Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy S4 AAV VIRUS & VECTOR BIOLOGY to transduction effi ciency and per cell express
Trang 1Molecular Therapy Volume 20, Supplement 1, May 2012
RNA VIRAL VECTORS
(SIN) Long Terminal Repeats (LTRs) and harboring the strong
Spleen Focus Forming Virus (SF) enhancer/promoter sequences in
internal position driving GFP expression (SIN.LV.SF.GFP) caused
a signifi cant acceleration (p<0.006) in the time of hematopoietic
tumor onset (N=17, 186 days median survival) with respect to
mock-controls (N=80, 233 d) Treatment with a vector without
Open Reading Frames (ORFs) downstream the SF promoter (LV
SF) induced a dramatic acceleration of tumor onset (N=9, 112 d; Vs
Mock p<0.0001), similar to the acceleration induced by the treatment
with the highly genotoxic LV containing the SF sequences within
the LTR (LV.SF.LTR, N=11, 138 d, p<0.0001) These data confi rm
that the presence of the ORF downstream the enhancer/promoter
sequences has a major role in increasing the safety profi le of the
SIN design Importantly, we confi rmed these results applying the
LV systemic delivery in newborn heterozygous Cdkn2a+/- mice,
which do not display background oncogenesis unless treated with
genotoxic vectors To unravel the mechanisms of genotoxicity, we
performed integration analyses and we retrieved >7500 unique LV
integration sites from tumors derived from the different mouse strains
and vectors tested We identifi ed >60 Common Insertion Sites (CIS),
such as the well-known cancer genes Braf, Sfi 1, Mef2c and Map3k8
Thus, the vectors tested promoted both in Cdkn2a-/- and Cdkn2a+/-
tumor onset acceleration by insertional mutagenesis Interestingly,
in both mouse strains, LV.SF.LTR and the LV.SF vectors induced
oncogenesis by Braf activation through a
read-through/splicing-capture mechanism On the other hand, the oncogenesis induced by
SIN.LV.SF.GFP was the result of the overexpression of Map3k8 and
other CIS genes by enhancer mediated and truncation-activation
mechanisms Importantly, by performing multivariate analyses on a
collection of 229 Acute Myeloid Leukemia patients, we found that
the gene expression level of several of the newly identifi ed CIS genes
is an independent predictor of patients’ survival Hence, this in vivo
model is resourceful since it can be used to study the genotoxicity
of integrative vectors and as LV-based insertional mutagenesis tool
for the discovery of clinically relevant cancer genes
6 In Vivo Gene Delivery to the Striatum
Using a Novel Entry-Targeted Lentivirus: An Entry
Restricted Lentivirus Vector for Neurodegenerative
Diseases
Scott A Heldt,2 John V Cox,1 Robin L Krueger,1 Lorraine M
Albritton.1
1 Microbiology, Immunology and Biochemistry, College of
Medicine, University of Tennessee Health Science Center,
Memphis, TN; 2 Anatomy and Neurobiology, College of Medicine,
University of Tennessee Health Science Center, Memphis, TN.
Degeneration of neurons in the striatum and substantia nigra
are the hallmarks of the neurodegenerative diseases Parkinson’s,
Alzheimer’s and Huntington’s Diseases Neurons in the hippocampus
also degenerate in Alzheimer’s Disease Affected neurons express
several regulatory receptors including somatostatin receptors (SSTR)
Here we report in vivo demonstration of entry-targeted gene delivery
to the striatum by an entry-targeted lentiviral vector (LV) Entry
targeted LV was produced using a novel chimeric glycoprotein that
directs attachment and gene delivery specifi cally to cells expressing
SSTR The entry capacity of this chimeric glycoprotein via SSTR is
comparable to the capacity of the wildtype viral glycoprotein from
which it was engineered and approaches within fi ve-fold the capacity
of VSV G protein via the VSV receptor In in vivo studies, control VSV
G coated LV or experimental SSTR-targeted LV were microinjected
into the dorsal region of the caudate putamen of immune competent
C57BL/6 mice (AP= -0.5, ML= +2.0, DV= -4.0) G coated LV
transduced striatal cells along the end of the injection needle track and
some striatal neurons whose ascending axons projected into the region
at the end of the needle track In addition G coated LV transduced
cells in the corpus callosum through which the track descended SSTR-targeted LV transduced cells along the end of the injection track and numerous striatal projection neurons but not cells in the corpus callosum Synthetic somatostatin peptide is an agonist that causes internalization of SSTR from the cell surface Microinjection of somatostatin peptide followed by microinjection of LV into the dorsal region of the caudate putamen inhibited transduction of by SSTR-targeted LV but not by G coated LV indicating that the entry-SSTR-targeted
LV used SSTR as its receptor in vivo The ability of the SSTR-targeted glycoprotein to restrict gene delivery to SSTR-positive cells in vivo
demonstrates its potential as a new tool for targeting entry that when combined with transcriptionally targeted vector genomes should be usedful in gene therapy of neurodegenerative diseases, including Parkinson’s, Alzheimer’s and Huntington’s Since the entry targeting sequence in the chimeric glycoprotein is recognized by SSTR from a wide range of mammals including man, primates, mouse, rat, guinea pig and dog, the SSTR-targeting glycoprotein can be used unchanged
in man and all common disease models to target lentiviral vectors for delivery of multiple neuroprotective and corrective genes
Performance in the CL20 SIN HIV Lentivector Used To Correct Human X-Linked Chronic Granulomatous Disease as Evaluated Ex Vivo and
in NSG Mice Transplanted with Transduced Patient CD34+ Stem Cells
Uimook Choi,1 Sherry Koontz,1 Harry L Malech.1
1 Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, Bethesda, MD.
