Hemophilia Gene Therapy and the Immune System

The immune system is comprised of two temporally modulated responses:

• The innate immune response^2,5
• The adaptive immune response^2,5

Innate immune response
Innate immunity is responsible for recognizing and controlling infections during the initial hours and days after exposure to a new pathogen (rapid response).^5,6 This response is not specific to a particular pathogen (antigen-independent), but rather acts as the body’s first line of defense against any invading pathogen, aiming to prevent its spread and movement throughout the body.^2,6 The innate immune system has no immunologic memory and, therefore, does not specifically recognize reinfection with the same pathogen in the future.⁷

A number of non-specific inflammatory cells are involved in the innate immune response, including macrophages, natural killer cells and dendritic cells.⁸ These cells have the ability to recognize conserved molecular features of pathogens, allowing them to be rapidly activated.⁶ Many of the cells in the innate immune system produce cytokines or interact with other cells directly in order to activate the adaptive immune system.^5,6

Activation of the innate immune system relies on pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), which allow innate immune cells to recognize foreign particles.^5,8,9 PRRs recognize common structures on pathogens, termed pathogen-associated molecular patterns (PAMPs);⁵ for example lipopolysaccharides,⁷ microbial unmethylated cytosine-phosphate-guanosine (CpG) DNA sequences (recognized by TLR9)¹⁰, single-stranded RNA from viruses (recognized by TLR7 and 8),¹⁰ and double-stranded DNA produced during infection (recognized by TLR3).¹⁰ Activation of PRRs results in the stimulation of transcription factors that induce the expression of pro-inflammatory cytokines (e.g. tumor necrosis factor and interleukin), or the type I interferon cascade.^2,5 Cytokines in turn initiate immune cell recruitment and induce inflammation, which are essential for the clearance of many pathogens.⁷ In parallel, TLR activation also triggers the induction of increased antigen-presenting capacity, therefore directing adaptive immune responses.⁹

Activation of the complement system is also an important component of the innate immune response. The complement system consists of a network of plasma and membrane bound proteins, and the activation of this pathway drives inflammatory and cytolytic responses to infectious organisms.¹¹

Adaptive immune response
The adaptive immune response is highly specific to the pathogen or foreign particle that has entered the body and provides long-lasting protection.¹ Adaptive immune responses are carried out by lymphocytes and there are two broad classes of responses: humoral and cellular immunity.¹

Both humoral and cellular immunity are key immunological considerations that exist within gene therapy.^2,12

• In the humoral immune response, B cells are activated to secrete antibodies which bind to the specific foreign particle.¹ These neutralizing antibodies can bind to vector capsids when administered via the peripheral circulation and may affect transduction efficiency, and in turn the efficacy of the gene therapy⁴
• In the cellular immune response, T cells are activated when they react directly with epitopes from the foreign particle.¹ Activated T cells cause the cytotoxic attack of transduced cells and are correlated with a loss of protein expression.⁵ The cellular immune response results in immunologic memory after first contact with the specific foreign antigen, which allows for a more rapid and effective response to reinfection⁵

During the process of gene therapy, and in vivo administration of the gene-carrying vector to the patient, the ‘foreign’ nature of the vector (in current hemophilia gene therapy approaches this is recombinant adeno-associated virus [rAAV]) can elicit an immune response.^2,13

In general, rAAV has gained wide acceptance as gene therapy transfer vectors due to their mild pro-inflammatory profile.² However, initial innate immune responses to rAAV, in addition to pre-existing immunity to AAV are possible.²

Innate immune response to rAAV gene therapy
Interaction of the rAAV vector components (transgene and capsid) with the innate immune system have the potential to determine the efficacy of gene therapy.² Studies show that the single-stranded DNA genome of rAAV can interact with the innate immune system via TLR9/MyD88 and type I interferon cascade, as well as triggering nuclear factor κB-dependent production of cytokine and chemokine release.² Another study suggested that CpG enrichment in the transgene via codon optimization may elicit an innate immune response, possibly through TLR9, which results in loss of transgene expression over the short term.¹⁴

