| Human Molecular Genetics | Pages |
Retrovirus-mediated enzymatic correction of Tay-Sachs defect in transduced and non-transduced cells
Introduction
Results
Construction and characterization of the pG.HEXA vector
Gene expression in transduced fibroblasts
Partial restoration of HexA activity in non-transduced cells through internalization of the enzyme secreted by transduced cells
[beta]-Hexosaminidase isoenzymatic patterns in treated or non-treated cells and culture media
Impact of [alpha]-subunit overexpression on HexB activity
Discussion
Materials And Methods
Cell culture
Retroviral vector
Molecular hybridization
Immunoblotting
Enzyme assays
Thermostability of [beta]-hexosaminidases
[beta]-Hexosaminidase electrophoresis
Uptake of [beta]-hexosaminidases
Acknowledgements
References
Retrovirus-mediated enzymatic correction of Tay-Sachs defect in transduced and non-transduced cells
Tay-Sachs disease is a severe neurodegenerative disorder due to mutations in the HEXA gene coding for the [alpha]-chain of the [alpha]-[beta] heterodimeric lysosomal enzyme [beta]-hexosaminidase A (HexA). Because no treatment is available for this disease, we have investigated the possibility of enzymatic correction of HexA-deficient cells by HEXA gene transfer. Human HEXA cDNA was subcloned into a retroviral plasmid generating to G.HEXA vector. The best [Psi]-CRIP producer clone of G.HEXA retroviral particles was isolated, and murine HexA-deficient fibroblasts derived from hexa -/- mice were transduced with the G.HEXA vector. Transduced cells overexpressed the [alpha]-chain, resulting in the synthesis of interspecific HexA (human [alpha]-chain/murine [beta]-chain) and in a total correction of HexA deficiency. The [alpha]-chain was secreted in the culture medium and taken up by HexA-deficient cells via mannose-6-phosphate receptor binding, allowing for the restoration of intracellular HexA activity in non-transduced cells.
INTRODUCTION
[beta]-Hexosaminidase A deficiency leads to a lysosomal storage disorder known as Tay-Sachs disease. This deficiency induces accumulation of GM2 gangliosides, predominantly in neurons. The clinical spectrum of this autosomal recessive disease varies from the common infantile lethal neurodegenerative form to less severe adult onset forms, often characterized by motor neuron impairments (1).
The [beta]-hexosaminidases occur as three isoenzymes due to different associations of the [alpha]- and [beta]-subunits. The [beta]-hexosaminidase A (HexA), a heterodimer [alpha]-[beta], and the [beta]-hexosaminidase B (HexB), a homodimer [beta]-[beta], represent the two major forms (2). The [beta]-hexosaminidase S (HexS), a homodimer [alpha]-[alpha], a minor form, appears to have negligible catalytic activity (3). The [alpha]- and [beta]-subunits are encoded by two different genes, HEXA and HEXB, which are located on chromosomes 15 and 5, respectively. Tay-Sachs disease is due to mutations affecting the [alpha]-subunit gene (4), and there is no treatment for this fatal disease.
In fact, gene therapy could be a therapeutic approach for this type of disease for a number of reasons. Firstly, several mutations have been described in the HEXA gene which give rise to a pseudodeficiency characterized by residual HexA enzyme activity ~10% of the normal activity and a complete lack of symptoms (5,6). Secondly, like other lysosomal hydrolases, HexA is modified by glycosylation and phosphorylation of mannose residues, which enables lysosome targeting via the binding to mannose 6-phosphate receptor (M6Pr) (7). However, some HexA molecules escape from this binding, are secreted and can then be taken up by other cells after binding to M6Pr located at their plasma membrane, even given that this phenomenom seems to be weaker for [beta]-hexosaminidase than for other lysosomal hydrolases (8). The HexA-M6Pr complexes are internalized and targeted to lysosomes. Therefore, grafting of a pool of cells which overexpress and highly secrete HexA could lead to a measurable HexA activity in defective cells via a secretion-recapture mechanism. However, due to the existence of the blood-brain barrier and the neurological symptoms of Tay-Sachs disease, we are aware that an efficient treatment of this disease would require either secretion of an enzymatic form able to cross the blood-brain barrier, or in situ secretions of the enzyme in the central nervous system and its diffusion to the different neuronal cells. These prerequisites do not constitute minor challenges, and our strategy could first be applied to other types of lysosomal diseases without predominant neurological symptoms.
