Adenovirus gene transfer vector toxicity to mouse embryos: implications for human IVF

Jon W. Gordon1

Department of Obstetrics/Gynecology, Mt Sinai Medical Center, New York, USA


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: The promulgation and diversification of micromanipulation procedures which open the zona pellucida of the oocyte or early embryo is steadily increasing the chance that zygotes will encounter infectious viral agents or gene transfer vectors derived from these agents. Such interactions could lead to toxic effects on the embryo or to insertion of foreign genes into the germ line. Adenovirus is a ubiquitous human viral pathogen that is commonly used as a gene therapy vector. However, the toxicity of this virus or its vector derivatives to embryos has not been extensively investigated. METHODS: In the present study, a mouse model was used to investigate the time of appearance of embryo toxicity, the manifestations of that toxicity, and the mechanism by which adenoviruses exert toxicity. The effects of exposure to adenovirus on in-vivo embryo development was also examined. RESULTS: These vectors exerted no deleterious effects until after the 2-cell stage, where they caused developmental delay and disorganized cleavage. Toxicity was associated with expression of a lacZ reporter gene cloned into the vectors, and with the proportion of infectious particles within the virus preparations: preparations with a high proportion of `empty capsids', including a preparation of wild-type virus, were less toxic than a replication-defective vector with a high proportion of functional, infectious particles. Subzonal insertion of adenovirus vector followed by embryo transfer led to a dramatic reduction in the number of embryos which developed to term—findings which establish the physiological significance of the findings from in-vitro culture. CONCLUSIONS: These results indicate that adenoviruses and their vector derivatives are not likely to insert genetic material into the germ line, but they may pose a significant threat to the viability of human embryos undergoing opening of the zona pellucida during IVF.

Key words: adenovirus/mouse/preimplantation development/toxicity


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The enormous efficacy of IVF for treatment of infertility has been further augmented by a variety of procedures that employ direct micromanipulation of gametes and embryos. A clinical study which demonstrated that zona opening could allow fertilization by sperm from infertile men (Gordon et al., 1988Go) was quickly followed by successful clinical use of partial zona dissection (Malter and Cohen, 1989Go) subzonal sperm insertion (Fishel et al., 1991Go) and ICSI (Palermo et al., 1992Go). The zona pellucida has been opened in an effort to improve hatching (Cohen et al., 1992Go), and embryo biopsy, in conjunction with ICSI, had been employed for preimplantation genetic diagnosis (Handyside et al., 1992Go). More recently, ooplasmic transplantation has been explored as an approach to improving oocyte quality (Cohen et al., 1997Go). A common feature of all micromanipulation procedures is that the zona pellucida, which normally protects the embryo from infectious agents, is breached.

Because the zona pellucida presents a formidable barrier to infection, these micromanipulation procedures create the opportunity for infectious agents, particularly viruses, to access the zygote. Infection could have toxic effects or lead to transfer of viral genetic material into the germ line. This risk increases as viruses are modified to serve as gene therapy vectors, which may be administered in very high quantities to tissues that are not normally targets for infection by the wild-type virus. Retroviruses and their vector derivatives have already been shown to infect mouse embryos (Jaenisch, 1976Go; van der Putten et al., 1985Go) bovine or primate mature oocytes (Chan et al., 1998Go, 2000Go) or mouse spermatogonia (Nagano et al., 2001Go) and insert their proviral DNA genomes into the host chromosomal material.

A class of gene therapy vector that is very widely used today is derived from adenovirus. Although these ubiquitous DNA viruses normally infect the upper respiratory tract, they are capable of infecting a wide variety of cell types (Yeh and Perricaudet, 1997Go). These vectors have cytotoxic effects related to their replication (Shenk, 1996Go) and can also rarely insert their genetic material into the host chromosomal DNA (Harui et al., 1999Go). These biological features of adenovirus vectors, as well as their increasing use in gene therapy, make it important to characterize their effects on early embryos, both with respect to toxicity and germ line gene transfer potential. Adenovirus vectors encoding the bacterial ß-galactosidase (lacZ) gene have been injected under the zonae pellucidae of mouse, rat and cow embryos, and been observed to reduce development to the blastocyst stage (Kubisch et al., 1997Go). However, the stage of development at which toxicity first appears, the manifestations of this toxicity, and the mechanism of toxicity have not been examined. Moreover, the potential of vector exposure to interfere with development in vivo has never been explored.

