Titration of AAV-2 particles via a novel capsid ELISA: packaging of genomes can limit production of recombinant AAV-2


We demonstrate the rapid and reliable quantification of physical AAV-2 (adeno-associated virus type 2) particles via a novel ELISA based on a monoclonal antibody which selectively recognizes assembled AAV-2 capsids. Titration of a variety of recombinant AAV-2 (rAAV) preparations revealed that at least 80% of all particles were empty, compared with a maximum of 50% in wild-type AAV-2 stocks, indicating that the recombinant genomes were less efficiently encapsidated. This finding was confirmed upon titration of CsCl gradient fractions from recombinant and wild-type AAV-2 stocks. ELISA-based measurement of capsid numbers revealed a large number of physical particles with low densities corresponding to empty capsids in the recombinant, but not in the wild-type AAV-2 preparations. Moreover, additional expression of VP proteins during rAAV production was found to result in an excessive capsid formation, whilst yielding only minor increases in DNA-containing or transducing rAAV particles. We conclude that encapsidation of viral genomes rather than capsid assembly can be limiting for rAAV production, provided that a critical level of VP expression is maintained. The feasibility of quantifying AAV-2 capsid numbers via the ELISA allows determination of physical to DNA-containing or infectious particle ratios. These are important parameters which should help to optimize and standardize the production and application of recombinant AAV-2.


Gene therapy vectors derived from the human parvovirus AAV-2 (adeno-associated virus type 2) have gained attention owing to a unique combination of attractive features. Wild-type AAV-2 is nonpathogenic in humans and naturally defective, requiring coinfection with a helpervirus (eg adenovirus) for a productive infection.1 Recombinant AAV-2 can infect both dividing and nondividing cells in vitro and in vivo2,3,4,5,6 and have the potential for site-specific integration into chromosome 19.7,8,9 Long-term expression of heterologous genes transduced by rAAV has been observed in a variety of human cells and tissues, such as muscle,10,11 lung,12 central nervous system,13,14,15 retina16 and liver.17

An important prerequisite for the testing of AAV-2 vectors in preclinical and clinical studies is the accurate and reliable titration of the recombinant virus particles. Precise information on rAAV titers is not only crucial for the careful planning and execution of such studies, but also for comparing results amongst laboratories. In brief, presently available methods for rAAV titration can be divided into biological and physical assays. Biological assays rely on infection of cultured cells followed by events that depend on the biological functionality of the rAAV vector, ie either replication of the recombinant genomes in cells in the presence of AAV-2 and adenoviral helper functions, or expression of the transduced heterologous gene.18,19,20 These two types of assay yield titers of infectious or transducing particles, respectively. In contrast, physical methods are independent of biological functions of the recombinant viruses. Typically, viral DNA is extracted from the rAAV particles via enzymatic digestion of the capsids and quantified using serially diluted plasmid DNA bearing the transgene as standard.18,21 The number of DNA-containing particles is then calculated assuming that each virion carries one single-stranded DNA molecule with a defined size.

In spite of being rather simple in concept and execution, all common rAAV titration methods display distinct disadvantages. Biological methods are highly dependent on the properties of the vectors and the particular assay conditions, which are prone to differ among various laboratories. For example, quantification of functional transducing rAAV particles is influenced by the cell type used, the promotor driving the transgene or the transgene itself.22,23 Another crucial parameter is the helper virus, which can increase the efficiency of rAAV-mediated gene transduction and consequently the functional rAAV titers by three orders of magnitude.23,24,25 Similarly, physical methods to measure DNA-containing particles involve a number of steps, in particular exhaustive enzymatic treatment of the virions, which are possibly inconsistent in different laboratories and thus may provoke variable results. As a consequence, rAAV titers as well as the ratios of physical to functional particles, which are considered as important indices for the quality of rAAV preparations, show a great variation, thus making direct comparisons of results from different studies quite difficult. Moreover, whilst quantification of encapsidated viral genomes is usually considered to yield total particle titers, this method of course fails to detect empty virions which are devoid of DNA. In fact, there is a lack of methods for quantification of total numbers of physical rAAV particles, ie the sum of assembled packaged or empty virions. Yet, in view of the possibility that AAV-2 capsid proteins can elicit a humoral host immune response, methods to determine exact numbers of total AAV-2 capsids in a given rAAV preparation are urgently needed.