We previously had reported that an adenoviral vector allowed for restoration of HexA activity in Tay-Sachs fibroblasts (9). However, the transient expression of genes transferred by this type of vector (10-12) makes them unsuitable for the stable correction required in human patients with chronic genetic diseases such as lysosomal diseases. We have therefore developed HexA retroviral vectors that permit long-term expression by integration into the genome of target cells.
Here we demonstrate that retroviral transfer of the human [alpha]-subunit gene in mouse HexA-deficient fibroblasts leads to overexpression of the [alpha]-subunit. Transduced cells showed restoration of the enzymatic activity following heterodimer formation between the human [alpha]-chain and the murine [beta]-chain. Moreover, high levels of HexA were secreted from transduced cells and some enzyme could be taken up by mouse HexA-deficient fibroblasts. This demonstrates that our retroviral vector is able to correct HexA-deficient cells by directly delivering the gene coding for the missing subunit of the HexA heterodimer and by directing secretion of large amounts of recombinant enzyme that can be taken up to a certain extent by non-transduced cells. This experiment represents the initial step before testing HEXA retroviral gene transfer in the murine Tay-Sachs model (13-15).
RESULTS
Construction and characterization of the pG.HEXA vector
[Psi]-CRIP amphotropic packaging cells were co-transfected with both pG.HEXA (Fig. 1) and pRSV.neo (to confer neomycin resistance to transduced cells). Sixty stable transfected clones were analysed for the production of retroviral particles, and the best producer clone G.HEXA 10 was retained. The titre of the G.HEXA 10 clone was 105 c.f.u./ml, as evaluated by quantitative Southern blot technique (16) after transduction of 5×105 NIH 3T3 cells (data not shown).
| Figure 1. Structure of the p.G.HEXA retroviral vector and analysis of the integration event by Southern blot. (A) The location of the SmaI sites used for Southern blot analysis and the sizes of the predicted RNA transcripts are shown in kilobases (kb) below the vector. LTR, long terminal repeat; [psi], packaging signal; [Delta]gag, deleted and mutated Moloney gag gene; PGK, mouse phosphoglycerate kinase promoter; HEXA cDNA, cDNA coding for the human [alpha]-subunit of [beta]-hexosaminidase; [Delta]U3, deleted enhancer domain in the 3[prime] LTR. (B) Confirmation of the presence of an unrearranged G.HEXA provirus in the genome of HexA-deficient murine fibroblasts [denoted HexA(-)M.F] by Southern blot. SmaI-digested genomic DNA from HexA(-)M.F and HexA(-)M.F transduced with the M48.[beta]-Gal or G.HEXA retroviral vectors were electrophoresed, transferred onto a nylon membrane and hybridized with the HEXA cDNA as a probe. For positive controls, 0.2 and 0.5 copies of p.G.HEXA were used. A 4.1 kb band co-migrating with the proviral band was present only in HexA(-)M.F transduced with the G.HEXA vector. |
|
To characterize the pG.HEXA retroviral vector, primary cultures of HexA-deficient murine fibroblasts, derived from hexa -/- mice (13), were transduced with retroviral supernatant of the G.HEXA 10 clone. Several rounds of transduction were necessary to obtain ~90% of transduced cells. The proportion of positive cells was determined by immunofluorescence staining using a goat anti-human [beta]-hexosaminidase [alpha]-subunit as primary antibody (data not shown).
The integration of an unrearranged G.HEXA provirus into the genome of HexA-deficient murine fibroblasts was demonstrated by the presence of a 4.1 kb specific band in transduced cells after Southern blotting using SmaI-digested genomic DNA (Fig. 1).
