Differences in sensitivity of monkeys and humans to antisense oligonucleotide (ASO)–induced complement alternative pathway (AP) activation were evaluated in monkeys, humans, and in serum using biochemical assays. Transient AP activation was evident in monkeys at higher doses of two 2′-O-methoxyethyl (2′-MOE) ASOs (ISIS 426115 and ISIS 183750). No evidence of AP activation was observed in humans for either ASO, even with plasma ASO concentrations that reached the threshold for activation in monkeys. The absence of complement activation in humans is consistent with a query of the Isis Clinical Safety Database containing 767 subjects. The in vivo difference in sensitivity was confirmed in vitro, as monkey and human serum exposed to increasing concentrations of ASO indicated that monkeys were more sensitive to AP activation with this class of compounds. The mechanistic basis for the greater sensitivity of monkeys to AP activation by 2′-MOE ASO was evaluated using purified human or monkey factor H protein. The binding affinities between a representative 2′-MOE ASO and either purified protein are similar. However, the IC50 of fluid-phase complement inhibition for monkey factor H is about 3-fold greater than that for human protein using either monkey serum or factor H–depleted human serum. Interestingly, there is a sequence variant in the monkey complement factor H gene similar to a single nucleotide polymorphism in humans that is correlated with decreased factor H protein function. These findings show that monkeys are more sensitive to 2′-MOE ASO–mediated complement activation than humans likely because of differences in factor H inhibitory capacity.
The potential attributes of developing antisense drugs for multiple molecular targets and disease states on the basis of a common platform chemistry have been realized over the years in many ways (Crooke, 2001). Although the diversity of antisense mechanisms and chemical modifications makes it difficult to generalize across divergent chemical classes, it is possible to generalize the properties of sequences within a chemical class that have been selected for nonclinical assessment and clinical development. A broad assessment of the overall safety and tolerability profiles across multiple sequences can be particularly useful in translating the nonclinical to clinical experiences to provide the best understanding and ensure the safe use of these drugs. This type of analysis is possible within 2′-O-methoxyethyl (2′-MOE) gapmer antisense oligonucleotides (ASOs) that are approximately 20 nucleotides in length, as this is the class of ASOs that has accumulated the greatest amount of experience in clinical trials, and continues to be the most broadly studied of the chemically modified ASOs.
The nonclinical safety assessment of 2′-MOE ASOs has relied heavily on cynomolgus monkeys to identify potential toxicities due to remarkable similarities with humans in pharmacokinetics, tissue distribution, and elimination (Yu et al., 2007), and a similar level of sensitivity to pathogen-associated molecular pattern–associated innate immune activation (Hartmann et al., 1999; Klinman et al., 2002). Still, as with any nonclinical toxicology evaluation, there is uncertainty about how well the effects observed in animals translate to humans. Activation of the complement alternative pathway (AP) by 2′-MOE–modified ASOs is a well known class effect at high doses in monkeys (Henry et al., 1997a,b, 2002). As such, it is important to understand the transient AP activation and its relevance to humans.
Acute complement AP activation was initially encountered in nonclinical studies reported in the mid-1990s, largely because of the dramatic hemodynamic changes associated with this event in monkeys following intravenous bolus at high doses (Galbraith et al., 1994; Henry et al., 2002). However, upon more thorough characterization of the threshold nature of the plasma ASO exposure-effect relationship, and better understanding of the mechanism of activation, complement activation has been of relatively little concern in clinical trials (Crooke, 2001). In part, the concern has been addressed by clinical dose regimens designed to keep plasma concentrations below the threshold concentration for complement AP activation in monkeys. These include relatively short-duration intravenous infusions or administering the ASO by subcutaneous injection.
Although the concern for complement activation in clinical trials has been effectively addressed through dosing paradigms (Glover et al., 1997; Nemunaitis et al., 1999; Sewell et al., 2002), it is still of interest to better understand complement activation in monkeys and the potential relevance to other laboratory animals and humans. Over the years, it has become clear that complement activation does not occur in mice, rats, pigs, or dogs with 2′-MOE ASOs (Isis internal data). Furthermore, the doses of 2′-MOE ASOs for some indications in humans, such as oncology, have measured plasma concentrations that approach or exceed the threshold concentrations identified in monkeys, without causing complement activation (Chi et al., 2005). These data suggest humans are actually less sensitive than monkeys to complement activation, and are corroborated by in vitro complement activation models using human or monkey serum (Henry et al., 2014).
