Novel strategies to target PCSK9 in patients with hypercholesterolemia
Novel strategies to target PCSK9: beyond monoclonal antibodiesLiterature - Seidah NG, Prat A, Catapano AL et al. - Cardiovasc Res 2019;115(3): 510–518
In 2003, two gain-of-function (GOF) mutations in the PCSK9 gene were found to be associated with autosomal dominant hypercholesterolemia. In 2004, the capacity of PCSK9 to trigger hepatic LDL receptor (LDLR) degradation was described. When loss-of-function (LOF) mutations were discovered in individuals with lifelong low levels of LDL-c and low risk of coronary heart disease, PCSK9 became an attractive therapeutic target to reduce LDL-c levels.
When PCSK9 is inhibited, degradation of LDLR is prevented, thus allowing uptake of LDL particles from the blood to continue, resulting in lower LDL-c levels. Monoclonal antibodies (mAbs) were the first strategy developed to reduce PCSK9 levels. This is a summary of a review of other strategies targeting PCSK9 that are currently in development, as well as other promising innovative strategies.
Inhibition of the PCSK9-LDLR interaction by monoclonal antibodies
The first approved pharmacological strategy to reduce PCSK9 levels is injection of anti-PCSK9 mAbs. The mAbs evolocumab and alirocumab target circulating PCSK9 and significantly reduce LDL-c levels by ~60% and Lp(a) levels by 20-30%. Evolocumab and alirocumab significantly reduce the incidence of CV events by 15%, as shown by the FOURIER and ODYSSEY trials, respectively. Anti-PCSK9 mAbs are self-injected monthly or biweekly, which may help adherence, as compared with daily oral statins. Lower variability in LDL-c levels is obtained with every-2-week dosing than with every-4-week dosing. Two years of treatment with evolocumab provided persistent tolerability, adherence, safety, and efficacy in statin-intolerant patients who had statin-associated muscle symptoms. Overall, the currently available data have shown that anti-PCSK9 mAbs are safe, even in subjects who achieve very low LDL-c levels.
PCSK9 gene silencing and CRISPR editing
PCSK9 expression can be blocked by gene silencing, a physiological post-transcriptional process in which selected genes are turned off. A pharmacotherapeutic example of gene silencing is treatment with small interfering RNA (siRNA) that can control the expression of specific genes, including those involved in lipid and lipoprotein metabolism. Inclisiran is a double strand siRNA directed at PCSK9. Data from a phase I study in healthy volunteers with LDL-c ≥100 mg/dL found that a single dose of 300 mg subcutaneously injected inclisiran reduced PCSK9 levels by ~75% and LDL-c levels by ~50% for 6 months. Moreover, a two-dose regimen of inclisiran, 300 mg each, reduced LDL-c levels by 52.6% in the ORION-1 phase 2 trial, which is similar to levels achieved by mAbs for 6 months.
The siRNA inclisiran may reduce the liver intracellular PCSK9 levels, but drops in apoB, non-HDL-c, VLDL-c and triglycerides were similar to those obtained with mAbs, suggesting a limited effect of liver intracellular PCSK9 in cholesterol metabolism. The long-term effect of inclisiran on clinical outcomes and its effects in hypercholesterolemic and heterozygous familial hypercholesterolemia (FH) patients will be examined in the ORION-4 and a clinical phase 3 trial, respectively. So far, no side effects have been observed with inclisiran.
An alternative approach is the delivery of CRISPR-Cas9 to the liver for the in vivo base editing of PCSK9. Although in vivo and in utero studies in mice have shown a reduction in plasma and serum cholesterol levels, without apparent evidence of off-target mutagenesis, more research is needed to evaluate if translation into the clinic is appropriate.
Inhibition of PCSK9 mRNA translation
Another attractive strategy might be blocking PCSK9 synthesis by translational inhibition of PCSK9 mRNA. Indeed, an orally active compound could efficiently interrupt translation of PCSK9. However, further development of this strategy was halted due to the lack of PCSK9 specificity.
//PCSK9 // mRNA might also be post-transcriptionally downregulated by microRNA mimetics, such as miR-191, miR-222 and miR-224. However, it should be noted that these miRs are not specific for PCSK9.
Targeting the autocatalytic processing of proPCSK9
PCSK9 expression can also be inhibited by interfering with the autocatalytic processing of proPCSK9 into PCSK9. PCSK9 can only leave the endoplasmic reticulum (ER) after the autocatalytic cleavage of the zymogen proPCSK9 and the generation of a heterodimer of mature PCSK9 with its prodomain. Natural and engineered PCSK9 mutants could not undergo this auto-processing, thus stayed in the ER, and this works in a dominant-negative manner.