Gene expression in transduced fibroblasts
Northern blot analysis showed specific transcripts only in transduced G.HEXA cells when compared with non-transduced cells or with cells transduced with M48 retroviral vector (which encodes the [beta]-galactosidase gene under the control of the PGK promoter) (Fig. 2A). Two types of HEXA mRNA were synthesized: the 2.1 kb major form initiated from the PGK promoter, representing ~95% of the total HEXA mRNA, and a 4.2 kb minor form initiated from the mutated Moloney long terminal repeat (LTR) promoter, representing ~5% of the total HexA mRNA. Multiple transcription initiation sites of the housekeeping PGK promoter are responsible for the fuzzy appearance of the 2.1 kb band. The abundance of HEXA mRNA was at least one order of magnitude higher in transduced murine fibroblasts than in non-transduced normal human fibroblasts.
| Figure 2. HEXA mRNA expression, protein expression and restoration of enzymatic activity in retrovirus-transduced deficient mouse fibroblasts. (A) Northern blot analysis of 20 µg of total RNA from non-transduced HexA(-)M.F or from HexA(-)M.F transduced with the M48.[beta]-Gal or G.HEXA vectors. Normal human fibroblasts [denoted HexA(+)H.F] were used as positive controls. Two bands of 1.9 and 2.4 kb were present in normal human fibroblasts. The two specific bands of 2.1 and 4.2 kb, present only in HexA(-)M.F transduced with G.HEXA vector, correspond to the two transcripts using either the transcriptional start site at the internal PGK promoter or that at the Moloney 5[prime] LTR promoter, respectively. (B) Immunodetection of human [beta]-hexosaminidase [alpha]-chain in protein extracts and in culture supernatants from ex vivo transduced HexA(-)M.F. Protein extracts (20 µg) or filtered supernatant (25 µl) were electrophoresed. [beta]-Hexosaminidase [alpha]-chain immunodetection was performed with a goat anti-human [alpha]-chain primary antibody used at a 1:750 dilution, and an HRP-conjugated rabbit anti-goat IgG followed by ECL. (C) Hexosaminidase activities in cell extracts or in the culture medium of HexA(-)M.F transduced with the G. | 
|
HEXA vector. The histograms show the hexosaminidase activities measured with the [alpha]-chain-specific artificial fluorogenic substrate, 4-MUGS. The enzymatic values of the different cell extracts or cell supernatants are means ± SD of triplicate samples.
Synthesis of the [alpha]-chain was analysed by western blot (Fig. 2B) with a goat anti-human [alpha]-chain-specific antibody. In normal human fibroblasts, only the 54 kDa mature form was detected. Conversely, in G.HEXA-transduced HexA-deficient fibroblasts, two different bands of 54 and 67 kDa at a ratio of ~1:1 were found, corresponding to the mature and precursor forms of the [alpha]-chain, respectively. The combined level of mature and precursor forms synthesized in G.HEXA-transduced fibroblasts was ~7-fold higher than in normal human fibroblasts. Additionally, the immunoblot demonstrated that the [alpha]-chains synthesized by G.HEXA-transduced fibroblasts were secreted into the culture medium only in the precursor form, as previously reported (17). The [alpha]-chain secretion of normal human cells was too weak to be detected in this experiment.
The HexA activity in G.HEXA-transduced murine fibroblasts, as measured with the [alpha]-chain-specific artificial substrate 4-methylumbelliferyl-7-(6-sulfo-2-acetamido-2-deoxy-[beta]-d-glucopyranoside (4-MUGS), was 6-fold higher than that of normal human or murine fibroblasts (Fig. 2C). In culture media, the enzymatic activity of the HexA secreted by G.HEXA-transduced fibroblasts was 100-fold higher than that of HexA secreted by non-transduced human normal fibroblasts (Fig. 2D).
