Downloaded from www.microbiologyresearch.org by IP: 54.70.40.11 On: Sat, 15 Dec 2018 16:21:57 Journal of General Virology (2000), 81, 2605–2609. Printed in Great Britain .......................................................................................................................................................................................................... SHORT COMMUNICATION Blood clearance rates of adenovirus type 5 in mice Ramon Alemany, Kaori Suzuki and David T. Curiel Division of Human Gene Therapy, University of Alabama at Birmingham, Room 614, Wallace Tumor Institute, 1824 Sixth Avenue South, Birmingham, AL 35294, USA Persistence of adenovirus type 5 in blood has implications for the pathogenicity of the virus infection and for the use of this virus in oncolysis and gene therapy. In this study, the kinetics of adenovirus clearance from blood in mice has been evaluated. After a single inoculation of concen- trated virus into the vena cava, virus half-life was less than 2 min. Depletion of Kupffer cells (KC) resulted in increased viraemia. After tail-vein in- jection, virus and latex beads co-localized within KC. An important factor in clearance by KC is the negative charge of particles. Deletion of the hexon hypervariable region 1 acidic stretch decreased the negative charge of the virion but it did not increase blood persistence. Coating with PEG (‘ PEGylation ’) reduced the clearance rate but also reduced in- fectivity. Adenovirus type 5 (Ad5) infects the respiratory epithelium and can reach the bloodstream through lymphatics (Horwitz, 1996). Persistence in the circulation can affect virus patho- genesis and the outcome of therapies that use recombinant adenoviruses (Hitt et al., 1997; Kirn & McCormick, 1996). Interventions that slow clearance would favour tissue- or tumour-specific targetting approaches based on systemic delivery. We studied the kinetics of blood clearance of Ad5 in mice to obtain a better understanding of the limitations of adenovirus as a therapeutic agent. Different viruses are cleared from the blood by Kupffer cells (KC) (Kirn et al., 1982). Blockage of macrophages has indicated the interaction of adenovirus and KC (Lieber et al., 1997 ; Wolff et al., 1997; Worgall et al., 1997). When KC are not blocked, 90 % of the adenovirus genome is cleared from the liver in 24 h (Worgall et al., 1997). To demonstrate the direct interaction of adenovirus with KC, we injected Ad-CMV-LacZ virus (Alemany et al., 1996) labelled with red fluorophore (Cy3, Amersham; Leopold et al., 1998) into the tail vein of BALBc Author for correspondence : Ramon Alemany. Fax 1 205 975 7949. e-mail ramon.alemanyccc.uab.edu mice (10 virions per mouse). Ten minutes later, we injected fluorescein-labelled latex beads (2 μm diameter ; Sigma) as a KC tracer. After a further 10 min, livers were resected and embedded for cryosection. The co-localization of virus and latex beads within KC was readily observed as patchy aggregates all over the sinusoids (Fig. 1 C). In most instances, co-localization was evident as yellow fluorescent regions containing several latex beads. Some red fluorescence was also observed with no beads, probably due to incomplete marking of all KC. In order to confirm further the role of KC in virus uptake, KC were depleted with GdCl  (three tail-vein injections of 200 μl of a 2 mgml solution at 54, 30 and 6 h before injection of adenovirus and latex beads). The distribution of adenovirus in the liver changed from the patchy pattern to a regular, diffuse pattern enriched towards the perivenous region of the liver (Fig. 1D). Whether this periportal to perivenous enrichment reflects differences in the levels of coxsackievirus adenovirus receptor (CAR) in parenchymal or endothelial cells or the effect of other factors requires further investigation. The patchy distribution of adenovirus capsids taken up by KC reflects the main deposition site of the inoculum. More sensitive techniques to detect viral capsid proteins would probably reveal the distribution of smaller amounts of virus in other cell types. An attempt has recently been made to quantify the distribution of an adenovirus vector among different liver cell types after intravenous injection in rats (Davern et al., 1999). Although hepatocytes, stellate cells, KC and endothelial cells are transduced, the measurements of gene expression probably do not reflect the amount of virus adsorption. Indirect evidence that each liver cell type interacts with adenovirus comes from studies of liver toxicity (Lieber et al., 1997). Viraemia varies among different viruses (Kirn et al., 1982). One of the most important determinants of blood clearance is size. Rapid clearance ( 99% of an intravenous inoculum within 1 h) occurs with large viruses or small viruses opsonized with antibodies or complement. The net charge of the viral particle also affects the clearance kinetics (Jahrling & Eddy, 1977). To measure blood clearance, we flanked the GFP- expression cassette from pTracer-SV40 (Invitrogen) with Ad5 sequences (left, bp 27040–28045 ; right, bp 30863–31948) and inserted it into E3 by homologous recombination (Chartier et al., 1996). GFP transduction provides a faster and more 0001-7118  2000 SGM CGAF