Published Ahead of Print on November 3, 2016, as doi:10.3324/haematol.2016.153312.
Copyright 2016 Ferrata Storti Foundation.
Small-molecule Factor D inhibitors selectively block the
alternative pathway of complement in paroxysmal nocturnal
hemoglobinuria and atypical hemolytic uremic syndrome
by Xuan Yuan, Eleni Gavriilaki, Jane A. Thanassi, Guangwei Yang, Andrea C. Baines,
Steven D. Podos, Yongqing Huang, Mingjun Huang, and Robert A. Brodsky
Haematologica 2016 [Epub ahead of print]
Citation: Yuan X, Gavriilaki E, Thanassi JA, Yang G, Baines AC, Podos SD, Huang Y, Huang M,
and Brodsky RA. Small-molecule Factor D inhibitors selectively block the alternative pathway of complement
in paroxysmal nocturnal hemoglobinuria and atypical hemolytic uremic syndrome.
Haematologica. 2016; 101:xxx
doi:10.3324/haematol.2016.153312
Publisher's Disclaimer.
E-publishing ahead of print is increasingly important for the rapid dissemination of science.
Haematologica is, therefore, E-publishing PDF files of an early version of manuscripts that
have completed a regular peer review and have been accepted for publication. E-publishing
of this PDF file has been approved by the authors. After having E-published Ahead of Print,
manuscripts will then undergo technical and English editing, typesetting, proof correction and
be presented for the authors' final approval; the final version of the manuscript will then
appear in print on a regular issue of the journal. All legal disclaimers that apply to the
journal also pertain to this production process.
Small-molecule Factor D inhibitors selectively block the alternative pathway of
complement in paroxysmal nocturnal hemoglobinuria and atypical hemolytic
uremic syndrome
Xuan Yuan1*, Eleni Gavriilaki1*, Jane A. Thanassi2, Guangwei Yang2, Andrea C.
Baines1, Steven D. Podos2, Yongqing Huang2, Mingjun Huang2 and Robert A.
Brodsky1
1
Division of Hematology, Department of Medicine, Johns Hopkins University School
of Medicine, Baltimore, MD
2
Achillion Pharmaceuticals, New Haven, CT
Running Head: Factor D inhibitors in PNH and aHUS
Word Count: 3873, Tables/Figures: 2/5, References: 43
*XY and EG contributed equally
Acknowledgments
This work was supported by a grant from the Aplastic Anemia and MDS International
Foundation and R01HL133113 (R.A.B). We thank the Chemistry, Complement
Biology, and ADME groups at Achillion Pharmaceuticals (New Haven, CT) who
provided factor D inhibitors and characterized their pharmacological activities and
pharmacokinetics. E.G. was a Johns Hopkins-Libra fellow during the performance of
these studies.
Corresponding author
Dr. Robert A. Brodsky
1
Ross Research Building, Room 1025,
720 Rutland Avenue, Baltimore, MD 21205-2196,
Email:
[email protected]
2
Abstract
Paroxysmal nocturnal hemoglobinuria and atypical hemolytic uremic syndrome are
diseases of excess activation of the alternative pathway of complement that are treated
with eculizumab, a humanized monoclonal antibody against the terminal complement
component C5. Eculizumab must be administered intravenously, and moreover some
patients with paroxysmal nocturnal hemoglobinuria on eculizumab have symptomatic
extravascular hemolysis, indicating an unmet need for additional therapeutic
approaches. We report the activity of two novel small-molecule inhibitors of the
alternative pathway component factor D using in vitro correlates of both paroxysmal
nocturnal hemoglobinuria and atypical hemolytic uremic syndrome. Both compounds
bind human factor D with high affinity and effectively inhibit its proteolytic activity
against purified factor B in complex with C3b. When tested using the traditional Ham
test with cells from paroxysmal nocturnal hemoglobinuria patients, the factor D
inhibitors significantly reduced complement-mediated hemolysis at concentrations as
low as 0.01
μΜ. Additionally the compound ACH-4471 significantly decreased C3
fragment deposition on paroxysmal nocturnal hemoglobinuria erythrocytes, indicating
a reduced potential relative to eculizumab for extravascular hemolysis. Using the
recently described modified Ham test with serum from patients with atypical
hemolytic uremic syndrome, the compounds reduced the alternative pathwaymediated killing of PIGA-null reagent cells, thus establishing their potential utility for
this disease of alternative pathway of complement dysregulation and validating the
modified Ham test as a system for preclinical drug development for atypical
hemolytic uremic syndrome. Finally, ACH-4471 blocked alternative pathway activity
when administered orally to cynomolgus monkeys. In conclusion, the small-molecule
factor D inhibitors show potential as oral therapeutics for human diseases driven by
3
the
alternative
pathway
of
complement
including
paroxysmal
nocturnal
hemoglobinuria and atypical hemolytic uremic syndrome.
Keywords: Factor D; complement inhibitors; paroxysmal nocturnal hemoglobinuria;
atypical hemolytic uremic syndrome; modified Ham test
Introduction
The complement system provides an important defense against bacteria, fungi,
and viruses1, 2. Mutations (e.g, PIGA, factor H, factor I, etc.) and autoantibodies (e.g,
factor H autoantibodies) that lead to unregulated complement activation are
implicated in the pathogenesis of many disorders including paroxysmal nocturnal
hemoglobinuria (PNH) and atypical hemolytic uremic syndrome (aHUS)3-5.
Complement activation is initiated by: 1) the lectin pathway, 2) the classical
pathway, and 3) the alternative pathway of complement (APC) which converge on the
common central component C3. The APC also serves as an amplification loop for the
lectin and classical pathways. It has been estimated that the APC may account for 80
percent of complement activation products even when the complement cascade is
initiated by the classical and lectin pathways6.
The APC is constitutively activated at low levels via slow spontaneous
hydrolysis of an internal thioester within C3 that generates C3(H2O). This activated
C3(H2O) in solution phase binds factor B to generate the proconvertase C3(H2O)B,
which is processed by the serine protease factor D to the APC C3 convertase,
C3(H2O)Bb. This C3 convertase then cleaves additional C3 molecules to generate
C3a and C3b, the latter of which can covalently attach to available surfaces7.
Deposited C3b can then elicit a rapid localized amplification, especially when
4
complement regulation is impaired, described as follows in a process designated the
amplification loop. Deposited C3b can pair with factor B, which is cleaved by factor
D to generate a second form of the APC C3 convertase, C3bBb. Membrane-bound
C3bBb then cleaves additional C3 to generate further C3b deposits, which pair with
additional factor B molecules to repeat the cycle. The end result of this amplification
is C3b opsonization, release of the anaphylatoxins C3a and C5a, and assembly of the
terminal complement complex (also known as the membrane attack complex, MAC)
on the target surface. The entire process is depicted in Figure 1A.
PNH is caused by a somatic PIGA mutation that prevents expression of
glycosylphosphatidylinositol (GPI) anchored proteins on the surfaces of affected
cells, among them the complement regulators CD55 and CD59 which normally
protect host cells from complement-mediated lysis. CD55 interferes with C3
convertase formation following C3b attachment, preventing spurious amplification of
initial activation events8, whereas CD59 prevents localized MAC assembly following
amplification9. Due to the absence of both regulators from erythrocyte membranes,
PNH manifests with chronic hemolytic anemia primarily due to APC activation on
mutant erythrocyte membranes. PNH is effectively treated with the humanized
monoclonal antibody, eculizumab10, 11. Eculizumab binds C5 and blocks its cleavage
to C5a and C5b and thus protects mutant erythrocytes from MAC formation and lysis.