Although it is not known why the monkey is uniquely sensitive to AP activation, we do know the mechanism of action. In monkeys, ASO-mediated activation of the AP is directly related to a nonspecific interaction between the ASO and factor H protein, a regulatory component of the AP (Henry et al., 1997a, 2002, 2014). Factor H is the predominant fluid-phase regulator of the AP. Impairment of factor H function or decreased quantity of available factor H leads to the potential for uncontrolled AP activation (Pickering et al., 2002; de Cordoba and de Jorge, 2008).
To better document the differences in sensitivity to 2′-MOE ASO–induced complement AP activation between monkeys and humans, this paper provides a detailed comparison between these species following both intravenous and subcutaneous administration, and reports a broader analysis of complement split product measurements from our monkey and human safety databases that comprises data from 14 unique 2′-MOE ASOs. This paper also demonstrates differences in the relative inhibitory activity of factor H on the AP activation between the species that likely explains the atypical sensitivity of monkeys to this effect.
Materials and Methods
Characterization of ASOs
ISIS 426115, ISIS 183750, ISIS 104838, and ISIS 393929 are 20-nucleotide phosphorothioate (PS) ASOs with 2′-MOE modification. Their sequences and modifications are as follows:
ISIS 426115: GCAGCCATGGTGATCAGGAG
ISIS 183750: TGTCATATTCCTGGATCCTT
ISIS 104838: GCTGATTAGAGAGAGGTCCC
ISIS 393929: GCTTCAGTCATGACTTCCTT
The underlined bases contain 2′-MOE modifications, and all cytosines are 5-methyl cytosine. All four ASOs were synthesized at Isis Pharmaceuticals, Inc. (Carlsbad, CA).
Monkey studies were conducted at an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)–accredited facility. Male and female cynomolgus monkeys (Vietnamese or Chinese origins) were individually housed in cages in a temperature- and humidity-monitored environment, providing a 12-hour light/dark cycle. Animals were fed twice daily with ad libitum access to water at any time. All study protocols were approved by the testing facility's Institutional Animal Care and Use Committee (IACUC) prior to dose administration. Both studies were conducted in compliance with the Food and Drug Administration Good Laboratory Practice for Nonclinical Laboratory Studies regulations.
In a 6-week monkey study, animals (3–5/sex per group) were dosed subcutaneously on Days 1, 3, 5, and 7, and weekly thereafter up to Day 42, at dose levels of 0, 4, 8, 12, or 40 mg/kg.
In a 4-week monkey study, animals (3–6/sex per group) were given a loading dose of 0, 3, 15, or 50 mg/kg by continuous intravenous infusion for 72 hours starting on Day 1, followed by maintenance doses of 0, 2, 4, or 15 mg/kg by 3-hour intravenous infusion on Days 8, 11, 15, 18, 22, 25, 29, and 32.
Healthy volunteers (age 18–55 years) were enrolled in the phase I study. This was a double-blind, placebo-controlled, dose-escalation study conducted at a single center. Subjects were randomized to ISIS 426115 or placebo prior to administration of their first dose. The study consists of four single-dose and four multiple-dose subcutaneous cohorts. Subjects were given single or multiple doses by subcutaneous injection at 60, 120, 210, or 420 mg (approximately 0.9, 1.7, 2, or 6 mg/kg body weight in a 70-kg subject, respectively) for up to 6 weeks. Data presented were the Bb and C3a profiles from the multiple-dose cohorts (n = 6–10) on Days 1 and 36.
In a phase I dose-escalation study, patients (1–10 patients per group) with advanced cancer were given multiple cycles of treatment of various duration by 3-hour intravenous infusion at 200, 400, 600, 800, 1000, or 1200 mg (approximately 2.9, 5.7, 8.6, 11.4, 14.3, or 17.1 mg/kg body weight in a 70-kg patient, respectively). Given the variability of the treatment cycles among patients, data presented here only contained Bb and C3a levels at screen, Day 1 predose and end of infusion (EOI). All protocols were approved by the institutional review board, and the studies were in compliance with Good Clinical Practice and all applicable laws and regulations.
