JOURNAL
OF SURGICAL
RESEARCH
62, 209-213 (19%)
A Model for Directed Foreign Gene Delivery to Rat Liver Cells in KVO JACEK ROZGA,
M.D., PH.D., ALBERT D. MOSCIONI, PH.D., DANIEL NEUZIL, M.D., AND ACHILLES A. DEMETRIOU, M.D., PH.D.
Departments of Surgery, Vanderbilt University School of Medicine and Department of Veterans Affairs Medical Center, Nashville, Tennessee 37212 Submitted for publication June 26, 1991
MATERIALS A novel technique for directed delivery of retroviral genes to rat liver cells in vivo is described. Vascular isolation of the liver was achieved in situ and perfusate containing retrovirus expressing the bacterial gene conferring resistance to Hygromycin-B was delivered selectively to the posterior liver lobes. After 16 min, normal blood flow to the liver was restored. The portal venous branch supplying the two anterior liver lobes was ligated either at the same time (Group I, n = 4) or 20 hr prior to perfusion (Group II, n = 4) to stimulate DNA synthesis in the posterior lobes. Controls (Group III, n = 4) were perfused with retrovirus without portal branch ligation. Hepatocyte transduction was assessed 7 days later by isolating the cells and assessing their viability in a selection medium. In Group I and II rats, 9.2 + 0.6 and 16.0 rf: l.O%, respectively, of harvested hepatocytes, expressed the Hygromycin-B gene. In contrast, a significantly smaller number of hepatocytes (2.8 + 0.9%, P < 0.003) expressed the gene in the absence of stimulation of DNA synthesis. o ISSZ. Academic Press,
AND METHODS
Adult male Sprague-Dawley rats (250-350 Animals. g) were purchased from Harlan Sprague-Dawley, Inc. (Prattville, AL) and were maintained on standard laboratory chow (Rodent 5001, Ralston Purina, St. Louis, MO) and tap water ad libitum in a 12-hr light/dark cycle. Rats were acclimatized to our laboratory conditions for 1 week prior to use in the experiment. Operations were carried out under ether anesthesia using sterile technique and magnification. All experiments were performed according to protocols approved by the Institutional Animal Experimentation Review Committee. Chemicals. Chemicals used for perfusion and cell isolation, collagenase (type IV), antibiotics, bovine serum albumin, and tissue culture media were purchased from Sigma (St. Louis, MO).
An amphotropic virus containing Viral supernatant. the bacterial “hph” gene encoding for the enzyme hygromycin-B-phosphotransferase which confers resistance to Hygromycin-B (HB) toxicity was used.
Inc.
INTRODUCTION
A large number of inborn errors of liver metabolism are due to a single gene defect. Gene therapy for the correction of defective liver function would require either transplantation of genetically corrected autologous hepatocytes or in uiuo targeted delivery of DNA coding for synthesis of a normal or missing gene product. Recently, several investigators succeeded in delivering a functional gene to hepatocytes in vitro [l-4], but a number of formidable problems remain to be resolved to achieve stable, long-term expression of the gene. These include: improved transduction efficiency, expansion of the population of transduced cells, and development of techniques of retransplantation in the original donor [2, 31.Attempts to deliver genes directly to cells in uiuo have been reported with short-term success [5, 61. Here, we report a novel technique of selective in situ perfusion of the rat liver which was successfully used to carry out retrovirus-mediated transduction of hepatocytes in uiuo.
Operative technique. The following technique was used (Fig. 1): 1. The abdomen was entered through a long midline incision. 2. The ligamentous attachments of the liver were divided. We found the use of an operating microscope helpful at this early stage. 3. Both costal margins were lifted using a U-shaped “retractor” made out of a paper clip. Next, while applying pressure downward on the liver with two fingers, the inferior caval vein was dissected between the liver and the diaphragm using curved microforceps. A 4-O silk (Ethicon) tie was placed around the inferior caval vein and both ends were passed through a 5-cm long polyethylene tubing (PE-50) and secured with a hemostat. A tourniquet was thus created which would be used to occlude the suprahepatic inferior caval vein. 4. The intestine was exteriorized and wrapped in a gauze soaked in warm normal saline. The infrahepatic inferior caval vein was dissected above the right renal
209
0022-4804/92 $1.50 Copyright 0 1992 hy AcademicPress,Inc. All rights of reproduction in any form reserved.