In the present study, mouse embryos were used to examine more extensively the effects of embryo exposure to adenoviruses and related gene transfer vectors. The timing of appearance of toxic effects, the manifestations of toxicity, and the mechanism of toxicity were evaluated. Embryo transfers were also performed after subzonal insertion of these vectors to determine if fetal development is impaired. The results indicate that the toxicity of these vectors could reduce pregnancy rate, but would not likely lead to germ line gene insertion.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Vectors
The following adenovirus vectors were obtained from the vector core facility, Institute of Gene therapy and Molecular Medicine, Mt Sinai School of Medicine. An adenovirus ß-galactosidase (Adß-gal) vector used for embryo exposure was replication defective, with deletions of the E1 and E3 early genes, and carried the bacterial ß-galactosidase (lacZ) gene driven by the Rous sarcoma virus (RSV) early promoter (Hall et al., 2000Go). This vector had a titre of 7x1012 particles/ml (p/ml), 3x1011 infectious particles/ml (PFU/ml). A `null' vector was also deleted for the E1 and E3 genes, but contained no added reporter genes; this had a titre of 2.5x1012 p/ml, 2.25x1011 PFU/ml. A preparation of wild-type adenovirus was also used; this had a titre of 4x1012 p/ml, 2x1011 PFU/ml. An additional Adß-gal vector was obtained from the Vector Core of the Institute for Human Gene Therapy of the University of Pennsylvania Health System. In this vector, the lacZ gene was driven by the human cytomegalovirus (CMV) early promoter; viral titres of the preparation were 5.2x1012 p/ml, 5x1010 PFU/ml. These vectors were diluted in KSOM (Erbeck et al. 1994Go) mouse embryo culture medium (Specialty media, Phillipsburg, NJ, USA; #MR-023-D), for exposure to embryos.

Production of mouse embryos
FVB/N and BDF1 mice were obtained from Taconic Farms (Germantown, NY, USA). CD-1 females were obtained from Charles River Laboratories (Wilmington, MA, USA). Animals were maintained under conventional conditions in a 14:10 h light:dark cycle. Immature (4- to 5-week-old) FVB/N females were superovulated by intraperitoneal injection of 5 IU of pregnant mares' serum (PMS; Sigma, St Louis, MO, USA; # G-4877) followed 48 h later by 5 IU of hCG (Sigma; #CG-10), and then placed with FVB/N males to mate. On the following morning, mating was documented by the presence of a vaginal plug, and mated animals were killed by cervical dislocation. Fertilized eggs were then incubated at 37°C for 5 min with 2 mg/ml hyaluronidase (Sigma; # H-3884) in KSOM medium for removal of cumulus cells, washed three times in KSOM, and incubated with adenovirus vectors.

Exposure of embryos to adenovirus vectors
For all experiments wherein embryos were exposed to vectors, the zona pellucida was first removed by brief exposure to acid Tyrode's solution as previously described (Talansky et al., 1987Go). Two different incubation protocols were used for exposure of embryos to adenoviruses and adenovirus-derived vectors. In one set of experiments, embryos were placed in 1x107 PFU/ml of the vectors at the 1-cell stage and cultured continuously in the vector to observe the stages of development at which toxicity became manifest. In all other experiments, 2-cell embryos were exposed to varying amounts of vector for 2 h at 37°C with continuous shaking. Culture medium used for these 2 h incubations consisted of KSOM supplemented with bovine serum albumin (BSA; Sigma #A-3311) to a final concentration of 4 mg/ml. The BSA was added and the embryos were maintained on a shaking platform for the 2 h incubation period in order to reduce adherence of embryos to the culture dish surface, which was greatly increased by the presence of adenovirus in the culture medium. Once adhered to the surface, the embryos could not be removed for washing without inducing lysis. After 2 h of incubation, the embryos were washed in three changes of KSOM and cultured in microdrops of KSOM under mineral oil (Squibb) which was equilibrated at 10% (v/v) with Earle's Balanced Salt Solution (EBSS; Sigma, #E-2888). With each wash, the transfer pipette was discarded so as to eliminate the possibility of contaminating embryos with residual adenovirus associated with the pipette's inside surface.

Control embryos were treated identically to those exposed to virus, with zona removal and 2 h of exposure to high-BSA medium on a shaking platform. These embryos were then cultured individually or in pairs in microdrops and evaluated for development to the blastocyst stage.

LacZ staining of embryos
This was performed as reported previously (Hall et al., 2000Go; Gordon, 2001Go). Briefly, embryos were pooled in 10 µl microdrops of KSOM under mineral oil. Excess medium was drained with a mouth pipette and replaced with 1.25% glutaraldehyde in phosphate-buffered saline. The draining and replacement procedure was repeated twice more, after which the process was reversed in order to replace the glutaraldehyde with KSOM. Embryos were then collected and placed in organ culture dishes which contained 1 ml of LacZ staining solution prepared as follows: 5.5 ml H2O, 10 µl 1 mol/l MgCl2, 28 µl 4 mol/l NaCl, 333 µl 1 mol/l HEPES pH 7.3, 750 µl 0.03 mol/l potassium ferrocyanide, 750 µl 0.03 mol/l potassium ferricyanide, 7 µl saturated NaOH. To 1 ml of this mixture was added 26 µl of X-gal solution (20 mg/ml dissolved in dimethyl formamide). This staining solution was prepared freshly immediately before each test, and the ferrocyanide and ferricyanide solutions were prepared freshly every 3 days. To minimize evaporation during staining, the sidewell of the organ culture dish was filled with 3 ml of distilled H2O. Embryos were stained overnight at 37°C in a standard tissue culture incubator. Staining of experimental and positive controls was performed simultaneously using the same preparation of staining solution.