Samples for complement and toxicokinetic analyses were collected at various time points throughout the nonclinical and clinical studies. Blood samples were processed to EDTA plasma within 60 minutes of collection per test facility standard operating procedures and stored at −60°C or lower until analysis. The complement split products Bb or C3a were measured using human enzyme-linked immunosorbent assay (ELISA) kits (Quidel Corporation, San Diego, CA) which are cross reactive with monkey samples. Plasma ASO concentrations were measured by Isis Pharmaceuticals, Inc. by a hybridization-dependent nuclease ELISA method.
In Vitro Complement Activation Assay
Normal monkey (n = 3) or human (n = 3) serum was incubated with increasing concentrations of ISIS 426115 or ISIS 183750 at 0, 31.25, 62.5, 125, 250, or 500 μg/ml at 37°C for 30 minutes. The level of the complement split product C3a was evaluated using the human C3a ELISA kit. Zymosan was used as a positive control for complement AP activation in serum.
In Vitro Factor H Protein Functional Assay
Factor H Protein Fluid-Phase Function.
Normal monkey serum (n = 4) was treated with increasing concentrations of human or monkey factor H protein at up to 200 μg/ml, immediately followed by the addition of 250 μg/ml ISIS 104838. Samples were then incubated at 37°C for 30 minutes before measuring for C3a using the human ELISA assay. In another in vitro model, factor H–depleted human serum was incubated with various concentrations of human or monkey factor H protein up to 500 μg/ml (n = 3), mixed well, then immediately followed by 10 µl of MgEGTA (final concentration at 1 mM) to allow the initiation of AP activation. After 30 minutes of incubation at 37°C, samples were measured for C3a using the human ELISA assay. The monkey factor H protein was purified by M.K.P. at University of Texas Health Science Center using pooled serum collected from Asian-sourced cynomolgus monkeys. The human factor H protein and factor H–depleted human serum were purchased from Complement Technology, Inc. (Tyler, TX).
In Vitro Factor H Protein Cell Surface Functional Assay (Modified AP Complement Activity Assay)
Serum from a single patient with low factor H activity was incubated with rabbit red blood cells (RBCs) with or without the presence of human or monkey factor H protein up to 183 μg/ml. Pooled serum from healthy donors with normal factor H activity was used as a negative control. All in vitro factor H protein functional studies were conducted at the Complement Laboratory, National Jewish Health, Denver, CO.
Clinical Safety Database Analysis
Bb levels from 250 placebo-treated subjects and 767 ASO-treated subjects or patients enrolled in 14 different clinical trials were included in the query. All 14 compounds were 2′-MOE ASOs and were delivered by subcutaneous administration. The upper limit of normal (ULN) for complement activation was defined as the mean ± 3 S.D. A positive signal was defined as a Bb level >2× ULN.
ASO–Factor H Protein Binding Assay
Solutions of human or monkey factor H protein were prepared over a range of concentrations (100 nM to 15 μM) with 100 nM fluorescein-labeled ISIS 393929 (20-mer, 2′-MOE ASO). Factor H protein and ISIS 393929 were allowed to equilibrate at 37°C for at least 30 minutes before analysis by size exclusion chromatography. In brief, a 25-μl sample was loaded onto a Zenix SEC-300 4.2 × 300–mm column (Sepax Technologies, Newark, DE) cooled to 8°C and separated with 1× phosphate-buffered saline (Gibco, Life Technologies, Grand Island, NY) at 350 μl/min. Protein migration was monitored by UV absorption at 280 nm. Migration of fluorescein-labeled ASO was monitored with fluorescence detection using Ex520/Em480 and peak area determined for protein-bound and -unbound ASO. The fraction bound was defined as the peak area protein-bound ASO/unbound ASO. The concentration of free factor H protein was defined as the total concentration of protein − (100 nM × fraction ASO bound). The binding affinity (Kd) was determined using GraphPad Prism (GraphPad Software Inc., La Jolla, CA).
Cynomolgus Monkey Complement Factor H Gene Amino Acid Sequence
The cynomolgus monkey complement factor H (CFH) gene sequence was generated by P.R.R. at National Jewish Health. The amino acid sequence is as follows:
The cynomolgus CFH gene amino acid sequence was blasted against published rhesus monkey (Macaca mulatta) (accession XP_001111875) and human CFH sequences (accession AAI42700).
