo.
By ectopic and transient PDX-1 expression in liver in vivo,
using the first generation recombinant adenoviruses, we demonstrate that PDX-1
induces in liver a wide repertoire of both exocrine and endocrine pancreatic
gene expression. Moreover, PDX-1 induces its own expression (auto-induction),
which in turn may explain the long lasting nature of the "liver to
pancreas" transdifferentiation. Insulin as well glucagon-producing cells
are mainly located in the proximity of hepatic central veins, possibly
allowing direct hormone release into the bloodstream, without affecting normal
hepatic function. Importantly, we demonstrate that hepatic insulin production
triggered by Ad-CMV-PDX-1 recombinant adenovirus administration is
functional and prevents streptozotocin-induced hyperglycemia in Balb/c mice
even 8 months after the initial treatment. We conclude that PDX-1 plays an
important instructive role in pancreas differentiation, not only from
primitive gut endoderm but also from mature liver. Transconversion of liver to
pancreas may serve as a novel approach for generating endocrine-pancreatic
tissue that can replace malfunctioning -cells in diabetics. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-cell phenotype. The strategy used for this purpose is
ectopic expression of a specific pancreatic master regulator gene. Our
hypothesis was that such a developmental redirection was most likely to occur
between tissues that are developmentally related, such as liver and pancreas
(1–4).
Transdifferentiation of pancreas to liver has been described in both
experimental models and human pathology
(5–10).
Pancreatic and duodenal homeobox gene-1
(PDX-1,1 also known as
IDX-1, IPF-1, STF-1, or IUF-1) possesses a dual role in the pancreas; it is
essential for normal pancreas development during embryogenesis and for
maintenance of -cell function in the adult
(11,
12). During organogenesis
PDX-1 is expressed in all cells differentiating toward the exocrine
and endocrine components of the pancreas
(11–13).
In the adult, its expression is restricted to -cells and to 20% of
somatostatin producing 
-cells. PDX-1 is neither expressed in
glucagon producing 
-cells nor in mature exocrine pancreas
(11,
12,
14).
We demonstrated previously that transdifferentiation of liver to pancreas could be induced by ectopic expression of PDX-1; expression of PDX-1 in liver induced expression of the otherwise silent endogenous insulin genes and was sufficient to direct the production and secretion of mature, biologically active insulin from a restricted population of cells in the liver in vivo. Furthermore, insulin secreted from liver of PDX-1-treated mice ameliorated streptozotocin (STZ)-induced diabetes (1).
In the present study PDX-1 cDNA is delivered in vivo by systemic administration of recombinant adenoviruses that carry the rat PDX-1 gene under the cytomegalovirus promoter (Ad-CMV-PDX-1). This gene delivery system results in efficient transgene expression, mainly in the liver (15). Because genes delivered in vivo by recombinant adenoviruses do not integrate into the host genome, ectopic expression is transient, lasting from days to a few weeks (15–17). Therefore, our previous study monitored the effects of the ectopic PDX-1 over a short period (up to 2 weeks), to parallel the predicted time frame of the PDX-1 transgene expression (1).
In order to determine whether PDX-1 fulfills the role of a differentiation
factor in mature hepatic tissues, we sought to determine whether it induces,
in addition to insulin, a broad repertoire of pancreatic markers, especially
these that are not considered its immediate targets. Moreover, if indeed PDX-1
acts as a differentiation factor also in mature liver and not only in the
primitive gut endoderm, then it may trigger a cascade of events that do not
require continuous expression of the transgene. Therefore, we sought to
determine whether the transient ectopic PDX-1 induces a long lasting process
of liver to pancreas transconversion. Moreover, bona fide master
regulators such as MyoD have the capacity to activate their own expression
(18). This has been proven
also for PDX-1 but only in pancreatic -cells
(19,
20). Because additional
transcription factors such as hepatic nuclear factor 1 and 3 that
are known to control PDX-1 expression in pancreatic -cells are
also expressed in liver and could cooperate with the PDX-1 transgene,
we sought to analyze whether the endogenous otherwise silent homologue of
PDX-1 is induced also in liver in response to the transgene
expression, thus in turn directing a long lasting process of liver to pancreas
transdifferentiation, also when the transgene is no longer expressed.