Following endocytosis, rAAV capsids can be degraded in endosomes, resulting in the transgene or capsid being exposed to PRRs such as TLR9 or TLR2, triggering an innate immune response.¹⁵ Further to this, the capsid of rAAV (specifically serotype 2) may interact with the innate immune system via TLR2.² Although there is evidence that this immune recognition occurs, the implications of these interactions is not fully understood.²

Pre-existing immunity
During our lifetime, we can have natural exposure to wild-type AAV.¹⁶ AAV itself cannot replicate and cause an infection; it is dependent on co-infection with helper viruses to replicate (e.g. adenovirus or herpes simplex virus).² This exposure to AAV results in the generation of memory B and T cells.^2,13,16 Upon re-exposure to AAV, the innate immune responses are triggered by antigen-presenting cells, initiating the release of pro-inflammatory cytokines and the formation of neutralizing antibodies (nAbs) against various AAV serotypes,^13,16 in addition to the expansion of a pool of pre-existing CD8⁺ memory T cells.¹⁶

The recombinant capsid of the rAAV vector is a close mimic of a viral capsid (although it is not a virus and is not capable of inducing synthesis of viral proteins).² Immune responses to the vector can therefore be influenced by prior exposure to wild-type AAV from which the vector was engineered.² This pre-existing immunity against AAV serotypes may inhibit rAAV transduction of target cells following administration of the vector, thereby impacting the efficiency and limiting the delivery of rAAV-based gene therapy.²

Further to this, there is a high amount of similarity in the amino acid sequence and structural homology across AAV capsids of some of the different AAV serotypes.² Anti-AAV2 antibodies display the highest prevalence; however, anti-AAV antibodies show cross-reactivity over a wide range of serotypes.^2,5 Rates of seroprevalence for various AAV serotypes can differ with age, type of AAV, geographical location, testing method and other factors.^2,5,17

Pre-existing immunity against AAV may therefore impact the applicability of subsequent rounds of gene therapy with the same AAV, and at present, this suggests that re-administration of closely related AAV-derived vectors may not be successful due to AAV cross-reactivity.¹⁸

Immune responses against the administered vector can potentially impact the expected therapeutic effect of gene therapy¹ with the following implications:  

• Prevention of initial transduction – humoral immune response
  • Pre-existing AAV antibodies (from previous exposure to wild-type AAV) can bind to AAV-derived capsids and prevent initial transduction.^5,13,16 This may result in clearance of the vector before it reaches the target cell, preventing the expression of the therapeutic gene^5,13
• Loss of expression – cellular immune response
  • T cells can activate as part of the cell-mediated adaptive immune response, resulting in their expansion and subsequent destruction of transduced cells.¹³ As transduced cells are destroyed, transgene expression is lost resulting in the reduction of expression of the therapeutic gene^4,5,18 The enzyme-linked immune absorbent spot (ELISpot) is used to detect the cytotoxic T lymphocyte (CTL) response once gene therapy is administered²⁰
• Elevated liver enzymes
  • Elevated transaminases (alanine aminotransferase and aspartate aminotransferase) are commonly observed in the first months post-infusion with gene therapy.²¹ Increased liver enzymes have been seen to precede or coincide with a loss of transgene expression.²¹ A possible cause may be the interactions between the rAAV vector and host as a result of an immune response, which may include CTL responses to the AAV capsid, resulting in destruction of the transduced hepatocytes^2,21
• Gene therapies with rAAV infusion are currently limited to a ‘one-time’ treatment
  • Following gene therapy administration, nAbs to the AAV capsid may develop resulting in a long-lasting humoral response.^16,18 Co-prevalence of nAbs against multiple serotypes may limit options for switching to another AAV gene therapy vector.^18,20 This may preclude retreatment with gene therapy, limiting current gene therapies to a single administration^{18, 22,23}

In order to optimize therapeutic efficacy of gene therapy, immunogenicity management strategies must consider both humoral and cellular immune responses. At present, many of these proposed strategies are theoretical or under investigation in clinical trials.