However, due to the absence of CD55, the mutant erythrocytes in PNH patients on
eculizumab treatment continue to be opsonized by C3b and thus are susceptible to
extravascular hemolysis12. In most PNH patients on eculizumab this leads to
asymptomatic chronic extravascular hemolysis; however, a subset of PNH patients on
eculizumab have symptomatic hemolysis and require red blood cell transfusions10, 13.
Moreover, rare patients (~3.5 percent of the Japanese population) carry a genetic C5
5
variant (2654G>A, Arg885His) that prevents eculizumab binding and causes
eculizumab resistance14. Another limitation of eculizumab is that it must be
administered intravenously indefinitely every two weeks to block intravascular
hemolysis and prevent thrombosis. Thus, novel complement inhibitors are needed to
address these limitations.
aHUS is also caused by APC dysregulation. aHUS presents with signs and
symptoms of thrombotic microangiopathy (TMA) including thrombocytopenia, nonimmune hemolytic anemia, peripheral blood schistocytes, and often end-organ
damage to the kidneys and central nervous system. Most cases are caused by
mutations in APC genes or autoantibodies directed against APC regulatory proteins1517
. These mutations either disable the regulatory proteins that help degrade cell
surface C3b, including haploinsufficient mutations of factor H (most commonly),
factor I, MCP, and thrombomodulin, or enhance the function of proteins that drive the
APC such as gain-of-function mutations in C3 or factor B15,
17-19
. The result is
activation of terminal complement on various target tissues, including renal and
vascular endothelium, and cell death mediated by the MAC. Distinguishing aHUS
from other TMAs such as thrombotic thrombocytopenic purpura (TTP) is challenging.
Recently, we described a serum-based assay that distinguishes most cases of aHUS
from TTP20. This assay (modified Ham test) measures the ability of human serum to
kill PIGA-null reagent cells. Due to APC dysregulation in the serum of most aHUS
patients, the PIGA-null cells that fail to express CD55 and CD59 are more readily
killed by aHUS serum than by serum from healthy individuals or TTP patients.
Complement factor D is a promising therapeutic target to treat PNH and
aHUS. Factor B is its only natural substrate, and of all complement proteins, factor D
has the lowest abundance in serum and is the rate-limiting component of the APC21.
6
Its crystal structure has been elucidated in a series of studies 22-24. Here, we describe
two novel, small-molecule inhibitors that bind factor D and inhibit its cleavage of
factor B. We demonstrate in vitro that these inhibitors block PNH cell hemolysis,
mitigate C3 fragment accumulation on PNH cell surfaces, and block dysregulated
APC activity in aHUS serum. We also show that the modified Ham test can be used to
evaluate novel complement inhibitors in aHUS, and finally we demonstrate APC
suppression in serum obtained from non-human primates following oral
administration of one inhibitor.
Methods
Human samples: Three PNH patients and four aHUS patients were enrolled in this
study (Table 1 and Table 2). Patients were diagnosed with aHUS as previously
defined25, 26. All patients gave written informed consent. This study was approved by
the Institutional Review Board and conducted in accordance with the Declaration of
Helsinki. Blood was collected in EDTA and serum separation tubes (Supplement).
Factor D inhibitors: Compounds ACH-3856 and ACH-4471 were synthesized by
Achillion Pharmaceuticals and fully characterized by proton nuclear magnetic
resonance (1H-NMR), high performance liquid chromatography (HPLC), and mass
spectrometry.
Assays for factor D binding, factor D enzymatic activity, and hemolysis mediated by
APC and the classical pathway: Binding kinetics and affinities of compounds to
human factor D were determined by surface plasmon resonance. Inhibition of Factor
D enzymatic activity was evaluated with 80 nM purified human factor D and the
small synthetic thieoster substrate Z-Lys-SBzl, or with 0.8 nM purified human factor
D and its natural substrate C3bB. Inhibition of APC-mediated hemolytic activity was
7
assessed with 8% normal human serum and rabbit erythrocytes; inhibition of
hemolysis by the classical pathway was assessed with 0.5% normal human serum and
antibody-sensitized sheep erythrocytes. Methods for these assays are described in the
Supplement.
Inhibition of APC-mediated cell killing using PNH and aHUS patient samples (Ham
test and modified Ham test): Inhibition of hemolytic assay by factor D inhibitors was
assessed using PNH erythrocytes at 1% hematocrit in GVB0/MgEGTA (pH 6.4) and
20% acidified human serum (Ham test), as previously described28. Based on the same
principle, the modified Ham test was performed to assess the efficacy of factor D
inhibitors in APC-mediated killing caused by aHUS patient serum20. Both assays are
described in the Supplement.
Inhibition of C3 fragment deposition: C3 fragment deposition on PNH erythrocytes
from PNH patient 2 (blood group O) by 60% acidified C5-depleted human serum was
assessed by flow cytometry. Erythrocytes were washed and prepared as in the
hemolytic assay. Sample preparation and C3 fragment deposition measurement by
flow cytometry are described in detail in the Supplement.
APC activity in cynomolgus monkeys: The in-life portion of these studies was
performed at Ricerca Laboratories (Concord, OH) in accordance with guidelines in
Guide for the Care and Use of Laboratory Animals (National Academy Press,
Washington, DC, 2011) and the USDA Animal Welfare Act (9 CRF, Parts 1 through
3). Experimental procedures were reviewed and approved by Ricerca’s Institutional
Animal Care and Use Committee (IACUC) prior to initiation. ACH-4471 was
formulated in solution at 80 mg/mL in PEG400:water (70:30) (w:w). Three male
cynomolgus monkeys were dosed by oral gavage with ACH-4471 at 200 mg/kg twice,
8
12 hours apart. Serial blood samples were collected at specified time points through
30 hours. Plasma ACH-4471 concentration was determined by LC-MS/MS with a
lower limit of quantitation of 2.44 ng/mL. Pharmacokinetic parameters were
calculated with non-compartmental analysis using Phoenix 6.2 WinNonlin (Pharsight,
Princeton, NJ) from individual plasma concentration versus time using the lineartrapezoidal method. Serum APC activity was measured in duplicate using the APCspecific Wieslab assay (Euro Diagnostica, Malmö, Sweden). Activity at each time
point was normalized to pre-dosing activity in the same animal. Serum factor D
concentrations were determined using the Human Complement Factor D Quantikine
ELISA Kit (R&D Systems).
Results
Inhibition of factor D proteolytic activity
The potency of the small-molecule factor D inhibitors ACH-3856 and ACH4471 was investigated by biophysical and biochemical methods. Both compounds
showed high binding affinity to human factor D, shown by surface plasmon resonance
studies with recombinant enzyme (Table 3 and Supplemental Figure 1), with
respective KD values of 0.00036 µM and 0.00054 µM. Additionally, the compounds
inhibited the proteolytic activity of purified factor D against its natural substrate
factor B in complex with C3b, blocking production of Bb fragment in a dosedependent manner with respective IC50 values of 0.0058 ± 0.0005 µM and 0.015 ±
0.003 µM (Table 3 and Figure 1B, C). The compounds act as catalytic inhibitors,
established by inhibition of factor D protease activity against a small synthetic
α
substrate (N- -Cbz-L-lysine thiobenzyl ester; Table 3 and Supplemental Figure 2),
distinguishing them from the anti-factor D monoclonal antibody inhibitor
9
lampalizumab which binds an exosite and sterically blocks factor B access to the
active site27. Additionally the compounds showed selectivity as they inhibited in a
standard assay for rabbit erythrocyte hemolysis upon APC activation but not in a
counterpart assay for the complement classical pathway (Table 4), and moreover did
not inhibit a panel of 12 human serine proteases (Supplemental Table 1), thus
establishing their limited potential for off-target effects.