Partial restoration of HexA activity in non-transduced cells through internalization of the enzyme secreted by transduced cells
The uptake of [beta]-hexosaminidases secreted from the G.HEXA-transduced fibroblasts by HexA-deficient murine fibroblasts was analysed by incubating them with a 10-fold concentrated serum-free culture medium of the G.HEXA-transduced fibroblasts for 24 h in the presence or absence of d-mannose-6-phosphate (M6P). Concentrated culture media of HexA-deficient non-transduced murine fibroblasts or of HexA-deficient murine fibroblasts transduced with the M48.[beta]-Gal retroviral vector were used as negative controls. Partial restoration of the HexA enzymatic activity was seen only in the cells maintained in concentrated culture medium of the G.HEXA-transduced fibroblasts and was specifically inhibited in the presence of M6P, thus demonstrating the endocytosis of some secreted enzyme via the M6Pr (Fig. 3). The HexA activity observed in these cells represents ~5% of the intracellular activity found in normal human fibroblasts.
Figure 3. Partial restoration of enzymatic activity of non-transduced HexA(-)M.F by uptake of the HexA enzyme secreted into the medium by transduced cells. Culture media from HexA(-)M.F transduced with the M48.[beta]-Gal or the G.HEXA vector were harvested after 48 h of culture, filtered on a 0.22 µm filter and concentrated 10-fold on an ultrafiltration column. Then, HexA(-)M.
F were incubated during 24 h with this concentrated medium in the presence or absence of 5 mM d-mannose-6-phosphate. Histograms represent the hexosaminidase activities obtained in cell extracts with the [alpha]-chain-specific fluorogenic substrate, 4 MUGS. The enzymatic values of the different cell extracts are means ± SD of triplicate samples.
[beta]-Hexosaminidase isoenzymatic patterns in treated or non-treated cells and culture media
To analyse the different forms of [beta]-hexosaminidases synthesized by the transduced or non-transduced HexA-deficient fibroblasts, cell protein extracts were loaded on cellulose acetate strips (Fig. 4A). Figure 4. A shows the positions of human HexS ([alpha]-[alpha]) from Sandhoff fibroblasts (lane 1), human HexB ([beta]-[beta]) and HexA ([alpha]-[beta]) from normal human fibroblasts (lane 2), and murine HexB from HexA-deficient murine fibroblasts (lanes 4, 5 and 8). In HexA-deficient murine fibroblasts transduced with the G.HEXA vector, the murine HexB band seems less intense, and two strong faster bands appeared, most likely corresponding to the expected human/murine [alpha]-[beta] interspecific HexA and to human HexS (lane 3). This interpretation is confirmed by the effect of anti-human [alpha]- and [beta]-chain antibodies that do not precipitate murine HexB (lanes 8 and 9), but precipitate the other two human [alpha]-containing forms (i.e, interspecific HexA and HexS) (lane 8).
|
Figure 4. Electrophoretic separation of [beta]-hexosaminidase forms in cell extracts (A) or in culture media (B) of HexA(-)M.F transduced with the G.HEXA vector. (A) Before loading on cellulose acetate gel, cell extracts were incubated or not with polyclonal antibody raised against the human HexA recognizing specifically both the human [alpha]- and [beta]-subunits. On the left are indicated the positions of human HexB, A and S (h.Hex), and the arrow indicates the origin of migration. On the right is indicated the position of the murine HexB. -Lane 1, Sandhoff human fibroblasts (S.H.F), with presence of HexS only (i.e. human [alpha]-[alpha] heterodimers); lane 2, normal human fibroblasts [HexA(+)H.F] with the two normal isozymes, HexB and HexA (i.e. [beta]-[beta] and [alpha]-[beta] dimers); lane 3, fibroblasts from HexA-deficient mice transduced with the G.HEXA vector [HexA(-)M.F-G.HEXA]. Three bands can be detected plus an unidentified fast migrating smear. The slow migrating band is murine HexB, well detected in HexA-deficient fibroblasts (lanes 4 and 5). The intermediate band most likely corresponds to human/murine [alpha]-[beta] interspecific HexA, and then we detect human HexS (human [alpha]-[alpha] homodimers). Lane 4, as lane 3, but transduced with the M48.[beta]-Gal control vector; lane 5, as lane 3, but non-transduced; lane 6, as lane 2; lane 7, as lanes 2 and 6 + anti-HexA antibodies: both HexA and HexB isoforms are precipitated; lane 8, as lane 3 + anti-HexA antibodies: all [alpha]-chain-containing isoforms (i.e. interspecific HexA and human HexS) are precipitated, murine HexB is the only non-precipitated form; lane 9, as lane 5 + anti-HexA antibodies: no modification. (B) Lane 1, normal human serum, with bands corresponding to precursor forms of HexB and HexA; lane 2, culture medium of HexA-deficient murine cells transduced with the G.HEXA vector: precursor forms of interspecific HexA and HexS are detected. |
