Dettenhoffer and Yu [1] were the first to report the use of discontinuous 6-18% (w/v) iodixanol gradient in a sedimentation velocity mode to purify HIV-1 virions without affecting the infectivity of the virus (see Application Sheet V34). 

The technique was subsequently extended to the purification of Mason-Pfizer monkey virus by Gottwein et al [2] and the protocol is described in Section 2; an alternative buoyant density method [3,4] is described in Section 3. In all comparative studies between CsCl and iodixanol, it has been shown that the recovery of virus infectivity is much higher and the particle:infectivity ratio much lower when viruses are purified in iodixanol. 

Although sucrose is generally less deleterious to viral infectivity than CsCl, it can nevertheless also have serious effects on certain important aspects of viral function; in particular the loss of surface glycoproteins from retroviruses has been noted [5].

This may be related to its viscosity, which is much higher than that of iodixanol. Like CsCl, sucrose must be dialyzed before infectivity can be measured. In contrast, many add-on techniques can be performed and cells infected with virus, without dialysis of iodixanol.