Vaccine strategies for lowering LDL by immunization against proprotein convertase subtilisin/kexin type 9

Abstract

Purpose of review

mAbs targeting proprotein convertase subtilisin/kexin type 9 (PCSK9) have the potential to become groundbreaking therapies for the treatment of hypercholesterolemia. However, one major drawback of mAb-based therapy for a chronic condition like dyslipidemia is its relatively high cost. This review summarizes two recent studies describing novel vaccine approaches for lowering LDL-cholesterol by active immunization against PCSK9.

Recent findings

PCSK9 is a plasma protein secreted by the liver that controls cholesterol homeostasis by enhancing endosomal and lysosomal degradation of the LDL receptor. Two PCSK9 inhibitory mAbs (evolocumab and alirocumab) have recently been approved by the Food and Drug Administration and a third mAb (bococizumab) is in late stage clinical trials. Treatment with PCSK9 mAbs, in combination with statins, reduces LDL-cholesterol levels by as much as 40–60%. As an alternative to mAbs, there have been two recent studies describing the development of vaccines that target PCSK9. These studies have shown that PCSK9 vaccines can effectively induce high-titer antibody responses that reduce proatherogenic lipoproteins in animal models.

Summary

A PCSK9 vaccine-based approach could serve as a more widely applicable and a more cost-effective approach than mAb therapy for controlling hypercholesteremia and associated cardiovascular disease.

INTRODUCTION

There are significant and well established health benefits to lowering levels of circulating LDL-cholesterol (LDL-C) for the prevention of cardiovascular disease (CVD). A meta-analysis conducted by the Cholesterol Treatment Trialists’ Collaboration found that there is a ~20% decrease in CVD morbidity and mortality for every ~40 mg/dl reduction in LDL-C [1]. Currently, treatment with 3-hydroxy-3-methylglutaryl CoA reductase inhibitors (statins) is the standard of care for hypercholesterolemic patients. Although statins are usually quite effective at reducing LDL-C and well tolerated, intensive statin therapy has some risks, most commonly myopathy, and rarely more serious side-effects [2]. In addition, a significant fraction, approximately 20%, of high-risk patients with hypercholesterolemia do not achieve adequate control of LDL-C levels with only statin treatment [3]. Long-term compliance with daily statin therapy is also as low as 50% for some populations [4].

LDL-C in plasma is principally removed from circulation when it interacts with LDL receptors (LDL-R) that are largely expressed on hepatocytes in the liver (Fig. 1). Upon LDL-R binding, LDL-containing particles are endocytosed and then undergo lysosomal catabolism. Following this process, LDL-R is recycled back to the cell surface. Proprotein convertase subtilisin/kexin type 9 (PCSK9) is a serum protein produced by the liver that acts as a negative regulator of the LDL-R by blocking the recycling of the receptor to the cell surface. PCSK9 directly binds to LDL-R and mediates the internalization and degradation of the receptor, thus increasing the circulating levels of LDL-C by preventing its cellular uptake [5,6].

FIGURE 1.

Diagram of interaction of PCSK9 with LDL-R pathway. As shown on the left, when LDL binds to the LDL-R it causes the internalization and degradation of LDL but LDL-R is recycled back to the surface. As shown on the right, the binding of PCSK9 to LDL-R leads to the internalization of the receptor complex and the intracellular degradation of both LDL and LDL-R. PCSK9 is recycled and secreted back into the plasma, resulting in a further depletion of LDL-R, and thereby raising LDL plasma levels. LDL-R, LDL receptor; PCSK9, proprotein convertase subtilisin/kexin type 9.