PNH and aHUS patient characteristics
PNH and aHUS are diseases of complement dysregulation that could be
amenable to factor D inhibitors. As proof of concept, we assessed the factor D
inhibitors in vitro using erythrocytes and serum obtained from PNH patients and
serum from aHUS patients. Patient characteristics are shown in Table 1 (PNH) and
Table 2 (aHUS). Erythrocytes were recovered from three PNH patients: two on longterm eculizumab treatment (patients 1 and 2) and one treatment-naïve (patient 3), and
were used for the hemolytic assay and the C3 fragment deposition assay. Serum was
also recovered from PNH patients 1 and 3 (this time after she was started on
eculizumab) following eculizumab administration for use in the C3 fragment
deposition assay. Five serum samples used in the modified Ham assay were recovered
from four aHUS patients: one in acute phase before treatment and later in remission
while on eculizumab treatment (patient 2, two samples), one in remission while on
eculizumab treatment (patient 3, one sample), and two in remission after
discontinuation of eculizumab (patients 1 and 4, one sample each).
Inhibitory effects on hemolysis of PNH erythrocytes
The Ham test identifies PNH erythrocytes by their hemolytic susceptibility to
acidified serum. We evaluated ACH-3856 and ACH-4471 by incubating erythrocytes
10
from PNH patients with acidified ABO blood group-compatible normal human serum
containing the inhibitors in half-log dilution series (Figure 2). As C3 convertase
formation is Mg++-dependent, EDTA inhibition was used to quantitate complementindependent background in each experiment. Above this background, the acidified
serum in the presence of Mg++ mediated the complement-dependent hemolysis of
80% (patient 1), 48% (patient 2), and 25% (patient 3) of patient erythrocytes (panel
A). The factor D inhibitors potently inhibited hemolysis with IC50 values ranging
from 0.0029 µM to 0.016 µM for ACH-3856 (IC90 values from 0.0082 µM to 0.064
µM) and from 0.0040 µM to 0.027 µM for ACH-4471 (IC90 values from 0.015 µM to
0.11 µM) (panels B, C). The factor D inhibitors therefore have the potential to inhibit
the APC-mediated hemolysis that underlies much of the morbidity and mortality in
PNH patients.
C3 fragment deposition on PNH erythrocytes
As eculizumab blocks terminal complement activity only, the continued
accumulation of C3 fragments on mutant erythrocytes in treated patients often results
in extravascular hemolysis. As factor D inhibitors should inhibit this opsonization, we
tested the ability of ACH-4471 to prevent C3 fragment deposition on PNH
erythrocytes as detected by flow cytometry. Erythrocytes from PNH patient 2 were
incubated with acidified human serum depleted of C5 to prevent hemolysis, yet allow
continued C3 fragment deposition (Figure 3) during the incubation period. While
serum acidification led to C3 fragment deposition on CD59-negative PNH
erythrocytes, EDTA inhibited this event, confirming that it was a complementspecific activity. ACH-4471 inhibited this deposition in a dose-dependent manner,
with IC50 and IC90 values of 0.031 µM and 0.089 µM, respectively. Inhibition was
also observed with erythrocytes from PNH patients 1 and 3 when assessed using their
11
own serum collected immediately after eculizumab infusion; the percentage of C3positive erythrocytes was 37.6% (patient 1) and 5.5% (patient 3) without ACH-4471,
decreasing to 11.9% (patient 1) and 1.8% (patient 3) with 0.1 µM ACH-4471 (data
not shown). Furthermore, in addition to anti-C3c antibody used above, we also
examined the phenomenon with anti-C3b and anti-C3d antibodies and we observed
similar results on erythrocytes from PNH patient 2 except that approximately 10% of
erythrocytes stained positive for C3d in the presence of EDTA, indicating that these
C3 fragments were deposited in vivo before harvesting from patients (Supplemental
Figure 3). These results suggest that the extravascular hemolysis that occurs in PNH
patients on eculizumab treatment might be blocked by factor D inhibition.
Factor D inhibitors block dysregulated APC in aHUS patient serum
An important distinction between aHUS and other TMAs is the underlying
APC dysregulation observed in aHUS. The modified Ham test can distinguish aHUS
from other TMAs based on the sensitivity of a PIGA-null reagent cell line to APC
activation; cell killing greater than 20% by patient serum has been shown to serve as a
sensitive and specific indicator of aHUS5, 20. We evaluated ACH-3856 and ACH-4471
in the modified Ham test by incubating PIGA-null cells with aHUS patient sera
containing the factor D inhibitors (Figure 4). In the absence of inhibitor (0.000 µM),
all five aHUS serum samples promoted a degree of cell killing similar to previously
described levels5, 20. Notably, the presence of therapeutic eculizumab in two of the
five sera (patient 2 sample b and patient 3) did not minimize the observed killing; this
finding is consistent with previous observations and has been attributed to the
eculizumab dilution that accompanies the serum dilution necessary for appropriate
assay conditions20. PIGA-null cell killing activity in aHUS patient sera was heatsensitive and therefore complement-dependent. ACH-3856 (left panel) and ACH12
4471 (right panel) both blocked the APC-mediated killing by aHUS patient sera at
sub-micromolar concentrations. These results suggest that small-molecule factor D
inhibitors have the potential to mitigate complement-mediated damage in aHUS
patients, and support the utility of the modified Ham assay for preclinical aHUS drug
assessment.
Efficacy of ACH-4471 following oral delivery in non-human primates
To profile the effects of continuous factor D inhibition over a period of at least
24 hours in vivo, we administered ACH-4471 orally to three monkeys in two serial
200 mg/kg doses separated by 12 hours. This deliberately high dose was chosen based
on an initial assessment of pharmacokinetic properties in monkeys which revealed
lower oral exposure due to higher clearance than in other animal species including
rats and dogs (not shown). The monkeys tolerated the compound well and showed no
clinical abnormalities. Figure 5A shows ACH-4471 concentrations in plasma at time
points from 0 hours (pre-dosing) to 30 hours (18 hours after the second dose). In
parallel, serum was collected for determination of APC activity by Wieslab assay and
of factor D concentrations. APC activity was suppressed continuously by >95%
continuously through the 30 hour time period (Figure 5B) with no significant increase
in factor D concentrations (Figure 5C). The observed stability of serum factor D
concentrations was expected because, as a small molecule, ACH-4471 should not
interfere with renal factor D clearance, in marked contrast to the effect of systemic
delivery of the humanized monoclonal anti-factor D antibody lampalizumab29. These
results demonstrate the potential utility of oral delivery of ACH-4471 for therapeutic
inhibition of APC activation.
Discussion
13
The present study describes the novel compounds ACH-3856 and ACH-4471
which potently inhibit factor D proteolytic activity. Factor D is a highly specific
serine protease with factor B as its only natural substrate30, making it a favorable
target for the largely APC-driven manifestations of both PNH and aHUS. We
demonstrate that these inhibitors effectively block APC-mediated cell killing using
erythrocytes from PNH patients (Ham test) or serum from aHUS patients (modified
Ham test). Furthermore, ACH-4471 inhibited the elevated deposition of C3 fragments
that proceeds on PNH cells when terminal complement activity is blocked such as by
therapeutic
eculizumab.
In
addition,
ACH-4471
demonstrated
good
oral
bioavailability and inhibited APC activity in serum samples recovered following oral
administration to non-human primates.