210
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 52, NO. 3, MARCH
1992
nous branch supplying the two anterior liver lobes (tourniquet in controls perfused with the viral supernatant but not subjected to regenerative stimulus), (4) infra-hepatic inferior caval vein (tourniquet), and (5) supra-hepatic inferior caval vein (tourniquet) after injection of 3-5 ml of lactated Ringer solution to wash the posterior liver lobes out of blood. Recirculating perfusion with viral supernatant (1 X lo6 virions/ml, 40 ml) or vehicle alone (controls) was carried out at the rate of 3 ml/min using a roller pump (Masterflex, Barnant Co, Barrington, IL), as shown in Figs. 1 and 2A, for 15 min. Subsequently, the tourniquets and microclips were removed (in reverse order of their placement), inflow and outflow tubings were withdrawn, and the gastroduodenal and right iliac veins were ligated. Normal blood flow to the posterior liver lobes was reestablished (Fig. 2B). Experimental design. Sixteen Sprague-Dawley rats C underwent in situ perfusion of the posterior liver lobes. They were divided into four experimental groups: Group FIG. 1. Diagrammatic depiction of the in uiuo perfusion system (ICV, inferior caval vein; RIV, right iliac vein; HA, hepatic artery; I rats (n = 4) underwent portal branch ligation and perGDV, gastroduodenal vein; PV, portal vein). fusion with “hph” virus at the same time. Group II rats (n = 4) underwent portal branch ligation and perfusion with hph virus 20 hr later. Group III rats (n = 4) were vein and a branch of the phrenic vein where a second perfused with hph, without portal branch ligation. tourniquet was placed and left untied. Group IV rats (n = 4) underwent portal branch ligation 5. The hepatic artery was dissected with great care to followed, 20 hr later, by perfusion with vehicle alone. avoid contraction of the vessel. Subsequently, the gasOne rat from each group was killed at 24 hr with ether troduodenal vein, which is the most proximal tributary overdose, for histological evaluation of the perfused liver of the portal vein, was dissected and tied off distally uslobes. All remaining rats were evaluated 7 days after pering a 6-O silk. A second 6-O silk tie was placed at the orifice of the gastroduodenal vein. The two ligatures fusion by isolating hepatocytes from the perfused liver were then stretched, the lumen of the vein was opened as lobes, culturing them and selecting with Hygromycin-B. Hepatocyte isolation. Under ether anesthesia, the close as possible to the distal ligature, and silastic tubing (Medical-Grade Tubing, Dow Corning, o.d. 0.025 in.) abdomen was entered through a midline incision, and prefilled with heparinized saline was inserted into the the infra-hepatic caval vein was cannulated using a 16gauge catheter (Flash Cath, Travenol, Dearfield, IL) gastroduodenal vein and thereafter into the main trunk of the portal vein. The tubing was secured by tying the which was secured above the right renal vein. The portal proximal ligature. It is important to note that, in spite of vein was then divided and the chest was opened to ligate the relatively large size of the tubing, it is soft and easy to the inferior caval vein. The liver was perfused at the rate manipulate. Cannulation is bloodless and can be accom- of 25 ml/min in a retrograde fashion, first with EDTA solution according to Wang et al. [8] (10 min), and then plished rapidly, without interrupting the portal flow. with calcium-enriched 0.05% collagenase type IV (6-10 6. The liver hilum was exposed and the portal venous branch supplying the two anterior liver lobes was dis- min). The digested liver parenchyma was filtered sected and ligated. This was done to stimulate DNA syn- through a 100 pm mesh and after three washings in Dulthesis (regeneration) in the remaining perfused liver becco’s modified Eagle’s medium, cell viability was delobes [7]. In controls, the portal venous branch was not termined using trypan blue exclusion. ligated. Instead, a tourniquet was used to isolate the anCells were plated on Selection with Hygromycin-B. terior liver lobes during perfusion. 35-mm collagen-coated plates (1 X lo6 cells/plate). Dur7. The right iliac vein was dissected, distally ligated, ing the attachment period, Dulbecco’s modified Eagle’s and punctured with IS-gauge catheter (Flash Cath., medium without glutamine but with 10% fetal bovine Travenol, Dearfield, IL). The site of puncture was se- serum was used. After 3 hr, the medium was changed to cured with a sling and the tip of the catheter was ad- hormonally defined serum-free RPM1 1640 which was vanced into the inferior caval vein above the second then replaced every 24 hr. Cultures were maintained at tourniquet, i.e., in the vicinity of the hepatic veins. 37°C in a humidified atmosphere (5% CO, and 95% air). Perfusion. The vessels were occluded in the follow- After 3 days in culture, cells were selected with Hygroing order: (1) hepatic artery (microclip), (2) portal vein mycin B (LD, = 500 pg/ml) and their viability was below the gastroduodenal vein (microclip), (3) portal ve- assessedafter 4 days, following fixation in methanol and
ROZGA
ET AL.: IN
VZVO GENE
THERAPY
MODEL
FIG. 2. Photograph of an actual perfusion of the posterior liver lobes. On (A), the right liver lobes are blanched during perfusion supernatant. (B) Shows the normal gross appearance of the same liver lobes immediately after revascularization.