In this report, we present a novel AAV-2 capsid ELISA that fits in this gap by allowing the rapid and reliable quantification of assembled, empty or full, AAV-2 particles. Given this property, the ELISA not only permits a better characterization of rAAV preparations, but also a more profound investigation of the AAV-2 genome packaging process in general. Indeed, we obtained and present first evidence that encapsidation of recombinant AAV-2 genomes occurs rather inefficiently and can be a limiting step for rAAV vector production.

A sandwich ELISA allowing the quantification of assembled AAV-2 particles was developed based on a previously described monoclonal antibody, A20, that specifically binds to assembled AAV-2 capsids, but not to single capsid proteins.26 A scheme depicting further details is shown in Figure 1a. To calibrate the ELISA, stocks of empty AAV-2 capsids or full AAV-2 virions were used as particle standards. The empty capsids were prepared from 293 cells infected with a recombinant adenovirus which expressed the AAV-2 cap gene under the control of a human CMV promoter (rAdVP) and purified as described.27 To determine the maximum number of AAV-2 capsids in this empty particle standard, the protein concentration of an aliquot was determined using a commercial protein quantification assay (BioRad, München, Germany). Assuming a stoichiometry of the three VP proteins of 1:1:10 in the assembled capsid (Figure 1b), an approximate number of 1.2 × 1012 capsids/ml was deduced from the total protein concentration of 7.8 ng/μl measured for the empty particle standard. In addition, the capsid number was determined by mixing equal amounts of the empty particle stock and an adenovirus 5 (Ad-5) stock of known titer (5 × 1010 pfu/ml) and by counting the AAV-2 and Ad-5 particles in 40 randomly chosen electron microscopy pictures of this mixture (Figure 1c). A 34.6-fold excess of AAV-2 capsids relative to Ad-5 capsids was counted, and since the Ad-5 stock contained 2.3 × 1011 capsids/ml (determined according to Ref. 28), the particle standard was calculated to have 8 × 1012 empty AAV-2 capsids/ml (assuming that AAV-2 and Ad-5 particles bind equally well to the EM grids). Likewise, a preparation of CsCl-purified, UV-inactivated wild-type AAV-2 particles was established as full particle standard, having a similar titer of 7.9 × 1012 capsids/ml.

Figure 1

Development of an AAV-2 capsid ELISA. (a) Scheme depicting the principle of the AAV-2 ELISA. 96-well cell culture plates were coated at 4°C overnight with purified monoclonal antibody A20 (200 ng in 100 μl per well) in 13 mM Na2CO3, 35 mM NaHCO3, pH 9.6. Purification of the A20 antibody has been described.20 The plates were washed three times with PBS, 0.05% Tween-20 (wash buffer) and blocked with 0.2% casein, 0.05% Tween-20 in PBS (blocking reagent, 200 μl per well) for 2 h at room temperature. For AAV-2 titration, virus stocks were added to the wells in five- or 10-fold serial dilutions (in wash buffer, final volume 100 μl) and incubated for 1 h at room temperature. As negative control, 100 μl wash buffer were applied to a separate well. The plates were washed as described above and then incubated with biotinylated A20 antibody diluted in blocking reagent (100 ng A20 per well) for 1 h at room temperature. Biotinylation of A20 was performed using the ECL protein biotinylation module from Amersham (Braunschweig, Germany). Biotinylated A20 was purified using G25 sephadex columns (Pharmacia, Freiburg, Germany) and tested for successful biotinylation via dot blot assay using streptavidin-coupled peroxidase (Dianova, Hamburg, Germany) and tetramethylbencidine (TMB) as substrate. The plates were washed and incubated with streptavidin(StAv)-coupled peroxidase (PO) diluted in blocking reagent to a final concentration of 1 μg/ml (100 μl per well) for 1 h at room temperature. After washing three times with wash buffer, the plates were incubated with 100 μl substrate (0.01% TMB, 0.1 M sodium acetate, 0.003% peroxide) per well for 10 min at room temperature. The reaction was stopped by adding 100 μl 1 M H2SO4 to each well, and absorption at 450 nm was measured using an EMax Elisa reader (MWG, Ebersberg, Germany). The background absorption obtained for the negative control (usually less than 0.1) was subtracted from all other values. The calibration of the AAV-2 capsid ELISA is described in the text. (b) An aliquot of a preparation of empty AAV-2 capsids that was used as standard for the ELISA was separated via SDS polyacrylamide gel electrophoresis (lane c) and stained with Coomassie blue. Lane M contained several polypeptides that served as size markers (from top to bottom phosphorylase b, bovine serum albumin, ovalbumin and carbonic anhydrase; the sizes in kilodalton are indicated on the left). The purity of the particle standards was usually greater than 95%. (c) An aliquot of the full wild-type AAV-2 particle standard was mixed with an aliquot of an Ad-5 stock of known capsid titer and analyzed via electron microscopy. Shown is one out of 40 pictures taken of this virus mixture. The ratio of AAV-2 (small) relative to Ad-5 capsids (large) was quantified and the number of physical particles in the AAV-2 preparation was calculated (see text). (d) Typical ELISA titration curves obtained with empty AAV-2 capsids or UV-inactivated wild-type AAV-2 as particle standards. Absorbance values at 450 nm (OD450) measured for the serially diluted AAV-2 particle standards were plotted against particle numbers (as determined by protein quantification and particle counting, see text), and standard titration curves were derived by ‘eye-fitting’. Since empty AAV-2 capsids can easily be produced in abundant amounts (see text), they were preferred as standard for all experiments described here.