Heterozygote individuals who carry the PCSK9-Q152H LOF mutation that prevents the autocatalytic cleavage, indeed showed reduced circulating PCSK9 and LDL-c levels. That these patients are in good health suggests that complete retention of proPCSK9 in the ER does not result in unwanted side-effects. Thus, while inhibiting of the autocatalytic processing appears to be an attractive approach, engineering a small molecule inhibitor turns out to be challenging, because of the non-linear kinetics of the conversion of proPCSK9 into PCSK9 and because both plasma and ER membranes need to be crossed to reach the ER lumen.
Alternatively, inhibitors that prevent the interaction of the recently reported ER-resident cargo receptor SURF4 with mature PCSK9, could possibly reduce PCSK9 secretion.
Other inhibitors of PCSK9-LDLR binding
Other strategies to inhibit the interaction between PCSK9 and the LDLR include blockade with EGF-A-like peptides and small molecule inhibitors. Several engineering steps have been made towards development of an EGF-A-like peptide that effectively antagonizes PCSK9 activity. There is still room for improvement of the structure to generate a potent orally active small molecule inhibitor. Engineered adnectins are fragments of the fibronectin Type III domain that bind the catalytic subunit of PCSK9. While a ~50% reduction in LDL-c levels was achieved with a single injection in cynomolgus monkeys, this strategy has been abandoned as it could not favorably replace the mAb approach.
An albumin-binding domain-fused Anticalin protein has a somewhat longer half-life (~120 h) in plasma compared to mAbs (~60-120 h) and adnectins (~74-108 h). It reduces LDL-c levels with ~50-60% up to 21 days in cynomolgus monkeys. However, clinical trials are awaited to inform on efficacy and safety of these biologicals.
Blockade of PCSK9-LDLR sorting to lysosomes by CHRD antibodies
The C-terminal cysteine-histidine rich domain (CHRD) of PCSK9 is crucial for directing the PCSK9-LDLR complex to lysosomes/endosomes for degradation. Deletion of CHRD as a strategy failed because it resulted in an inactive secreted form of PCSK9 that still binds to LDLR, but does not induce its degradation. This research suggested that an as yet unidentified ‘protein X’ is needed to bind CHRD and/or LDLR to escort the PCSK9-LDLR complex to lysosomal degradation. Preventing complex formation of PCSK9-LDLR with protein X may therefore inhibit PCSK9 function. A bulky Fab that binds to CHRD indeed inhibited ~50% of the extracellular PCSK9’s ability to enhance the degradation of LDLR. Moreover, mAbs directed at CHRD decreased LDL-c levels by ~40% in cynomolgus monkeys.
Three single domain antibodies (sdAbs) have recently been generated that recognize exclusively the C-terminal M1/M3 domains of the CHRD of PCSK9. These sdAbs, when fused to a mouse Fc-sequence, reduced LDL-c levels with ~50% during more than 17 days in mice expressing exclusively human PCSK9. In contrast to mAbs, these sdAbs did not inhibit formation of the PCSK9-LDLR complex, nor did they increase levels of circulating PCSK9, but rather prevented PCSK9 activity on the LDLR. sdAbs are not as effective as mAbs, but they can be used to dissect out the sorting mechanisms of the PSK9-LDLR complex to endosomes/lysosomes and to identify critical residues regulating such trafficking.
Alternatively, the immune system may be instructed to eliminate endogenous circulating PCSK9 by means of PCSK9-peptide-based vaccines. Preclinical studies have shown a strong an long-lasting immune response, and consequently reduced plasma levels of PCSK9, total cholesterol, and non-HDL-c, as well as systemic inflammation in mice. Also, immunized mice even showed reduced atherosclerotic lesion area and aortic inflammation.
A phase I trial showed safety and tolerability of the vaccine in healthy subjects, with a PCSK9-specific antibody response in the vast majority of immunized individuals. LDL-c was reduced with a mean of 13.3% at week 70 and persisted for at least 30 weeks after the boost immunization. The possibility of serious unsuspected side-effects due to the absence of PCSK9 expression in adult livers needs to be excluded, especially in those with decreased liver function.
It is though that PCSK9 has other functions in the developing liver than in extrahepatic tissues. Alirocumab and evolocumab only reduce circulating hepatocyte-derived PCSK9. Although a few individuals lack functional PCSK9, it remains to be seen whether siRNA silencing of liver intracellular and secreted PCSK9 or lifelong deletion of PCSK9 in liver and extrahepatic tissues with CRISPR-Cas9 is equally beneficial. In conclusion, a paradigm shift in the treatment of hypercholesterolemia has occurred since PCSK9 and its role in the regulation of LDL-c was discovered. Although some of the injectable PCSK9-targeting agents are rapidly evolving, safe, orally active PCSK9 inhibitors may be developed in the future. These orally active PCSK9 inhibitors are expected to be cheaper and may therefore have a more widespread use worldwide in the treatment of various pathologies benefiting from low levels of PCSK9.