The present study analyzes PDX-1 capacity to function as a pancreatic differentiation factor in mature liver and investigates part of the mechanism of liver to pancreas transconversion. Moreover, it examines whether the PDX-1-induced liver to pancreas transdifferentiation process carries a therapeutic benefit long after the initial ectopic PDX-1 gene expression decays.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-galactosidase
(generously contributed by C. B. Newgard) serves as a control for nonspecific
effect of adenovirus-mediated gene delivery.
Cell Culture—The mouse pancreatic derived cell lines
-TC-1 and -TC-1 and the rat pancreatic cell line RIN1046-38 were
cultured according to conditions published previously
(21,
22).
Animals and Recombinant Adenoviruses—Mice were housed in an air-conditioned environment, under a 12-h light/dark cycle, and handled according to institutional animal welfare regulations. 8–9-week-old (18–19 g) BALB/c mice were treated by 1–5 x 1010 plaque-forming units of the indicated recombinant adenoviruses by systemic injection into the tail vein (in a volume of 200–300 µl of physiological saline). Blood was drawn from the tail, for determination of glucose concentration (Accutrend® GC, Roche Applied Science). Liver was harvested for immunohistochemical staining (fixed in 4% formaldehyde and embedded in paraffin), for analysis of gene expression (total RNA), and for determination of pancreatic hormone content in liver. For the last two analyses, hepatic specimens were immediately frozen in liquid nitrogen and stored at –70 °C.
RNA Isolation and RT-PCR Analysis—Total RNA was isolated
from frozen tissues using Tri-Reagent (Molecular Research Center). RNA samples
were treated by 10 units of RQ1 RNase-free DNase I (Promega) for 60 min. cDNA
was prepared by reverse transcription (native avian myeloblastosis virus
reverse transcriptase, Chimerx) by using 4 µg of DNA-free total RNA and 0.5
µg of oligo(dT)15. PCR was performed using T3 Thermocycler
(Biometra, Göttingen, Germany), and products were separated on 1.8%
agarose gels and visualized with ethidium bromide. The sequence of the primers
used for PCR and reactions conditions were as listed in
Table I. Note that in order to
discriminate between expression of the endogenous mouse PDX-1 and the
ectopic rat homologue, two sets of specific oligonucleotide primers were
designed (see Table I). RNA
isolated from the -TC-1 cell line was used as a control for insulin and
somatostatin genes expression, and the -TC-1 cell line was used for
glucagon.
|
Quantitative Analysis for Pancreatic Gene Expression Using Real Time PCR—RT-PCR was performed on the LightCycler (Roche Applied Science) using SYBR-Green I dye.
Amplification conditions included initial denaturation at 95 °C for 10
min, followed by 55 cycles for both mouse and rat PDX-1, or 30 cycles
for -actin. For both PDX-1 homologues, each cycle included
denaturation at 95 °C for 15 s, annealing at 59 °C, and extension at
72 °C for 15 s. Annealing for -actin was performed at 56
°C for 10 s. The fluorescent signal was monitored 5 s after each cycle at
90 °C for -actin, 87 °C for mouse PDX-1, and
88 °C for rat PDX-1. Melting curve program was carried out at 68
°C for 40 s to analyze the specificity of the generated products.
The RT-PCR for mouse PDX-1 was performed 3 times and for rat PDX-1
2 times. Both rat and mouse PDX-1 levels were normalized to the
respective -actin mRNA levels in the same samples.
Pancreatic Hormones Analysis by Immunohistochemistry—Slides of 4-µm paraffin-embedded sections were deparaffinized and incubated in 3% H2O2 and in blocking solution (for both insulin and glucagon detection), using the commercially available HistomouseTM-SP Kit (Zymed Laboratories Inc., South San Francisco, CA). Sections were then incubated for 1 h at 37 °C with monoclonal antibodies against human insulin and against human glucagon (Sigma), both at a dilution of 1:200. Because both insulin and glucagon are detected within the same tissue, at the same time using the same dilution of distinct monoclonal antibodies, they are considered extraneous to each other. Slides were exposed to the secondary biotinylated IgG for 30 min at room temperature and then incubated in streptavidin-peroxidase followed by a chromogen peroxide solution. A control using only secondary antibody without primary antibodies followed by streptavidin-peroxidase and a chromogen peroxide solution was performed to rule out possible background of the system.