Serial two-fold dilutions of the empty or full particle standard were used to obtain ELISA titration curves (examples are shown in Figure 1d). In the experiments reported below, only the linear part of the standard curves, ie the part between optical densities of 0.4 and 1.6 (corresponding to 107 to 108 particles per 100 μl), was used to derive values for unknown samples. The coefficient of variation in this part of the curves was 3–25%, but increased up to 67% at lower optical densities (calculated from 10 standard curves). Therefore, we consider the optical density of 0.4, which corresponds to approximately 107 assembled particles per sample, as the quantification limit of the AAV-2 capsid ELISA, while the actual detection limit is somewhat lower (approximately 106 particles). This renders the AAV-2 ELISA sensitive enough to be used for titration of any rAAV preparation, assuming an average generation of at least 104 assembled particles per cell during rAAV production.

The finding that the curves derived from empty or full particle standard samples with similar titers were nearly identical proved that the ELISA can be applied to the titration of both types of AAV-2 particles. This was expected since the ELISA relies on antibody(A20)-mediated recognition of an epitope which is supposed to be present on any assembled AAV-2 capsid, irrespective of being empty or full. Furthermore, earlier experiments have already shown that the monoclonal antibody A20 recognizes both wild-type and recombinant particles.20,26 Thus, the AAV-2 capsid ELISA not only provides a high grade of versatility, but is also highly suitable for standardization between laboratories, as it allows the titration of any type of AAV-2 particle via one invariant protocol.

Finally, spiking experiments were performed to address the question if sample matrix effects would influence the outcome of the ELISA measurement. Aliquots of a stock of purified empty capsids were mixed with different aliquots of cell culture medium, CsCl or MgCl2 and titrated via ELISA. Except for MgCl2 at concentrations above 20 mM, particle titers were found to vary less than 20%, which was within the range that could be expected from the variation coefficient of the standard curves (see above). From these findings, it was concluded that ELISA-based titration is not influenced by sample matrix effects.

Following the general characterization of the ELISA, we analyzed how the assembled particle titers obtainable via the ELISA would relate to titers determined by common methods for AAV-2 titration. Therefore, we first prepared a series of wild-type AAV-2 stocks (Table 1) and titrated them via the ELISA to gain numbers of assembled particles. In parallel, these stocks were analyzed via commonly used methods to determine corresponding numbers of DNA-containing, infectious or transducing particles. For details and results see Table 1. As expected, the assembled particle titers determined via ELISA were always highest, reaching up to approximately 1012 capsids/ml for the AAV-2 stocks derived from infection, and being only slightly lower for those generated by transfection. Importantly, the titers of DNA-containing particles were generally two- to four-fold lower than the assembled particle titers. This was clearly in line with the assumption that only a portion of the total particles present in an AAV-2 stock has packaged DNA or is infectious. Assuming that the DNA quantification method yielded reliable DNA-containing particle titers, the data suggest that a minimum of 25 up to 50% of the AAV-2 capsids were packaged or somehow associated with viral genomes. The numbers of infectious or transducing AAV-2 particles were one to three orders of magnitude lower than the numbers of assembled particles and at least 10-fold lower than those of the DNA-containing particles (Table 1 and Figure 2a-c).