A number of genetic studies have shown that mutations that modulate PCSK9 activity can have profound effects on serum LDL-C levels. Gain of function mutations in PCSK9 are associated with autosomal dominant hypercholesterolemia, a disease that is characterized by increased serum LDL-C levels (>300 mg/dl) and a corresponding increased risk of CVD [7]. In contrast, humans with loss-of-function PCSK9 mutations are hypocholesterolemic (15–25% decrease in LDL-C) and have disproportionately lower incidence of CVD (~50% reduction), most likely because they have a life long reduction of LDL-C [8]. Strikingly, individuals with compound heterozygote loss-of-function mutations, which largely abolish PCSK9 secretion and/or function, have exceptionally low serum LDL-C (<20 mg/dl) and appear healthy despite having no detectible circulating PCSK9 [9].

PROPROTEIN CONVERTASE SUBTILISIN/KEXIN TYPE 9 mAbs

Given the important role of PCSK9 in regulating LDL metabolism and the fact that loss-of-function mutations appear not to be associated with adverse effects, PCSK9 has emerged as an extremely attractive therapeutic target for lowering LDL-C. Clinical trials of PCSK9-specific mAbs, including evolocumab (Amgen), bococizumab (Pfizer), and alirocumab (Regeneron) have shown that these drugs, in combination with statins, reduce LDL-C levels by about 60% [10^■-12^■]. Although there is no definitive evidence yet that PCSK9 inhibitors will reduce CVD events, results from the Alirocumab trial [11^■] are suggestive that this will be a good strategy for reducing CVD events, particularly for high-risk patients with elevated LDL-C while on statins. A potential major advantage of PCSK9 inhibition is that it appears to work by a complementary mechanism to statins and significantly reduces LDL-C beyond what is possible with statin alone. It has been shown that statins increase circulating levels of PCSK9 by more than 30%, making them somewhat self-limiting in their ability to reduce LDL-C [13-15]. This likely occurs because a transcription factor that is indirectly upregulated by statins (Sterol Element Binding Protein-2) activates both the LDL receptor and Pcsk9 genes [16]. Indeed, statins are much more effective when there are low PCSK9 levels, as first shown in PCSK9 knockout mice [17]. PCSK9-targeted therapeutics may, therefore, have value in preventing and treating CVD in combination with statins, particularly in those patients resistant to statin therapy or who have discontinued statins because of adverse side-effects. In 2015, the Food and Drug Administration approved evolocumab and alirocumab as second-line treatment for high LDL-C for adults whose cholesterol is not adequately controlled by diet or statin treatment.

RATIONALE FOR ALTERNATIVES TO mAb THERAPY

mAb-based immunotherapies have radically improved the treatment of many different chronic diseases. Nevertheless, mAbs have significant shortcomings that can limit patient access and clinical utility. mAbs are relatively expensive to manufacture and these costs are passed on to patients. Anti-PCSK9 mAbs are currently priced at more than US$14 000/year [18]; thus their use can be cost-prohibitive, particularly for the long-term treatment of elevated LDL-C, a common chronic condition. Another limitation is that anti-PCSK9 mAbs need to be injected frequently (once or twice a month) and at high doses (~140 mg for the 2 ~ month regimen). This can result in tolerability issues and poor compliance. Moreover, it has been shown that a substantial proportion of patients treated with other types of mAbs lose responsiveness over time because of the induction of antidrug antibodies, a phenomenon that occurs even with fully humanized mAbs. For example, in a study that followed rheumatoid arthritis patients treated with the TNFa inhibitor adalimumab for 3 years, over a quarter of patients eventually developed anti-mAb antibodies, and development of these antibodies correlated with a lower likelihood of clinical remission [19].

ALTERNATIVE APPROACH FOR TARGETING PROPROTEIN CONVERTASE SUBTILISIN/KEXIN TYPE 9-ACTIVE VACCINATION

Active vaccination against self-antigens involved in chronic diseases is an appealing new alternative to mAb-based therapies [20,21]. The goal of active vaccination is to induce therapeutic effects similar to those provided by passive administration of mAbs, but with far fewer administrations and lower doses, and without the possibility of inducing drug-neutralizing immune responses.