Radioimmunoassay (RIA) of Pancreatic Hormones—Frozen tissues were homogenized in 0.18 N HCl, 35% ethanol. The homogenates were extracted overnight at 4 °C with continuous stirring, and the supernatants were lyophilized. Samples were dissolved in 0.8 ml of RIA Assay Buffer, supplemented by a mixture of protease inhibitors (Sigma). Hepatic insulin and glucagon levels were determined using rat radioimmunoassay (RIA, catalog numbers SRI-13K and GL-32K, Linco, MO, and Coat-A-Count, Diagnostic Products Corp.). Somatostatin concentrations were determined by RIA (Euro-Diagnostica, Sweden). Hepatic content of pancreatic hormones was normalized to the wet weight of extracted tissues.
Analyses of Hepatic Function—Serum biochemistry profile consisting of albumin, aspartate aminotransferase, alanine aminotransferase, and total bilirubin and amylase levels was determined using Olympus AU 2700 Apparatus (Olympus, Germany) in serum samples.
Statistical Analysis—Statistical analysis was preformed using Two-sample Assuming Unequal Variances t test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
In the developing pancreas PDX-1 serves as an early molecular marker that temporally correlates with pancreatic commitment (13, 23, 24). Our data suggest that PDX-1 recapitulates to some extent its role in pancreas organogenesis, also in a mature fully differentiated tissue, such as liver.
PDX-1 Triggers a Long Lasting Process of Liver to Pancreas Developmental Shift—In order to explore the mechanism underlying the PDX-1-induced developmental shift, and to specifically determine whether cells in liver that ectopically express PDX-1 exhibit several pancreatic developmental options that persist for long periods of time, we carefully followed insulin, glucagon, and somatostatin gene expression and protein production for 6 months after the initial, single adenovirus-mediated PDX-1 administration to mice livers in vivo.
8–9-Week-old mice were treated by systemic injection of
Ad-CMV-PDX-1, a recombinant adenovirus that carries the rat
PDX-1 gene (STF-1) under the control of the CMV promoter
(1). Pancreatic gene expression
in liver was analyzed in comparison to age-matched control mice (treated by
either Ad-CMV--galactosidase or untreated).
Despite the expected transient PDX-1 expression achieved by adenovirus-mediated delivery of the gene to liver, expression of insulin and somatostatin persisted for 6 months, at both the mRNA (Fig. 2) and protein levels (Fig. 4). Glucagon gene expression was evident during the first 4 months (Figs. 2 and 3). Importantly, insulin I and insulin II genes expression was evident in 80–100% of PDX-1-treated mice even 6–8 months after the initial PDX-1 treatment. The increase in insulin and glucagon expression in PDX-1 versus control treated mice livers was further confirmed using real time PCR (data not presented).
|
|
|
The temporal differences between insulin and glucagon gene expression could
reflect a unique phenomenon that characterizes pancreas organogenesis in
mature liver, and may suggest a more stable transconversion toward the -
and -cell phenotype. Glucagon gene is not a direct PDX-1
target gene, and its persistent expression in liver suggests that PDX-1 is
functioning as a differentiation factor in this organ.
Quantitative Analysis of Insulin, Glucagon, and Somatostatin Hormones Production in PDX-1-treated Livers as a Function of Time after Initial Ad-CMV-PDX-1 Administration—Our data demonstrate that the pancreatic genes expressed in liver as a consequence of ectopic PDX-1 expression are successfully translated into the respective proteins.
Immunohistochemical analysis (Fig. 3) localizes the insulin-producing cells mainly in the proximity of central veins even 4–6 months after PDX-1 ectopic gene delivery (Fig. 3, A and C). Although glucagon-positive cells are also localized in the proximity of central veins (Fig. 3B), immunohistochemical analysis of these two hormones performed on sequential slides suggest that these hormones do not co-localize within the same cell. Liver cells present in areas close to the central vein in liver are known to correspond to mature cells (25).