Table 1 Titration of wild-type AAV-2 stocks
Figure 2

Excess of empty capsids in rAAV stocks. (a–c) Ratios of physical to DNA-containing, infectious or transducing particles. Wild-type AAV-2 stocks (left of the broken line) were prepared either via infection of 293 cells with an aliquot of an AAV-2 virus stock (MOI 10) or via transfection of 293T cells with pSSV9 or pTAV2–0 (see Table 1). The cells were additionally infected with Ad-5 (MOI 10). Recombinant AAV-2 (right of the broken line) were generated via cotransfection of 293T or 293 (marked with asterisks) cells with pΔTR or pDG and one of the indicated vector plasmids (see Table 2). pΔTR-transfected cells were infected with Ad-5 (MOI 10). Shown are mean values with standard deviations of at least three independently generated and titrated virus stocks each. As an exception, the cotransfections of 293 cells with pDG and pTRUFlacZ or pACVlacZ were carried out only twice. (d–f) Titration of wild-type AAV-2 and rAAV(lacZ) particles fractionated via CsCl gradients. A wild-type AAV-2 stock was made via coinfection of 293 cells with AAV-2 and Ad-5 (MOI 10 each), while rAAV(lacZ) particles were generated via cotransfection of 293T cells with either pTRUFlacZ or pACVlacZ and the pDG helper plasmid. For CsCl purification, the supernatants were first centrifuged at 38 000 g and 4°C for 3 h. The pelleted virus particles were then resuspended in 1 ml 1% desoxycholate and incubated for 30 min at 37°C. CsCl was added to a final volume of 8 ml and adjusted to a density of 1.40 g/cm3. A CsCl gradient was formed by centrifugation at 211 000 g and 20°C for 24 h using a fixed angle rotor (TFT 65.13, Kontron Instruments, Neufahrn, Germany). Fractions of 0.5 ml each with densities between 1.30 and 1.50 g/cm3 (plotted on the X axis) were collected and titers of physical (d), DNA-containing (e) and transducing (f) particles (Y axis) determined (as described in legends to Tables 1 and 2). Titration of the physical wild-type AAV-2 particles revealed two peaks for empty or full capsids, respectively, as indicated by the open and filled capsid symbols and arrows. The rAAV peaks corresponding to full particles decreased and were shifted to higher densities of approximately 1.45 g/cm3, whereas the peaks of empty particles increased (top picture). In each experiment, so-called defective-interfering particles, ie capsids which were only partially loaded with viral genomes or contained deleted genomes,38 were found at intermediate densities of 1.35 to 1.41 g/cm3, thus forming a capsid ‘smear’ throughout the gradients. Titration of DNA-containing or transducing particles (middle and bottom picture) showed distinct peaks for wild-type or recombinant AAV-2, respectively, at different densities (marked by arrows).

In the course of these first titration experiments, the AAV-2 ELISA proved to be very reliable. Repeated titration of one distinct AAV-2 stock under consistent conditions (as described in Figure 1a) gave capsid titers which were highly reproducible. The inter-assay variation was actually found to be less than 20%, which is most likely due to the fact that ELISA-based AAV-2 titration, in contrast to common methods, neither involves steps which are prone to influence the particle measurement, such as prior enzymatic treatment of the virions, nor any biological components, eg cells or a helper virus.

Subsequent experiments aimed at confirming that the ELISA could also be used for the titration of recombinant AAV-2 particles. Table 2 lists a variety of rAAV stocks that were generated and then analyzed in a manner identical to the wild-type AAV-2 stocks. In general, most of the various titers measured for the rAAV stocks prepared using the pDG helper plasmid were several-fold higher than those obtained via transfection and infection of the cells with the pΔTR helper construct and Ad-5, respectively. This confirms and extends our own previous findings that rAAV production using pDG is more efficient than using a standard Rep-/VP-expressing packaging plasmid.20