STRATEGIES FOR DEVELOPING VACCINES AGAINST SELF-ANTIGENS

Induction of antibody responses against self-antigens, such as PCSK9, are normally limited by the mechanisms of B-cell tolerance. However, potentially self-reactive B cells are surprisingly common in circulation [22], and antiself antibody responses can be readily elicited by immunizing with vaccines that have features that provoke the efficient activation of self-reactive B cells. The first key component that is required for a vaccine to induce self-reactive antibodies is the presence of foreign T helper epitopes. B cells require help from CD4⁺ T cells to become fully activated (T help mediates class switching and differentiation into long-lived plasma and memory cells). Although B-cell tolerance mechanisms are leaky, the mechanisms of central T-cell tolerance are highly effective at preventing the egress of self-reactive T cells from the thymus. Thus, even when potentially autoreactive B cells do encounter self-antigen, they fail to receive T help, because T cells specific for self-antigens are not normally present. Thus, effective vaccines for self-antigens must be physically linked to a source of foreign T-cell helper epitopes. A second feature that can dramatically enhance the ability to induce antiself antibody responses is antigen multivalency. It has long been recognized that antigens, which have highly dense, repetitive (multivalent) structures, such as icosahedral virions, can activate B cells and induce antibody responses at much lower concentrations than monomeric antigens [23-26]. Increasing antigen valency can be valuable in enhancing the induction of high-titer, long-lasting antibody responses against foreign antigens found in pathogens, but it is particularly critical for inducing strong antibody responses against self-antigens [26-28].

VACCINES TARGETING PROPROTEIN CONVERTASE SUBTILISIN/KEXIN TYPE 9

In the first report on the vaccination against PCSK9 in 2012, mice were immunized with human recombinant PCSK9, along with a DNA oligonucleotide as an adjuvant [29]. High-titer antihuman PCSK9 antibodies were detected after four vaccinations, but the titers were considerably lower against the mouse PCSK9 protein. Nevertheless, total cholesterol levels were reduced by approximately 40% after the vaccination. The human and mouse PCSK9 proteins are only about 78% identical so differences in the amino acid sequence between the mice and human proteins most likely accounted for the formation of cross-reactive antibodies. Two recent studies have extended these findings by showing that vaccines targeting human PCSK9 peptides that are nearly identical to mouse sequences can also induce a strong antimouse antibody response and reduce plasma cholesterol. AFFiRiS AG, an Austrian biotechnology company, developed a panel of PCSK9 vaccines in which short peptides that mimicked PCSK9 epitopes were conjugated to a keyhole limpet hemocyanin, a carrier protein that provides a source of foreign T helper epitopes [30^■■]. In mice and rats, the vaccine elicited high-titer anti-PCSK9 antibody responses that significantly reduced total cholesterol levels. The anti-PCSK9 antibodies persisted for at least a year after immunization. Anti-PCSK9 antibodies had a half-life of about 4 months, which is longer than the half-life of passively infused mAbs. Importantly, a booster vaccination a year after the initial series of immunizations effectively increased antibody levels to near peak titers. The vaccine failed to elicit detectable T-cell responses against PCSK9 and was not associated with any other toxicities.

We have developed a candidate PCSK9 vaccine in which a human PCSK9 peptide is displayed at high-valency on the surface of a bacteriophage virus-like particle by chemical conjugation (VLP) [31^■■]. VLPs are formed upon overexpression of viral structural proteins, which then spontaneously self-assemble into particles that resemble the virus from which they were derived but are noninfectious because they lack viral nucleic acid. VLPs have been used as one of the most common approaches for producing multivalent vaccines. Using a variety of chemical or genetic tools, antigens can be readily displayed at high valency on the surface VLPs. This approach has been successfully used to target self-molecules that are involved in the pathogenesis of a diverse variety of chronic diseases mediated by self-antigens, including Alzheimer’s disease, hypertension, inflammatory diseases, and certain cancers [21]. Many of these vaccines have been shown to have clinical efficacy in animal models and several have been tested in human clinical trials. For example, clinical trials of a bacteriophage VLP-based vaccine targeting angiotensin II, a regulator of blood pressure, showed that this vaccine was highly immunogenic and significantly reduced blood pressure in hypertensive patients [32].