211
with \ riral
FIG. 3. Microphotograph of cultured cells harvested from livers perfused with culture medium alone (A) and viral supernatant (B) and subsequently exposed to challenge with Hygromycin B. (A) Shows dead cells in Group IV controls and (B) demonstrates viable, successfully transduced (resistant to Hygromycin B toxicity) cells from Group II animals.
212
JOURNAL
OF SURGICAL
RESEARCH:
Giemsa staining. Percentage viability and transduction were calculated. Data were analyzed statistically using one-way analysis of variance. RESULTS
Liver perfusion for hepatocyte isolation, consistently produced high viability (>85%) cell populations. Vascular isolation of the liver in situ without systemic and splanchnic venous decompression was routinely carried out in approximately 15-16 min. At 24 hr following liver perfusion, there was no histologic evidence of hepatocyte injury by light microscopic morphologic criteria. In Group I, II and IV rats, the nonligated, previously perfused liver lobes appeared normal histologically and numerous cells in mitosis were observed. After transduction with the hph virus, 9.2 +- 0.5% cells in Group I rats survived Hygromycin B treatment. In controls perfused with vehicle alone (Group IV), no resistance to Hygromycin B selection was seen (0.6 f 0.2% viable cells; Fig. 3A). Similar results were obtained in nonligated Group III rats perfused with the hph virus (2.8 + 0.9% viable cells; P < 0.05 compared with Group IV controls). In contrast, significant transduction was achieved in Group II rats in which the nonligated regenerating liver lobes were perfused with the virus at 20 hr, the expected high DNA synthesis. In the livers of these animals, 16.0 + 1.0% hepatocytes were resistant to Hygromycin B (statistical significance was P -c 0.003 when compared to the other groups; Fig. 3B). DISCUSSION
The isolated perfused liver has been used to study hepatic metabolism and as a model for regional cancer chemotherapy [g-13]. We have now utilized selective liver lobe perfusion in uiuo to carry out targeted delivery of foreign genes to hepatocytes. Several animal models of in uiuo liver perfusion have been described [g-16]. Most of them, however, are technically demanding and require oxygenation and/or shunting to prevent splanchnit venous stasis [g-12,14-16]. In contrast, our model is simple, rapid, safe, and with some practice, it is associated with neither perioperative mortality nor damage to the liver. The main reason for the simplicity of this technique is that the perfusion period is kept short (not longer than 20 min). Apart from technical details, another major difference with other methods [lo, 11, 16, 171 is the ability in the described model to carry out selective isolation and perfusion of the posterior liver lobes followed by reperfusion and establishment of normal blood flow. Use of portal branch ligation to induce DNA synthesis was found to be less invasive and equally effective as partial hepatectomy [7]. Hepatocyte cultures prepared from rats infused with an amphotropic retrovirus carrying the hph gene, dem-
VOL.
52, NO. 3, MARCH
1992
onstrated transduction of a significant number of hepatocytes. It is important to note that livers perfused with the vector at the time close to expected peak DNA synthesis (20 hr) yielded hepatocyte cultures with significantly higher resistance to Hygromycin B toxicity when compared to those transduced at the time portal branch ligation was carried out. In the absence of portal branch ligation, almost no resistance to Hygromycin B was seen (~1%). We, therefore, demonstrated that anterior liver lobes ligation is an effective technique of enhancing perfused liver cell transduction, presumably by stimulating DNA synthesis. This is in agreement with our recent experience with retroviral transduction in partially hepatectomized rats in uiuo [ 181 as well as with the results of in vitro studies by Miller et al. [ 191,which suggest that active DNA synthesis is needed for efficient uptake, integration and expression of retroviral DNA in hepatocytes. Use of selective in situ perfusion rather than direct intraportal infusion of viral particles was carried out because in pilot experiments, using this virus, we were unable to obtain hepatocyte infection following direct infusion of virus particles into the portal vein. This is probably due to inadequate uptake of virus particles by the hepatocytes following a single passage through the liver. Once through the liver, the virus can then be taken up by cells in other tissues. It would be possible, however, theoretically to achieve hepatocyte infection by direct infusion of virus into the portal vein by infusing very large number of virions n a constant infusion over a long period of time. Infusion of such large numbers of virus particles in the whole body may be detrimental. We believe that for future clinical therapy, it would be advantageous to avoid the need for partial hepatectomy to induce enhanced hepatocyte DNA synthesis. Selective (lobar) occlusion, either by ligation of portal venous branches or insertion of balloon-tipped catheters, can be achieved in humans (or large animals) to effectively induce enhanced DNA synthesis. In addition, growth factor infusion through catheters in portal vein branches could be carried out alone or in combination with occlusion. We therefore believe that the technique of portal vein ligation as a means of stimulating hepatocyte DNA synthesis, has greater clinical relevance than partial hepatectomy. In conclusion, we developed a novel method of selective liver lobe perfusion combined with portal vein branch ligation which results in successful retrovirusmediated transduction of adult rat hepatocytes in uiuo. Isolated organ perfusion could be adapted for use in large animals or humans and, therefore, it may be a potentially useful tool for treating specific inborn errors of hepatic metabolism. ACKNOWLEDGMENT We would like to express our appreciation to Immunex Co., Seattle, Washington, for providing us with the viral probes and for valuable technical advise.