Eculizumab improves life for PNH patients by eliminating intravascular
hemolysis, decreasing the need for blood transfusions, and reducing the risk of
thrombosis10, 11. Yet, the factor D inhibitors present a potential advantage as some
eculizumab-treated PNH patients experience symptomatic extravascular hemolysis
due to dysregulated C3 fragment deposition that proceeds even with inhibition of
terminal complement activity31. Several upstream complement inhibitors under
investigation may address this limitation, including the engineered complement
receptor 2/factor H fusion protein TT3032, peptidic C3 inhibitors (such as Cp40 and
APL-2)33, and the C1 esterase inhibitor C1INH (Cinryze)28. Suitable systemic
exposure however may be difficult to achieve with some of these investigational
compounds, and unlike the factor D inhibitors these molecules will not be available in
oral formulation.
APC-mediated hemolysis of erythrocytes from all three PNH patients was
sensitive to factor D inhibitors, albeit over an approximately seven-fold range in IC50
14
values. The reason for this variation remains unknown yet the inhibitors showed
excellent potency against even the least susceptible PNH patient cells. Also of note,
C3 fragment opsonization proved less susceptible than hemolysis, with respective
IC50 values of 0.031 µM and 0.004 µM for ACH-4471 against cells from patient 2.
This distinction is likely derived in part from operational differences between the
assays, as PNH cell opsonization requires substantially more serum (60%) than
hemolysis (20%); moreover, the detection thresholds for APC activity could differ
widely if, for example, less activity is required for hemolysis than for C3 fragment
detection.
Terminal complement inhibition with eculizumab is also highly effective for
treating aHUS. Until recently, there was no functional assay with adequate sensitivity
and specificity to distinguish aHUS from other TMA syndromes, or to evaluate
potential therapeutic compounds. The experiments presented here use the recently
reported modified Ham test to demonstrate that factor D inhibitors effectively
overcome the characteristic APC activation on target membranes by serum of aHUS
patients. Moreover, although the modified Ham test does not replicate the complex
pathophysiology of aHUS observed in vivo, this assay may serve as a useful
preclinical system for testing novel complement inhibitors in aHUS.
Importantly for systemic delivery, factor D levels remained unchanged in the
monkeys following ACH-4471 administration. The factor D inhibitors thus differ
from the humanized monoclonal anti-factor D antibody lampalizumab, which is in
clinical study as an intravitreal treatment for age-related macular degeneration34.
Lampalizumab delivered intravenously to cynomolgus monkeys elicited nearly tenfold increases in serum factor D levels within five hours, effectively neutralizing its
inhibitory activity29. Similar increases have been reported for other therapeutic
15
antibodies against proteins that, like factor D, have high turnover rates mediated by
processes including renal filtration35. In contrast, the stability of serum factor D levels
in the present study indicates that the small molecule inhibitors will likely maintain
their potency following systemic delivery.
A significant concern with complement inhibition is the potential for increased
susceptibility to certain bacterial infections. Complete or partial factor D deficiency in
humans is associated with reduced bacterial phagocytosis in vitro38,39, and complete
factor D deficiency is associated with increased risk for recurrent infections with
Neisseria or other encapsulated bacteria which is comparable to the lifetime risk
observed in the setting of terminal complement deficiencies.36-41. Hence, vaccination
for subjects on a factor D inhibitor will be warranted as it is for subjects on
eculizumab. We anticipate that immunological protection following vaccination
against Neisseria will be better preserved by factor D inhibitors than by eculizumab.
Studies have demonstrated for example that opsonophagocytosis of Neisseria by
granulocytes depends on antibody-mediated activation of the complement classical
pathway, and that serum bactericidal activity depends additionally on the complement
terminal pathway42. As factor D inhibitors selectively target the APC and preserve the
classical, lectin, and terminal pathways, the protection elicited by antibodies through
vaccination should be at least partially preserved in the presence of factor D
inhibition. In contrast, eculizumab is expected to block the serum bactericidal activity
conferred by vaccination, leading to diminished protection. Furthermore, the higher
clearance of small molecules relative to therapeutic biologics will allow for rapid
restoration of complement activity if dosing must be ceased in the event of infection.
Nevertheless the safety of ACH-4471 will need to be monitored closely, especially
given the contribution of the APC-dependent amplification loop, which by one
16
estimate may account for up to 80 percent of complement classical pathway
activation6.
In conclusion, factor D is a promising target for oral therapy of diseases driven
by APC dysregulation. Based on results presented here, and guided by additional
assessments of its pharmacology, pharmacokinetic properties, and safety and
toxicology, ACH-4471 has been selected for clinical development in PNH and is
currently in phase 1 clinical study.
Authorship
Contribution: X.Y., E.G., and R.A.B. designed and conducted the study, analyzed
data, drafted and corrected the manuscript. J.A.T., G.Y., A.C.B., S.D.P. Y.H. and
M.H. contributed in designing and performing research, drafting and correcting the
manuscript.
Conflict-of-interest disclosure: R.A.B. is a member of the Scientific Advisory Board
of Achillion Pharmaceuticals, Alexion Pharmaceuticals, and Apellis Pharmaceuticals.
The remaining authors declare no competing financial interest.
17
References
1.
Varela JC, Tomlinson S. Complement: An Overview for the Clinician. Hematol Oncol
Clin North Am. 2015;29(3):409-427.
2.
Walport MJ. Complement. First of two parts. N Engl J Med. 2001;344(14):1058-1066.
3.
Brodsky RA. Paroxysmal nocturnal hemoglobinuria. Blood. 2014;124(18):2804-2811.
4.
Cataland SR, Wu HM. Diagnosis and management of complement mediated
thrombotic microangiopathies. Blood Rev. 2014;28(2):67-74.
5.
Brodsky RA. Complement in hemolytic anemia. Blood. 2015;126(22):2459-2465.
6.
Harboe M, Mollnes TE. The alternative complement pathway revisited. J Cell Mol
Med. 2008;12(4):1074-1084.
7.
Pangburn MK, Muller-Eberhard HJ. Initiation of the alternative complement pathway
due to spontaneous hydrolysis of the thioester of C3. Ann N York Acad Sci. 1983;421:291298.
8.
Lublin DM, Atkinson JP. Decay-accelerating factor: biochemistry, molecular biology,
and function. Ann Rev Immunol. 1989;7:35-58.
9.
Rollins SA, Sims PJ. The complement-inhibitory activity of CD59 resides in its capacity
to block incorporation of C9 into membrane C5b-9. J Immunol. 1990;144(9):3478-3483.
10.
Hillmen P, Young NS, Schubert J, et al. The complement inhibitor eculizumab in
paroxysmal nocturnal hemoglobinuria. N Engl J Med. 2006;355(12):1233-1243.
11.
Brodsky RA, Young NS, Antonioli E, et al. Multicenter phase 3 study of the
complement inhibitor eculizumab for the treatment of patients with paroxysmal nocturnal
hemoglobinuria. Blood. 2008;111(4):1840-1847.
12.
Risitano AM, Notaro R, Marando L, et al. Complement fraction 3 binding on
erythrocytes as additional mechanism of disease in paroxysmal nocturnal hemoglobinuria
patients treated by eculizumab. Blood. 2009;113(17):4094-4100.
13.
DeZern AE, Dorr D, Brodsky RA. Predictors of hemoglobin response to eculizumab
therapy in paroxysmal nocturnal hemoglobinuria. Eur J Haematol. 2013;90(1):16-24.
14.
Nishimura J, Yamamoto M, Hayashi S, et al. Genetic variants in C5 and poor
response to eculizumab. N Engl J Med. 2014;370(7):632-639.
15.
Maga TK, Nishimura CJ, Weaver AE, Frees KL, Smith RJ. Mutations in alternative
pathway complement proteins in American patients with atypical hemolytic uremic
syndrome. Hum Mutat. 2010;31(6):E1445-1460.
16.