|
The same technique was used to analyse the different [beta]-hexosaminidase forms secreted in the culture media (Fig. 4B). No forms are detected in the culture media of murine HexA-deficient fibroblasts, untransduced or transduced with the M48.[beta]-Gal retroviral vector, or in the culture media of human or murine non-deficient fibroblasts (data not shown). Conversely, the culture medium of the G.HEXA-transduced fibroblasts reveals two different bands which correspond to the precursor forms of the interspecific HexA and to the human HexS (lane 2) as determined by comparison with the bands obtained for non-deficient human serum (lane 1) (18).
Impact of [alpha]-subunit overexpression on HexB activity
Overexpression of the [alpha]-chain could lead to a HexB deficiency in transduced fibroblasts by displacement of the [beta]-[beta] association in favour of the [alpha]-[beta] association. To evaluate a possible depletion in [beta]-[beta] complexes in the G.HEXA-transduced fibroblasts, we tested the HexB enzymatic activity of protein extracts after a 2 h incubation at 52°C to eliminate the thermolabile HexA and HexS forms (19). HexB activity in the G.HEXA-transduced fibroblasts was reduced to ~30% of the HexB activity in normal murine fibroblasts (Fig. 5).
Figure 5. Impact of the overexpression of the [alpha]-chain on HexB activity. Cell extracts from HexA(+)M.F or from HexA(-)M.F transduced with the G.HEXA vector were incubated for 2 h at either 4 or 52°C to eliminate the thermolabile HexA and HexS. Histograms represent the hexosaminidase activities obtained in cell extracts with the [beta]-subunit-specific artificial fluorogenic substrate, 4-MUG. The hexosaminidase activities found after incubation at 52°C represented the HexB activities. The enzymatic values of the different cell extracts are means ± SD of triplicate samples.
DISCUSSION
We have obtained the G.HEXA retroviral vector encoding the human [beta]-hexosaminidase [alpha]-chain that transduces, at very high efficiency (obtaining 90% of transduced cells), murine HexA-deficient fibroblasts.
By comparison with normal human fibroblasts, the HexA-deficient fibroblasts transduced with the G.HEXA vector overexpressed the HEXA mRNA. The resulting human [alpha]-chain was produced both in the mature and precursor forms, whereas only the mature form was detectable by immunoblotting in protein extracts from normal human fibroblasts. The [alpha]-chain overexpression led to both a 3.5-fold increase of the mature lysosomal form, and to a cytoplasmic accumulation of the precursor form. The presence of the mature form of the human [alpha]-chain proves that the [alpha]-subunit synthesized from G.HEXA provirus is post-translationally modified by glycosylation, phosphorylation and protein cleavage in the different cell compartments (endoplasmic reticulum, Golgi and lysosome), allowing for a correct targeting to the lysosomes. The accumulation of the precursor form is known to take place essentially in the endoplasmic reticulum and in the Golgi compartments not in the cytosol. This was confirmed by immunofluorescence using an anti-human [alpha]-chain on the G.HEXA-transduced HexA-deficient fibroblasts, which showed a periplasmic expression pattern (endoplasmic reticulum, Golgi, lysosomes and vesicles) specific to the lysosomal hydrolases (20) (data not shown). The accumulation of precursor forms is the consequence of either the limiting amount of proteins implicated in maturation and lysosomal targeting of the [alpha]-chain (glycosylation or phosphorylation enzymes, or M6Pr) or of the excess [alpha]-chain production limiting the possibilities of [alpha]-[beta] heterodimerization. Indeed, in Sandhoff fibroblasts, lacking the [beta]-chain, only the precursor form of the [alpha]-chain is produced, leading to the presence of HexS in the cytoplasm (17).