ROZGA ET AL.: IN VZVO GENE
REFERENCES
2.
3.
4. 5.
6.
7.
8.
9.
213
MODEL
10.
Mulcare, R. J., Solis, A., and Fortner, J. G. Isolation and perfusion of the liver for cancer chemotherapy. J. Surg. Res. 15: 87, 1973.
11.
Ryan, C. J., Ryan, S., Wood, C. B., and Blumgart, L. H. Technique for in vivo perfusion of chemotherapeutic agents into the rat liver. J. Surg. Res. 26: 142, 1979.
12.
Boddie, A. W., Booker, L., Mullins, J. D., Buckley, C. J., and Bride, C. M. Hepatic hyperthermia by total isolation and regional perfusion in vivo. J. Surg. Res. 26: 447, 1979.
13.
Stone, R. T., Jabour, A., Wilson, S. E., and Rangel, D. M. Uptake of 5-fluorouracil during isolated perfusion of the canine liver. J. Surg. Res. 13: 347, 1980.
14.
Skibba, J. L., Almagro, U. A., Condon, R. E., and Petroff, R. J. A technique for isolation of the canine liver with survival. J. Surg. Res. 34: 123, 1983.
15.
Sindelar, W. F. Isolation-perfusion cil. Ann. Surg. 201: 337, 1984.
16.
Radnell, M., Jeppsson, isolated liver perfusion cytotoxic drug perfusion 394, 1990.
17.
Skibba, J. L. and Collins, F. G. Effect oftemperature on biochemical functions in the isolated perfused rat liver. J. Surg. Res. 24: 435, 1978.
18.
Moscioni, A. D., Rozga, J., Egli, E., Neuzil, D. F., Weisser, K. E., Overell, R. W., Holt, J. T., and Demetriou, A. A. In vivo retrovirally mediated foreign gene delivery to rat hepatocytes. Gastroenterology 100: 777, 1991. [Abstract]
19.
Miller, D. G., Adam, M. A., and Miller, A. D. Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol. Cell. Biol. 10: 4239, 1990.
1. Wilson,
J. M., Jefferson, D. M., Chowdhury, J. R., Novikoff, P. M., Johnston, D. E., and Mulligan, R. C. Retrovirus-mediated transduction of adult hepatocytes. Proc. Natl. Acad. Ski. USA 85: 3014,1988. Wilson, J. M., Johnston, D. E., Jefferson, D. M., and Mulligan, R. C. Correction of the genetic defect in hepatocytes from the Watanabe heritable hyperlipidemic rabbit. Proc. Natl. Acad. Sci. USA 85: 4421, 1988. Anderson, K. D., Thompson, J. A., DiPietro, J. M., Montgomery, K. T., Reid, J. M., and Anderson, W. F. Gene expression in implanted rat hepatocytes following retroviral-mediated gene transfer. Somat. Cell. Mol. Genet. 15: 215, 1989. Pasco D. S., Fagan J. B. Efficient DNA-mediated gene transfer into primary cultures of adult rat hepatocytes. DNA 8: 535,1989. Wu, C. H., Wilson, J. M., and Wu, G. Y. Targeting genes: Delivery and persistent expression of a foreign gene driven by mammalian regulatory elements in vivo. J. Biol. Chem. 264: 16,985, 1989. Wolff, J. A., Malone, R. W., Williams, P., Chong, W., Acsadi, G., Jani, A., and Felgner, P. L. Direct gene transfer into mouse muscle in vivo. Science 247: 1465, 1990. Rozga, J., Jeppsson, B., and Bengmark, S. Portal branch ligation in the rat. Re-evaluation of a model. Am. J. Pathol. 125: 300, 1986. Wang, S., Renaud, G., Infante, J., Catala, F., and Infante, R. Isolation of rat hepatocytes with EDTA and their metabolic functions in primary culture. In Vitro Cell. Dev. Biol. 21: 526, 1985. Hems, R., Ross, B. D., Berry, M. N., and Krebs, H. A. Gluconeogenesis in the perfused rat liver. Biochem. J. 101: 284, 1966.