2′-MOE ASO–Induced Complement Activation Is Common in Monkeys but Not in Humans.
In the 6-week study with ISIS 426115, monkeys were dosed subcutaneous at 0, 4, 8, 12, or 40 mg/kg, and plasma samples for assessment of complement activation were collected at multiple time points following dosing on Days 1 and 42 (Fig. 1, A and B). Transient increases in Bb concentrations above predose baseline were observed on Day 1 at 12 (3.7-fold) and 40 mg/kg (22-fold). Peak levels of Bb were reached approximately 4 hours postdose, and then returned to baseline by 24 hours postdose. A 2- to 3-fold increase in Bb was common in control monkeys, suggesting that level of activation was secondary to animal handling or the dosing procedure. On Day 42, a similar increase (3.3-fold) was observed at 12 mg/kg, but a lower peak Bb level (10-fold) was noted at 40 mg/kg, as compared with Day 1. The elevation in Bb corresponded to ASO plasma Cmax values of 29 and 72 μg/ml for 12- and 40-mg/kg groups on Day 42, respectively (Fig. 1C). The absence of complement activation at the lower doses in monkeys is attributed to lower plasma drug concentrations that are below the threshold for activation.
In comparison, for the phase I study in healthy volunteers (Fig. 1, D–F), doses of ISIS 426115 were 60, 120, 210, or 420 mg (approximately 0.9, 1.7, 2, or 6 mg/kg body weight in a 70-kg subject, respectively) and were delivered by subcutaneous administration. No complement activation was measured after either a single dose on Day 1 or multiple doses on Day 36 (Fig. 1, D and E). ASO plasma Cmax values in humans ranged from approximately 4 to 9 μg/ml at the highest dose of 6 mg/kg on Day 36 (Fig. 1F), which were well below the Cmax associated with complement AP activation in monkeys.
In the 4-week monkey study with ISIS 183750, animals were given loading doses of 0, 3, 15, or 50 mg/kg by continuous intravenous infusion for 72 hours starting on Day 1, followed by maintenance doses at 0, 2, 4, or 15 mg/kg by 3-hour intravenous infusion twice weekly thereafter starting on Day 8. Complement activation was evaluated prestudy (baseline) and then at 3 hours (EOI) and 24 hours postdose on Day 18 (Fig. 2, A and B). Increases in complement split products, approximately 5- and 24-fold over prestudy baseline for Bb and C3a, respectively, were observed at 15 mg/kg, with peak response noted at the end of the 3-hour infusion and corresponding to an average peak plasma ASO concentration of approximately 75 μg/ml (Fig. 2C). No meaningful increases in Bb or C3a were measured at doses ≤4 mg/kg, where peak plasma concentrations were ≤29 μg/ml.
The complement split product profile in the phase I oncology trial with ISIS 183750 following a 3-hour infusion of doses up to 1200 mg (approximately 17.1 mg/kg body weight in a 70-kg patient) is presented in Fig. 2, D and E. As shown, peak plasma ASO concentrations (at EOI) on Day 1 ranged from 40 to 70 μg/ml at the higher doses (Fig. 2F), which approached the apparent threshold for complement activation in monkeys. However, none of the patients in the study had clinically meaningful increases in Bb or C3a (≤2-fold of predose) at EOI on Day 1, as compared with preinfusion levels. Furthermore, no evidence of complement activation was observed in patients after multiple doses of ISIS 183750 (data not shown).
No Evidence of Treatment-Related Complement Activation with 2′-MOE ASOs in Humans Based on the Isis Clinical Safety Database.
The Bb profile from 250 placebo and 767 2′-MOE ASO–treated subjects in 14 different clinical trials is summarized in Table 1, with all ASOs being delivered by subcutaneous administration. Overall, only a small percentage of subjects in the ASO-treated group (5.3%, or 41 out of 767) had Bb levels ≥2× ULN (1.6521 μg/ml). This incidence is comparable to the incidence of Bb increases observed in the placebo group (4.4%, or 11 out of 250). In addition, there was no dose-dependent increase in Bb associated with subcutaneous ASO treatment. The incidences were 7.6, 4.9, and 0.7% for dose ranges of <2.1 mg/kg, 2.1–4.3 mg/kg, and >4.3 mg/kg, respectively.