Table 2 Titration of recombinant AAV-2 virus stocks

Similar to wild-type AAV-2, the assembled particle titers obtained via the ELISA were highest in each of the rAAV stocks analyzed, being on average only slightly below those calculated for the wild-type virus particles (Tables 1 and 2). The numbers of DNA-containing rAAV particles, however, were markedly lower than the particular wild-type AAV-2 titers, resulting in higher average ratios of physical to DNA-containing rAAV particles of approximately 5 up to 60:1. These ratios depended mainly on the vector plasmid used but seemed not to be influenced by the choice of cell line or helper plasmid (Table 2 and Figure 2a–c). This indicated that, as compared with wild-type AAV-2, fewer capsids were loaded with recombinant genomes, namely a maximum of only 20% (eg pTRUFlacZ) down to 1.7% (pAVJE). The numbers of infectious or transducing rAAV showed an even stronger decrease not only compared with wild-type AAV-2, but in particular to the rAAV capsid titers, indicating that only a very small portion of less than 0.1% of the assembled rAAV particles were actually infectious or capable of transduction in these assays (Table 2 and Figure 2a–c).

Caution was taken, however, when interpreting these initial findings which were based on comparisons of capsid and genome numbers. This was because the DNA quantification method used is highly dependent upon a number of parameters and thus possibly yielded DNA-containing particle titers that were not fully reliable. In particular, it can not be excluded that numbers of DNA-containing particles were overestimated due to replicated genomes outside the capsids that were protected against DNaseI digestion (eg by Rep or cellular proteins). The preponderance of empty capsids in rAAV stocks thus needed additional confirmation. Therefore, a wild-type AAV-2 and two rAAV(lacZ) stocks were subjected to CsCl density gradient centrifugation and gradient fractions having densities of 1.30 to 1.50 g/cm3 were analyzed using the same titration methods as before (Figure 2d-f). The quantification of either DNA- containing or transducing wild-type AAV-2 particles revealed one distinct peak at a density of 1.40–1.43 g/cm3, where full and infectious particles were expected to be found.18,37,38 Importantly, ELISA-based measurement of assembled particles resulted in a peak which was not only the strongest one, but also matched the peaks of the DNA-containing or transducing particles. A second, but smaller culmination of physical particles was found at a density of around 1.33 g/cm3, where empty particles were supposed to accumulate.18,37,38 Two conclusions are drawn: firstly, the fact that measurement of capsid concentrations gave two peaks at the expected densities which complemented the data obtained using common titration methods strongly proved that the ELISA is dependable. Second, the appearance of a distinct peak of assembled particles being paralleled by culminations of DNA-containing and transducing particles reinforced the initial finding that a large portion of wild-type AAV-2 virions is actually packaged and capable of transduction.

Analyses of the two rAAV(lacZ) stocks gave distinct peaks for DNA-containing and transducing particles that matched each other, although as compared with wild-type AAV-2, a shift to a higher density of 1.43–1.46 g/cm3 was noticed. However, ELISA-based titration revealed a predominant peak of assembled rAAV particles at a low density of 1.33–1.36 g/cm3, indicating a large portion of empty capsids. Vice versa, an ELISA peak matching the majority of DNA-containing and transducing rAAV particles at a density of about 1.44–1.45 g/cm3 was barely detectable (Figure 2d–f). A larger portion of empty versus full particles was also observed after CsCl gradient fractionation of rAAV stocks that were generated using the pΔTR plasmid (data not shown). Taken together, all results clearly confirmed the preponderance of non- packaged and non-transducing particles in rAAV stocks. Moreover, this finding is also clearly in line with our recently published electron microscopy analysis of rAAV stocks in which the majority of the observed rAAV particles were found to be empty.20 Two reasons for this phenomenon can be speculated: recombinant AAV-2 genomes could either be replicated less efficiently in cells than wild-type genomes, yielding fewer single stranded DNA molecules ready for encapsidation. Alternatively, recombinant AAV-2 genomes might also be lacking some yet undefined elements required for efficient DNA encapsidation additional to the terminal repeats. However, discriminating between these possibilities would have required additional investigation which was beyond the range of this report.