Quantitative analysis of hepatic insulin stored in the liver of PDX-1-treated mice indicates that even 4–6 months after treatment, hepatic insulin content is about 30–75 ng/g tissue, compared with 1–9 ng/g tissue in age-matched control livers (Fig. 4A). Although this represents a substantial increase compared with control liver, it is about 1.3–3% of pancreatic content (51280 ng/g tissue), even after we take in consideration that liver is about 20-fold larger than pancreas. A significant 2-fold increase in hepatic pro/glucagon and somatostatin content was observed up to at least 4 months after initial Ad-CMV-PDX-1 administration (Fig. 4, B and C).
Despite hepatic insulin production, serum insulin and glucose levels in PDX-1-treated mice bearing normal pancreatic function were normal throughout the duration of the experiment (insulin, 1.0 ± 0.5 versus 0.9 ± 0.4 ng/ml, and glucagon, 0.16 ± 0.08 versus 0.12 ± 0.05 ng/ml in PDX-1-treated compared with controls, respectively).
The persistent production of pancreatic hormones in liver suggests that ectopic PDX-1 triggers a cascade of events that may not require the continuous presence of the PDX-1 transgene.
Ectopic PDX-1 Triggers the Expression of the Endogenous, Otherwise Silent PDX-1 Gene in Liver; a Possible Mechanism of Sustained Liver to Pancreas Transdifferentiation—In order to explain the sustained developmental shift in liver, triggered by the transient ectopic PDX-1 expression, we analyzed whether the transgene induces the expression of otherwise silent pancreatic transcription factors, which in turn dictate the continuous expression of pancreatic genes. An obvious candidate is the endogenous PDX-1 gene itself.
To analyze the induction of the endogenous and otherwise silent PDX-1 gene in liver by the ectopic gene, we treated mice by systemic delivery of recombinant adenovirus that directs expression of the rat PDX-1 homologue, and we used specific oligonucleotide primers to distinguish between the ectopic PDX-1 transgene (rat) mRNA (cDNA) and the endogenous mouse mRNA, by RT-PCR.
PCR analysis of DNA samples isolated from liver of Ad-CMV-PDX-1-treated mice confirms that the virally encoded transgene disappears between 30 and 56 days after adenovirus injection (Fig. 5A).
|
Fig. 5B
demonstrates that the ectopic rat PDX-1 expression parallels the
observed presence of delivered viral DNA in liver and also extinguishes after
1 month (Fig. 5A). The
only homologue of PDX-1 expressed in treated livers for the whole
duration of the experiment is the endogenous and otherwise silent mouse
homologue (Fig. 5B).
Endogenous PDX-1 expression was exclusive to mice that received the
rat PDX-1 transgene, was evident in 75% of ectopic PDX-1-treated mice
(21 of 28 mice), and in none of the 25 control treated livers analyzed. By
using real time PCR, we further analyzed the identity and quantities of the
relative levels of mouse versus rat PDX-1 gene expression in
liver as a function of time after the initial treatment, using identical
conditions (but different primers), and we normalized it to -actin
within the same samples.
As demonstrated in Fig. 5C, the mRNA encoding the ectopic rat PDX-1 is maximal at 5 days, drops by 85% at day 30, and disappears thereafter. By contrast, the endogenous mouse PDX-1 is expressed at substantial levels for the whole duration of the experiment. Unfortunately, due to the high level of homology between the two nuclear proteins, no specific antibody could distinguish between the rat and the mouse homologues of PDX-1, although it seems highly unlikely that the nuclear protein detected only in mouse livers 2–6 months after the initial adenovirus delivery corresponds to the ectopic transiently expressed transgene (data not presented). These data taken together suggest an auto-induction of the endogenous and otherwise silent PDX-1 in liver, which in turn may suggest a mechanistic explanation for the long lasting mode of liver to pancreas transconversion process.