Subzonal insertion and embryo transfer
To test whether adenovirus exposure impaired embryo development, embryos were recovered at the 1-cell stage, cultured to the 2-cell stage, and subjected to subzonal insertion of the null adenovirus vector, which was inserted at a concentration of 1x109 PFU/ml. On the day of embryo recovery, mature CD-1 females were mated to vasectomized males and examined the following morning for vaginal plugs. For subzonal insertion, holding pipettes were pulled by hand from vaccine capillary tubing (Mercer Glass Works, Brooklyn, NY, USA; No. MX-999) and finished on a DeFonbrune microforge. Microinjection pipettes were produced from filament-containing glass capillary tubing (World Precision Instruments, Sarasota, FL, USA; No. TW100F-4) and pulled on a vertical pipette puller (David Kopf Instruments model 700C). The microneedle tubing was filled from the base.

Subzonal insertion (SUZI) was performed on a Leitz microscope using Leitz micromanipulators. Two-celled embryos were held in place and the microneedle was inserted into the perivitelline space, after which virus suspension was expelled until swelling of the perivitelline space was readily visible. About 100 pl of virus suspension was inserted. These manipulations were performed in HEPES-buffered M2 medium (Quinn et al., 1982Go) (Sigma No. M-7167). As a control for the effects of the manipulation procedure, embryos not exposed to adenovirus were subjected to subzonal insertion using medium alone.

Immediately after SUZI, embryos exposed to the vector and control embryos were transferred together into the oviducts of pseudopregnant CD-1 females using previously reported methods (Gordon et al., 1980Go). This procedure provided an internal control for differences in the ability of the foster females to become pregnant. To distinguish the exposed from the unexposed embryos, coat colour markers were used as follows: FVB/NxFVB/N (albino) embryos were subjected to SUZI with the vector and co-transferred with FVB/NxBDF1 (black Agouti) controls. In order to control for possible differences in developmental success between the albino and Agouti embryos, the reverse experiment was also performed, such that the embryos with the Agouti gene received the vector and were co-transferred with albino controls. Pseudopregnant females were implanted with 30 embryos (15 virus-exposed and 15 controls) and were allowed to deliver litters. Coat colour markers were then used to determine the percentage of virus-exposed embryos born as compared with controls.

Statistical analyses
For statistical analysis of successful preimplantation development, groups of experimental and control embryos were scored against each other using a 2x2 contingency table and the chi-square test. Sample sizes were sufficient to eliminate the need for a Yates correction. Differences were considered statistically significant at P < 0.05. These tests were selected after consultation with the department of Biomathematics, Mt Sinai School of Medicine.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Embryos cultured in adenovirus vectors die rapidly after the 2-cell stage
In order to determine the stage of development at which adenovirus exerts toxic effects, 1-celled zygotes were cultured in the RSV-lacZ vector at 1x107 PFU/ml, and the percentage of embryos which reached each subsequent stage of cleavage development was compared against the starting number. When embryos were cultured continuously in the presence of adenovirus vector, cleavage to the 2-cell stage did not differ from control embryos cultured in medium alone (Table IGo). However, cleavage to the 4-cell stage was markedly impaired, and no normal embryos were seen beyond the 4-cell stage.


View this table:
[in this window]
[in a new window]
 
Table I. Development of mouse embryos in the presence of adenovirus vector in comparison with controls.
 
Embryos exposed for 2 h to adenovirus at the 2-cell stage exhibit a variety of abnormalities
When 2-celled embryos were exposed to the RSV-lacZ adenovirus vector for 2 h, washed and then cultured, many were able to develop beyond the 2-cell stage, but the percentage of embryos that reached the blastocyst stage was significantly reduced (Table IIGo). Many of these embryos exhibited several different kinds of developmental abnormalities. Developmental delay was clearly observed in embryos which otherwise cleaved normally. To assess developmental delay, two readily identifiable developmental landmarks were evaluated: cleavage from the 2-cell stage to the 3- or 4-cell stage, and development of a blastocyst cavity by compacted morulae. In order to examine developmental progression beyond the 2-cell stage, embryos were exposed to the adenovirus for 2 h, washed and then examined 53–56 h after hCG administration for cleavage to the 3- or 4-cell stage. When these embryos were compared with controls, cleavage beyond the 2-cell stage at this time point was significantly less frequent (Table IIIGo). Embryos which progressed to the compacted morula stage were then compared with controls for the frequency of appearance of a blastocyst cavity at 92 h after hCG. In these experiments, only compacted morulae which later successfully formed a blastocyst cavity were scored for blastocyst formation. This eliminated from consideration any embryos that underwent developmental arrest at the morula stage. As demonstrated in Table IIIGo, blastocyst formation was clearly delayed in embryos exposed at the 2-cell stage to the adenovirus. These findings indicated that one toxic effect of adenovirus vector exposure was slowed, but otherwise normal development occurred.