ASO-Induced Complement Activation Is Recapitulated in Monkey Serum but Not in Human Serum.
Serum collected from human or monkey was incubated with increasing concentrations of ISIS 426115 or ISIS 183750, ranging from 31.2 to 500 μg/ml, for 30 minutes at 37°C. A dose-dependent increase in C3a, up to 7- to 10-fold of saline control, was measured in monkey serum with both ASOs, whereas no or minimal complement activation (as indicated by C3a) was present in human serum under the same conditions (Fig. 3, A and B). These data suggest differential sensitivities of human and monkey serum in response to ASO-induced complement activation.
Monkey Factor H Protein Is a Weaker Regulator of Fluid-Phase Complement Activation than Human Factor H Protein.
The potency for fluid-phase inhibition of AP was compared using purified monkey and human factor H proteins. Test systems were set up to compare inhibition of the AP following activation by either an ASO in monkey serum or MgEGTA in human factor H–depleted serum. In monkey serum, in vitro AP activation occurred following incubation with ISIS 104838 at 250 μg/ml, leading to a 3- to 4-fold increase in C3a without the presence of the purified factor H protein (data not shown). Concentration-dependent inhibition was measured for both monkey and human factor H proteins; however, a weaker inhibition was found with the monkey protein, as the IC50 was about 3.3-fold greater than with the human protein (Fig. 4A).
Since human serum is typically not responsive to ASO-induced complement activation in vitro, factor H–depleted human serum was used to further investigate the relative function of monkey and human factor H proteins (Fig. 4B). Once the factor H–depleted human serum was prepared, MgEGTA was used to allow the initiation of the AP activation, while inhibiting the classic pathway activation by chelating Ca2+. An approximately 15-fold increase in C3a (over baseline) was measured following MgEGTA treatment and 30 minutes of incubation, suggesting rapid complement activation (data not shown). Control of the AP was restored by supplementing monkey or human factor H protein to the physiologic concentration of 500 μg/ml. Notably, monkey factor H protein had weaker inhibition in this test system compared with human factor H protein, as the IC50 was about 2.9-fold greater than the human protein. This is consistent with the findings in monkey serum as described earlier.
No Difference in Cell-Surface Complement Inhibition Was Found for Both Monkey and Human Factor H Proteins.
The cell-surface inhibitory effect of factor H protein was tested using a modified AP complement activity (AH50) assay. Serum from a patient with low factor H protein activity (n = 1) was incubated with rabbit RBCs with or without the presence of human or monkey factor H protein. Pooled serum from healthy donors was used as a negative control. As shown, a dose-dependent inhibition of complement-mediated rabbit RBC lysis was observed with both monkey and human factor H proteins with similar potency (Fig. 5).
Similar Binding Affinity of ASO to Monkey and Human Factor H Proteins.
The relative binding affinity (Kd) of fluorescein-labeled ISIS 393929, a representative 2′-MOE ASO, to monkey or human factor H proteins is presented in Table 2. The Kd values for monkey and human factor H proteins were 2.6 and 2.9 μM, respectively. This indicates a comparable binding affinity between factor H protein and a 2′-MOE ASO in both species, and appears to have much tighter binding compared with other abundant plasma proteins, such as albumin, where the Kd values for the same 2′-MOE ASO (ISIS 393929) were about 45 and 24 μM for monkey and human proteins, respectively (Isis internal data).
Single Nucleotide Polymorphisms in Cynomolgus Monkey CFH Gene Sequence May Affect Protein Function.
The CFH gene amino acid sequence from cynomolgus monkey had 98% homology compared to the published rhesus monkey sequence, but only 88% homology compared to the human CFH gene amino acid sequence (Isis internal data). Among the nonconserved amino acid changes, a Y402N single nucleotide polymorphism (SNP) (tyrosine/asparagine polymorphism at amino acid 402 position) is found in cynomolgus and rhesus monkey CFH sequences, when compared with the human sequence (Fig. 6). SNPs or mutations of the CFH gene sequence could potentially have an influence on protein function.