Insulin Produced in Liver in Response to PDX-1 Transgene Expression Is Functional and Prevents STZ-induced Hyperglycemia—In order to determine whether PDX-1 gene delivery induces long lasting production of biologically active insulin, we analyzed whether it provides protection against STZ-induced diabetes, even 8 months after the initial Ad-CMV-PDX-1 administration. Mice were treated by 220 mg/kg STZ, and the incidences of hyperglycemia were compared with those in age-matched controls. Sixty percent of the control BALB/c mice developed hyperglycemia (6 of 10) within 3–5 days of STZ injection. By contrast, only 1 of 5 PDX-1-treated mice developed hyperglycemia in response to STZ treatment (20%). In order to analyze whether hepatic insulin production contributed to such a protection, we analyzed the hepatic and the pancreatic insulin content in PDX-1-treated and control mice.
Immunohistochemical studies and insulin content analyses revealed that in
response to STZ treatment, pancreatic -cells were mostly destroyed, and
pancreatic insulin content in both control diabetic mice and in PDX-1-treated
mice (that remained normoglycemic) dropped by 95 ± 1%. By contrast,
hepatic insulin content in PDX-1-treated mice liver remained unaffected by the
STZ treatment and was 40-fold increased compared with PDX-1-untreated mice.
Fig. 6 demonstrates the ratio
between hepatic and pancreatic insulin content upon STZ treatment in control
versus PDX-1-treated mice. Whereas in both groups pancreatic insulin
content was similar, and corresponded to 5% of normal pancreas, only mice
treated by PDX-1 remained normoglycemic. Moreover, because hepatic insulin
content remained unaffected by STZ, the relative contribution of hepatic
insulin production increased to 25.6% that of STZ-treated pancreata (instead
of 1.3–3% when compared with normal pancreas). Although these results
demonstrate that the PDX-1-induced developmental shift is both long lasting
and functional, they may also suggest that developmentally shifted cells in
liver resist -cell-specific toxins.
|
The profound biological activity in face of a relatively modest hepatic insulin production may suggest an apparent preferential efficacy of insulin produced in liver, which remains to be further investigated.
Importantly, despite the ongoing local insulin production in liver even 6–8 months after initial viral infection, hepatic functions were not perturbed and bilirubin levels remained unaltered (Table II). Transient alterations in hepatic functions occurred in response to adenovirus administration; however, hepatic function returned back to normal levels within 1–2 months. Moreover, serum amylase levels were not affected by ectopic PDX-1 expression in liver, suggesting that the pancreatic exocrine function in liver was not induced (Table III). PDX-1-treated mice consistently gained weight with a slightly increased rate compared with that of age-matched controls (Fig. 7).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-cell-specific
toxins.
Although the PDX-1 role in pancreas organogenesis is mainly studied during
embryogenesis, our data demonstrate that PDX-1 possesses an instructive role
in pancreas differentiation also when ectopically expressed in mature and
fully differentiated tissue such as liver. PDX-1 triggers the expression of
pancreatic genes that are not its known immediate targets, such as glucagon,
p48, and elastase (Fig. 1).
PDX-1 is not expressed in mature pancreatic -cells, and
several lines of evidence suggest that glucagon gene expression is indirectly
repressed by PDX-1 (26). The
fact that ectopic PDX-1 expression leads to activation of glucagon as
well as acinar gene expression in liver suggests that PDX-1 acts in this
tissue as a differentiation factor, mediating its effect via downstream
induced or pre-existing transcription factors in liver
(26,
27). The development of
distinct cell types characteristic of endocrine pancreas in mature liver could
also be explained by the formation of a gradient of PDX-1 protein levels that
may differentially affect the endocrine pancreatic lineage identity in liver
(21,
28). For example, transient
PDX-1 expression followed by a decrease in PDX-1 levels could promote
the activation of glucagon gene expression, as suggested for INS-1 cells
(26).
The capacity of PDX-1 to induce its own expression in an extra-pancreatic
tissue, such as liver, strengthens the view that it functions as a bona
fide master regulator
(18). Several transcription
factors including PDX-1 itself
(20,
29) have been suggested as
regulators of PDX-1 gene expression in pancreatic -cells, most
also present in liver (19,
20,
30). Indeed, liver may possess
a unique predisposition to activate endogenous PDX-1 gene expression
due to pre-existing hepatocyte nuclear factor 3 and hepatocyte nuclear
factor 1 in this tissue.