THERAPY
of the liver with 5-fluoroura-
B., and Bengmark, S. A technique for in the rat with survival and results of on liver tumor growth. J. Surg. Res. 49:
OF SURGICAL
RESEARCH
62, 209-213 (19%)
A Model for Directed Foreign Gene Delivery to Rat Liver Cells in KVO JACEK ROZGA,
M.D., PH.D., ALBERT D. MOSCIONI, PH.D., DANIEL NEUZIL, M.D., AND ACHILLES A. DEMETRIOU, M.D., PH.D.
Departments of Surgery, Vanderbilt University School of Medicine and Department of Veterans Affairs Medical Center, Nashville, Tennessee 37212 Submitted for publication June 26, 1991
MATERIALS A novel technique for directed delivery of retroviral genes to rat liver cells in vivo is described. Vascular isolation of the liver was achieved in situ and perfusate containing retrovirus expressing the bacterial gene conferring resistance to Hygromycin-B was delivered selectively to the posterior liver lobes. After 16 min, normal blood flow to the liver was restored. The portal venous branch supplying the two anterior liver lobes was ligated either at the same time (Group I, n = 4) or 20 hr prior to perfusion (Group II, n = 4) to stimulate DNA synthesis in the posterior lobes. Controls (Group III, n = 4) were perfused with retrovirus without portal branch ligation. Hepatocyte transduction was assessed 7 days later by isolating the cells and assessing their viability in a selection medium. In Group I and II rats, 9.2 + 0.6 and 16.0 rf: l.O%, respectively, of harvested hepatocytes, expressed the Hygromycin-B gene. In contrast, a significantly smaller number of hepatocytes (2.8 + 0.9%, P < 0.003) expressed the gene in the absence of stimulation of DNA synthesis. o ISSZ. Academic Press,
AND METHODS
Adult male Sprague-Dawley rats (250-350 Animals. g) were purchased from Harlan Sprague-Dawley, Inc. (Prattville, AL) and were maintained on standard laboratory chow (Rodent 5001, Ralston Purina, St. Louis, MO) and tap water ad libitum in a 12-hr light/dark cycle. Rats were acclimatized to our laboratory conditions for 1 week prior to use in the experiment. Operations were carried out under ether anesthesia using sterile technique and magnification. All experiments were performed according to protocols approved by the Institutional Animal Experimentation Review Committee. Chemicals. Chemicals used for perfusion and cell isolation, collagenase (type IV), antibiotics, bovine serum albumin, and tissue culture media were purchased from Sigma (St. Louis, MO).
An amphotropic virus containing Viral supernatant. the bacterial “hph” gene encoding for the enzyme hygromycin-B-phosphotransferase which confers resistance to Hygromycin-B (HB) toxicity was used.
Inc.
INTRODUCTION
A large number of inborn errors of liver metabolism are due to a single gene defect. Gene therapy for the correction of defective liver function would require either transplantation of genetically corrected autologous hepatocytes or in uiuo targeted delivery of DNA coding for synthesis of a normal or missing gene product. Recently, several investigators succeeded in delivering a functional gene to hepatocytes in vitro [l-4], but a number of formidable problems remain to be resolved to achieve stable, long-term expression of the gene. These include: improved transduction efficiency, expansion of the population of transduced cells, and development of techniques of retransplantation in the original donor [2, 31.Attempts to deliver genes directly to cells in uiuo have been reported with short-term success [5, 61. Here, we report a novel technique of selective in situ perfusion of the rat liver which was successfully used to carry out retrovirus-mediated transduction of hepatocytes in uiuo.
Operative technique. The following technique was used (Fig. 1): 1. The abdomen was entered through a long midline incision. 2. The ligamentous attachments of the liver were divided. We found the use of an operating microscope helpful at this early stage. 3. Both costal margins were lifted using a U-shaped “retractor” made out of a paper clip. Next, while applying pressure downward on the liver with two fingers, the inferior caval vein was dissected between the liver and the diaphragm using curved microforceps. A 4-O silk (Ethicon) tie was placed around the inferior caval vein and both ends were passed through a 5-cm long polyethylene tubing (PE-50) and secured with a hemostat. A tourniquet was thus created which would be used to occlude the suprahepatic inferior caval vein. 4. The intestine was exteriorized and wrapped in a gauze soaked in warm normal saline. The infrahepatic inferior caval vein was dissected above the right renal
209
0022-4804/92 $1.50 Copyright 0 1992 hy AcademicPress,Inc. All rights of reproduction in any form reserved.