Irrespective of the exact reasons, the pure finding that rAAV preparations are characterized by an excess of empty capsids is of particular importance with respect to rAAV production. Several reports have recently shown that an efficient rAAV production depends on a strong VP protein expression,39,40 which, as demonstrated by others, itself is a prerequisite for a high rate of AAV-2 capsid assembly.20,41 Based on these correlations, we finally investigated whether further increasing VP protein expression and thus the rate of AAV-2 capsid assembly during rAAV production would result in a concomitant increase in assembled and functional rAAV particle titers. Therefore, several rAAV stocks were generated via cotransfection of 293 cells with pDG and pTRUFlacZ and by additionally overinfecting the cells with the rAdVP virus (see above). Control rAAV stocks were prepared under identical conditions, but leaving out the rAdVP infection. Titration of the stocks derived from the rAdVP-infected cells revealed a two- to five-fold increase in DNA-containing and only marginal changes of transducing rAAV particle numbers as compared with the control stocks (Figure 3a). However, the capsid numbers were up to 25-fold increased, indicating the generation of a massive excess of non-packaged and non-transducing assembled particles as a result of the rAdVP infection (Figure 3a).

Figure 3

rAAV production under conditions of increased AAV-2 VP expression. (a) 293 cells were transfected with pDG and pTRUFlacZ (left side, control) as described in Table 2. In parallel, cells were transfected and additionally infected with a recombinant adenovirus (rAdVP) having the E1 region replaced by the AAV-2 cap gene under the control of a human CMV promoter at a MOI of 10 (right side, +rAdVP). The generation of the rAdVP will be described in detail elsewhere (Ferrari et al, manuscript in preparation). Shown are mean titers of physical (black bars), DNA-containing (white bars) and transducing (grey bars) rAAV particles obtained in three independent experiments including standard deviations (for details of titration methods see Tables 1 and 2). (b) Western blot analysis. 293T were transfected with pDG and pTRUFlacZ in molar ratios of 10:1, 5:1, 1:1, 1:5 and 1:10 (indicated by the triangles). Total cell extracts were prepared and separated via SDS polyacrylamide gel electrophoresis, and AAV-2 Rep and VP proteins were detected using the monoclonal antibodies 303.9 and B1 as described.20 (c) rAAV particles were extracted from the transfected cells (see b) 2 days after transfection and titers of physical (black bars), DNA-containing (white bars) and transducing (grey bars) particles determined (as described in Tables 1 and 2). Shown are mean values and standard deviations of three independent experiments.

Similar observations were made when VP protein expression was elevated by transfection of 293T cells with increasing molar amounts of pDG relative to pTRUFlacZ (Figure 3b/c). A 1:1 molar ratio of helper and vector plasmid resulted in the production of a seven- to eight-fold excess of assembled to DNA-containing particles, while yielding the highest transducing particle titers. As expected, transfection of a five- or 10-fold molar excess of pDG was followed by an elevated production of Rep and VP proteins (in a constant stoichiometry, Figure 3b) and consequently an increased generation of assembled capsids (as detected via ELISA, Figure 3c). On the other hand, the numbers of DNA-containing or transducing particle titers slightly declined, which led to increased ratios of assembled to DNA-containing or to transducing particles. After transfection of up to 10-fold lower molar amounts of pDG than pTRUFlacZ, AAV-2 protein expression as well as titers of assembled or transducing particles decreased, and only the numbers of DNA- containing particles remained constant (Figure 3b/c). Thus, the excess of assembled to DNA-containing and to a lesser extent, transducing particles was significantly reduced under those conditions, however, the absolute transducing particle titers were also lowest. Similar results were obtained using the 293 cell line except that the absolute particle titers were all several-fold lower (data not shown).

In sum, these two experiments show that VP expression is only limiting for rAAV production until a certain threshold of assembled capsids, ready to get packaged with recombinant genomes, is reached. Rather than improving yields of DNA-containing or transducing particles, further overexpression of VP proteins only results in the excessive production of empty capsids. This is of course highly undesirable since under these conditions host immune responses to VP proteins become more likely. Together these findings again strongly indicate that as yet undefined elements involved in DNA replication or encapsidation, which are missing on common rAAV vectors, but are present on wild-type genomes, are generally limiting rAAV production. With the feasibility of measuring capsid numbers and ratios of empty to full particles given now by the AAV-2 ELISA, future studies should aim at elucidating this obvious block in replication or encapsidation of rAAV genomes. The identification of these elements or mechanisms required for an optimum yield of packaged AAV-2 particles would certainly have a significant impact on the production of rAAV vectors for human gene therapy.