Our data are consistent with the notion that only a subpopulation of cells in the liver are predisposed to undergo a long lasting and comprehensive PDX-1-induced developmental shift. Despite the initial and relatively abundant ectopic PDX-1 expression in liver, uniformly spread among 30–50% of the cells, insulin- and glucagon-positive cells represent less than 1% of the total cells in liver. Immunohistochemical analysis localizes pancreatic hormone-producing cells mainly to the proximity of central veins (Fig. 3) (1). This localization is not attributable to blood circulation because the blood supply that originally brings the recombinant adenoviruses into the liver enters via the hepatic artery and drains through the central veins (31). Therefore, we suggest that liver cells located in the proximity of central veins may possess a certain predisposition that allows them to undergo PDX-1-induced developmental shift. Hepatocytes located in the proximity of central veins are definitely mature and were reported to express high levels of glutamine synthetase (25). A metabolic zonation in liver was suggested by many researchers, and oxygen pressure was suggested to develop and maintain gradients of gene expression in liver (25, 32–35). Therefore, the different populations of cells in liver may consist of slightly distinct levels of certain hepatic transcription factors expression that in turn may affect the ability of these cells to undergo a PDX-1-induced developmental shift (33). An additional possible reason for preferred central vein localization of PDX-1-induced developmentally shifted cells could be the proximity to signals released from endothelial cells around blood vessels. These were documented to induce essential steps in organ formation such as morphogenesis and cell differentiation of both liver and pancreas (36, 37). These cells may harbor signals that instruct PDX-1-expressing cells in liver epithelium to adapt pancreatic cell fate (36, 37). However, we do not know whether signals released from endothelial cells that surround central veins are different from these released from endothelial cells that surround other blood vessels. Additional, yet unknown factors may be involved in the process of liver to pancreas developmental shift that occurs in response to ectopic PDX-1 expression. The location of insulin-producing cells in the proximity of central veins may have a substantial functional advantage in allowing direct release of the produced hormones into the bloodstream, thus preventing intra-hepatic accumulation of pancreatic hormones that could adversely affect carbohydrate metabolism and glycogen storage.
Whereas PDX-1 expression alone in our study is sufficient to induce pancreatic hormones in mice liver, studies performed in transgenic Xenopus tadpoles suggest that PDX-1 should be further activated by VP-16 to exert profound effects and convert most of the liver into pancreas (38). This study confirms the pivotal role of PDX-1 in controlling the process of liver to pancreas fate redirection; however, the functional consequences of converting most of the liver into pancreas are in question and are possibly demonstrated in the study by Kojima et al. (39).
Additional pancreatic transcription factors
(40–42)
may possess the capacity to induce pancreatic repertoire of gene expression in
liver. Recently Kojima et al.
(39) demonstrated the capacity
of NeuroD to induce pancreatic markers. Interestingly, ectopic NeuroD combined
with ectopic betacellulin expression in liver induced both downstream and
upstream transcription factors that are part of the pancreatic transcriptional
network including the PDX-1, and corrected hyperglycemia in STZ
diabetic mice (39). In
addition, the same study suggests that ectopic PDX-1 expression is
toxic to liver due to the induction of the pancreatic acinar function. Hepatic
function analyses performed in our model do not demonstrate similar effects;
serum amylase levels in PDX-1-treated mice were normal, and mice consistently
gained weight (Tables II and
III and
Fig. 7). The discrepancy
between our approach and the approach by Kojima and co-workers is likely to be
caused by the distinct levels and the longer time frame of the ectopic
PDX-1 expression. The use of helper-dependent adenovirus combined
with the potent mammalian elongation factor-1 promoter used in the
study by Kojima et al.
(39) may result in substantial
increase in ectopic PDX-1 expression in liver that may span several
orders of magnitude, in close to 100% of cells in liver, in an almost
irreversible mode
(43–45).
Recombinant adenoviruses seem to be an optimal tool used for gene delivery in vivo for the purpose of inducing a transdifferentiation process. The ectopic PDX-1 serves only as a transient trigger for the persistent transdifferentiation process, leaving the host genome with no modification or insertions of foreign DNA (Fig. 5) and with normal hepatic function as demonstrated in our model (Table II).