210
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 52, NO. 3, MARCH
1992
nous branch supplying the two anterior liver lobes (tourniquet in controls perfused with the viral supernatant but not subjected to regenerative stimulus), (4) infra-hepatic inferior caval vein (tourniquet), and (5) supra-hepatic inferior caval vein (tourniquet) after injection of 3-5 ml of lactated Ringer solution to wash the posterior liver lobes out of blood. Recirculating perfusion with viral supernatant (1 X lo6 virions/ml, 40 ml) or vehicle alone (controls) was carried out at the rate of 3 ml/min using a roller pump (Masterflex, Barnant Co, Barrington, IL), as shown in Figs. 1 and 2A, for 15 min. Subsequently, the tourniquets and microclips were removed (in reverse order of their placement), inflow and outflow tubings were withdrawn, and the gastroduodenal and right iliac veins were ligated. Normal blood flow to the posterior liver lobes was reestablished (Fig. 2B). Experimental design. Sixteen Sprague-Dawley rats C underwent in situ perfusion of the posterior liver lobes. They were divided into four experimental groups: Group FIG. 1. Diagrammatic depiction of the in uiuo perfusion system (ICV, inferior caval vein; RIV, right iliac vein; HA, hepatic artery; I rats (n = 4) underwent portal branch ligation and perGDV, gastroduodenal vein; PV, portal vein). fusion with “hph” virus at the same time. Group II rats (n = 4) underwent portal branch ligation and perfusion with hph virus 20 hr later. Group III rats (n = 4) were vein and a branch of the phrenic vein where a second perfused with hph, without portal branch ligation. tourniquet was placed and left untied. Group IV rats (n = 4) underwent portal branch ligation 5. The hepatic artery was dissected with great care to followed, 20 hr later, by perfusion with vehicle alone. avoid contraction of the vessel. Subsequently, the gasOne rat from each group was killed at 24 hr with ether troduodenal vein, which is the most proximal tributary overdose, for histological evaluation of the perfused liver of the portal vein, was dissected and tied off distally uslobes. All remaining rats were evaluated 7 days after pering a 6-O silk. A second 6-O silk tie was placed at the orifice of the gastroduodenal vein. The two ligatures fusion by isolating hepatocytes from the perfused liver were then stretched, the lumen of the vein was opened as lobes, culturing them and selecting with Hygromycin-B. Hepatocyte isolation. Under ether anesthesia, the close as possible to the distal ligature, and silastic tubing (Medical-Grade Tubing, Dow Corning, o.d. 0.025 in.) abdomen was entered through a midline incision, and prefilled with heparinized saline was inserted into the the infra-hepatic caval vein was cannulated using a 16gauge catheter (Flash Cath, Travenol, Dearfield, IL) gastroduodenal vein and thereafter into the main trunk of the portal vein. The tubing was secured by tying the which was secured above the right renal vein. The portal proximal ligature. It is important to note that, in spite of vein was then divided and the chest was opened to ligate the relatively large size of the tubing, it is soft and easy to the inferior caval vein. The liver was perfused at the rate manipulate. Cannulation is bloodless and can be accom- of 25 ml/min in a retrograde fashion, first with EDTA solution according to Wang et al. [8] (10 min), and then plished rapidly, without interrupting the portal flow. with calcium-enriched 0.05% collagenase type IV (6-10 6. The liver hilum was exposed and the portal venous branch supplying the two anterior liver lobes was dis- min). The digested liver parenchyma was filtered sected and ligated. This was done to stimulate DNA syn- through a 100 pm mesh and after three washings in Dulthesis (regeneration) in the remaining perfused liver becco’s modified Eagle’s medium, cell viability was delobes [7]. In controls, the portal venous branch was not termined using trypan blue exclusion. ligated. Instead, a tourniquet was used to isolate the anCells were plated on Selection with Hygromycin-B. terior liver lobes during perfusion. 35-mm collagen-coated plates (1 X lo6 cells/plate). Dur7. The right iliac vein was dissected, distally ligated, ing the attachment period, Dulbecco’s modified Eagle’s and punctured with IS-gauge catheter (Flash Cath., medium without glutamine but with 10% fetal bovine Travenol, Dearfield, IL). The site of puncture was se- serum was used. After 3 hr, the medium was changed to cured with a sling and the tip of the catheter was ad- hormonally defined serum-free RPM1 1640 which was vanced into the inferior caval vein above the second then replaced every 24 hr. Cultures were maintained at tourniquet, i.e., in the vicinity of the hepatic veins. 37°C in a humidified atmosphere (5% CO, and 95% air). Perfusion. The vessels were occluded in the follow- After 3 days in culture, cells were selected with Hygroing order: (1) hepatic artery (microclip), (2) portal vein mycin B (LD, = 500 pg/ml) and their viability was below the gastroduodenal vein (microclip), (3) portal ve- assessedafter 4 days, following fixation in methanol and
ROZGA
ET AL.: IN
VZVO GENE
THERAPY
MODEL
FIG. 2. Photograph of an actual perfusion of the posterior liver lobes. On (A), the right liver lobes are blanched during perfusion supernatant. (B) Shows the normal gross appearance of the same liver lobes immediately after revascularization.