Frimat M, Tabarin F, Dimitrov JD, et al. Complement activation by heme as a
secondary hit for atypical hemolytic uremic syndrome. Blood. 2013;122(2):282-292.
18
17.
Noris M, Caprioli J, Bresin E, et al. Relative role of genetic complement abnormalities
in sporadic and familial aHUS and their impact on clinical phenotype. Clin J Am Soc Nephrol.
2010;5(10):1844-1859.
18.
Goicoechea de Jorge E, Harris CL, Esparza-Gordillo J, et al. Gain-of-function
mutations in complement factor B are associated with atypical hemolytic uremic syndrome.
Pro Natil Acado Sci U S A. 2007;104(1):240-245.
19.
Delvaeye M, Noris M, De Vriese A, et al. Thrombomodulin mutations in atypical
hemolytic-uremic syndrome. The New England journal of medicine. 2009;361(4):345-357.
20.
Gavriilaki E, Yuan X, Ye Z, et al. Modified Ham test for atypical hemolytic uremic
syndrome. Blood. 2015;125(23):3637-3646.
21.
Volanakis JE, Barnum SR, Giddens M, Galla JH. Renal filtration and catabolism of
complement protein D. New Engl J Med. 1985;312(7):395-399.
22.
Narayana SV, Carson M, el-Kabbani O, et al. Structure of human factor D. A
complement system protein at 2.0 A resolution. J Mol Biol. 1994;235(2):695-708.
23.
Kim S, Narayana SV, Volanakis JE. Catalytic role of a surface loop of the complement
serine protease factor D. J Immunol. 1995;154(11):6073-6079.
24.
Cole LB, Chu N, Kilpatrick JM, Volanakis JE, Narayana SV, Babu YS. Structure of
diisopropyl fluorophosphate-inhibited factor D. Acta Crystallogr D Biol Crystallogr.
1997;53(Pt 2):143-150.
25.
Coppo P, Schwarzinger M, Buffet M, et al. Predictive features of severe acquired
ADAMTS13 deficiency in idiopathic thrombotic microangiopathies: the French TMA
reference center experience. PloS One. 2010;5(4):e10208.
26.
Cataland SR, Yang S, Wu HM. The use of ADAMTS13 activity, platelet count, and
serum creatinine to differentiate acquired thrombotic thrombocytopenic purpura from
other thrombotic microangiopathies. Br J Haematol. 2012;157(4):501-503.
27.
Katschke KJ, Jr., Wu P, Ganesan R, et al. Inhibiting alternative pathway complement
activation by targeting the factor D exosite. J Biol Chem. 2012;287(16):12886-12892.
28.
DeZern AE, Uknis M, Yuan X, et al. Complement blockade with a C1 esterase
inhibitor in paroxysmal nocturnal hemoglobinuria. Exp Hematol. 2014;42(10):857-861.
29.
Loyet KM, Good J, Davancaze T, et al. Complement inhibition in cynomolgus
monkeys by anti-factor d antigen-binding fragment for the treatment of an advanced form
of dry age-related macular degeneration. J Pharmacol Exp Ther. 2014;351(3):527-537.
30.
Kim S, Narayana SV, Volanakis JE. Mutational analysis of the substrate binding site of
human complement factor D. Biochemistry. 1994;33(48):14393-14399.
31.
Hill A, Rother RP, Arnold L, et al. Eculizumab prevents intravascular hemolysis in
patients with paroxysmal nocturnal hemoglobinuria and unmasks low-level extravascular
hemolysis occurring through C3 opsonization. Haematologica. 2010;95(4):567-573.
19
32.
Risitano AM, Notaro R, Pascariello C, et al. The complement receptor 2/factor H
fusion protein TT30 protects paroxysmal nocturnal hemoglobinuria erythrocytes from
complement-mediated hemolysis and C3 fragment. Blood. 2012;119(26):6307-6316.
33.
Risitano AM, Ricklin D, Huang Y, et al. Peptide inhibitors of C3 activation as a novel
strategy of complement inhibition for the treatment of paroxysmal nocturnal
hemoglobinuria. Blood. 2014;123(13):2094-2101.
34.
Do DV, Pieramici DJ, van Lookeren Campagne M, et al. A phase ia dose-escalation
study of the anti-factor D monoclonal antibody fragment FCFD4514S in patients with
geographic atrophy. Retina. 2014;34(2):313-320.
35.
Davda JP, Hansen RJ. Properties of a general PK/PD model of antibody-ligand
interactions for therapeutic antibodies that bind to soluble endogenous targets. MAbs.
2010;2(5):576-588.
36.
Kluin-Nelemans HC, van Velzen-Blad H, van Helden HP, Daha MR. Functional
deficiency of complement factor D in a monozygous twin. Clin Exp Immunol. 1984;58(3):724730.
37.
Hiemstra PS, Langeler E, Compier B, et al. Complete and partial deficiencies of
complement factor D in a Dutch family. J Clin Invest. 1989;84(6):1957-1961.
38.
Weiss SJ, Ahmed AE, Bonagura VR. Complement factor D deficiency in an infant first
seen with pneumococcal neonatal sepsis. J Allergy Clin Immunol. 1998;102(6 Pt 1):10431044.
39.
Biesma DH, Hannema AJ, van Velzen-Blad H, et al. A family with complement factor
D deficiency. J Clin Invest. 2001;108(2):233-240.
40.
Sprong T, Roos D, Weemaes C, et al. Deficient alternative complement pathway
activation due to factor D deficiency by 2 novel mutations in the complement factor D gene
in a family with meningococcal infections. Blood. 2006;107(12):4865-4870.
41.
Densen P. Complement deficiencies and meningococcal disease. Clin Exp Immunol.
1991;86 Suppl 1:57-62.
42.
Hellerud BC, Aase A, Herstad TK, et al. Critical roles of complement and antibodies in
host defense mechanisms against Neisseria meningitidis as revealed by human complement
genetic deficiencies. Infect Immun. 2010;78(2):802-809.
43.
Lin Z, Schmidt CQ, Koutsogiannaki S, et al. Complement C3dg-mediated
erythrophagocytosis: implications for paroxysmal nocturnal hemoglobinuria. Blood.
2015;126(7):891-894.
20
Table 1. Clinical data on PNH patients studied.
Patient ID
Age,
Gender
Erythrocyte
clone size
Type II/ III (%)
Granulocyte
clone size
(%)
LDH
(U/L)
Hemoglobin
(g/dL)
Direct
Coombs
C3
Direct
Coombs
IgG
Absolute
reticulocyte
count (K/cu mm)
Months on
eculizumab at
q14daysdosing
at start of study
1
51, F
3/64
96
266
9.5
Pos
Neg
251.1
57
2
58, M
1/98
99
281
10.3
Pos
Neg
321.6
107
3*
72, F
1/13
81
842
9.7
Neg
Neg
138.3
0
PNH: Paroxysmal nocturnal hemoglobinuria; LDH: lactate dehydrogenase; NA: not applicable; * transfused within 60 days
21
Table 2. Clinical data on aHUS patients studied.