Importantly, our study demonstrates that despite the short term expression of the ectopic PDX-1 transgene, the triggered developmental shift remains functional for the whole duration of the experiment, which is at least 8 months, and possibly forever.
Moreover, our data suggest that the developmentally shifted cells in liver
may resist the deleterious effects of the -cell-specific toxins. The
distinct STZ effects on liver versus pancreatic -cells may stem
from the fact that -cells unlike liver cells possess a restricted
capacity to metabolize hydroxyl radicals and hydrogen peroxide. By contrast,
liver cells are equipped with high enzymatic activities of peroxidase,
catalase, and superoxide dismutase, which together prevent accumulation of
free radicals, thus rendering the cells resistant to many toxins and cytokines
(46,
47). The exact mechanism of
the therapeutic effect and the possible resistance of developmentally shifted
liver cells to -cell toxins should be further analyzed.
Additional studies are needed to fully understand the mechanism of PDX-1
and possibly additional pancreatic transcription and growth factors, mediated
liver to pancreas transdifferentiation process. Markers of predisposed cells
should be identified, and transdifferentiated cells should be better
characterized in order to determine to what extent they mimic normal
pancreatic -cell functions.
Islet cell implantation as a treatment for diabetic patients will be widely
available only when new sources of islets or -cells are found. Our study
suggests that liver could serve as an important source of tissue for
generating functional insulin-producing cells to be used in replacing
malfunctioning -cells in patients with diabetes.
|
FOOTNOTES |
|---|
¶ This work was performed in partial fulfillment of the requirements for a
Ph.D. degree. Both authors contributed equally to this work.
To whom correspondence should be addressed. Tel.: 972-3-5303152; Fax:
972-3-5302083; E-mail:
sferber{at}sheba.health.gov.il.
1 The abbreviations used are: PDX-1, pancreatic and duodenal homeobox gene-1;
Ad-CMV-PDX-1, replication-deficient recombinant adenovirus that
carries expression of PDX-1 cDNA under the control of the cytomegalovirus
promoter; STZ, streptozotocin; RT, reverse transcriptase; RIA,
radioimmunoassay.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Koya, S. Lu, Y.-P. Sun, D. L. Purich, M. A. Atkinson, S.-W. Li, and L.-J. Yang Reversal of Streptozotocin-Induced Diabetes in Mice by Cellular Transduction With Recombinant Pancreatic Transcription Factor Pancreatic Duodenal Homeobox-1: A Novel Protein Transduction Domain-Based Therapy Diabetes, March 1, 2008; 57(3): 757 - 769. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Amsalem, Y. Mardor, M. S. Feinberg, N. Landa, L. Miller, D. Daniels, A. Ocherashvilli, R. Holbova, O. Yosef, I. M. Barbash, et al. Iron-Oxide Labeling and Outcome of Transplanted Mesenchymal Stem Cells in the Infarcted Myocardium Circulation, September 11, 2007; 116(11_suppl): I-38 - I-45. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. C. Murtaugh Pancreas and beta-cell development: from the actual to the possible Development, February 1, 2007; 134(3): 427 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Dutton, N. L. Chillingworth, D. Eberhard, C. R. Brannon, M. A. Hornsey, D. Tosh, and J. M. W. Slack beta cells occur naturally in extrahepatic bile ducts of mice J. Cell Sci., January 15, 2007; 120(2): 239 - 245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
) or control (
)
mice are presented separately for each individual mouse. Hepatic glucagon
content (PDX-1-treated, n = 10; control, n = 10)
(B) and hepatic somatostatin (PDX-1-treated, n = 9; control,
n = 7) content (C) are presented as mean ± S.E. for
mice 2–4 months after recombinant adenovirus administration. Hormonal
levels are determined by radioimmunoassay and normalized to the weight of the
homogenized tissue (*, p < 0.05; **, p < 0.06).
) and rat PDX-1 (
) expression was analyzed at different
time points by LightCycler. The calculation is performed relative to
) as described under
"Materials and Methods." Body weight was monitored once a month
after4hof fasting. n = 6 at every time point analyzed; * indicates
statistical significant increase in body weight upon PDX-1 treatment;
p < 0.05.