Current considerations for the management of humoral immune responses include:

• Enrollment of nAb-naïve subjects: Most clinical studies have selected patients with low-to-undetectable anti-AAV against the recombinant vector being used in order to avoid immune memory responses to AAV^2,19
• Utilizing high vector doses to overcome nAbs: It is theorized that low capsid doses are more likely to be neutralized by anti-AAV antibodies, therefore affecting transfer efficacy, whereas higher capsid doses may be able to overcome this limitation.^2,24 However, there may be a theoretical limit to the vector dose above which transfer efficacy is diminished as capsid-specific T cells are activated and detect and clear transduced cells²
• Inclusion of empty capsids in vector preparations to adsorb anti-AAV antibodies: The empty capsids serve as decoys to adsorb circulating antibodies to AAV, allowing vector transduction.^2,19 However, this strategy increases antigen load (amount of antigen presented on Major histocompatibility complex [MHC] class I molecules) in the target organ, thus potentially triggering CTL immunity^2,19
• Administering immunosuppressive drugs to prevent or eradicate the humoral immune response to AAV: Selective immunosuppressive drugs have been shown to have some effect in inducing immune tolerance.^2,25 However, this strategy is not effective in the complete eradication of pre-existing high-titer nAbs²
• Changing the AAV serotype or engineering AAV capsids that are less receptive to nAbs: This approach is currently theoretical but has shown promise in laboratory testing.²⁶ Of note, modification of AAV vector tissue tropism may occur as a result of switching or altering the capsid and should be taken into consideration.^2,27 Additionally, the potential of this approach may be impacted by the high cross-reactivity  of anti-AAV capsid antibodies of some serotypes^2,28
• Isolation of target tissue to avoid vector dilution in blood and therefore exposure to nAbs: By isolating the target tissue from the systemic circulation (using techniques such as balloon catheters followed by saline flushing) it may be possible to avoid vector dilution in blood and exposure to nAbs.^19,25,29 This strategy limits systemic exposure to the vector; however, it is not useful if systemic gene transfer is required for therapeutic efficacy, and it is not feasible for all target tissues¹⁹
• Repeating plasma exchange to adsorb immunoglobulins: This is a non-invasive procedure that has been shown to be effective in reducing anti-AAV antibody titer and thereby minimizing contact between the vector and nAbs.^19,25 However, several cycles of plasmapheresis are required and high-titer AAV antibodies may not be completely eradicated using current approaches^30–32
• Non-specific cleavage of circulating immunoglobulins: Treatment with the cysteine protease IdeS (derived from Streptococcus pyogenes) may provide a nAb-free period for AAV vector delivery²⁵
• Masking vector epitopes: Using lipids or other cell-derived products to coat the surface of vectors may permit them to escape detection by nAbs²⁵

Current considerations for the management of cellular immune responses include:

• Selecting AAV-naïve patients: The selection of AAV-naïve patients with no prior infection from wild-type AAV avoids memory immune responses to AAV²
• Utilizing efficient AAV vectors of hyperactive variants to reduce the vector dose: Animal models suggest that this may be effective in preventing T-cell-mediated clearance of transduced cells;³³ however, for protocols where the expected therapeutic dose would be markedly higher than those tested to date in humans, it may not be possible to apply this approach.² Additionally, this approach requires the use of hyperactive variants and these are not available for every transgene²
• Administering immunosuppressive drugs to block T-cell responses to AAV: Steroids are known to block the T-cell responses directed against the AAV capsid.^2,12,23 Further to this, several immunosuppression drugs can be administered for prolonged periods. However, there are risks associated with systemic immunosuppression in patients, particularly when blocking the induction of regulatory T cells^2,34,35
• Using proteasome inhibitors or mutant capsids that are not efficiently ubiquitinated: Foreign proteins are targeted for degradation by ubiquitination, an essential step for their degradation by proteasomes.³⁶ Proteasomal degradation results in antigenic peptides that are presented on the MHC class I molecule on the surface of transduced cells.^2,19 The use of both proteasome inhibitors and mutant capsids that are not efficiently ubiquitinated have shown to have some effect in reducing presentation of AAV capsid antigen on MHC I molecules.^2,34 However, the effect may be limited, or may require pharmacotherapy for longer periods²