View this table:
[in this window]
[in a new window]
 
Table II. Development of 2-cell mouse embryos after exposure to equal numbers of infectious particles or total particles of four adenovirus vectors
 

View this table:
[in this window]
[in a new window]
 
Table III. Preimplantation development of virus-exposed and control embryos at two time points
 
In addition to developmental delay, embryos exposed to the adenovirus vector manifested several aberrant morphological features. These included premature compaction, with tight apposition of blastomeres as early as the 4-cell stage (Figure 1AGo), exclusion of blastomeres from the embryo proper (Figure 1BGo), and a characteristic `blebbing' of the cytoplasm (Figure 1CGo). That this latter anomaly was related to adenovirus infection was confirmed by staining embryos for LacZ activity. As shown in Figure 1CGo, the abnormal region of the embryo expresses large quantities of the reporter gene product. Another important feature of this figure is that the blastomeres are clearly mosaic for LacZ production, indicating that in embryos that survive beyond the 2-cell stage, the number of infectious viruses which entered the embryo was low.



View larger version (42K):
[in this window]
[in a new window]
 
Figure 1. Cleavage abnormalities of mouse embryos exposed at the 2-cell stage to 1x107 PFU/ml of the RSV-lacZ adenovirus vector for 2 h. (A) Abnormally tight association of blastomeres in a 4-cell embryo. (B) Exclusion of blastomeres from a developing morula. (C) Abnormal `blebbing' of the embryo at the 4-cell stage. This embryo was stained for LacZ, and heavy staining in a mosaic pattern is apparent. Original magnification, x200.

 
In addition to developmental delay, many embryos exposed to the vector failed to reach the blastocyst at all (Table IIGo). Developmental failure manifested as cell lysis, disorganized cleavage with separation of blastomeres, and developmental arrest. The severity of these abnormalities correlated with the duration of exposure to the adenovirus vector: When embryos were not washed after exposure, all of them were destroyed (data not shown). Higher titres of the virus also increased the frequency of complete developmental failure.

The toxicity of these vectors appeared to be associated with expression of the lacZ transgene inserted within them, in that exposure of 1-celled embryos followed by washing led to very low toxicity and expression, whereas exposure at the 2-cell stage resulted in expression of the transgene even in embryos that manifested toxicity (Figure 1Go). These findings were consistent with the notion that infection is a prerequisite for toxicity.

Investigations of the mechanism of adenovirus toxicity
Experiments were undertaken to determine if the lacZ reporter gene exerted toxicity independent of that resulting from exposure to the virus and expression of viral genes. To investigate this issue, vectors which express LacZ protein from the RSV promoter, or wild-type adenovirus, were compared for toxicity to a null vector which carried no reporter genes but which was deleted for the same adenovirus genes, E1 and E3. These two preparations were diluted such that the number of infectious particles (PFU) was the same for both preparations. In another set of experiments, the total numbers of particles were equalized between the two preparations. As shown in Table IIGo, the null vector was in fact far more toxic than the RSV-lacZ vector. These findings indicate that LacZ protein is not toxic to early embryos—an inference supported by published literature wherein otherwise normal transgenic animals have been produced which express the lacZ gene from the broadly expressed CMV promoter (Baskar et al., 1996Go).

The finding that the null vector was more toxic than the lacZ vector, even though the portion of the viral genome in each vector was identical, was further explored. In all adenovirus preparations, many viral capsids without functional genomes are produced. These particles are capable of interacting with viral receptor, but they cannot express viral genes inside the cell. It was hypothesized that empty capsids could interact with receptor on embryos and interfere with the binding of infectious particles. If infection were important to toxicity, such interference could reduce toxicity. Of note was the observation that in the null vector preparation, the ratio of total viral particles to infectious particles was lower than in the RSV-lacZ preparation (Table IIGo), a finding consistent with the notion that toxicity was related to infection rather than perturbance of the cell surface by interaction with the receptor.

In order to test this hypothesis further, embryos were exposed to a preparation of wild-type adenovirus. Although wild-type infectious particles might exert increased toxicity on embryos because of the expression of the full complement of viral genes, the virus cannot replicate in mouse cells (Ginsberg et al., 1991Go) and thus, toxicity due to cytolytic replication would not take place. The ratio of total particles to infectious particles in the wild-type preparation was far higher than for the null vector, and exposure of embryos to this preparation in fact led to less toxicity than with the null vector (Table IIGo). Preliminary tests comparing the null vector to another lacZ vector in which the reporter gene was driven by the CMV promoter were also performed. The latter vector preparation had a very low proportion of infectious particles to total particles (about 1:100). These experiments have thus far compared the null and CMV vectors when normalized for infectious particles, and again, as shown in Table IIGo, the toxicity of this vector was less than that of the null vector. The ratio of total particles to infectious particles for all vectors used is shown in Table IIGo; preparations with the lowest ratio (that is, the highest proportion of infectious particles, the null vector) were seen to be the most toxic.