We produced VLP-based vaccines that targeted five short peptide regions of human PCSK9 protein that were predicted to be directly involved in LDL-R binding. All of the vaccines elicited high-titer antibodies against the test peptides and several showed a good response against the intact human recombinant PCSK9 protein. The vaccine targeting PCSK9 amino acids (207–223), which was nearly identical in amino acid sequence between humans and mice, mediated the largest reduction in total cholesterol levels, ~30% relative to controls. This vaccine targeted a random coil region of PCSK9, which may have also accounted for its ability to stimulate high-titer antibodies against the intact PCSK9 protein. We also performed a small pilot study in rhesus macaques that were simultaneously treated with simvastatin. Immunization elicited high-titer anti-PCSK9 antibodies against the intact protein and vaccinated macaques had a marked reduction in LDL-C (30–40%) relative to control macaques treated with statin alone, but had no significant effect on HDL-cholesterol levels. Taken together, these studies demonstrate the feasibility of an active vaccination approach for lowering LDL-C levels by targeting PCSK9.

IS TARGETING PROPROTEIN CONVERTASE SUBTILISIN/KEXIN TYPE 9 SAFE?

Purposefully, eliciting antibodies against a self-antigen understandably raises safety concerns. However, accumulating clinical data suggests that targeting PCSK9 is likely to be safe. Theoretical concerns that PCSK9 inhibition might cause adverse neurocognitive events were not borne out in clinical trials of anti-PCSK9 mAbs. There are some data in mice linking PCSK9 knockout with a deficiency in liver regeneration [33], but humans with mutations that abolish PCSK9 expression are apparently healthy without any signs of liver disease [9]. T-cell tolerance against most self-antigens is effective, especially against serum-associated self-antigens, such as PCSK9, that are routinely exposed to the systemic immune system, and T-cell responses against PCSK9 were not detected in the AFFiRiS study [30^■■]. Because the response to a PCSK9 vaccine can likely be controlled by modulating the dose and the number of booster immunizations, and the fact that antibody responses to other self-antigen targeted vaccines wane over time [20], it should be possible to make an initial assessment of vaccine safety in a dose escalation trial, in which the titers in the initial vaccines would be low. There is already a regulatory precedent for this sort of trial design for vaccines targeting self-antigens [32,34,35].

CONCLUSION

A vaccine against PCSK9 for lowering LDL-C appears to be an attractive alternative to mAb therapy, but there are many future challenges that will have to be addressed. First of all, vaccination of human populations is known to result in a highly variable antibody response, so it will be important to formulate vaccination protocols that consistently induce high-titer anti-PCSK9 antibody responses. Our bias is of multivalent particle-based vaccines that induce high-titer antibody, but even VLP-based vaccines may need to be augmented with adjuvants to ensure maximum responsiveness. In addition, it will be important to have longitudinal data on the kinetics of antibody responses so that boosting schedules can be devised to maintain antibodies at levels that have the desired therapeutic effects. Finally, it will also have to be eventually shown in large-scale clinical trials that lowering LDL-C by a vaccine will decrease CVD events. If a PCSK9-based vaccine that safely and effectively lowers LDL-C can be developed, it would likely, however, represent a major therapeutic breakthrough, because of its potential applicability to the general population and the global impact it could have in reducing CVD.

KEY POINTS.

• Passive immunization by infusion with mAbs against PCSK9 lowers LDL-C in human clinical trials.

• Active immunization with a vaccine to induce autoantibodies against PCSK9 also lowers LDL-C in animal studies.

• Additional studies need to be performed to test safety and efficacy but a vaccine against PCSK9 may provide a low-cost alternative to mAb therapy.