Patient
Age,
ID
Gender
1
47, F
2a
59, M
2b
3
4
35, F
23, F
Disease
phase
(weeks)
ADAMTS13 ADAMTS13
activity (%) inhibitor (%)
PLT
LDH
Cr
count
(mg/dl) (mg/dl)
3
(x10 /μl)
Genetic
mutation
Previous
treatments
Current
treatment
Response to Serum
eculizumab tested
CFHR3-CFHR1
homozygous
deletion;
CFH
heterozygous
c.2850G>T,
p.Gln950His
-
-
-
No
PEx (x5)
Eculizumab
(x8)
-
Yes
Yes
PEx (x4)
-
-
Yes
Eculizumab
(x12)
Yes
Yes
-
-
-
No
Acute
75
NA
9
2505
2.3
Remission
(20 w)
NA
NA
314
228
1.9
Acute
15
67
39
676
2.3
Remission
(12 w)
NA
NA
330
170
1.4
Acute
18
76
10
1270
2.4
Remission
(82 w)
NA
108
75
1.1
-
Eculizumab
(x77)
Yes
Yes
Acute
102
NA
62
2820
6.3
-
-
-
No
Remission
(60 w)
NA
NA
205
178
3.5
Eculizumab
(x37)
Dialysis
Yes
Yes
None
None
NA
None
aHUS: atypical hemolytic uremic syndrome; F: female; M: male; CFHR: complement factor H related proteins; CFH: complement factor H; ADAMTS13: a
disintegrin and metalloprotease with thrombospondin type 1 motif, 13; NA: not available; PEx: Plasma exchange.
22
Table 3. Compound Binding to and Inhibition of Factor D Enzyme
fD Binding
fD Activity
On
Off
Affinity
C3bB
Thioester
Compound
ka (M-1 s-1)
kd (s-1)
KD (µM)
N
IC50 (µM)
N
IC50 (µM)
N
0.0058 ± 0.005
4
0.036 ± 0.003
83
ACH-3856
1.7e7 ± 7e6
0.0086 ± 0.0061
0.00048 ± 0.00016
2
0.015 ± 0.003
4
0.035 ± 0.001
3
ACH-4471
2.6e6 ± 4e5
0.0015 ± 0.0001
0.00057 ± 0.00004
2
fD: factor D. Mean ± standard deviation from N independent experiments. fD binding parameters were assessed by surface plasmon resonance; representative
sensorgrams are shown in Supplemental Figure 2. fD activity was measured both against its natural substrate C3bB and against a small synthetic thioester substrate.
23
Table 4. APC-Specific Inhibition of Complement Hemolytic Activity
APC Hemolysis
Compound
IC50 (µM)
IC90 (µM)
ACH-3856
0.0087 ± 0.0039
0.022 ± 0.0089
ACH-4471
0.017 ± 0.011
0.070 ± 0.023
CPC: classical pathway of complement. Mean ± standard deviation from N independent experiments.
N
40
6
CPC Hemolysis
IC50 (µM)
>200
>200
N
1
2
24
Figure 1. Inhibition of factor D proteolytic activity. (A) Depiction of factor D
inhibitors in the alternative complement pathway (APC). Factor D (fD) participates in
C3 convertase generation by cleavage of factor B (fB) at two steps in the APC
cascade: generation of the initial C3 convertase (C3(H2O)Bb) following spontaneous
APC activation (tickover); and, the production of surface-bound C3 convertase
(C3bBb) which mediates dramatic amplification of the initial activation (amplification
loop) and the consequent opsonization of target surfaces by C3b, formation of the
terminal complement complex (aka, membrane attack complex, MAC), and release of
the anaphylatoxins C3a and C5a. An erythrocyte is shown to depict the membranebound events. Additional regulatory proteins not shown here can promote (properdin)
or attenuate (factor H, factor I, multiple membrane-bound proteins) APC activity. (B)
Factor D proteolytic activity against factor B in complex with C3b. Two stained gels
from a single representative experiment show dose-dependent inhibition by
compounds of factor B (fB) cleavage to its products Ba and Bb. Control reactions
included one omitting C3b and fB (labeled “fD”) and one omitting fD (labeled “No
fD”). (C) Dose-response curves and IC50 values from the representative experiment of
panel B. Average IC50 values ± standard deviation for ACH-3856 and ACH-4471
were 0.0058 ± 0.0005 µM and 0.015 ± 0.003 µM in four independent experiments.
25
Figure 2. Factor D inhibitors block hemolysis of erythrocytes from PNH patients.
(A) Hemolysis of erythrocytes obtained from the three PNH patients in 20% acidified
normal human serum (Ham test). The control samples “Buffer”, “Serum + EDTA”,
and “H2O”, and the patient-specific samples without inhibition, “Serum”, were as
defined in Methods. (B) Inhibition by ACH-4471 (○) and ACH-3856 (□) of hemolysis
reactions using erythrocytes from the three PNH patients. Each inhibitor
concentration was tested in duplicate with mean ± SD shown. (C) IC50 and IC90 values
of ACH-4471 and ACH-3856 from the hemolysis reactions in panel B.
26
Figure 3. Factor D inhibitor ACH-4471 blocks C3 fragment deposition on
erythrocytes from PNH patient. (A) Representative flow cytometric scattergrams of
erythrocytes from PNH patient 2 treated with acidified C5-depleted serum (ACH4471-treated or EDTA-treated). C3 fragment deposition was assessed using antihuman C3c antibody (X-axis). PNH mutant erythrocytes were identified by the
absence of cell-surface CD59 expression (Y-axis). Respective gating thresholds are
indicated by vertical and horizontal lines. Note, a fraction of CD59-negative cells
(~10%) were positive when stained with anti-C3d antibody despite staining negative
for anti-C3b and anti-C3c antibodies when tested in the presence of EDTA to prevent
in vitro complement activation; this result indicates that some CD59-negative cells
had been coated with C3 fragments but were transformed into the terminal opsonin
C3dg in vivo before harvesting, a process known to be rather rapid (below and
Supplemental Figure 3)43. (B) ACH-4471 dose-response curve from the experiment in
panel A; IC50 and IC90 values were 0.031 µM and 0.089 µM respectively.
27
Figure 4. Factor D inhibitors block cell killing by sera from aHUS patients.
Significant killing of PIGA-null reagent cells by the five aHUS patient sera in the
modified Ham test is shown (0.000 µM ACH-3856, left; 0.000 µM ACH-4471, right).
The addition of factor D inhibitors caused a dose-dependent reduction of cell killing
at the indicated concentrations (ACH-3856, left; ACH-4471, right). Activity was also
abrogated by heat inactivation of serum complement (data not shown). Horizontal
dashed line indicates the 20% threshold that is considered indicative of cell killing
due to APC dysregulation in aHUS serum. The five patient samples are shown for
each condition with mean +/- standard deviation.
28
Figure 5. Factor D inhibitor inhibits APC activity following oral dosing in
cynomolgus monkeys. (A) ACH-4471 concentrations at the indicated time points in
plasma samples collected from three cynomolgus monkeys following oral dosing with
ACH-4471 at 200 mg/kg at 0 and 12 hours (arrows). Key pharmacokinetic parameters
were calculated as follows: maximum concentration (Cmax), 4,020 ± 1,480 ng/mL; time
at Cmax (Tmax), 15.3 ± 1.2 hours (3.3 ± 1.2 hours following the second dose); and 30hour exposure level (AUC0-30), 48,300 ± 19,100 ng/mL•h. Parameters were calculated
as mean ± SD from the three animals. (B) APC activity assessed by Wieslab assay in
serum samples collected at the indicated time points. Activity in each sample was
normalized to the activity in the same animal at 0 h (pre-dosing). Mean ± SD are
shown from duplicate assay values. (C) Serum fD concentrations at the indicated time
points. Circle, square, and triangle shapes; animals 1, 2, and 3 respectively.
29
Supplemental Materials
Supplemental Methods
Human samples
Blood samples were centrifuged, and erythrocytes were washed three times in
phosphate buffered saline (PBS) and resuspended at 30-40% hematocrit in GVB0 (gel
veronal buffer without Ca++ and Mg++, pH 7.3) supplemented with 10 mM MgEGTA
(GVB0/MgEGTA, pH 7.3). Blood from the three PNH patients was collected for serum
and erythrocytes before scheduled eculizumab infusion (patients 1 and 2 on long-term
treatment; patient 3, treatment naïve). For additional experiments, serum was also
collected immediately after eculizumab infusion. Serum from the aHUS patients on
eculizumab was collected immediately prior to dosing of eculizumab.