Exposure to adenovirus vectors impairs embryo development in vivo
In order to determine if exposure of embryos to adenovirus vectors interferes with embryo development in vivo, embryos were exposed to vector using SUZI and then cotransferred with genetically marked controls into pseudopregnant foster mice. Newborns were then scored to determine the numbers born after exposure to the vector in comparison with controls. There was a marked reduction in development of exposed embryos in comparison to controls (Table IVGo). Reduced birth of exposed embryos could not be due to differences between foster females, as exposed and control embryos were mixed and transferred into the same foster mice. Differences in birth rates between exposed and control embryos could also not have been due to genetic differences in developmental vigour between the groups of embryos, because two sets of experiments were performed in which the genotypes of the exposed and control embryos were reversed. In both sets of studies, embryos subjected with SUZI using adenovirus vectors were impaired in their development in comparison with controls.


View this table:
[in this window]
[in a new window]
 
Table IV. Full term development of embryos subjected to subzonal insertion (SUZI) with a null adenovirus vector and co-transferred with control embryos
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
In the present study, the toxicity of adenovirus and adenovirus-derived gene transfer vectors to mouse embryos was studied. Toxicity was found to become evident after the 2-cell stage, was related largely to entry of the virus into cells, disrupted preimplantation development in a variety of ways, and probably did not result from expression of the lacZ gene, which was cloned into the vectors. Furthermore, it was found that exposure to these vectors markedly impaired embryo development in vivo. These findings indicate that exposure of human embryos to these vectors would disturb development and reduce implantation rates, but would not likely result in children born with insertion of vector-derived genes into the germ line. When adenovirus vectors are administered systemically, or when adenovirus infection occurs, toxicity to host cells may occur by a variety of mechanisms. Interaction of virus with the surface receptor can alter the surface properties of cells and damage them, whilst viral capsid proteins and transgene products can elicit both humoral and cellular immune responses. Expression of viral genes can incite a cell-mediated immune response, or might damage cells directly. Cytolytic replication of replication-competent vectors or wild-type virus can destroy infected cells that support virus replication. In the mouse embryo, many of these potential causes of toxicity are eliminated. Both cell-mediated and humoral immunity are absent, and in mouse embryos, cytolytic replication cannot occur. Remaining possible causes of toxicity include disturbance of the embryo surface, toxicity of viral proteins within cells, and toxicity of transgene products. The latter two mechanisms are dependent upon virus entry and, consequently, on the presence of functional virus receptor on the cell surface. The adenovirus receptor has two major components: the coxsackie and Ad receptor (CAR) and the {alpha}vß3 or {alpha}vß5 integrin molecule (for a review see Nemerow and Stewart, 1999). While CAR promotes association of the virus with the cell surface, the integrin molecule mediates internalization of the virus (Wickham et al., 1993Go; Nemerow and Stewart, 1999). Definitive tests for the presence of mouse CAR on the mouse embryo cell surface have never been developed. Comprehensive analyses of integrin subunits on the mouse embryo are lacking, though one study reports the absence of the critical {alpha}v subunit on unfertilized eggs (Tarone et al., 1993Go). An attempt was also made to identify functional adenovirus receptor in early embryos, but the absence of highly sensitive reagents for the mouse CAR, the rapid changes in surface molecules as cleavage proceeds, and the altered conformation of cell surfaces as cleavage progresses precluded definitive identification of the timing of appearance of functional receptor. These tests also showed, however, absence of infection of embryos after exposure at the 1-cell stage followed by washing—findings which again support the notion that functional receptor is minimal or absent at the one-cell stage (Gordon, 2002Go).

An indirect but sensitive test for the presence of functional adenovirus receptor is expression within exposed cells of reporter genes cloned into recombinant vectors. In order for such genes to be active, the viral genome must enter the cell, shed capsid proteins, and be translocated to the nucleus. The extant literature documents no reporter gene activity in mouse embryos before the 2-cell stage, though one report of expression at this stage followed transient exposure of the 1-cell embryos to the virus (Tsukui et al., 1996Go). The present study and previous experiments (Hall et al., 2000Go; Gordon, 2001Go) have clearly demonstrated high expression of reporter genes in adenovirus vectors when these vectors are used to infect 2-cell embryos.

The results of the present study are consistent with the presence of functional receptor at the 2-cell stage but not at earlier stages. Continuous culture of 1-cell embryos in vector leads to no disturbance of development before the 2-cell stage. However, after this time development is severely affected, such that very few embryos reach the 4-cell stage, and none survive beyond four cells (Table IGo). Transient exposure of 2-cell embryos results in delay of the very next cleavage division (Table IIIGo), and it has been shown previously that exposure at the 2-cell stage leads to high reporter gene expression (Hall et al., 2000Go). These findings all support the notion that viral receptor first appears at the 2-cell stage.