The human [alpha]-chain expressed by G.HEXA-transduced HexA-deficient fibroblasts was able to associate with the murine [beta]-chain to form a human/murine [alpha]-[beta] interspecific HexA, as demonstrated by cellulose acetate electrophoresis. This result is not particularly surprising, since 85% identity and 92% similarity is shared between the amino acid sequences of the murine and human [alpha]-chains (21). Moreover, the putative active site residues of the human [alpha]-chain described by Fernandes et al. (22) and residues shown to be important by mutational analysis for [alpha]-[beta] association (residues R504 and G269) (23-25) are identical to those of the murine [alpha]-chain (21). Therefore, [alpha]-chains of these two different species are likely to be totally interchangeable.
The HexS isoform resulting from the association of two [alpha]-chains was also found in non-negligible amounts in the G.HEXA-transduced HexA-deficient fibroblasts. This is most likely due to overexpression of the [alpha]-chain and to the stronger affinity of the [beta]-chain for itself than for the [alpha]-chain (2). Thus, the presence of both [alpha]-chain and [beta]-chain in the endoplasmic reticulum, with an accumulation of [alpha]-chain, shifts the balance towards [alpha]-[beta] heterodimer formation. However, [beta]-[beta] complexes representing 30% normal HexB activity are explained by the highly effective [beta]-[beta] homodimerization. The remaining [alpha]-chains form the HexS. An alternative hypothesis implicates a factor required for heterodimeric association, as suggested by Proia et al. (17). The limitation of this factor could lead to a preferential [alpha]-[alpha] association of precursor forms (HexS) as described in Sandhoff disease fibroblasts (17), and to a persistent [beta]-[beta] association (HexB). As discussed previously, there is only a partial HexB depletion in the HexA-deficient fibroblasts transduced with G.HEXA vector, and the residual HexB level is certainly sufficient to prevent any disorders. Indeed Dreyfus et al. (19) previously described an asymptomatic adult with only a 10% HexB residual activity.
The synthesis of HexA and HexS in the G.HEXA-transduced HexA-deficient fibroblasts allowed for an over-correction of the enzyme defect, as determined by the [alpha]-chain-specific enzymatic activity obtained. This activity was ~7-fold higher than the enzymatic activity in normal fibroblasts. This increase correlates with the increase in [alpha]-chain expression in transduced fibroblasts. As the ratio between mature and precursor [alpha]-chain in HexA-deficient fibroblasts transduced with the G.HEXA vector was 1:1, with a mature [alpha]-chain 3- to 4-fold more abundant than in normal fibroblasts, our enzymatic assay represents the sum of the activities of mature and precursor forms. However, it is probable that only mature [alpha]-chains, heterodimerized with [beta]-subunits to generate HexA isozyme, are functionally relevant in vivo, allowing for degradation of the GM2 ganglioside, as the Vmax and Km of HexS are much lower those that of HexA (3). Therefore, retrovirus-mediated transduction of the [alpha]-chain cDNA in fibroblasts deficient in HexA ([alpha]-[beta]) activity allows for a more than complete restoration of enzyme activity. Our previous results using adenoviral vectors have demonstrated that such transduced Tay-Sachs fibroblasts completly restored GM2 ganglioside metabolism (9).