Binding Affinity to factor D
Binding kinetics and affinities of compounds to human factor D were
characterized by surface plasmon resonance using a BiacoreTM 3000 system (GE
Healthcare). Recombinant factor D protein (R&D Systems, Minneapolis, MN) was
immobilized on a CM5 sensor chip (GE Healthcare) using standard amine coupling
chemistry with HBS-N (10 mM HEPES, 0.15 M NaCl, pH 7.4) as the running buffer.
Briefly, the surface was activated with a 1:1 ratio of 0.4 M 1-ethyl-3-(3dimethylaminopropyl) carbodiimide hydrochloride/0.1 M N-hydroxy succinimide at a
flow rate of 10 μL/min. Protein was diluted to 0.02 mg/mL in 10 mM sodium acetate
pH 6.0 was captured by injecting onto the activated chip surface. Residual activated
groups were blocked with a 7 minute injection of 1:1 mixture of 1 M ethanolamine (pH
8.5) with HBS-N. Final surface density of protein was 4000 RU (resonance units).
Compound binding data were collected in 50 mM TrisHCl pH 7.5, 1 M NaCl, 0.005%
Page 1 of 12
Tween 20 supplemented with DMSO (1%) at 50 μL/min. Data were double referenced
and corrected for DMSO-excluded volume effects and fitted to a 1:1 binding model,
allowing for mass transport effects, using Scrubber software (BioLogic Software).
Factor D protease activity against a small synthetic substrate
Factor D protease activity was measured using a biochemical assay consisting
of purified factor D (Complement Technology, Inc., Tyler, TX), the synthetic
thioester substrate N-α-Cbz-L-lysine thiobenzyl ester (Z-Lys-SBzl; Sigma-Aldrich)
and the colorimetric reagent 5,5'-dithiobis-(2-nitrobenzoic acid) (DTNB, Ellman’s
reagent). Factor D-mediated esterolysis of substrate generates free sulfhydryl groups
that react with the colorimetric detection agent DTNB to produce yellow color.
Compounds were dissolved in DMSO and diluted in DMSO in half-log dilution series
to 50× final assay concentrations. Factor D was diluted in assay buffer (50 mM Tris,
1.0 M NaCl, pH 7.5) and added to compound in a polypropylene microtiter plate for a
5 minute room temperature preincubation at 2× final concentration. 50 μL factor D +
compound mixture was then transferred with gentle mixing to 50 μL assay buffer with
Z-Lys-SBzl and DTNB in a microtiter plate. Final concentrations were 2 μg/mL (80
nM) factor D, 100 μM Z-Lys-SBzl, and 100 μM DTNB. A405 was measured in a
Molecular Devices Spectramax Plus 384 plate reader at 30-second intervals for 30
minutes. Reaction rates (mA450 min-1) calculated by the Spectramax software were
used to calculate IC50 values using nonlinear regression analysis with the fourparameter sigmoidal dose-response equation (Prism, GraphPad Software, La Jolla,
CA).
Factor D proteolytic activity against its natural substrate C3bB
280 nM C3b, 400 nM factor B, and 0.8 nM factor D (Complement Tech) were
incubated in assay buffer (PBS with 10 mM MgCl) in the absence or presence of
Page 2 of 12
compound (in half-log dilution series from 0.000316 µM to 1 µM) at 37°C for 30
minutes. Factor B cleavage was visualized by non-reducing 4-20% Tris-glycine SDSPAGE followed by Coomassie blue staining. Bb product was quantified in parallel by
MicroVue Bb Plus ELISA (Quidel). Inhibitor IC50 values were calculated by
nonlinear regression analysis using the four-parameter sigmoidal dose-response
equation (Prism, GraphPad Software).
Inhibition of APC-mediated and CPC-mediated hemolysis
Alternative pathway of complement (APC) hemolytic assays with rabbit
erythrocytes were performed using standard procedures as described previously with
minor modifications. Specifically, a final concentration of 8% normal human serum
(NHS) was used as this amount consistently led to nearly complete lysis of rabbit
erythrocytes. Compounds were assayed in duplicate. Reactions consisted of 50 µL
GVB0 (gel veronal buffer without Ca++ and Mg++, pH 7.3, Complement Technology)
supplemented with 10 mM MgEGTA (GVB0/MgEGTA, pH 7.3), 50µL 20% NHS
(Complement Technology) in GVB0/MgEGTA and 20µL erythrocytes (5×108/mL,
Complement Technology) in GVB0/MgEGTA in the absence or presence of
compound (concentration range from 0.000316 to 1µM) or DMSO. A 0% hemolysis
control consisted of 100µL GVB0/MgEGTA plus 20µL erythrocytes and the 100%
hemolysis control consisted of 100µL water plus 20µL erythrocytes. After 30
minutes incubation at 37°C, erythrocytes were pelleted and optical absorbance of
supernatants was read at 405 nM. EC50 and EC90 values were determined by nonlinear regression analysis using the four-parameter sigmoidal dose-response equation
(Graphpad Prism, La Jolla, CA).
CPC hemolytic assays with antibody sensitized sheep erythrocytes were
performed using standard procedures as described previously with minor
Page 3 of 12
modifications 1. Specifically, a final concentration of 0.5% normal human serum
(NHS) was used as this amount consistently led to nearly complete lysis of
erythrocytes. Compounds were assayed in duplicate. Reactions consisted of 100 µL
GVB++ (gel veronal buffer with 0.15 mM Ca++ and 0.5 mM Mg++, pH 7.3),
Complement Technology), 0.6 µL NHS (Complement Technology) and 30 µL
erythrocytes (5×108/mL, Complement Technology) in GVB++ in the absence or
presence of compound (concentration range from 0.09 to 200 µM) or DMSO. A 0%
hemolysis control consisted of 100µL GVB++ plus 30µL erythrocytes and the 100%
hemolysis control consisted of 100µL water plus 30µL erythrocytes. After 60
minutes incubation at 37°C, erythrocytes were pelleted and optical absorbance of
supernatants was read at 541 nM. EC50 and EC90 values t were determined by nonlinear regression analysis using the four-parameter sigmoidal dose-response equation
(Graphpad Prism, La Jolla, CA).
Inhibition of APC-mediated lysis in PNH (Ham test)
The hemolytic assay was performed with PNH erythrocytes at 1% hematocrit
in GVB0/MgEGTA (pH 6.4) and 20% acidified human serum, as previously described
2
. Briefly, ABO blood group-compatible normal human serum (blood group B serum
for patient 1, blood group B; pooled serum for patients 2 and 3, blood group O; sera
from Complement Tech) was preincubated for 10 minutes on ice with inhibitor solution
in half-log dilution series in dimethyl sulfoxide (DMSO) (final concentrations: 0.001
µM to 1 µM inhibitor; 1% DMSO) prior to acidification. Samples containing serum and
inhibitor were acidified by addition of 0.2 N HCl to GVB0/MgEGTA (pH 6.4). 10 µL
PNH erythrocytes were added per well and incubated at 37°C for 1 hour. Controls:
“Buffer”, GVB0/MgEGTA (pH 6.4) alone; “Serum+EDTA”; acidified serum
inactivated by 5 mM EDTA to determine 0% APC-mediated hemolysis; “H2O”, H2O
Page 4 of 12
in place of buffer for complete osmotic hemolysis; and “Serum”, DMSO in place of
inhibitor to determine 100% APC-mediated hemolysis. Supernatants were collected by
centrifugation and optical absorbance was measured at 405 nM (A405) using an iMark
Microplate reader (Bio-Rad). Hemolysis in each well was normalized to the 0% and
100% controls, and IC50 and IC90 values were determined from duplicate wells by
nonlinear regression analysis using the four-parameter sigmoidal dose-response
equation (Prism).