X-linked chronic granulomatous disease (X-CGD) results from a defi ciency of gp91phox, the NADPH/fl avin/heme binding component
of the phagocyte oxidase Because murine retrovirus vectors have been associated with genotoxicity in some X-CGD clinical studies and have failed to provide long-term functional oxidase correction greater than 1% beyond years after gene therapy, we have turned to SIN lentivectors as a possible solution to these safety and effi cacy issues Safety considerations dictate that it would achieve the highest level of per neutrophils oxidase correction with the lowest insert copy number per stem cell SIN lentivectors use an internal promoter to drive production of therapeutic transgene We have explored several promoters in the CL20 SIN HIV lentivector, in which gp91phox cDNA was codon optimized to increase gene expression, fi nding to our surprise that some promoters that look very promising when evaluated in ex vivo culture, perform less well when evaluated in human neutrophils arising in vivo from human CD34+ hematopoietic stem cells (HSC) transplanted long term in the NSG mouse Transduction of X-CGD HSC with HIV lentiviral vector (MOI 10) was done as two overnight transductions following
24 hours pre-activation with growth factors Most of the transduced cells were transplanted into mice, but some cells were retained for prolonged ex vivo culture At 3 weeks in culture transduced X-CGD HSC using CL20 with shortened EF1a (267bp) or with mPGK promoter vector versus normal versus untransduced patient HSC were 52%, 76%, 75% and 0% gp91phox positive, respectively, with MFI as measure of gp91phox expression of 502, 562, 682
fl uorescence units for the transduced and normal, suggesting better performance for the mPGK promoter CL20 vector than for the EF1a short promoter CL20 vector At 8 to 16 weeks post-transplant, the CD13+ human neutrophils arising in mouse bone marrow were assessed for gp91phox expression Mice engrafted with shortened EF1a CL20, mPGK CL20 transduced, normal control or untransduced patient HSC had 58%, 15%, 79%, and 0% human CD13+ neutrophils expressing gp91phox, respectively, with MFI of 1093, 974 and 1409 for the transduced and normal control Interestingly, mPGK promoter vector transduced cells showed enhanced performance with respect
Trang 2Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy S4
AAV VIRUS & VECTOR BIOLOGY
to transduction effi ciency and per cell expression of gp91phox
compared to shortened EF1a promoter vector after 3 weeks in vitro
culture, yet signifi cantly inferior performance with respect to both
parameters in the xenograft models Transgene copy number per
gp91phox expressing human cell was consistently close to 1 copy
per cell with both CL20 vectors In conclusion, we have developed
a candidate lentivirus vector using CL20 with EF1a short promoter
that has acceptable transduction effi ciency and per cell expression
of gp91phox to justify bringing this vector to the clinic In addition
we demonstrate the importance of using the mouse xenograft model
to evaluate vectors to be used in human HSC before making fi nal
decisions about effi cacy
Targeted Gene Therapy
Scott R Witting,1 Priya Vallanda.1
1 Medical and Molecular Genetics, Indiana University School of
Medicine, Indianapolis, IN.
The cellular tropism of lentiviruses may be altered by swapping
envelope (env) proteins with those of other viruses Env proteins
mediate binding to a cell surface receptor and the subsequent fusion
of the virus lipid envelope with the cell for infection The most
common env used with lentiviruses is from vesicular stomatitis
virus (VSV-G) because it infects most cell types and high titers are
routinely obtained However, its wide tropism and instability in
human serum are not optimal for in vivo gene therapy applications
Nipah virus is a rare, extremely lethal paramyxovirus which is also
a distant relative of the measles virus Infection is thought to initiate
in mucus membranes Propagation and dissemination then occurs via
the bloodstream Consequently, it has a highly specifi c tropism for
endothelial cells due to the expression of the Nipah virus receptor,
ephrin B2 Thus, pseudotyping with Nipah virus envelope proteins
could be a valuable tool for gene therapy applications involving the
circulatory system Since a lentiviral vector was recently pseudotyped
with the related measles virus envelope proteins by truncating the
cytoplasmic domains, we hypothesized similar truncations in the
Nipah virus envelope proteins were possible We present data showing
the successful pseudotyping of a lentiviral vector with Nipah env
proteins The Nipah pseudotype (Np-L) yields high titers, can be
concentrated, and has superior stability in human serum compared
to VSV-G lentivirus Importantly, the successful in vitro transduction
of human primary endothelial cells and no signifi cant transduction of
bone marrow progenitor cells demonstrates its potential for vascular
gene therapy
9 Identifi cation of the Galactose Binding
Domain of the AAV9 Capsid
Christie L Bell,1 Brittney L Gurda,1 Kim Van Vliet,2 Mavis
Agbandje-McKenna,2 James M Wilson.1
1 Gene Therapy Program, Department of Pathology and
Laboratory Medicine, University of Pennsylvania, Philadelphia;
2 Department of Biology and Molecular Biology - The McKnight
Brain Institute, University of Florida, Gainesville.