Transient activation of the complement AP has been observed in studies in monkeys with 2′-MOE ASOs, and appears to be a monkey-specific phenomenon for which the magnitude is mostly dependent upon the ASO plasma Cmax. Complement activation is routinely monitored in clinical trials with 2′-MOE ASOs by measuring the same complement activation split products examined in monkeys (e.g., C3a, C5a, and/or Bb). Importantly, complement activation in clinical trials has largely been absent. For example, in a monkey study with subcutaneous administration of ISIS 426115, increases in complement split products were seen at doses of 12 mg/kg and were much more evident at 40 mg/kg, where the doses corresponded to a mean Cmax of 29.3 and 71.8 μg/ml, respectively. The absence of complement activation at clinically relevant doses of ISIS 426115 in humans was at least in part due to the ASO Cmax (up to ∼9 μg/ml at 6 mg/kg s.c.) that remained well below the threshold for complement activation found in monkeys. However, higher human doses (up to 17.1 mg/kg) in the ISIS 183750 oncology trial were comparable to the highest dose used in the corresponding monkey study. Even though there were significant increases in Bb and C3a in monkeys at 15 mg/kg with 3-hour intravenous infusion, no evidence of complement activation was found in patients, even with a Cmax of approximately 40–70 μg/ml, which achieved the threshold for activation in monkeys. These data suggest that monkeys are more sensitive than humans to 2′-MOE ASO–induced complement activation.
The published clinical experience to date confirms a lack of complement activation associated with 2′-MOE–modified ASOs following subcutaneous or short-term intravenous infusion (Sewell et al., 2002; Chi et al., 2005; Kastelein et al., 2006). This is consistent with a query of the Isis Clinical Safety Database that indicated no meaningful or dose-dependent increase in complement split products (e.g., ≥2× ULN for Bb) in 767 subjects with subcutaneous administration of 2′-MOE–modified ASOs at typical therapeutic doses (50–400 mg total dose). A phase I oncology study with a 2′-MOE ASO targeting the survivin protein using a 3-hour intravenous infusion of up to 750 or 1000 mg (10.7 or 14.3 mg/kg body weight in a 70-kg subject) reported minimal increase in Bb (<2-fold of baseline) but no C5a changes (Talbot et al., 2010; Tanioka et al., 2011). However, it is not uncommon to observe a low level of Bb elevation in placebo or 2′-MOE ASO–treated subjects or monkeys that is attributed to the dosing procedure, as a modest stress-induced complement activation has been shown to occur in humans (Burns et al., 2008). The absence of ASO-related complement activation with 2′-MOE ASO in humans, even at plasma concentrations at the activation threshold for monkeys, may also be partly attributed to the MOE substituents on the 2′ ribose position, which decreases the magnitude of complement activation by reducing nonspecific protein interactions (Henry et al., 2014), the most important of which may be decreasing the interaction with factor H protein.
The only clear exceptions to the absence of complement activation in clinical trials were following continuous 24-hour intravenous infusion of very high doses of first-generation full PS ASOs ISIS 3521 and ISIS 5132 (Rudin et al., 2001; Advani et al., 2005). Both cases were in oncology trials where patients were given up to the maximum tolerated doses of 24 or 30 mg/kg via a 24-hour intravenous infusion. Treatment-related increases in Bb and C3a were noted at higher doses and correlated with plasma ASO concentrations of approximately 12–19 μg/ml with 24-hour intravenous infusion, well below the threshold for activation in monkeys for the first-generation ASO following subcutaneous or short-term intravenous infusion. These data suggested that, in addition to the plasma Cmax,, the duration of drug exposure may also play a role in ASO-induced AP activation.
The differential response of humans and monkeys to ASO-induced AP activation was replicated in an in vitro model using monkey or human serum. The negligible complement activation in human serum with ISIS 426115 and ISIS 183750 was consistent with our experience with other ASO sequences, and suggested possible functional differences between human and monkey factor H in regulating the AP. A series of in vitro biochemical experiments were designed to address the mechanistic basis of the sensitivity of monkeys to ASO-induced complement activation, by directly comparing monkey and human factor H protein function in fluid-phase and cell-surface complement regulation, as well as ASO-protein interactions. Protein binding data suggested comparable high Kd for both monkey and human factor H proteins to a 2′-MOE ASO. However, the functional assays in monkey serum or factor H–depleted human serum demonstrated a weaker fluid-phase inhibition of AP activation by monkey factor H protein relative to the human factor H protein. These data suggest a possible functional difference in the monkey and human factor H proteins, and likely account for the differential sensitivity to in vivo complement activation.