211
with \ riral
FIG. 3. Microphotograph of cultured cells harvested from livers perfused with culture medium alone (A) and viral supernatant (B) and subsequently exposed to challenge with Hygromycin B. (A) Shows dead cells in Group IV controls and (B) demonstrates viable, successfully transduced (resistant to Hygromycin B toxicity) cells from Group II animals.
212
JOURNAL
OF SURGICAL
RESEARCH:
Giemsa staining. Percentage viability and transduction were calculated. Data were analyzed statistically using one-way analysis of variance. RESULTS
Liver perfusion for hepatocyte isolation, consistently produced high viability (>85%) cell populations. Vascular isolation of the liver in situ without systemic and splanchnic venous decompression was routinely carried out in approximately 15-16 min. At 24 hr following liver perfusion, there was no histologic evidence of hepatocyte injury by light microscopic morphologic criteria. In Group I, II and IV rats, the nonligated, previously perfused liver lobes appeared normal histologically and numerous cells in mitosis were observed. After transduction with the hph virus, 9.2 +- 0.5% cells in Group I rats survived Hygromycin B treatment. In controls perfused with vehicle alone (Group IV), no resistance to Hygromycin B selection was seen (0.6 f 0.2% viable cells; Fig. 3A). Similar results were obtained in nonligated Group III rats perfused with the hph virus (2.8 + 0.9% viable cells; P < 0.05 compared with Group IV controls). In contrast, significant transduction was achieved in Group II rats in which the nonligated regenerating liver lobes were perfused with the virus at 20 hr, the expected high DNA synthesis. In the livers of these animals, 16.0 + 1.0% hepatocytes were resistant to Hygromycin B (statistical significance was P -c 0.003 when compared to the other groups; Fig. 3B). DISCUSSION
The isolated perfused liver has been used to study hepatic metabolism and as a model for regional cancer chemotherapy [g-13]. We have now utilized selective liver lobe perfusion in uiuo to carry out targeted delivery of foreign genes to hepatocytes. Several animal models of in uiuo liver perfusion have been described [g-16]. Most of them, however, are technically demanding and require oxygenation and/or shunting to prevent splanchnit venous stasis [g-12,14-16]. In contrast, our model is simple, rapid, safe, and with some practice, it is associated with neither perioperative mortality nor damage to the liver. The main reason for the simplicity of this technique is that the perfusion period is kept short (not longer than 20 min). Apart from technical details, another major difference with other methods [lo, 11, 16, 171 is the ability in the described model to carry out selective isolation and perfusion of the posterior liver lobes followed by reperfusion and establishment of normal blood flow. Use of portal branch ligation to induce DNA synthesis was found to be less invasive and equally effective as partial hepatectomy [7]. Hepatocyte cultures prepared from rats infused with an amphotropic retrovirus carrying the hph gene, dem-
VOL.
52, NO. 3, MARCH
1992
onstrated transduction of a significant number of hepatocytes. It is important to note that livers perfused with the vector at the time close to expected peak DNA synthesis (20 hr) yielded hepatocyte cultures with significantly higher resistance to Hygromycin B toxicity when compared to those transduced at the time portal branch ligation was carried out. In the absence of portal branch ligation, almost no resistance to Hygromycin B was seen (~1%). We, therefore, demonstrated that anterior liver lobes ligation is an effective technique of enhancing perfused liver cell transduction, presumably by stimulating DNA synthesis. This is in agreement with our recent experience with retroviral transduction in partially hepatectomized rats in uiuo [ 181 as well as with the results of in vitro studies by Miller et al. [ 191,which suggest that active DNA synthesis is needed for efficient uptake, integration and expression of retroviral DNA in hepatocytes. Use of selective in situ perfusion rather than direct intraportal infusion of viral particles was carried out because in pilot experiments, using this virus, we were unable to obtain hepatocyte infection following direct infusion of virus particles into the portal vein. This is probably due to inadequate uptake of virus particles by the hepatocytes following a single passage through the liver. Once through the liver, the virus can then be taken up by cells in other tissues. It would be possible, however, theoretically to achieve hepatocyte infection by direct infusion of virus into the portal vein by infusing very large number of virions n a constant infusion over a long period of time. Infusion of such large numbers of virus particles in the whole body may be detrimental. We believe that for future clinical therapy, it would be advantageous to avoid the need for partial hepatectomy to induce enhanced hepatocyte DNA synthesis. Selective (lobar) occlusion, either by ligation of portal venous branches or insertion of balloon-tipped catheters, can be achieved in humans (or large animals) to effectively induce enhanced DNA synthesis. In addition, growth factor infusion through catheters in portal vein branches could be carried out alone or in combination with occlusion. We therefore believe that the technique of portal vein ligation as a means of stimulating hepatocyte DNA synthesis, has greater clinical relevance than partial hepatectomy. In conclusion, we developed a novel method of selective liver lobe perfusion combined with portal vein branch ligation which results in successful retrovirusmediated transduction of adult rat hepatocytes in uiuo. Isolated organ perfusion could be adapted for use in large animals or humans and, therefore, it may be a potentially useful tool for treating specific inborn errors of hepatic metabolism. ACKNOWLEDGMENT We would like to express our appreciation to Immunex Co., Seattle, Washington, for providing us with the viral probes and for valuable technical advise.
