J Sep Sci. 2016 Nov;39(22):4299-4304.

Manipulating adenoviral vector ion-exchange chromatography: hexon vs fiber


1University of Zagreb, Faculty of Science, Department of Biology, Marulićev trg 9a, 10000 Zagreb, Croatia

2Ruđer Bošković Institute, Division of Molecular Biology, Laboratory for Cell Biology and Signaling, Bijenička 54, 10000 Zagreb, Croatia

3BIA Separations, Mirce 21, 5270 Ajdovščina, Slovenia

4Univ Paris-Sud, 15 rue Georges Clémenceau, 91405 Orsay Cedex, France

5CNRS UMR 8203, Vectorologie et thérapeutiques anticancéreuses, Gustave Roussy, 114 rue Edouard Vaillant, 94805 Villejuif Cedex, France

*equal contribution

#corresponding author : mladen.krajacic@biol.pmf.hr, +385 1 4898 075, Marulićev trg 9a, 10000 Zagreb, Croatia

a present address: University College London, Department of Biochemical Engineering, Bernard Katz Building, Gordon Street, London, WC1H 0AH, UK

b present address: Singapore Centre on Environmental Life Sciences Engineering, Nanyang Technological University, SBS-01n-27, 60 Nanyang Drive, 637551 Singapore

c present address: Sandoz Biopharmaceuticals, Mengeš, Slovenia

Short title: Manipulating adenoviral vector chromatography


The serotype specificity of adenovirus ion-exchange chromatography has previously been studied using standard particle-based columns, and the hexon protein has been reported to determine retention time. In this study, we have submitted Adenovirus type 5 recombinants to anion-exchange chromatography using methacrylate monolithic supports. Our experiments with hexon-modified adenoviral vectors show more precisely that the retention time is affected by the substitution of amino acids in hypervariable region 5, which lies within the hexon DE1 loop. By exploring the recombinants modified in the fiber protein, we have proven the previously predicted chromatographic potential of this surface constituent. Modifications that preserve the net charge of the hexon protein, or those that cause only a small charge difference in the fiber protein, in addition to shortening the fiber shaft, did not change the chromatographic behavior of the adenovirus particles. However, modifications that include the deletion of just two negatively charged amino acids in the hexon protein, or the introduction of a heterologous fiber protein, derived from another serotype, revealed recognizable changes in anion-exchange chromatography. This could be useful in facilitating chromatography-approach purification by creating targeted capsid modifications, thereby shifting adenovirus particles away from particular interfering substances present in the crude lysate.


Anion-exchange chromatography, monolithic columns, recombinant adenoviral vectors, hexon protein, fiber protein



Adenoviruses have been used in almost 25 percent of human gene therapy trials, mostly in cancer treatment and vaccination. Human adenovirus serotype 5 (Ad5) is the most frequently used. Among three major structural proteins (Fig. 1), hexon protein is the most abundant, penton base protein is also significant, and the fibers extend from the penton bases [1].


The use of adenoviruses in clinical gene delivery has enormously increased the demand for developing new approaches to their purification, as well as for in-process control. Following column chromatography had emerged as an approach, monolithic chromatographic supports have been considered advantageous comparing to particle-based columns, particularly when dealing with large biomolecules and macromolecular complexes, including viruses [2-4]. A review on application of monoliths, exclusively focused on subjects of interest in virology, has recently been published [5].



Fig. 1. The main constitutive proteins of the adenoviral capsid, and modifications introduced in the recombinant viral vectors (as described precisely in J Sep Sci (2016) 39(22), 4299-4304.


In this work, recombinant adenoviral vectors, rearranged in the main structural proteins – hexon and fiber (Fig. 1), were analyzed to quantify changes required for the modification of viral chromatographic features. All the modifications have been created earlier with the purposes related to gene therapy (Fig. 2,3) [6-10]. The survey was extended by adding another vector, the unpublished “funny” double-recombinant with the antagonistic modifications, demonstrating that there is a range of manipulating strategies to be followed. The word-game used in the title “hexon vs fiber” stands particularly for compensation of the two opposite effects in a single adenovirus recombinant: the loss of negative charge in the hexon, and the gain of negative charge in the fiber (Fig. 4). Moreover, one can anticipate that manipulating potential might exist in some other hypervariable regions, across the capsid structure, that have been rarely exploited so far.


The figures below present a comparison of uninfected and Ad5 infected cell lysates by CIM QA chromatography in which E3 and E5 fractions were recognized as infection-specific. Following 10 min of centrifugation (16,060 g) and filtration (0.45 µm), the samples (300 µL) were loaded with 20 mM Tris, 5% glycerol, pH 7.5, and eluted with a gradient of loading/eluting buffer, where the eluting buffer additionally contains 2 M NaCl. The flow rate was 2 mL/min. FT–flow through, E1-8 – elution fractions.