Adeno-associated virus (AAV) serotype 9 vectors show promise for
gene therapy of a variety of diseases due to their ability to transduce
multiple tissues, including heart, skeletal muscle, and the alveolar
epithelium of the lung Additionally, AAV9 is unique compared to
other AAV serotypes in that it is capable of surpassing the
blood-brain barrier and transducing neurons in the blood-brain and spinal cord
It has recently been shown that AAV9 uses galactose as a receptor
to transduce many different cell types in vitro, as well as cells of
the mouse airway in vivo In this study, we aimed to identify the
specifi c amino acids of the AAV9 capsid necessary for binding to galactose To determine potential amino acids that may be involved
in galactose binding, we aligned the AAV9 capsid sequence to that of other AAV serotypes that do not bind galactose, including AAVs 1,
2, 6, 7, and 8, to allow identifi cation of amino acids unique to AAV9 that have either a charged or polar side chain These amino acids were then mutated to non-polar alanine by site-directed mutagenesis and the resultant capsid mutants were tested for their ability to bind to three different CHO cell lines- the parental CHO cell line Pro-5, as well as somatic cell glycosylation mutants of this cell line including Lec-2, which lack cell-surface sialic acid residues and therefore have exposed galactose residues, and Lec-8, which are devoid of cell-surface galactose residues From this, it was discovered that
an asparagine residue at position 470 (N470) was necessary for AAV9 binding to galactose because mutation of this residue led to
a signifi cant decrease in binding to Lec-2 cells By subsequently mutating amino acids in close proximity to N470 on the capsid surface, four additional amino acids were determined to be necessary for galactose binding- D271, N272, Y446, and W503 The importance of these amino acids was also confi rmed by in vivo studies in the mouse lung As shown in previous studies, AAV9 delivered intranasally following pre-treatment with neuraminidase demonstrates robust transduction of cells of the conducting airways due to the exposed galactose residues on the cell surface (Bell et al, 2011) We observed, however, that intranasal delivery of the fi ve AAV9 mutants defi cient
in galactose binding did not result in conducting airway transduction following neuraminidase pre-treatment, which further demonstrates their inability to bind galactose Together, the fi ve identifi ed amino acids necessary for galactose binding form a pocket at the base of the protrusions around the icosahedral 3-fold axes of symmetry of the capsid Using computational ligand docking, we observed that the most favorable interactions between a galactose residue and the AAV9 capsid occur in the defi ned binding pocket Identifying the interactions necessary for AAV9 binding to galactose may lead to advances in vector engineering
and Functional Signifi cance of the 381 Amino Acids in the C-Terminal Half of AAV9 Capsid
Kei Adachi,1 Tatsuji Enoki,2 Yasuhiro Kawano,1,2 Hiroyuki Nakai.1
1 Department of Molecular and Medical Genetics, Oregon Health and Science University, Portland, OR; 2 Center for Cell and Gene Therapy, Takara Bio Inc., Otsu, Japan.
Profound understanding of how AAV’s biological properties are determined by its capsid amino acid sequence would signifi cantly help design the next generation of AAV vectors with the most desirable biological properties in each application However, the conventional approaches in the AAV research have provided only limited information for the knowledge-based designing of new AAV vectors To overcome the limitations in the traditional approaches,
we have developed a new method that allows comprehensive and high-throughput structural and functional analyses, named “AAV Barcode-Seq” (Adachi et al., Abstract 397, ASGCT 2011) In this approach, we create AAV mutant libraries in which each AAV (wild type and mutants) has a unique set of two 12-nt DNA barcodes within its viral genome, and determine the relative quantities of each mutant and wild type AAV genomes in various samples by Illumina sequencing This approach makes it possible to analyze the biological consequences of multiple mutants in multiple samples of interest at one time by a single sequencing run Here we show the fi rst set of the “AAV Barcode-Seq” results in a comprehensive double alanine scanning of the carboxy-terminal half of the AAV9 capsid In the study, we produced a total of 191 double alanine-mutated AAV9 to scan the entire amino acid (aa) positions 356 through 736 These 191 mutants (i.e., a total of 382 AAV clone preparations, 2 clones per