The [alpha]-chain precursor forms accumulated in cellular compartments were not sequestered, but were secreted in large amounts in the culture medium in the form of HexS and HexA precursors, as previously described (26). Such HexA secretion is most likely due in part to saturation of one or several enzymes implicated in maturation of the HexA or to saturation of intracellular M6Prs in transduced cells. We have demonstrated that this secreted HexA can be recaptured by non-transduced HexA-deficient murine fibroblasts via M6Pr-mediated endocytosis, leading to an enzymatic activity of ~5% of that in normal fibroblasts. This correlates with previous results by Moullier et al. (27) showing a similar receptor-mediated uptake of [beta]-glucuronidase secreted in the culture medium of retrovirally transduced fibroblasts and, therefore, confirms the potential of our gene therapy approach for lysosomal diseases. Lacorazza et al. (28) have used the same type of retroviral vector coding for the human [alpha]-chain of [beta]-hexosaminidase to transduce non-deficient murine neuronal progenitor cells. The authors report that overexpression of the human [alpha]-chain in transduced neuronal cells is accompanied by an increase in endogenous thermostable enzyme HexB, whereas our data show a decrease to 30% of the normal HexB activity. The difference in the HexB activity after [alpha]-chain overexpression between the two studies may be due to intrinsic cell type differences (multipotent progenitor neuronal cells versus fibroblasts) in terms of turnover rate for [beta]-subunit mRNA or protein, or in terms of transcriptional regulation of the [beta]-subunit gene. In the present study, we show that HexB activity is decreased by partial titration by [alpha]-chains. In addition, Laccoraza et al. do not address the issue of the contribution of HexS to their HexA activity by using the electrophoresis separation method. In fact, it is probable that accumulation of recombinant [alpha]-chains in non-deficient neuronal cells already expressing their own hexa gene led to formation of a significant level of HexS, as shown here.
In conclusion, our results show that the G.HEXA retroviral gene transferred in murine deficient fibroblasts allows for synthesis, secretion and re-uptake of HexA, which confirms the potential of this gene therapy approach for Tay-Sachs disease and, most likely, for other lysosomal diseases.
Indeed, secretion of lysosomal enzymes by transduced cells and recapture by deficient non-transduced cells would in theory provide a minimal enzyme activity in most cells of the organism, provided that the enzyme can be delivered to both sides of the blood brain barrier. Delivery to neuronal cells would be especially crucial for Tay-Sachs disease in which neurological symptoms are predominant. Delivery of HexA to brain cells would require either targeting of the enzyme to the central nervous system (29), or deliverance into the central nervous system, i.e : by transduction of ependymal cells (30), or by implantation of transduced cells in the brain (31,32). Since individuals with a HexA pseudodeficiency (10% of residual activity) show no symptoms, a low level of correction could be sufficient to improve evolution of the Tay-Sachs disease.
We currently are testing this approach on the HexA-deficient mouse model obtained by targeted gene disruption of the hexa murine gene (13). Although these animals do not faithfully reproduce symptomatology of the Tay-Sachs disease, they exhibit detectable accumulation of GM2 ganglioside in some neuronal cells as early as 3 months of age (33). We will therefore be able to study the effect of our strategy on enzyme activity and GM2 ganglioside accumulation in differents organs, on both sides of the blood-brain barrier.
MATERIALS AND METHODS
Cell culture
[Psi]-CRIP packaging cells and NIH-3T3 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS). Primary fibroblast cultures were established from newborn hexa -/- or C57BL/6 mice skin biopsy fragments treated with 0.5% collagenase (Boehringer Mannheim, Germany) for 90 min at 37°C. Cells were plated in 60 mm culture plates and grown in DMEM supplemented with 10% FCS.
Retroviral vector
The human [beta]-hexosaminidase [alpha]-subunit cDNA obtained from the pAdRSV.Hexa plasmid (9) was modified by replacement of the original ATG by an NcoI site generating a consensus Kozak sequence to enhance the translation efficacy, and by elimination of the polyadenylation signal sequence. The 1.7 kb modified cDNA was subcloned into pSP72 vector to give the p.J.HEXA plasmid. This plasmid was then cut by BamHI and BglII and ligated into the Moloney-based retroviral vector pM48 (gift of O. Danos, Evry, France) (27) digested by BamHI. The HEXA cDNA was under the control of the internal mouse PGK promoter. The construction pG.HEXA was used to transfect the [Psi]-CRIP amphotropic packaging cell line (gift of A. Weber-Benarous, Paris, France) (34) along with pRSV-neo, and a high titre producer clone was isolated (G.HEXA.10 vector; Fig. 1).