Inhibition of dysregulated APC in aHUS
The modified Ham test was performed to assess the efficacy of factor D
inhibitors in aHUS patient serum3. Briefly, PIGA-null reagent cells were plated in 80
µL GVB++ (gel veronal buffer with 0.15 mM Ca++ and 0.5 mM Mg++, pH 7.3), then
20% serum preincubated with DMSO or inhibitors was added, and reactions were
incubated for 30 minutes at 37°C. Heat-inactivated serum was used as a negative
control. Next, cells were washed and incubated with WST-1 (Roche) at 1:10 dilution
for 2 hours at 37°C. Optical absorbance was measured in an iMark Microplate
Absorbance Reader (Bio-Rad) at 490 nm with a reference wavelength at 595nm. Mean
absorbance was calculated for each sample based on triplicates and background
absorbance (from a blank containing WST-1) as previously described.
Inhibition of C3 fragment deposition
C5-depleted human serum (Complement Technology) was preincubated for 10
minutes at 37°C with DMSO or compound solution (half-log dilution series in DMSO;
final concentrations 0.0001 µM to 1 µM compound, 1% DMSO). Serum was acidified
by addition of 0.2 N HCl to GVB0/MgEGTA (pH 6.4). 10 μL erythrocytes was added
per well and reactions were incubated at 37°C for 24 hours. Negative control reactions
Page 5 of 12
included: serum inactivated by 20 mM EDTA; and, opsonization of normal blood group
O erythrocytes. After incubation, erythrocytes were collected by centrifugation and
washed with 1% BSA/PBS. Erythrocytes were stained with 1:200 FITC-conjugated
anti-C3c antibody (Abcam) and 1:100 allophycocyanin-conjugated anti-CD59
(Invitrogen) and analyzed by flow cytometry. The phenomenon was also examined with
1:50 FITC-conjugated anti-C3b (anti-Human C3/C3b/iC3b, Cedarlane, catalog #
CL7632F) and 1:200 AlexaFluor488-conjugated anti-C3d antibodies (Novus
Biologicals, catalog #NB100-65510AF488) with the same methodology. Ghost and
intact erythrocytes were gated using forward and side scatter; analysis was limited to
intact cells. Gating for FITC-conjugated (or AlexaFluor488-conjugated) and
allophycocyanin-conjugated antibodies was based on isotypic control antibodies. Data
were analyzed using FlowJo software (Tree Star, Inc.). IC50 and IC90 values were
determined by nonlinear regression analysis using the four-parameter sigmoidal doseresponse equation (Prism). C3 fragment deposition on PNH erythrocytes from patients
1 and 3 (obtained before eculizumab infusion) was similarly assessed using the patient’s
own serum collected immediately after eculizumab infusion.
Off-target serine proteases
Inhibition of purified serine proteases by ACH-3856 and ACH-4471 was
determined by Cerep Laboratories (Eurofins Pharma Discovery Services) using
purified enzymes paired with specific substrates under standardized reaction
conditions. Reaction progress in duplicate reactions was measured following
incubation by fluorometry or photometry. Percent inhibition was calculated from
measured specific activity in the presence and absence of inhibitor. Reactions were
validated with reference inhibitors in parallel reactions
Supplemental Results
Page 6 of 12
Binding kinetics and affinities of ACH-3856 and ACH-4471 to recombinant
factor D were assessed by surface plasmon resonance with immobilized recombinant
protein: parameters are presented in the main manuscript (Table 3) and in
Supplemental Figure 1. Inhibitory activities of these compounds against purified
factor D enzyme with a small synthetic thioester substrate are presented in the main
manuscript (Table 3) and Supplemental Figure 2. Inhibitory activities in APC and
complement classical pathway hemolysis assays are presented in the main manuscript
(Table 4).
Inhibition of APC-mediated lysis of PNH erythrocytes (Ham test) is presented
in the main manuscript (Figure 2). Inhibition of C3 fragment deposition on PNH
erythrocytes as detected by anti-C3c staining is presented in the main manuscript
(Figure 3); the staining of approximately 10% of cells by anti-C3d but not by anti-C3c
when treated with serum in the presence of EDTA is presented in Supplemental
Figure 3. Inhibitory activities against a panel of serine proteases are presented in
Supplemental Table 1.
Page 7 of 12
Supplemental Table 1. Inhibition of off-target serine proteases
% inhibition at 10 µM
Enzyme
Source
ACH-3856 ACH-4471
Thrombin
human plasma
3.3
1.2
Urokinase
human kidney
-7.6
ND
Chymase
human skin
27.6
-1.2
Elastase
human leukocytes
-3.5
0.7
cathepsin G
human leukocytes
7.6
0.4
dipeptidyl peptidase IV(DPP-IV) human recombinant
-1.9
-1
HNE human neutrophil elastase)
human serum
-4.9
ND
Tryptase
human lung
-4.5
1.4
Trypsin
human pancreas
-6
0.1
Kallikrein
Human plasma
2.8
20.8
PLAU (Urokinase)
Human Urine
ND
2.5
ELA2 (Neutrophil Elastase 2)
Human neutrophils
ND
14.5
Page 8 of 12
Supplemental Figure 1. Factor D inhibitor binding to factor D protein.
Representative surface plasmon resonance (SPR) sensorgrams with fitted curves are
shown for compound association with (0 through 60 seconds) and dissociation from
(60 seconds to end) immobilized factor D. ACH-3856 (top) or ACH-4471 (bottom)
was added in two-fold dilution series up to 100 nM. SPR response is expressed in
arbitrary units. Table 3 in the main manuscript presents kinetic parameters.
Page 9 of 12
Supplemental Figure 2. fD esterolytic activity. Dose-response curves are shown
from a representative experiment demonstrating enzymatic inhibition of purified
factor D by ACH-3856 (○) and ACH-4471(□). Table 3 in the main manuscript
presents IC50 values.
2.5
∆(mA405) min-1
2.0
1.5
1.0
0.5
0.0
0.001
0.01
0.1
1
µM
Page 10 of 12
Supplemental Figure 3. Staining of a subset of erythrocytes recovered from PNH
patient 2 with anti-C3d antibody but not anti-C3c antibody.
Flow cytometric scattergrams of erythrocytes from PNH patient 2 treated with
acidified C5-depleted serum (EDTA treated). C3 fragment deposition was assessed
using anti-human C3c or anti human C3b antibody (left panel, X-axis) or anti human
C3d antibody (right panel, X-axis). PNH mutant erythrocytes were identified by the
absence of cell surface CD59 expression (Y-axis). Respective gating thresholds are
indicated by vertical and horizontal lines.
Page 11 of 12
Supplemental References
1.
Morgan BP. Measurement of complement hemolytic activity, generation of
complement-depleted sera, and production of hemolytic intermediates. Methods Mol Biol.
2000;150:61-71.
2.
DeZern AE, Uknis M, Yuan X, et al. Complement blockade with a C1 esterase
inhibitor in paroxysmal nocturnal hemoglobinuria. Exp Hematol. 2014;42(10):857-61.
3.
Gavriilaki E, Yuan X, Ye Z, et al. Modified Ham test for atypical hemolytic uremic
syndrome. Blood. 2015;125(23):3637-46.
Page 12 of 12