Blood, 1 November 2001, Vol. 98, No. 9, pp. 2595-2596
Globin vectors: control without the locus
Hereditary hemoglobinopathies have long been favorites of gene
therapists, and a genetic cure would be a wonderful extension of the
decades of research spent understanding the molecular biology of
hemoglobin gene expression. The most realistic approaches rely on
retroviral vectors, which have the potential to integrate in hematopoietic stem cells, albeit at disappointingly low frequencies in
primates. In order to achieve high-level, erythroid-specific expression, scientists have turned to the 
-globin locus for
transcriptional control elements, especially globin promoters in
combination with parts of the locus control region (LCR) and intron
sequences. The low vector titers and frequent rearrangements caused by
these elements highlight the difficulties associated with including sequences that are normally not transcribed, or are now transcribed in
reverse orientation, in an RNA vector. Despite these problems, globin
gene therapists have stubbornly persisted, and recent results show that
optimized cassettes including -globin LCR, promoter, and intron
sequences can be incorporated into a lentiviral vector (instead of a
conventional murine leukemia virus vector) and lead to potentially
therapeutic hemoglobin protein levels in mice receiving transplants
(May et al, Nature. 2000;406:82-86).
An alternative approach to globin vector design is to abandon
problematic regulatory elements from the -globin locus and achieve
high-level, erythroid-specific expression some other way. Bodine and
colleagues showed that a double-copy murine leukemia virus vector
containing the promoter for the red cell membrane protein ankyrin
produced a respectable 8% 
-globin transgene mRNA expression level
(relative to endogenous 
-globin) in mice receiving transplants
(Sabatino et al, Proc Natl Acad Sci U S A. 2000;97:13294-13299). Here,
Moreau-Gaudry and colleagues (page 2664) extend these findings by
combining nonglobin, erythroid-specific promoters such as ankyrin with
different erythroid enhancers and a viral posttranscriptional regulatory element in a self-inactivating lentiviral vector. Using a
green fluorescent protein transgene, they find that the vectors can be
produced at high titers, that they are stable, and that erythroid-specific expression can be obtained in a majority of erythroid cells in mice receiving transplants. A similar vector with a
-globin transgene produced 11% to 28% mRNA expression levels
(relative to endogenous -globin) in MEL cells. Unfortunately, the
globin vectors were not tested in transplantation experiments; so it is
not clear if the same expression levels can be achieved in vivo. Still,
these experiments demonstrate that the problems of including -globin
locus regulatory elements in retroviral vectors can be avoided, and
there is no reason to believe that the optimal vector design has been
achieved yet. The next test will be to see whether these vectors can
surpass the impressive performance of lentiviral vectors based on the
-globin locus and produce therapeutic protein levels after transplantation.
David Russell
University of Washington