ROZGA ET AL.: IN VZVO GENE
REFERENCES
2.
3.
4. 5.
6.
7.
8.
9.
213
MODEL
10.
Mulcare, R. J., Solis, A., and Fortner, J. G. Isolation and perfusion of the liver for cancer chemotherapy. J. Surg. Res. 15: 87, 1973.
11.
Ryan, C. J., Ryan, S., Wood, C. B., and Blumgart, L. H. Technique for in vivo perfusion of chemotherapeutic agents into the rat liver. J. Surg. Res. 26: 142, 1979.
12.
Boddie, A. W., Booker, L., Mullins, J. D., Buckley, C. J., and Bride, C. M. Hepatic hyperthermia by total isolation and regional perfusion in vivo. J. Surg. Res. 26: 447, 1979.
13.
Stone, R. T., Jabour, A., Wilson, S. E., and Rangel, D. M. Uptake of 5-fluorouracil during isolated perfusion of the canine liver. J. Surg. Res. 13: 347, 1980.
14.
Skibba, J. L., Almagro, U. A., Condon, R. E., and Petroff, R. J. A technique for isolation of the canine liver with survival. J. Surg. Res. 34: 123, 1983.
15.
Sindelar, W. F. Isolation-perfusion cil. Ann. Surg. 201: 337, 1984.
16.
Radnell, M., Jeppsson, isolated liver perfusion cytotoxic drug perfusion 394, 1990.
17.
Skibba, J. L. and Collins, F. G. Effect oftemperature on biochemical functions in the isolated perfused rat liver. J. Surg. Res. 24: 435, 1978.
18.
Moscioni, A. D., Rozga, J., Egli, E., Neuzil, D. F., Weisser, K. E., Overell, R. W., Holt, J. T., and Demetriou, A. A. In vivo retrovirally mediated foreign gene delivery to rat hepatocytes. Gastroenterology 100: 777, 1991. [Abstract]
19.
Miller, D. G., Adam, M. A., and Miller, A. D. Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol. Cell. Biol. 10: 4239, 1990.
1. Wilson,
J. M., Jefferson, D. M., Chowdhury, J. R., Novikoff, P. M., Johnston, D. E., and Mulligan, R. C. Retrovirus-mediated transduction of adult hepatocytes. Proc. Natl. Acad. Ski. USA 85: 3014,1988. Wilson, J. M., Johnston, D. E., Jefferson, D. M., and Mulligan, R. C. Correction of the genetic defect in hepatocytes from the Watanabe heritable hyperlipidemic rabbit. Proc. Natl. Acad. Sci. USA 85: 4421, 1988. Anderson, K. D., Thompson, J. A., DiPietro, J. M., Montgomery, K. T., Reid, J. M., and Anderson, W. F. Gene expression in implanted rat hepatocytes following retroviral-mediated gene transfer. Somat. Cell. Mol. Genet. 15: 215, 1989. Pasco D. S., Fagan J. B. Efficient DNA-mediated gene transfer into primary cultures of adult rat hepatocytes. DNA 8: 535,1989. Wu, C. H., Wilson, J. M., and Wu, G. Y. Targeting genes: Delivery and persistent expression of a foreign gene driven by mammalian regulatory elements in vivo. J. Biol. Chem. 264: 16,985, 1989. Wolff, J. A., Malone, R. W., Williams, P., Chong, W., Acsadi, G., Jani, A., and Felgner, P. L. Direct gene transfer into mouse muscle in vivo. Science 247: 1465, 1990. Rozga, J., Jeppsson, B., and Bengmark, S. Portal branch ligation in the rat. Re-evaluation of a model. Am. J. Pathol. 125: 300, 1986. Wang, S., Renaud, G., Infante, J., Catala, F., and Infante, R. Isolation of rat hepatocytes with EDTA and their metabolic functions in primary culture. In Vitro Cell. Dev. Biol. 21: 526, 1985. Hems, R., Ross, B. D., Berry, M. N., and Krebs, H. A. Gluconeogenesis in the perfused rat liver. Biochem. J. 101: 284, 1966.
THERAPY
of the liver with 5-fluoroura-
B., and Bengmark, S. A technique for in the rat with survival and results of on liver tumor growth. J. Surg. Res. 49:
