We have developed novel SIVagm vectors pseudotyped with SeV F and HN envelope glycoproteins. SeV mutant F and HN proteins were incorporated efficiently into SIVagm vector particles and would be applicable to an extensive range of tissues in animals and humans, as sialic acid, which is a receptor for SeV, is ubiquitously present in several tissues.
In lentiviruses, the mechanism of incorporation of heterologous envelope proteins into virus particles is not well understood. Results from various genetic approaches indicate that the interactions that take place between MA and the C terminus of TMP (5
), as well as mutations in Env that affect its incorporation into mature virions, inhibit envelope-matrix interaction (5
). Another study showed that the cytoplasmic domain of gp41 is not required for the incorporation of Env into virions (9
). Reports on Env of SIVmac indicated that truncation of the cytoplasmic domain enhances envelope incorporation into SIVmac or HIV virions (25
In the case of SIVagm, the cytoplasmic tail of TMP is shorter than that of SIVmac (8
), and there was no report on the importance of the interaction between MA protein and Env. In this study, when SIVagm was pseudotyped with SeV wild-type and mutant F and HN proteins, the chimeric HN proteins SIVct/HN and SIVct+HN were efficiently incorporated into SIVagm particles. Although wild-type HN protein was efficiently expressed at the cell surface of transfected cells, the protein was not incorporated into pseudotyped virions. As the cytoplasmic tail of wild-type HN was 8 amino acids shorter in length than that of SIVct/HN (35 amino acids versus 43 amino acids), it appears that a specific domain in the cytoplasmic tail of TMP in SIVagm plays a major role in incorporation of HN proteins into virions.
SIVct+HN, with a longer cytoplasmic tail (78 amino acid) than SIVct/HN, was the most efficiently incorporated into virions and showed the highest titer. One explanation for this phenomenon is that the N terminus of SIVct+HN is located in close proximity to the viral matrix layer. In contrast to incorporation of HN proteins, chimeric F proteins carrying the cytoplasmic tail of SIVagm TMP were not expressed in producer cells (data not shown). There is as yet no logical explanation for why these chimeric F proteins are not expressed in the same way as chimeric HN proteins.
One report has demonstrated that a SIVmac Env glycoprotein with a short cytoplasmic tail is incorporated into particles in an MA-independent manner (7
). In addition, there have been several reports of incorporation of distantly related retroviral glycoproteins bearing a short cytoplasmic tail into HIV-1 particles (15
). In pseudotyping SIVagm with truncated F mutants, the Fct4, Fct14, and Fct27 proteins were expressed efficiently. However, the Fct27 protein was not incorporated into SIVagm particles. These results may support the suggestion that a short cytoplasmic tail in envelope glycoproteins avoids the steric constraints imposed by virus structure, leading to incorporation into viral particles. In the case of HN protein, however, cytoplasmic tail-truncated mutants were not incorporated into particles in spite of expression at the cell surface (data not shown).
Further analysis of the localization of these proteins and their interaction with lipid rafts will be required. Fct14 was more efficiently incorporated into virions than Fct4, but Fct14- and mutant HN (SIVct/HN and SIVct+HN)-pseudotyped vectors were more than 10-fold less infectious than Fct4- and mutant HN-pseudotyped vectors. Since these truncated F proteins had fusion activity, we conclude that interaction between F and HN, which is necessary for fusion promotion, is deficient, and the vector becomes less infectious in the case of the Fct14 and mutant HN proteins on vector particles.
SeV mutant F and HN proteins were sufficiently incorporated into SIVagm virions as examined by Western blot analysis. However, even under optimal conditions, the SIVagm vector pseudotyped with Fct4 and SIVct+HN was less infectious than the VSV-G-pseudotyped vector. Two explanations may account for this lower infectivity. First, because of the independent incorporation of each envelope glycoprotein, Fct4 and SIVct+HN, there are mixtures of various particles (Envless Fct4 or SIVct+HN vector particles; noninfectious Fct4 and SIVct+HN vector particles; and infectious Fct4 and SIVct+HN vector particles), resulting in a lower rate of functional vector. Second, the ratios of the Fct4 and SIVct+HN proteins on the virion surface are not optimum, and the interaction between the Fct4 and SIVct+HN proteins is inadequate.
To characterize this vector, we investigated the transduction of various cell types compared to transduction with the VSV-G-pseudotyped vector. An SeV mutant F- and HN-pseudotyped vector was able to transduce a panel of cell lines as observed with the VSV-G-pseudotyped vector. These results demonstrated that an SeV mutant F- and HN-pseudotyped vector had expanded cell tropism compared with SIVagm with the original envelope glycoprotein. Airway epithelial cells are targets for gene therapy of respiratory diseases such as cystic fibrosis. The VSV-G-pseudotyped vector, however, is unable to transduce these cells without causing them injury. In other reports, adeno-associated virus vector-mediated gene transfer to the airway was inhibited by a physical barrier at the apical surface (1
) and required a high titer (30
Localization of receptors is an important factor for virus infection. In polarized epithelia, expression of virus receptors on the apical or basolateral membrane was reported for several viruses. In the case of the VSV-G receptor, a distribution or uptake mechanism on differentiated airway cells has not been intensively studied. In contrast, a recombinant SeV has been reported to transduce a reporter gene to the airway epithelial cells efficiently from the apical surface (28
). The envelope proteins derived from pneumotropic viruses are expected to break through this barrier and to be usable as potential pseudotypes. Indeed, our data demonstrated that an SeV mutant F- and HN-pseudotyped SIVagm vector transduced rat trachea polarized airway epithelial cells.
In conclusion, we have developed a novel pseudotyped SIV vector that has envelope glycoproteins derived from paramyxovirus and shows wide cell tropism. In addition, this vector was able to transduce polarized rat airway epithelial cells from the apical and basolateral surfaces. SeV glycoproteins-pseudotyped SIVagm vectors are likely to be a powerful tool for gene therapy of various diseases, including respiratory diseases.