The present data provide some insight into the mechanisms whereby adenoviruses interfere with preimplantation development. In an effort to determine if the LacZ protein was toxic, a vector expressing the lacZ gene from the RSV promoter was compared with a `null' vector. The lacZ vector was in fact less toxic than the null vector (Table IIGo), findings consistent with the conclusion that the LacZ protein is not toxic. However, further investigations with a wild-type vector and a small number of experiments with another CMV-lacZ vector indicate that non-infectious viral particles, which are present in all vector and virus preparations, may interfere with virus entry and consequent toxicity. Because the null vector has the highest proportion of infectious particles of all preparations used (Table IIGo), it was concluded tentatively that entry of virus into the embryo is a key element of toxicity. If this is true, then some toxicity of the LacZ protein cannot be excluded, since the lacZ vectors used had higher proportions of non-infectious particles. If these vectors did not infect the embryo with equal efficiency to the null vector, then relative toxicity of the reporter gene cannot be formally assessed. However, the present data, as well as published data on transgenic mice expressing LacZ (Baskar et al., 1996Go), indicate that this protein is not toxic.

Analysis of data shown in Table IIGo is most consistent with the notion that vector toxicity relates to entry into the embryo. If expression of the reporter gene is not a direct cause of toxicity, then the other adenoviral genes must be considered. The E1 and E3 genes, deleted in most adenovirus vectors, induce unscheduled DNA synthesis and inhibit apoptotic cell death. Deletion of E1 and E3, which dramatically attenuates virus replication, is the most common approach to the engineering of adenovirus-based gene therapy vectors (Harvey et al., 1999Go), and the vectors used here had these genes deleted. Some laboratories have found that deletion of E4, which mediates virus assembly and which is present on our vectors, leads to longer-term expression of gene therapy vectors in vivo (e.g. Gao et al., 1996Go), presumably as a lessened immune response to the vector. Other laboratories have not observed significant changes with E4 deletion (Lusky et al., 1998Go). Taken together, these findings indicate that E4 would not exert a significant toxic effect on embryos, since an immune response to the protein is not possible in this experimental paradigm. Deletion of E2, which mediates viral chromosome replication (Yeh and Perricaudet, 1997Go) does not lessen the toxicity of adenovirus vectors in vivo (Ginsberg et al., 1990Go). The present results with the E1,E3 deleted vectors used here raise the possibility that other early genes such as E2 and E4 may have direct toxic effects on cells that would not be readily discerned when the vectors are administered to intact animals.

It is also possible that simply entry of virus capsid proteins, even in the absence of gene function, could be toxic. It has been observed (Kafri et al., 1998Go) that adenovirus inactivated by psoralen–UV crosslinking, when injected into animals, elicits a cell-mediated immune response. This finding indicates that viral proteins present in capsids of the vector preparation enter the cell in sufficient quantities to affect recipient cells, even in the absence of production of new protein by virus gene expression. If these proteins can provoke a cell-mediated immune response they might also exert direct toxic effects on cells. Although the present results strongly support the notion that entry of the virus into cells is a key factor in the appearance of toxic effects, disturbance of surface interactions between blastomeres during the course of preimplantation development is certainly not ruled out as another mechanism of embryo toxicity. Within 5 min of exposing 2-cell embryos to adenovirus vectors, the adhesive properties of the embryos are dramatically changed such that they attach firmly and irreversibly to the culture dish. These changes necessitated supplementation of the culture medium with BSA and the use of constant shaking during the period of virus exposure (see Materials and methods). This observation indicates that the cell surface is modified by adenovirus exposure. It is relevant in this regard that the integrin molecules which form part of the adenovirus receptor mediate cell adhesion. Moreover, many of the abnormalities of cleavage involve cell interactions, with inappropriate `compaction' of embryos at early stages, and exclusion of blastomeres from the embryo (Figure 1Go). Exclusion of cells was not always due to cell demise, as indicated by the fact that may of the cells that failed to be incorporated into the embryo went on to produce trophoblastic vesicles with continued culture (data not shown). Therefore, perturbance of surface interaction between blastomeres is implicated.

The present experiments wherein embryos were subjected to subzonal insertion of adenovirus vectors and transferred to females establish the physiological significance of the toxic effects observed during in-vitro culture. Abnormal cleavage in vitro is reflected in vivo as a failure of embryos to develop to term (Table IVGo). This observation, in conjunction with the fact that these vectors are derived from a human pathogen, have implications for IVF. Adenoviruses are likely to be present in the IVF laboratory environment, and they can readily pass through the 0.2 µm filters used to sterilize IVF media. The presence of adenoviruses or related vectors in patients or laboratory personnel as a result of infection or administration of gene therapy vectors could thus pose a significant threat to manipulated embryos. The present results would suggest that human embryos exposed to these viruses or their related vectors would die if exposed for prolonged periods, and exhibit a variety of cleavage abnormalities and slowed development if exposed briefly to low amounts of these agents (Figure 1CGo). These findings, in addition to the observation that adenovirus DNA integrates infrequently (10-3 to 10-5 integrations per exposed cell) even after high exposures (Harui et al., 1999Go), indicate that embryos which survive exposure to yield pregnancies would contain few virus particles and would be very unlikely to sustain an integration event. This prediction is supported by the author's own investigations, in which 94 mouse offspring were studied after adenovirus exposure at the 2-cell stage and transfer, and not a single example was found of an integration event that could be confirmed by both PCR and Southern hybridization (Gordon, 2002Go). In conclusion, the present results indicate that significant exposure to adenovirus or related vectors could reduce pregnancy rates but would not be likely to insert DNA into the germ line.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The author thanks G.Pohorenic for excellent technical assistance. J.W.G. is Mathers Professor.