Mutations and SNPs in the human CFH gene have been reported to cause functional changes in the protein, and have been implicated in a variety of conditions, including age-related macular degeneration (Edwards et al., 2005; Hageman et al., 2005; Haines et al., 2005), atypical hemolytic uremic syndrome (Taylor, 2001; Atkinson and Goodship, 2007; Jalanko et al., 2008; Lehtinen et al., 2009), and membranoproliferative glomerulonephritis type II/dense deposit disease (Abrera-Abeleda et al., 2006; Licht et al., 2006; Sethi et al., 2009). Among these, the Y402H (tyrosine/histidine polymorphism at amino acid position 402) SNP in the human CFH has been shown to be highly correlated with increased risk of developing age-related macular degeneration as well as diseases in kidney, brain, and vascular tissues (Edwards et al., 2005; Hageman et al., 2005; Haines et al., 2005; Donoso et al., 2010). The pathogenesis of these diseases is attributed to decreased function of the human factor H protein, which leads to increased complement activation. Interestingly, the sequence variants found in monkey (rhesus and cynomolgus) CFH genes, such as the Y402N, are similar to the human Y402H SNP, and may contribute to the weaker complement fluid-phase inhibition of monkey factor H protein seen in the in vitro models. The similar cell-surface complement regulation capability of monkey and human factor H protein observed in the modified AH50 assay suggests that the cell-surface regulation of the complement pathway may also involve other proteins or factors.
The potential for ASO-induced complement activation has been studied in several species, including mice, rats, guinea pigs, rabbits, and dogs; however, complement activation has only been observed in rhesus and cynomolgus monkeys. Differences in species sensitivity to ASO-induced complement activation have been clearly evident in the various animal models tested. No complement activation was measured in dogs following ASO dose regimens associated with a plasma Cmax of 250 μg/ml, which is much greater than the activation threshold in monkeys (70–100 μg/ml) (Henry et al., 2014). This observation was further supported by the measurement of total complement activity (CH50) in rats and the absence of acute anaphylactic-like reactions in rodent species treated with high doses (≥100 mg/kg) of ASOs (Isis internal data). Relative sensitivity to ASO-induced activation of complement has also been investigated in other species of nonhuman primates, such as the marmoset, a New World monkey. No complement activation (C3a and/or CH50) was evident after either a single dose of 30 mg/kg s.c. or in vitro using marmoset serum (Isis internal data). Possible explanations of these phenomena include differences in protein-ASO binding affinities among the different species or possible differences in stringency of regulation of the complement AP.
As mentioned previously, chemical modifications have been shown to play a pivotal role in decreasing the magnitude of complement activation. In addition to the 2′-MOE modification for the second-generation ASOs, future chemical iterations, such as constrained ethyl PS ASOs and ASOs that are either shorter in length and/or have a mixture of PS and phosphodiester backbone, are expected to further decrease protein-binding interactions and the potential for complement AP activation in monkeys, and provide further assurance of the safety of ASOs for human use.
Finally, the functional difference in monkey and human factor H protein in fluid-phase complement regulation provides a plausible mechanistic basis for the activation, and provides additional evidence that monkeys are uniquely sensitive to the phenomenon of ASO-induced complement activation.
Participated in research design: Shen, Frazer-Abel, Giclas, Henry.
Conducted experiments: Shen, Reynolds, Chappell.
Contributed new reagents or analytic tools: Pangburn.
Performed data analysis: Shen, Frazer-Abel, Younis.
Wrote or contributed to the writing of the manuscript: Shen, Henry.
- Received August 18, 2014.
- Accepted October 8, 2014.
L.S. and S.P.H. contributed equally to this work.
- alternative pathway
- antisense oligonucleotide
- complement factor H
- enzyme-linked immunosorbent assay
- end of infusion
- red blood cell
- single nucleotide polymorphism
- upper limit of normal
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics