Evolution of Gene Therapy, Historical Perspective

Abstract

The earliest conceptual history of gene therapy began with the recognition of DNA as the transforming substance capable of changing the phenotypic character of a bacterium and then as the carrier of the genomic code. Early studies of oncogenic viruses that could insert into the mammalian genome led to the concept that these same viruses might be engineered to carry new genetic material into mammalian cells, including human hematopoietic stem cells (HSC). In addition to properly engineered vectors capable of efficient safe transduction of HSC, successful gene therapy required development of efficient materials, methods, and equipment to procure, purify, and culture HSC. Increased understanding of the preparative conditioning of patients was needed to optimize engraftment of genetically modified HSC. Testing concepts in pivotal clinical trials to assess efficacy and determine the cause of adverse events has advanced the efficiency and safety of gene therapy. This chapter is a historical overview of the separate threads of discovery that joined together to comprise our current state of gene therapy targeting HSC.

Introduction to Theoretical Concepts and Early Background History Impacting HSC Gene Therapy

The history of gene therapy comprises the advance of theoretical concepts, understanding the human genome, availability of critical materials and instruments, design of vectors and chemical tools to manipulate and change genomic DNA, improvements in the procurement and culture/maintenance of stemness of HSC in culture, improvements in myeloid conditioning, the outcomes of conduct of clinical trials, observing successes and problems occurring in clinical trials, and deep study and elucidation of the mechanisms of problems that arise in clinical trials to seek and incorporate corrective measures. The evolution of our understanding of ethical issues impacting gene therapy, and the logistics of access to and cost of successful gene therapy treatments are also important elements of this history.¹

In this broad brush and somewhat unconventional view of the history of gene therapy, we address general principles; key experiments, basic science, and clinical trials that illustrate some general principle; and the evolution of materials and instrumentation that make current clinical approaches to gene therapy of HSC possible. We aim to complement rather than duplicate the extensive discussion of the background studies of gene therapy and the march of the many published clinical trials in specific disorders or categories of disorders that are the subject of the other chapters in this series, as well as excellent recent reviews.²

The earliest experiments that laid the foundation for gene therapy began with experiments on the transforming properties of bacteria.³ Alloway reported in 1932 that non-virulent (R type) pneumococci became lethal by adding cell free extracts from virulent (S type) pneumococci. When injected with these “transformed” pneumococci, the mice developed pneumonia and died.

In our view, the key conceptual background to all gene therapy emerged in the 1940’s with the seminal work by Avery and colleagues on bacterial transformation (which one could perhaps very loosely call gene therapy of bacteria). They identified DNA as the transforming factor that could change the physiology of a bacterial strain,⁴ and more specifically, showed that the ‘transforming substance’ was precipitated out by alcohol and later confirmed to be DNA. This was one of the key background elements to Watson and Crick in identifying the structure of DNA,⁵ postulating its role as the genomic code of all prokaryotic and eukaryotic organisms, and thus demonstrating that nucleic acid sequences, rather than proteins, carry genetic information. The next critical discovery was that of Marshall Nirenberg, who in 1961 discovered the “triplet” code by which DNA encodes for assembly of the twenty amino acids that serve as the building blocks of proteins.⁶

In parallel with this elucidation of the biochemical basis of heredity, were emerging concepts from early transformation studies in mammalian cells, for example the early reports that transformation of 8-azaguanine sensitive cells with nuclei and chromosomes from 8-azaguanine resistant cells rendered the transformants resistant due to transfer of a mutated hypoxanthine-guanine phosphoribosyltransferase gene.^7,8 An early review of mammalian cell transformation studies conducted over the following 18 years was reviewed in 1980 by Shows and Sakaguchi.⁹ This body of work further established that newly acquired biochemical traits from DNA transformation experiments in mammalian cells can be heritable.

Many other key concepts that evolved into current methods of viral vector mediated gene therapy were developed in the 1970s, during a period of active investigation of viruses capable of transforming normal tissues into cancers. From this work the concept emerged that perhaps these DNA and RNA tumor viruses known to insert into the genome of target cells could be modified in some way to remove the tumor causing elements, but retain their genome insertion capabilities to deliver a therapeutic payload. Some of the earliest published reviews of the history of gene therapy incorporating these essential concepts were those written in series of reports over time by Theodore Friedmann^10-13 who shared the 2015 Japan Prize with Alain Fischer for “For the Proposal of the Concept of Gene Therapy and its Clinical Applications.”

More generally, the term “gene therapy” now broadly includes the introduction or manipulation of DNA or RNA sequences in human cells to treat disease. There is general consensus among the U.S. Food and Drug Administration (FDA),¹⁴ the European Medicines Agency (EMA),¹⁵ and the American Society of Gene and Cell Therapy (ASGCT)¹⁶ defining gene therapy as changes in gene expression, achieved by replacing or correcting a disease-causing gene, inactivating a target gene, or inserting a new or modified gene, using a vector or delivery system of genetic sequence or gene, genetically modified microorganisms, viruses, or cells.

By the late 1970’s, while our understanding of the molecular basis of human diseases was advancing through cloning and sequencing of genes, there were major technical challenges to implement gene transfer. Exogenous DNA could be introduced to target cells by transformation or transfection, but the overall efficiency was low. Additionally, if the introduced gene(s) did not provide a survival advantage, the durability of gene transfer was also low. The resulting gene transfer efficiency at that time was about one in 100,000 cells, but nonetheless was proposed as a method to achieve genetic correction.¹⁷

Intense interest in inherited hemoglobinopathies such as sickle cell disease and beta-thalassemia fueled work on beta-globin, one of the first genes to be cloned and then studied with the intent of gene transfer for clinical application. Mulligan and colleagues replaced the viral capsid protein (VP1) of the SV40 genome with complementary DNA of rabbit beta-globin in a monkey kidney cell line, which produced large quantities of rabbit beta-globin mRNA and protein.¹⁸ Since there was no inherent advantage for beta-globin gene transformed cells, several labs worked on selectable genes to be co-transferred. Pellicer et al successfully inserted beta-globin and thymidine kinase (TK) genes into murine teratocarcinoma cells.¹⁹ The Cline lab inserted dihydrofolate reductase (DHFR) or TK in murine marrow cells.²⁰

Cline and colleagues from UCLA then applied these results and tested them clinically.^21-23 An experimental protocol to insert genetically modified marrow cells from beta-thalassemia patients, inject the cells in the femur after local irradiation, and treat with a selecting agent was submitted to the human research review committee at their home institution. Because the first two patients to be treated were receiving their care in other countries (in a hospital in Naples, Italy and at Hadassah Hospital, Jerusalem, Israel), not covered by the UCLA review committee, the team sought in parallel and secured permission in Naples and Hadassah for the clinical study. Both patients were informed of the experimental nature and the low likelihood of success in this approach. After femur irradiation and infusion of modified marrow cells, the patients reported no adverse events, and though selective agents were not used. Three months later, there was no demonstrable clinical benefit in both patients. Although safety of this clinical gene transfer was undebated, many controversial issues were brought forth.^24-26 Can a clinical protocol proceed with permission from some but not all institutions? How many preclinical experiments (in vitro or animal), and what degree of ‘success’ are needed to garner approval? While the responses to these issues are much more straightforward today, various review committees at that time were caught off guard and the consensus was that this was a rather premature and in retrospect problematic initial attempt at clinical application of gene therapy.²⁷

These first two attempts at human gene therapy generated much media attention and scrutiny by regulatory committees. The remainder of the decade into the early 1990’s, scientists were quietly working on recombinant DNA methods, in vitro and animal models for testing, and strategies to enhance transgene expression. It quickly became clear that using viral vectors was more efficient in gene transfer than the previous methods of physical entry by transfection, fusion, or even electroporation. Much of gene transfer experiments then focused on vector optimization and design, and brought this background discussion into the early modern era of gene therapy.

The following sections of this review will provide historical background of a number of parallel developments that provided the laboratory and clinical tools and materials that facilitate our current approaches to gene therapy targeting blood cells including HSC.

Design of Integrating Vectors Used for HSC Gene Therapy

Vectors engineered from gamma retroviruses,²⁸ long under study as the cause of a variety of cancers in mice, had the desired property of efficient insertion into the genome of target cells. Murine gamma retroviruses and their derivatives were the first of the genome integrating vectors to be applied to T lymphocytes and HSC in the clinical setting.

Gamma retroviruses are RNA viruses, that upon entry into a cell, are “reverse transcribed” (hence “retro”Virus) into a DNA sequence. It is the DNA virus sequence that ultimately inserts itself into the host cell’s genomic DNA, becoming a “provirus” that in turn generates RNA virus sequences and viral mRNAs encoding virus proteins required for the replication phase of the virus life cycle. The critical issue was how to turn these viruses that efficiently insert provirus DNA genomic sequence into mammalian cell genomes, but are also efficient at causing tumors, into safe tools for gene therapy. The solution was to remove and/or inactivate as many elements of the virus genome as possible, while still retaining the ability of the highly engineered provirus sequence to insert efficiently into the mammalian cell genome. The goal was a functioning single-cycle virus capable of cell entry, uncoating, reverse transcription into provirus DNA and insertion into the genome, but incapable of generating infectious virus. The solution involved separating the key elements required to generate replication-incompetent viral vector into three separate ‘production plasmids’: (1) an envelope (env) producing element (the vector virus coat also serving the purpose of binding to target cell and facilitating virus payload entry); (2) a gag-pol producing element (gag protein important for vector RNA packaging and polymerase for reverse transcribing the RNA); and (3) the vector sequence (retaining the psi element needed for packaging and the long terminal repeat (LTR) sequences at both ends of the vector sequence, which serves both as the internal strong promoter driving production of a therapeutic protein and containing initiation elements binding the two ends of the vector for the circle formation required for reverse transcription). Where possible the env and gag-pol codons were changed to avoid recombination events that could reconstruct a replication competent virus. To simplify the process of making different gamma retrovirus vectors, permanent packaging lines were devised that constitutively produce env and gag-pol, and when a specific vector sequence is added, clones could be assessed and chosen that constitutively produced vector in adequate titers. Many laboratories contributed to this technology and created a large array of different “flavors” of therapeutic gene therapy gamma retrovirus vectors. Many of these continue to be used for production of some CAR-T lymphocytes or therapeutic cloned T cell receptors. This tour-deforce of engineering involving the contribution of many labs has served as the core technology used in the first generation of gene therapy targeting HSC or lymphocytes.

The LTRs of gamma retroviruses were retained in the engineered vectors as convenient, very strong promoters to drive high levels of production of downstream inserted therapeutic protein coding sequences. However, these same LTR elements contain strong enhancer elements that can activate nearby genes. The engineered vectors by design retained the insertion targeting elements of the parent virus required to insert the DNA provirus into the mammalian genome. While insertion of vector seems to be random, it is actually stochastic in that the mechanism used by the vector couples to cellular elements, resulting in preferred sites of insertion into the genome. These preferred sites (also known as integration sites) are often located near the start of genes and in enhancer elements, and may in turn strongly interact with enhancer elements in the LTR.^29,30 While the odds of any one insert occurring in a sensitive site are very low, gene therapy for a human subject may involve tens to hundreds of millions of insertions. Depending on the vector, the LTR and the host human subject disease substrate, we now know from adverse leukemic insertional mutagenesis events occurring in a number of clinical trials, that gamma retroviral vectors can transactivate oncogenes such as LMO2, the MECOM complex, and other oncogene targets to initiate development of leukemia. These insertional mutagenesis events will be further discussed in greater detail in the last section of this historical review. Curiously, insertional mutagenesis leading to leukemic events have not been observed when the target of gamma retroviral vector gene therapy are T lymphocytes.

Well before the first insertional mutagenesis oncogenic events were observed in clinical trials of gene therapy using gamma retroviral vectors, certain limitations of this class of vectors (e.g. limits of therapeutic payload size, limits on use of alternate promoter elements instead of the LTR, absolute requirement for cell division for vector insertion into the genome) encouraged development of gene therapy vectors derived from human immunodeficiency virus (HIV). HIV is part of a different group of retroviruses called lentiviruses and the vectors engineered from HIV are referred to here as “lentivectors”. HIV and other lentiviruses have a more complex structure, and have a number of required functional elements not present in gamma retroviruses, such as rev, that needed to be considered while engineering HIV into a safe gene therapy tool.^31,32 As with gamma retroviruses, determining how much could be removed from the virus and whether addition of elements from other viruses might enhance function and efficiency of the vector was an iterative discovery process. From a historical perspective, some key advantages of lentivector function and engineering, and the insertional mutagenesis oncogenic events noted above have resulted for the most part in the abandonment of gamma retroviral vectors for transduction of HSC for clinical trials.

As with gamma retrovirus vectors, the production of lentivectors that are functional, but replication incompetent, required the separation of packaging elements into plasmids separate from the transfer vector. Almost all lentivector production for clinical application uses the membrane fusion G protein derived from vesicular stomatitis virus (VSV-G) as the vector envelope element, rather than the natural env component of HIV. The cell membrane target of the VSV-G protein is ubiquitous to all cells with high efficiency of binding and vector membrane fusion. Almost from the start, lentivector engineering strategies incorporated a self-inactivating (SIN) feature, modifying the LTR element that contains strong enhancers with transactivating potential and using safer promotors with little enhancer activity instead. This was accomplished by creating a deletion in the 3’ LTR of the vector production plasmid. During vector production, the intact 5’ LTR assists in the important packaging biochemistry needed to produce infective but replication incompetent lentivirus vector. During transduction, the SIN 3’ LTR binds to the 5’ LTR in the circularization and priming step that retrotranscribes the insertional provirus DNA from the lentivector RNA, and is incorporated into the 5’ end of the provirus DNA, thus “self-inactivating” the 5’ LTR. This safety feature removes enhancer and activator elements, and allows the therapeutic payload transgene(s) to be transcribed from a promoter of choice. The use of the SIN feature with alternate internal promoter is not limited to lentivectors. A group of investigators incorporated the SIN feature into the gamma retroviral vector that was efficacious in a clinical trial targeting hematopoietic stem cells to treat infants with X-linked severe combined immunodeficiency (X-SCID). This SIN gamma retroviral vector showed clinical efficacy,³³ with no insertional oncogenesis after median 9 years of follow-up (personal communication S.-Y. Pai). Importantly, this maneuver does not alter pattern of insertion characteristic of gamma retroviruses that tends to target enhancer elements and the 5’ end of genes.³³

The larger payload capacity of lentivectors, SIN design, and potential to incorporate tissue-specific promoters and enhancers catalyzed concerted efforts by several investigators to develop lentivectors that would drive high level beta-globin expression at specific stages of erythroid precursor development. In the 1990’s, the locus control region of beta-globin was discovered to contain hypersensitive sites (HS) that were important for high level expression. After a series of in vitro experiments, a lentivector containing optimized regulatory elements, the TNS9 vector, was shown to drive high level erythroid specific expression of adult beta-globin in transduced murine HSC, successfully correcting a murine model of beta-thalassemia.³⁴ This seminal work was followed shortly by the same group to correct another more severe phenotype of beta-thalassemia in Berkeley mice.³⁵

In parallel, another group designed a modified adult beta-globin in which the 87^th amino acid was switched from threonine to glutamine, to mimic the anti-sickling effect of gamma-globin. This T87Q version successfully corrected two sickle mouse models³⁶ and was expressed at high levels when transduced human sickle cord blood stem cells were transplanted into immunodeficient mice.³⁷ These reassuring preclinical studies ultimately led to treatment of a patient with compound beta-E/ beta-0 thalassemia, who achieved transfusion independence.³⁸

From those early efforts, an increasing number of successful clinical trials have emerged to treat thalassemia and sickle cell disease, described in significant detail in other chapters in this series. Some of these studies will be revisited later in this chapter in comments relating to transduction enhancers and certain types of adverse events associated with gene therapy.

Gene Editing of HSC

Gene editing is the most recently evolving technology to be applied to gene therapy of HSCs, having the capability of targeting a specific sequence within the genome, in contrast to the stochastic semirandom insertion of integrating viruses throughout the genome. The four main types of gene editing systems are meganucleases, zinc finger nucleases, transcription activatorlike effector nucleases (TALENs),³⁹ and clustered regularly interspaced short palindromic repeats-Cas (CRISPR-Cas).⁴⁰ Of particular near term historical note are recent clinical reports of unequivocal clinical benefit from successful application of the CRISPR technology applied to HSCs for correction of hemoglobinopathies.^41,42

CRISPR is derived from an anti-virus system that evolved in bacteria to “copy and memorize” virus sequence, thereby allowing the bacterium to target that same virus sequence for cleavage when subsequently infected by similar virus.⁴³ The discovery of the use of the CRISPR system for gene editing was the basis for the 2020 Nobel Prize in Chemistry (https://www.nobelprize.org/prizes/chemistry/2020/press-release/). The CRISPR system has revolutionized the field, due to its ease of application and versatility. CRISPR-based derivative methods such as base editing and prime editing use the core element of the CRISPR system to find the genome target sequence to enzymatically convert a single base pair or reverse transcribe a short sequence change into the genome, respectively.⁴⁴

Most current approaches to editing use electroporation to introduce the editing elements (CRISPR-Cas9 mRNA or protein, guide RNA, and any additional factors to augment editing) into HSC. There have been a number of small scale non-GMP research grade electroporation instruments available for gene editing development work. Fortuitously, clinical scale and high throughput GMP compliant commercial instruments in development in the past few years have become available just in time for the current initiatives in gene editing of HSC.

Hematopoietic Stem Cells: sourcing, selecting, culturing, transducing

Parallel to the development of vectors and editing tools for gene modifying HSC, advances in the procurement, culture, transduction and engraftment of HSC has had an important impact on the field.

HSC occupy niches in the marrow that facilitate retention of pluripotent potential to give rise to all hematopoietic lineages and asymmetric proliferation of some progeny into lineage specific progenitors.⁴⁵ Sourcing HSC for gene therapy or conventional allogeneic transplants initially was restricted to harvesting of bone marrow with needles. That HSC are constantly translocating at a slow rate from marrow to the circulation and back to the marrow was a key discovery that ultimately led to alternate sourcing of HSC.⁴⁶ There is a steady state presence of CD34+ HSC or progenitors in the peripheral blood of healthy humans of about 1400 cells per ml (1.4 per μl). This low baseline frequency of CD34+ HSC and progenitors in the circulation can be increased by daily injections of granulocyte colony-stimulating factor (G-CSF or filgrastim), which induces a transient release of CD34+ HSC and progenitors from the marrow into the peripheral blood, peaking at 5-6 days at an average of 76 per μl (a 50 fold increase), then declining, even with additional daily injections.^47,48 Subsequent studies indicated that CXCR4 (the receptor for stromal cell derived factor-1 [SDF-1], also known as CXCL12) tethers HSC within the marrow and that G-CSF breaks that tether by increasing granulocyte proliferation and release of granule enzymes in the marrow, thereby enhancing release of HSC into the circulation. More recently a small molecule inhibitor of the binding site of CXCR4, plerixafor (previously called AMD3100), was also shown to release HSC from the marrow to the circulation,⁴⁹ and when administered in the combination with G-CSF results in a synergistic mobilization of HSC to the peripheral blood.⁵⁰

The ease of apheresis collection of stem cells using continuous flow instruments following mobilization has resulted in this method becoming the preferred sourcing of HSC for many gene therapy clinical studies. Some patients with some inherited blood disorders that impair marrow proliferation (Fanconi anemia for example) will not mobilize, and infants who are too small for standard apheresis procedures still require bone marrow aspiration for sourcing HSC for gene therapy. Patients with sickle cell disease are at high risk of adverse events including vaso-occlusive crises when treated with G-CSF. Fortunately, several groups have shown that efficient and safe mobilization of HSC from marrow to peripheral blood can be accomplished using plerixafor alone as the mobilization agent in sickle cell patients.^51-54

Apheresis

The first apheresis device for separating blood components was developed in a collaboration between the National Cancer Institute and IBM,⁵⁵ and was appreciated at the time as a critical breakthrough in instrumentation for separation and collection of blood components.⁵⁶ A number of companies developed a series of progressively efficient continuous flow instruments to harvest of fractions enriched in HSC from G-CSF and/or plerixafor mobilized donors. Without this instrumentation now commonplace in blood collection centers, and simply viewed by the gene therapy community as “background” standard banking technology, much of the current progress in gene therapy would not have been possible.

Selection of HSC from Marrow or Apheresis Products

Multipotent permanently repopulating HSC express the CD34 surface antigen that was originally called My-10 as detected by a murine monoclonal antibody, raised against the KG-1a human tumor cell line.⁵⁷ Studies later demonstrated that immunomagnetic beads coated with anti-CD34 antibody could form complexes with HSC in a marrow or apheresis product, which in turn could be purified with magnets. These initial studies used incubation with chymopapain to release HSC from the beads.⁵⁸ This magnetic bead selection approach became the basis for commercial development of instruments for selective enrichment of HSC, two of which reached the late commercialization phase.

Baxter International, Inc (then called Baxter Healthcare Corporation) together with its “spin-off” subsidiary Nexell, Inc was the first to develop a fully automated instrument called the Isolex 300i that relied on binding of HSC to magnetic beads coated with the anti-CD34 antibody. Following magnetic separation, the washed product was exposed to an octapeptide that directly competed for the binding site of the anti-CD34 antibody to the CD34 antigen, thus removing the antibody complexed magnetic beads from the cells. This allowed the beads to be retained by a second pass through the magnetic field, yielding a cell product free of the beads and antibody.⁵⁹ A precommercial manual version of the Isolex system was used in one of the earliest clinical studies of gene therapy for CGD.⁶⁰

A very similar system developed in parallel by Miltenyi Biotec GmbH became their CliniMACS® CD34 Reagent System. This system uses anti-CD34 antibody chemically conjugated to dextran beads with an iron oxide/hydroxide core. A binding column with a magnetic gradient is used to separate the HSC bound to the anti-CD34 magnetic dextran bead from marrow or apheresis product. Unlike the Isolex system, there is no maneuver to separate the cells from the antibody conjugated beads, which are presumably degraded in culture or in vivo following transplantation. In October 2003 exclusive rights to market the Isolex system was acquired by Miltenyi and not further developed by them leaving only the CliniMACS® and its derivative devices available currently for clinical selection of HSC for gene therapies.⁶¹

Culture and Transduction of HSC

As noted previously, successful gamma retrovirus vector transduction requires one cell division cycle to complete integration into the genome of a cell. While lentivector transduction does not absolutely require cell division, HSC must enter at least the G1 phase of the cell cycle, and exposure of HSC to growth factors for a period of time in culture appears to enhance transduction efficiency. While it was initially hoped that gene editing methods might not require activation of HSC, a similar improvement of gene editing when HSC are cultured with growth factors has been observed by many investigators. Defining ex vivo culture conditions and growth factor combinations that enhance transduction or editing while minimizing loss of long-term marrow repopulating potential has been intensely studied. The discovery of each of the many critical HSC growth factors important for both efficient vector transduction and gene editing approaches will not be reviewed here; suffice it to say that without the work to identify, clone, and provide GMP compliant growth factors, gene therapy of HSC could not have advanced.

While there is no consensus about “best” conditions, the minimum combination of three growth factors, stem cell factor (SCF), FLT3-L, and thrombopoietin (TPO) is used by the majority of investigators for clinical gene therapy of HSC applications. Additional factors used by some investigators include interleukin (IL)-3 and/or IL-6. A key limiting issue is that culture in these growth conditions beyond 3 to 4 days results in significant loss of long-term repopulating potential. Outside the scope of this review are the discovery of other biochemical factors and conditions that may prolong the period during which HSC can be maintained in culture while delaying loss of long-term engraftment potential. This is an important emerging field for the future impact on HSC gene therapy.

While culture and transduction of HSC may be performed in standard tissue culture flasks, gas permeable flexible plastic culture bags are increasingly used to achieve more “closed system” handling. The earliest application of such systems in gene therapy clinical trials suggest better gas exchange, more consistent high viability and yield, and transduction efficiency.^60,62

Transduction enhancers

Maneuvers to achieve the highest efficiency transduction by gamma retroviral and lentiviral vectors have been an important aspect of the history of gene therapy. Quite early on, it was shown that addition of certain charged polymers to the transduction culture, such as polybrene or protamine sulfate, would enhance transduction by gamma retrovirus vectors; of the two, protamine sulfate continues to be used to enhance transduction of lentivectors.⁶³ Physical maneuvers such as centrifuging the culture plate or gas permeable bag have also been shown to enhance transduction,⁶⁴ but the practicalities of application at clinical scale have limited translation of this maneuver into the clinic. It has been presumed but not proven that ionic polymers and centrifugation worked in part by enhancing binding of vector to cells. A major advance was achieved when David Williams discovered that a fragment of fibronectin, when bound to the surface of a culture vessel, would greatly enhance transduction by gamma retrovirus vectors, likely both by binding vector and via some signaling effect on the target cells.⁶⁵ This fragment was commercialized as RetroNectin® and became a standard element applied into the clinic for gamma retroviral gene therapy transductions at clinical scale.^66,67 While RetroNectin® is used by some for lentivector transduction, it may not have as strong effect in enhancing lentivector transduction as it does for gamma retroviral transduction.

Several very potent enhancers of lentivector transduction have been described recently and is an area of very active investigation. A group of related poloxamers,⁶⁸ the first of which has been developed into a commercial product available GMP compliant called LentiBOOST®, can enhance transduction efficiency 2- to 5-fold. Prostaglandin E2 (PG-E2) alone enhances transduction by lentivectors several fold,⁶⁹ and in combination with LentiBOOST® or other poloxymer enhances lentivector transduction 10- to 20-fold.^70,71 The combination of LentiBOOST® plus low dose (1 uM) PG-E2 has been applied to a clinical trial of lentivector gene therapy for X-linked SCID with more than 10-fold increase in efficiency of transduction of HSC.⁷² Another recently discovered transduction enhancer is cyclosporin H.⁷³ The use of transduction enhancers has had dramatic impact on the success of gene therapy using lentivectors to treat sickle cell disease; the ability to consistently achieve vector copy numbers >3 in autologous transduced HSC of sickle cell patients greatly enhanced the degree of clinical benefit and durability of disease correction⁷⁴ (and as reviewed in greater detail in other chapters in this series).

Conditioning regimens

Choice and intensity of marrow ablative conditioning for HSC targeted ex vivo gene therapy and determining in which situations it may also be appropriate to target T, B and/or NK cells remains as contentious a topic with as little broad consensus as the choice of conditioning for allogeneic bone marrow transplantation. Much has been “borrowed” from studies and experience with what is needed to achieve multilineage engraftment in the allogeneic transplant setting, including the concept that for transplant of some disorders, very modest non-myeloablative conditioning may be sufficient to achieve the desired clinical goal. There is emerging evidence that busulfan may be a better conditioning agent for targeting long-term repopulating HSC than melphalan, but the comparative toxicity profiles of these two agents have influenced some investigators to use melphalan over busulfan in some HSC gene therapy settings.

Many of the earliest trials of HSC gene therapy for treatment of immune deficiencies with gamma retrovirus transduced HSC did not use any conditioning, assuming that the infused autologous transduced HSC or T lymphocytes would have ‘advantage’ and therefore conditioning was not needed. The contrast between outcomes observed when no conditioning was used in trials for different diseases is striking. In early trials of gene therapy for chronic granulomatous disease (CGD), a disorder in which gene corrected HSC or differentiated progeny have no survival or growth advantage over unmodified cells, engraftment of transduced myeloid cells was very low and transient, despite infusion of high numbers of transduced cells.⁶⁰ In severe combined immunodeficiency caused by deficiency of adenosine deaminase (ADA-SCID), a survival and growth advantage was known to result in reconstitution of T cells in the setting of allogeneic transplant in infancy without conditioning. The very first study of retroviral gene therapy for ADA-SCID targeting only T lymphocytes did result in long-term persistence and even some expansion of gene corrected T cells. But the production of adenosine deaminase was insufficient to achieve adequate detoxification of deoxyATP metabolites and there was no correction of B cell function.⁷⁵ The initial study and similar studies of gamma retrovirus transduced HSC gene therapy for ADA-SCID⁷⁶ did not use any conditioning. Though low level correction of T lymphocytes was seen, the gene marked HSC did not persist in the marrow and the clinical benefit overall was very modest.⁷⁷ In contrast, the landmark first gamma retrovirus HSC gene therapy treatment of infants with X-SCID caused by mutations in the IL2RG gene conducted in Paris by Alain Fischer and Marina Cavazzana in 2000⁷⁸ was extraordinarily successful at restoring T cell immunity to almost all of the infants treated without conditioning. Myeloid and HSC gene marking was very low and did not persist, and for the most part these patients had modest to nil long-term restoration of B and NK cell function.

A critical conceptual breakthrough was achieved by Alessandro Aiuti and colleagues who prepared pediatric patients with ADA-SCID with very modest doses of busulfan conditioning before infusing autologous gamma retrovirus transduced HSC.⁷⁹ The outcome of this maneuver was achievement of robust T and B cells function in the majority of patients and associated significant detoxification of deoxyATP metabolites, in the face of low levels of gene marking of bone marrow HSC and myeloid cells. Other investigators began and have continued to use busulfan conditioning to enhance engraftment of transduced autologous HSC gene therapy for a number of other disorders. Of note was the first use of low dose busulfan to condition older children and young adults with X-SCID before infusion of corrective lentivector transduced autologous HSC. This resulted for the first time in significant long-term restoration of B cell function and production of NK cells in addition to the production of functional T cells.⁸⁰ This was followed by demonstration of robust restoration of functional T, B and NK cells in newly diagnosed infants with X-SCID undergoing autologous gene therapy with the same vector after conditioning with low dose busulfan.⁸¹ Thus low dose busulfan appears necessary and sufficient to achieve immune reconstitution in forms of SCID where lymphocytes have a selective survival and growth advantage. However, for treatment of disorders affecting myeloid cells or platelets, or for correction of metabolic disorders requiring substantial production of the detoxifying enzyme, higher levels of myeloid engraftment of gene corrected HSCs is required. For that reason higher doses of busulfan into the sub-myeloablative level were applied in both the early studies of using gamma retroviruses and more recent studies using lentivector transduced HSC gene therapy of X-linked CGD and Wiskott-Aldrich syndrome.^82-87

Alessandra Biffi conducted seminal studies defining the optimal conditioning agent for transduced HSC gene therapy in early-onset metachromatic leukodystrophy, where delivery of the corrective gene and enzyme production in the brain by microglia appears critical to therapeutic corrective benefit.⁸⁸ In preclinical studies, she showed unequivocally that high dose busulfan was uniquely capable of targeting and clearing brain microglia, which was an essential maneuver to achieving rapid replacement of this compartment by incoming gene corrected microglia precursors.⁸⁹ Other conditioning agents such as melphalan were not capable to achieve this goal of detoxification of harmful metabolites in the brain of patients with metabolic and storage disease disorders.

Beyond the scope of this historical review, but likely to have significant impact in the near future for gene therapy of HSC are the emerging developments in monoclonal antibodies directly targeting HSC. At least one of these antibodies targeting cKit (CD117) is already in the clinic for allogeneic transplant (ClinicalTrials.gov identifier: NCT04429191 and NCT02963064), and likely will be used for HSC gene therapy in the near future.

Learning from Adverse Events in Gene Therapy

In this last section, we review adverse events observed in clinical trials of gene therapy. These important events, while unfortunate for the participants and sobering to the field, have informed the science of gene therapy and have resulted in corrective measures going forward.

Early discussions about the potential hazards of gene therapy mostly centered around the possibility that recombination events during vector production or after transduction could result in the restoration of replication competence to the vector. Much of the engineering of both the transfer plasmid that is the core element of the treatment vector, and the packaging elements (separating these elements into different plasmids and altering codons to avoid recombination) for both gamma retroviral and lentiviral vectors was intended to prevent recombination events. Molecular and virus replication assays were developed to high sensitivity that could detect very small numbers of replication competent virus in the final vector production product, in transduced target cells and their culture media, and in the patients treated with gene therapy. The US FDA and European regulatory agencies established strict criteria for steps in the process and for when and how such testing for replication competent virus should be conducted.⁹⁰ Perhaps because of these strict testing criteria and careful engineering, no case of replication competent vector has been reported to occur in any patient treated with either gamma retroviral nor lentiviral vector gene therapy.

Since the integrating retrovirus vectors insert throughout the genome of HSC, which are long-lived with differentiating progeny that are highly proliferative, it was surmised that an insertional event could activate an oncogene or growth program, or inhibit a cancer protection gene or other gene critical for cell function. Because vector-related oncogenesis was rarely if ever seen in preclinical animal studies, insertional mutagenesis was presumed to be unlikely. It was thus both surprising and concerning when the first cancer, a T lymphocytic leukemia, was detected about three years after treatment in one of the patients in the otherwise clinically very beneficial trial in Paris of gamma retrovirus gene therapy in infants with X-SCID.⁷⁸ The T lymphocytic leukemia cells had vector insertion within the first intron of the LMO2 protooncogene such that the enhancer elements in the LTR of the vector activated the LMO2 promoter driving aberrant overexpression.^91,92 Through examination of stored blood samples, the investigators showed that this insertion was followed by other oncogenic events as this clone evolved progeny clones, eventually giving rise to the cancer. Subsequent studies of other patients demonstrated that activation of protooncogenes other than LMO2 could give rise to the same clonal evolution process to cancer. Of the first 18 infants with X-SCID treated with gamma retrovirus gene therapy who achieved long term gene marking correction of T lymphocytes in clinical trials at sites in Paris and London, 5 developed an insertional mutagenesis related leukemia in the first 5 years. A 6^th patient developed a leukemia at about 15 years after treatment, indicating that while the greatest risk occurs in the first few years after treatment, there remains a long-term risk of oncogenesis as well.

These insertional mutagenesis oncogenic events led to exploration of SIN modifications to gamma retroviruses, and accelerated the ongoing development of SIN lentivectors. It also led to development of specialized cell lines that could be used to assess the potential of any candidate vectors to activate the LMO2 gene.⁹³ While initially, recipients of gamma retrovirus transduced HSC gene therapy for ADA-SCID appeared to be spared, very recently the first case of insertional mutagenesis mediated cancer was observed to occur in a patient with ADA-SCID treated with the commercialized gamma retrovirus (Strimvelis®) (See: https://ir.orchard-tx.com/news-releases/news-release-details/orchard-statement-strimvelisr-gammaretroviral-vector-based-gene).

Gamma retrovirus insertional mutagenesis induced cancers were subsequently in patients undergoing autologous HSC gene therapy for X-linked CGD⁹⁴ and Wiskott-Aldrich syndrome.⁹⁵ In these patients insertions in or near the MECOM oncogene complex or other oncogenes more associated with the development of myelodysplastic syndrome (MDS) and myeloid leukemias was often the instigating event, followed by clonal evolution involving additional oncogenic events leading to the oncogenic clone. As with the development of a pre-clinical test to assess potential of activation of the LMO2 gene, a specific test assessing the potential of a given vector to activate MECOM and related protocol oncogenes in mouse bone marrow cells and drive immortalization was developed by Ute Modlich and Christopher Baum.⁹⁶ This in vitro immortalization pre-clinical assay assessing genotoxicity has become a standard component of the testing required by the US FDA and the European regulatory agencies for insertional gene therapy vectors.