    Notes
 
1 Address for correspondence: Department of Obstetrics/Gynecology, Mt Sinai Medical Center, 1 Gustave L. Levy Place, New York, NY 10029, USA. E-mail: jon.gordon{at}mssm.edu Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Baskar, J.F., Smith, P.P., Nilaver, G., Jupp, R.A., Hoffmann, S., Peffer, N.J., Tenney, D.J., Colberg-Poley, A.M., Tucker, C., Ghazal, P. et al. (1996) The enhancer domain of the human cytomegalovirus major immediate-early promoter determines cell type-specific expression in transgenic mice. J. Virol., 70, 3207–3215.[Abstract]

Chan, A.W., Homan, E.J., Ballou, L.U., Burns, J.S. and Bremel, R.D. (1998) Transgenic cattle produced by reverse-transcribed gene transfer in oocytes. Proc. Natl Acad. Sci. USA, 95, 14028–14033.[Abstract/Free Full Text]

Chan, A.W.S., Chong, K.Y., Martinovich, C., Simerly, C. and Schatten, G. (2000) Transgenic monkeys produced by retroviral gene transfer into mature oocytes. Science, 291, 309–312.[Abstract/Free Full Text]

Cohen, J., Alikani, M., Trowbridge, J. and Rosenwaks, A. (1992) Implantation enhancement by selective assisted hatching using zona drilling of human embryos with poor prognosis. Hum. Reprod., 7, 685–691.[Abstract]

Cohen, J., Scott, R., Schimmel, T., Levron, J. and Willadsen, S. (1997) Birth of infant after transfer of anucleate donor oocyte cytoplasm into recipient eggs. Lancet, 350, 186–187.[ISI][Medline]

Erbeck, G.T., Lawitz, J.A., Papaioannou, V.E. and Biggers, J.S. (1994) Differential growth of the mouse preimplantation embryo in chemically defined media. Biol. Reprod., 50, 1027–1033.[Abstract]

Fishel, S., Antinori, S., Jackson, P., Johnson, J. and Rinaldi, L. (1991) Presentation of six pregnancies established by sub zonal insemination. Hum. Reprod., 6, 124–130.[Abstract]

Gao, G.P., Ykang, Y. and Wilson, J.M. (1996) Biology of adenovirus vectors with E1 and E4 deletions for liver-directed gene therapy. J. Virol., 70, 8934–8943.[Abstract]

Ginsberg, H.S., Horswood, R.L. and Chanoch, R.M. (1990) Role of early genes in pathogenesis of adenovirus pneumonia. Proc. Natl Acad. Sci. USA, 87, 6191–6195.[Abstract]

Ginsberg, H.S., Moldawer, L.L., Seghkal, P.B., Redington, M., Kilian, P.L., Chanock, R.M. and Prince, G.A. (1991) A mouse model for investigating the molecular pathogenesis of adenovirus pneumonia. Proc. Natl Acad. Sci. USA, 88, 1651–1655.[Abstract]

Gordon, J.W. (2001) Direct exposure of mouse ovaries and oocytes to high doses of an adenovirus gene therapy vector fails to lead to germ cell transduction. Mol. Ther., 3, 557–564.[ISI][Medline]

Gordon, J.W. (2002) High toxicity, low receptor density and low integration frequency severely impede the use of adenovirus vectors for production of transgenic mice. Biol. Reprod., in press.

Gordon, J.W., Scangos, G.A., Plotkin, D.J., Barbosa, J.A. and Ruddle, F.H. (1980) Genetic transformation of mouse embryos by microinjection of purified DNA. Proc. Natl Acad. Sci. USA, 77, 7380–7384.[Abstract]

Gordon, J.W., Grunfeld, L., Garrisi, G.J., Talansky, B.E., Richards, C., Wiczyk, H., Rappaport, D. and Laufer, N. (1988) Fertilization of human oocytes by sperm from infertile males after zona pellucida drilling. Fertil. Steril., 50, 68–73.[ISI][Medline]

Hall, S.J., Bar-Chama, N., Ta, S. and Gordon, J.W. (2000) Direct exposure of mouse spermatogenic cells to high doses of adenovirus gene therapy vector does not result in germ cell transduction. Hum. Gene Ther., 11, 1705–1712.[ISI][Medline]

Handyside, A.H., Lesko, J.G., Tarin, J.J., Winston, R.M. and Hughes, M.R. (1992) Birth of a normal girl after in vitro fertilization and preimplantation diagnostic testing for cystic fibrosis. N. Engl. J. Med., 327, 905–909.[Abstract]