G.HEXA 10 producer cell supernatants were harvested, filtered on a 0.22 µm filter and used to transduce HexA-deficient murine primary fibroblasts in the presence of 8 µg/ml Polybrene.
Molecular hybridization
Southern and northern blots were performed as described previously (35) using a HEXA cDNA probe labelled with[[alpha]-32P]dCTP by random primer extension.
Immunoblotting
Cells were resuspended in water in the presence of 1% (v/v) NP-40. Protein extracts were obtained in the supernatant after four rounds of freezing-thawing and centrifugation for 1 h at 15 000 r.p.m. Twenty µg of protein extracts were denatured and fractionated by 10% (w/v) SDS-PAGE. After electrophoresis, proteins were transferred to Hybond C filters (Amersham) and incubated with a purified goat anti-human [beta]-hexosaminidase [alpha]-chain antibody (antiserum 803, a gift of E. Neufeld, Los Angeles, CA) (26). The specific immune complexes were detected using a rabbit anti-goat IgG horseradish peroxidase (HRP)-conjugated secondary antibody by ECL.
Enzyme assays
Protein extracts were prepared as described for immunoblotting. The [beta]-hexosaminidase activities were determined using 4-methylumbelliferyl-2-acetamido-2-deoxy-[beta]-d-glucopyranoside (4-MUG) as a substrate. HexA was specifically assayed on 4-MUGS, the [alpha]-chain specific substrate, as described (36). Specific enzymatic activities were calculated as nmol/h/mg of protein.
Thermostability of [beta]-hexosaminidases
Protein extracts (10 µg) were prepared in 10-3 M citrate phosphate buffer pH 4.5. Thermal inactivation was performed at 52°C for 2 h. HexB activity was assayed with 4-MUG.
[beta]-Hexosaminidase electrophoresis
Protein extracts (20 µg) were incubated overnight with rabbit anti-human HexA antibody or with phosphate-buffered saline (PBS) at 4°C. Electrophoresis was performed as previously described (37) for 2 h at 4°C under 200 V on cellulose acetate strips (Cellogel; Chemetron, Milan, Italy) in 0.04 M phosphate buffer, pH 6, followed by incubation with the 4-MUG substrate.
Uptake of [beta]-hexosaminidases
Transduced cells were maintained in serum-free medium for 48 h. Then, supernatants were filtered on a 0.22 µm filter, concentrated 10-fold on an ultrafiltration column (ultrafree-15 Biomax-30; Millipore, Molsheim, France) by centrifugation at 3000 r.p.m. for 45 min at 10°C. HexA-deficient mouse fibroblasts were incubated for 24 h with 2 ml of concentrated supernatant. Inhibition of the uptake was tested by adding 5 mM M6P in concentrated supernatant. Enzymatic assays were performed on protein extracts from these cells.
ACKNOWLEDGEMENTS
We thank O. Danos for the pM48 plasmid and M48 retroviral producer cells, A. Weber-Benarous for the [Psi]-CRIP packaging cell line, E. Proia for the human [alpha]-chain antibody, B. Viollet and J. Manicom for helpful discussions, and C. Caillaud and H. Gilgenkrantz for help in preparation manuscript. This work was supported in part by grants from Vaincre les Maladies Lysosomales. J.E.G received sponsorship from Rhône-Poulenc Rorer, the C.N.R.S and Vaincre les Maladies Lysosomales.
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 4 Apr 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Martino, C. Emiliani, B. Tancini, G. M. Severini, V. Chigorno, C. Bordignon, S. Sonnino, and A. Orlacchio Absence of Metabolic Cross-correction in Tay-Sachs Cells. IMPLICATIONS FOR GENE THERAPY J. Biol. Chem., May 31, 2002; 277(23): 20177 - 20184. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||