While insertional mutagenesis had not been observed with SIN lentivectors, very recently, a patient with adrenoleukodystrophy treated with corrective SIN lentivector transduced autologous HSC in an otherwise clinically beneficial clinical trial⁹⁷ was reported to have leukemia.^98,99 Of note, the internal promoter of this SIN lentivector (MND) was originally derived from a murine gamma retrovirus vector. The MND LTR had been modified to reduce enhancer activity,¹⁰⁰ but no insulator was added to the vector to reduce any ‘cross-talk’ of this gamma retrovirus derived promoter to nearby genes. While studies are still ongoing to understand this event, the vector insertional event likely responsible for the oncologic transformation is in or near the MECOM protooncogene complex, invoking an oncogenic process similar to that observed in patients with X-linked CGD or Wiskott-Aldrich syndrome after gamma retrovirus-based gene therapy described above.

Two patients from the otherwise very successful clinical trial of lentivector transduced HSC gene therapy for sickle cell disease⁷⁴ have also developed acute myeloid leukemia.¹⁰¹ In the first patient, the leukemic clone did not have any vector insert, indicating that the leukemic process was not driven by vector insertion. In the second patient, the leukemic clone had a vector insert within intron 4 of the VAMP4 gene, which has never been implicated in oncogenesis.¹⁰² There was no effect on VAMP4 gene expression and very low expression of the beta globin transgene in the leukemic blast cells. Over 70% of the blast cells had monosomy 7 with partial loss of chromosome, with other genetic changes associated with the evolution of the original clone into the leukemic clone. The investigators concluded that the insert was a passive passenger with respect to the initiation and clonal evolution of the oncogenic process leading to this patient's leukemia. As leukemia is increasingly recognized in patients with sickle cell who receive myeloid conditioning with an allogeneic HSC transplant who either have mixed chimerism or loss of donor graft with recovery of host marrow, the authors also speculated that marrow stress from their disease may place sickle patients with pre-existing higher risk of leukemic transformation.

One of the first patients to be treated with lentivector gene therapy for β-thalassemia had a dominant clonal expansion of with vector insert into the HMGA2 gene primarily in the myeloid lineage to >10% of the lineage.³⁸ High mobility group A2 (HMGA2) protein is a non-histone architectural transcription factor that may serve as a growth element in HSC and primitive myeloid progenitors. In this patient the insert appeared to result in production of a truncated HMGA2 protein lacking sequence responsive to LET-7, which normally enhances degradation of HMGA2. In this setting the clone, while persistent, appeared to be benign, not affecting hematopoiesis nor with any evidence for oncologic evolution, and slow decreased in the clonal dominance over a 10 year follow up. In a clinical study of lentivector transduced HSC gene therapy for X-SCID,⁸⁰ a very significant overall increase in lentivector inserts within HMGA2 relative to what would have been expected from “random” insertion, particularly in the myeloid lineage has been seen. The clonal expansions observed in these patients appear to be caused by a cryptic splice acceptor in the cHS4 insulator element of the vector, which generates an aberrant transcript encoding a functional but truncated HMGA2 protein when the vector inserts into intron 3 of the HMGA2 gene. A study in primates showed that forced expression of 3’ truncated HMGA2 using lentivector resulted in expansion of transduced HSCs and myeloid progenitors likely through the growth effect of this factor, but did not affect normal hematopoiesis nor result in any malignant transformation.¹⁰³ While to date this appears to be benign and does not affect hematopoiesis, it bears watching for the long term. These observations point out the importance of testing lentivectors for cryptic splice sites as part of the screening and design process.

Summary

Decades of discovery and invention of critical theories, materials, molecular engineering, and devices have led to the current emerging era of clinical benefit from gene therapy.¹⁰⁴ We are likely only beginning to reap the full clinical benefit of our current technology, and the field is in a rapid phase of new discovery of methods that will enhance our ability to efficiently make desired changes to HSC to treat and cure disease. All current HSC gene therapy in the clinic is performed ex vivo requiring collection, purification, and culture of HSC, but future developments may allow in vivo delivery of gene modification materials that will gene alter HSC in situ for disease treatment. There also will almost certainly also be unexpected adverse events, which will enhance future safety and efficacy of gene therapy if these are properly studied.

Key points:

Gene therapy by genetic modification of hematopoietic stem cells (HSC) has reached a stage of development that has resulted in substantial clinical benefit. This chapter explores the separate threads of knowledge, conceptual design, materials and equipment required to reach our current era of clinically beneficial gene therapy.

Clinics care points:

• Gene therapy is the change in gene expression, achieved by replacing or correcting a disease-causing gene, inactivating a target gene, or inserting a new or modified gene, using a vector or delivery system.

• Gene therapy clinical trials are benefitting a growing number of patients with immunodeficiency, beta-globin disorders, and brain disorders.