Harui, A., Suzuki, S., Kochanek, S. and Mitani, K. (1999) Frequency and stability of chromosomal integration of adenovirus vectors. J. Virol., 73, 6141–6146.[Abstract/Free Full Text]

Harvey, B.-G., Worgall, S., Ely, S., Leopold, P.L. and Crystal, R.G. (1999) Cellular immune responses of healthy individuals to intradermal administration of an E1-E3-adenovirus gene transfer vector. Hum. Gene Ther., 10, 2823–2837.[ISI][Medline]

Jaenisch, R. (1976) Germ line integration and Mendelian transmission of the exogenous Moloney leukemia virus. Proc. Natl Acad. Sci. USA, 73, 1260–1264.[Abstract]

Kafri, T., Morgan, D., Krah., T., Sarvetnick, N., Skherman, L. and Verma, I. (1998) Cellular immune response to adenoviral vector-infected cells does not require de novo viral gene expression: implications for gene therapy. Proc. Natl Acad. Sci. USA, 95, 11377–11382.[Abstract/Free Full Text]

Kubisch, H.M., Larson, M.A., Eichen, P.A., Wilson, J.M. and Roberts, R.M. (1997) Adenovirus-mediated gene transfer by perivitelline microinjeciton of mouse, rat, and cow embryos. Biol. Reprod., 56, 119–124.[Abstract]

Lusky, M., Christ, M., Rittner, K., Dieterle, A., Dreyer, D., Mourot, B., Schultz, H., Stoeckel, F., Pavirani, A. and Mehtali, M. (1998) In vitro and in vivo biology of recombinant adenovirus vectors with E1, E1/E2A, or E1/E4 deleted. J. Virol., 72, 2022–2032.[Abstract/Free Full Text]

Malter, H.E. and Cohen, J. (1989) Partial zona dissection of the human oocyte: a nontraumatic method using micromanipulation to assist zona pellucida penetration. Fertil. Steril., 51, 139–148.[ISI][Medline]

Nagano, M., Brinster, C.J., Orwig, K.E., Ryu, B.Y., Avarbock, M.R. and Brinster, R.L. (2001) Transgenic mice produced by retroviral transduction of male germ-line stem cells. Proc. Natl Acad. Sci USA, 98, 13090–13095.[Abstract/Free Full Text]

Nemerow, G.R. and Steward, P.L. (1999) Role of {alpha}v integrins in adenovirus cell entry and gene delivery. Microbiol. Mol. Biol. Rev., 63, 725–733.[Abstract/Free Full Text]

Palermo, G., Joris, H., Devroey, P. and Van Steirteghem, A.C. (1992) Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet, 340, 17–18.[ISI][Medline]

Quinn, P., Barros, C. and Whittingham, D.G. (1982) Preservation of hamster oocytes to assay the fertilizing capacity of human spermatozoa. J. Reprod. Fertil., 66, 161–168.[Abstract]

Shenk, T. (1996) Adenoviridae: the viruses and their replication. In Fields, B.N., Snipe, D.M. and Howley, P.M. (eds), Fields Virology. Lippincott-Raven, Philadelphia, pp. 2111–2148.

Talansky, B.E., Barg, P.E. and Gordon, J.W. (1987) Membrane ion pump inhibitors block the fertilization of zona-free mouse oocytes by acrosome reacted sperm. J. Reprod. Fertil., 79, 447–455.[Abstract]

Tarone, G., Russo, M.A., Hirsch, E., Odorisio, T., Altruda, F., Silengo, L. and Siracusa, G. (1993) Expression of beta 1 integrin complexes on the surface of unfertilized mouse oocyte. Development, 117, 1369–1375.[Abstract/Free Full Text]

Tsukui, T., Kanegae, Y., Saito, I. and Toyoda, Y. (1996) Transgenesis by adenovirus-mediated gene transfer into mouse zona-free eggs. Nature Biotechnol., 14, 982–985.[ISI][Medline]

van der Putten, H., Botteri, H., Miller, F.M., Rosenfeld, A.D., Fan, M.G., Evans, R.M. and Verma, I. (1985) Efficient insertion of genes into the mouse germ line via retroviral vectors. Proc. Natl Acad. Sci. USA, 82, 6148–6152.[Abstract]

Wickham, T.J., Mathias, P., Chkeresh, D.A. and Nemerow, G.R. (1993) Integrins {alpha}vß3 and {alpha}vß5 promote adenovirus internalization but not virus attachment. Cell, 73, 309–319.[ISI][Medline]

Yeh, P. and Perricaudet, M. (1997) Advances in adenoviral vectors: from genetic engineering to their biology. FASEB J., 11, 615–623.[Abstract/Free Full Text]

Submitted on January 21, 2002; accepted on May 8, 2002.





This Article
Abstract
FREE Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Gordon, J. W.
PubMed
PubMed Citation
Articles by Gordon, J. W.