  1. 1

    Berns KI, Bohensky RA . Adeno-associated viruses: an update Adv Vir Res 1987 32: 243–305

  2. 2

    Alexander IE, Russell DW, Spence AM, Miller AD . Effects of gamma irradiation on the transduction of dividing and nondividing cells in brain and muscle of rats by adeno-associated virus vectors Hum Gene Ther 1996 7: 841–850

  3. 3

    Kaplitt MG, Makimura H . Defective viral vectors as agents for gene transfer in the nervous system J Neurosci Meth 1997 71: 125–132

  4. 4

    McCown TJ et al. Differential and persistent expression patterns of CNS gene transfer by an adeno-associated virus (AAV) vector Brain Res 1996 713: 99–107

  5. 5

    Podsakoff G, Wong KK, Chatterjee S . Efficient gene transfer into nondividing cells by adeno-associated virus-based vectors J Virol 1994 68: 5656–5666

  6. 6

    Russell DW, Miller AW, Alexander IE . Adeno-associated virus vectors preferentially transduce cells in S phase Proc Natl Acad Sci USA 1994 91: 8915–8919

  7. 7

    Kotin RM et al. Site-specific integration by adeno-associated virus Proc Natl Acad Sci USA 1990 87: 2211–2215

  8. 8

    Samulski RJ et al. Targeted integration of adeno-associated virus (AAV) into human chromosome 19 EMBO J 1991 10: 3941–3950

  9. 9

    Balague C, Kalla M, Zhang WW . Adeno-associated virus Rep78 protein and terminal repeats enhance integration of DNA sequences into the cellular genome J Virol 1997 71: 3299–3306

  10. 10

    Kessler PD et al. Gene delivery to skeletal muscle results in sustained expression and systemic delivery of a therapeutic protein Proc Natl Acad Sci USA 1996 93: 14082–14087

  11. 11

    Xiao X, Li J, Samulski RJ . Efficient long-term gene transfer into muscle tissue of immunocompetent mice by adeno-associated virus vector J Virol 1996 70: 8098–8108

  12. 12

    Flotte TR et al. Stable in vivo expression of the cystic fibrosis transmembrane conductance regulator with an adeno-associated virus vector Proc Natl Acad Sci USA 1993 90: 10613–10617

  13. 13

    Kaplitt MG et al. Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain Nat Genet 1994 8: 148–154

  14. 14

    Klein RL et al. Neuron-specific transduction in the rat septohippocampal or nigrostriatal pathway by recombinant adeno-associated virus vectors Exp Neurol 1998 150: 183–194

  15. 15

    Peel AL et al. Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters Gene Therapy 1997 4: 16–24

  16. 16

    Flannery JG et al. Efficient photoreceptor-targeted gene expression in vivo by recombinant adeno-associated virus Proc Natl Acad Sci USA 1997 94: 6916–6921

  17. 17

    Snyder RO et al. Persistent and therapeutic concentrations of human factor IX in mice after hepatic gene transfer of recombinant AAV vectors Nat Genet 1997 16: 270–276

  18. 18

    Snyder RO, Xiao X, Samulski RJ . Production of recombinant adeno-associated viral vectors. In: Dracopoli N et al (eds) Current Protocols in Human Genetics John Wiley: New York 1996 pp 12.1.1–12.1.24.

  19. 19

    Atkinson EM, Debelak DJ, Hart LA, Reynolds TC . A high-throughput hybridization method for titer determination of viruses and gene therapy vectors Nucleic Acids Res 1998 26: 2821–2823

  20. 20

    Grimm D, Kern A, Rittner K, Kleinschmidt JA . Novel tools for production and purification of recombinant adeno-associated virus vectors Hum Gene Ther 1998 9: 2745–2760

  21. 21

    Flotte TR et al. Gene expression from adeno-associated virus vectors in airway epithelial cells Am J Respir Cell Mol Biol 1992 7: 349–356

  22. 22

    Salvetti A et al. Factors influencing recombinant adeno-associated virus production Hum Gene Ther 1998 9: 695–706