Fig. 2. Shortened retention time of both the unassembled hexon protein (E3) and the virion fraction (E5) obtained for hexon-modified adenoviral vector (as reported in J Sep Sci (2016) 39(22), 4299-4304). The hypervariable region 5 was replaced by a non-targeting peptide containing glycine-alanine residua (reduction of the negative charge was achieved by this substitution). The shifted recombinant (Ad5 recomb) missed just two negatively charged aminoacids in the hexon protein comparing to original virus particle displaying wild-type capsid (Ad5 wild type). The infectivity of the recombinant vector was not influenced by this substitution.



Fig. 3. Prolonged retention time of the virion fraction (E5), but not of the free hexon protein (E3), obtained for fiber-pseudotyped recombinant (Ad5 recomb), as compared to Ad5 displaying wild-type capsid (Ad5 wild type). The shaft and the knob domain of the fiber were replaced with elements originating from heterologous serotype Ad3 as reported in J Sep Sci (2016) 39(22), 4299-4304.



Fig. 4. Shift in the retention time of the hexon fraction (E3), but not of the virion fraction (E5), obtained for double recombinant (Ad5 recomb), that combines the respective features of the two previously presented recombinant vectors (Fig. 1,2) as reported in J Sep Sci (2016) 39(22), 4299-4304. A cancelation of the opposing effects, provoked by each respective modification, revealed for virion fraction, as compared to Ad5 displaying wild-type capsid (Ad5 wild type).


The aim of the work was not to improve the purification process. It was just a proof of principle demonstrating how easy would it be to create targeted capsid modifications in order to shift adenovirus particles away from interfering material, whatever it is. The key question was how much, and where one might modify the virus – a living, reproducible entity – in order to alter its chromatographic behavior, and to abolish neither its structural stability, nor its biological activity. In this way, the authors wanted to share an idea that there is a plenty of possibilities to “play” with an adenovirus vector, in order to improve its “chromatographic separability”, not only its therapeutic functionality.



[1]  Majhen D, Calderon H, Chandra N, Fajardo C A, Rajan A, Alemany R, Custers J (2014) Adenovirus-Based Vaccines for Fighting Infectious Diseases and Cancer: Progress in the Field. Hum. Gene Ther 25, 1–17.
[2]  Krajacic M, Ivancic-Jelecki J, Forcic D, Vrdoljak A, Skoric D (2007) Purification of plant viral and satellite double-stranded RNAs on DEAE monoliths. J Chromatogr A 1144, 111–119.
[3]  Oksanen H M, Domanska A, Bamford D H (2012) Monolithic ion exchange chromatographic methods for virus purification. Virology 434, 271–277.
[4]  Ruscic J, Gutiérrez-Aguirre I, Tušek Žnidarič M, Kolundžija S, Slana A, Barut M, Ravnikar M, Krajacic M (2015) A new application of monolithic supports: The separation of viruses from one another. J Chromatogr A 1388, 69–78.
[5]  Krajacic M, Ravnikar M, Strancar A, Gutiérrez-Aguirre I (2017) Application of monolithic chromatographic supports in virus research. Electrophoresis (DOI: 10.1002/elps.201700152)
[6]  Vigne E, Dedieu J-F, Brie A, Gillardeaux A, Briot D, Benihoud K, Latta-Mahieu M, Saulnier P, Perricaudet M, Yeh P (2003) Genetic manipulations of adenovirus type 5 fiber resulting in liver tropism attenuation. Gene Ther 10 (2), 153–162.
[7]  Majhen D, Nemet J, Richardson J, Gabrilovac J, Hajsig M, Osmak M, Eloit M, Ambriović-Ristov A (2009) Differential role of αvβ3 and αvβ5 integrins in internalization and transduction efficacies of wild type and RGD4C fiber-modified adenoviruses. Virus Res 139, 64–73.
[8]  Ambriović-Ristov A, Mercier S, Eloit M (2003) Shortening adenovirus type 5 fiber shaft decreases the efficiency of postbinding steps in CAR-expressing and nonexpressing cells. Virology 312, 425–433.
[9]  Vigne E, Mahfouz I, Dedieu J F, Brie A, Perricaudet M, Yeh P (1999) RGD inclusion in the hexon monomer provides adenovirus type 5-based vectors with a fiber knob-independent pathway for infection. J Virol 73, 5156–5161.
[10]  Vigant F, Descamps D, Jullienne B, Esselin S, Connault E, Opolon P, Tordjmann T, Vigne E, Perricaudet M, Benihoud K (2008) Substitution of hexon hypervariable region 5 of adenovirus serotype 5 abrogates blood factor binding and limits gene transfer to liver. Mol Ther 16, 1474–1480.