  23. 23

    Clark KR, Voulgaropoulou F, Johnson PR . A stable cell line carrying adenovirus-inducible rep and cap genes allows for infectivity titration of adeno-associated virus vectors Gene Therapy 1996 3: 1124–1132

  24. 24

    Ferrari FK, Samulski T, Shenk T, Samulski RJ . Second-strand synthesis is a rate-limiting step for efficient transduction by recombinant adeno-associated virus vectors J Virol 1996 70: 3227–3234

  25. 25

    Fisher KJ et al. Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis J Virol 1996 70: 520–532

  26. 26

    Wistuba A et al. Subcellular compartmentalization of adeno-associated virus type 2 assembly J Virol 1997 71: 1341–1352

  27. 27

    Steinbach S, Wistuba A, Bock T, Kleinschmidt JA . Assembly of adeno-associated virus type 2 capsids in vitro J Gen Virol 1997 78: 1453–1462

  28. 28

    Mittereder N, March KL, Trapnell BC . Evaluation of the concentration and bioactivity of adenovirus vectors for gene therapy J Virol 1996 70: 7498–7509

  29. 29

    Chen CA, Okayama H . Calcium phosphate-mediated gene transfer: a highly efficient transfection system for stably transforming cells with plasmid DNA Biotech 1988 6: 632–638

  30. 30

    Samulski RJ, Berns KI, Tan M, Muzyczka N . Cloning of adeno-associated virus into pBR322: rescue of intact virus from the recombinant plasmid in human cells Proc Natl Acad Sci USA 1982 79: 2077–2081

  31. 31

    Samulski RJ, Chang LS, Shenk T . Helper-free stocks of recombinant adeno-associated viruses: normal integration does not require viral gene expression J Virol 1989 63: 3822–3828

  32. 32

    Heilbronn R, Bürkle A, Stephan S, zur Hausen H . The adeno-associated virus rep gene suppresses herpes simplex virus-induced DNA amplification J Virol 1990 64: 3012–3018

  33. 33

    Zolotukhin S et al. A ‘humanized’ green fluorescent protein cDNA adapted for high-level expression in mammalian cells J Virol 1996 70: 4646–4654

  34. 34

    de Wet JR et al. Firefly luciferase gene: structure and expression in mammalian cells Mol Cell Biol 1987 7: 725–737

  35. 35

    MacGregor GR, Caskey CT . Construction of plasmids that express E. coli beta-galactosidase in mammalian cells Nucleic Acids Res 1989 17: 2365

  36. 36

    Rittner K, Stoppler H, Pawlita M, Sczakiel G . Versatile eucaryotic vectors for strong and constitutive transient and stable gene expression Meth Mol Cell Biol 1991 2: 176–181

  37. 37

    Fisher KJ et al. Recombinant adeno-associated virus for muscle directed gene therapy Nature Med 1997 3: 306–312

  38. 38

    Myers MW, Carter BJ . Assembly of adeno-associated virus Virol 1980 102: 71–82

  39. 39

    Xiao X, Li J, Samulski RJ . Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus J Virol 1998 72: 2224–2232

  40. 40

    Vincent KA, Piraino ST, Wadsworth SC . Analysis of recombinant adeno-associated virus packaging and requirements for rep and cap gene products J Virol 1997 71: 1897–1905

  41. 41

    Weger S, Wistuba A, Grimm D, Kleinschmidt JA . Control of adeno-associated virus type 2 cap gene expression: relative influence of helper virus, terminal repeats, and Rep proteins J Virol 1997 71: 8437–8447

Download references


We are grateful to Dr Anna Salvetti for providing the HeLaRC32 cell line and to Dr Michael Chapman for supplying CsCl purified wild-type AAV-2. Andrea Hörster and Birgit Teichmann are thanked for their help with the FACS analyses. Thorsten Belz was involved in initial development of the ELISA. Dirk Grimm was supported by the BMBF grant 01KV9517/6.

Author information



Corresponding author

Correspondence to J A Kleinschmidt.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Grimm, D., Kern, A., Pawlita, M. et al. Titration of AAV-2 particles via a novel capsid ELISA: packaging of genomes can limit production of recombinant AAV-2. Gene Ther 6, 1322–1330 (1999). https://doi.org/10.1038/sj.gt.3300946

Download citation


  • AAV-2
  • recombinant AAV-2
  • AAV-2 titration

Further reading