A Remyelination-Promoting Human Antibody ... - ACS Publications

Subscriber access provided by University of Sunderland

Article

Surface plasmon resonance sensing on naturally derived membranes: A remyelination-promoting human antibody binds myelin with extraordinary affinity Milan Vala, Luke R. Jordan, Arthur E. Warrington, L. James Maher III, Moses Rodriguez, Nathan J Wittenberg, and Sang-Hyun Oh Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02664 • Publication Date (Web): 19 Sep 2018 Downloaded from http://pubs.acs.org on September 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Surface plasmon resonance sensing on naturally derived membranes: A remyelination-promoting human antibody binds myelin with extraordinary affinity

Milan Vala,1 Luke R. Jordan,1,2 Arthur E. Warrington,*,3 L. James Maher III,4 Moses Rodriguez,3 Nathan J. Wittenberg,*1,2 and Sang-Hyun Oh*,1

1

Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, Minnesota 55455, United States

2

Department of Chemistry, Lehigh University, Bethlehem, Pennsylvania 18015, United States

3

Departments of Neurology and Immunology, Mayo Clinic, Rochester, Minnesota 55905, United States

4

Department of Biochemistry and Molecular Biology, Mayo Clinic, Rochester, Minnesota 55905, United States

*Address correspondence to: [email protected], [email protected], [email protected]

ACS Paragon Plus Environment

1

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

Abstract rHIgM22 is a recombinant human monoclonal IgM designed to promote remyelination, and it is currently in Phase I clinical trials in patients with multiple sclerosis (MS). In animal models of demyelination, a single low dose of rHIgM22 stimulates oligodendrocyte maturation, induces remyelination, preserves axons, and slows the decline of locomotor deficits. Natural autoantibodies like rHIgM22 typically bind to multiple antigens with weak affinity. rHIgM22 binds to oligodendrocytes and myelin. Because the antigen(s) for rHIgM22 is prevalent within and exclusive to central nervous system (CNS) myelin, we used CNS myelin particles in combination with surface plasmon resonance to determine the kinetic and affinity constants for the interaction of rHIgM22 to myelin. We found that both the serum and recombinant forms of the antibody bind to myelin with very small dissociation constants in the 100 pM range, which is highly unusual for natural autoantibodies. The extraordinary affinity between rHIgM22 and myelin may explain why such a low effective dose can stimulate CNS repair in animal models of demyelination and underlie the accumulation of rHIgM22 in the CSF in treated MS patients by targeting myelin. Keywords. Multiple sclerosis, myelin, natural membrane sensing, human antibody, surface plasmon resonance (SPR), binding kinetics. TOC graphic.


2

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Multiple Sclerosis (MS) is a disease of the central nervous system affecting 2 million people worldwide. The hallmark of MS is demyelination including progressive axonal loss and neuron death, resulting in neurological deficits such as vision loss, and paralysis 1. Even though there have been advancements in decreasing inflammatory exacerbations, there is no cure for MS. Stimulating myelin repair is a key therapeutic goal in the treatment of MS, and treatment with monoclonal antibodies that bind myelin and oligodendrocytes and initiate remyelination has emerged as one promising approach 2. In particular, it has been shown that human monoclonal antibodies that bind myelin and oligodendrocytes can initiate dramatic increase in remyelination in animal models of demyelination 2. One such antibody is recombinant human IgM22 (rHIgM22) 3,4. This antibody has completed a Phase 1a/b clinical trial in adults with both stable and active MS. This therapeutic approach has potential to impact the repair of damaged myelin in multiple diseases such as MS, neonatal white matter injury, stroke, and spinal cord and traumatic brain injury. To better understand the in vivo targeting capability of rHIgM22 it is important to characterize its binding kinetics and affinity to CNS myelin. Surface plasmon resonance (SPR), which is the ‘gold standard’ for measuring the affinity and kinetics of recombinant antibodies 5–7 has previously been used to determine the binding kinetics of mouse monoclonal IgM antibodies that bind to known glycolipids found in oligodendrocytes and myelin membranes. In these studies the binding of O4 and O1 antibodies, which bind the myelin glycolipids sulfatide and galactocerebroside respectively, were characterized 8. Both O4 and O1 bind to purified antigens in a supported lipid bilayer (SLB) membrane with KD values on the order of 10-9 M, which indicates surprisingly high affinity for natural autoantibodies, which typically have KD that range from 10-7 to 10-3 M 9–11. A human IgM


3


(rHIgM12) that binds GD1a and GT1b gangliosides 12 and PSA-NCAM

Page 4 of 25

13

on the surface of

neurons has also been characterized. This IgM promotes and guides neurite outgrowth

14

,

prolongs lifespan in mouse models of amyotrophic lateral sclerosis 12, and preserves locomotor activity in a mouse models of demyelination 15. rHIgM12 also displays especially tight binding for a natural autoantibody, with its KD ranging from 25 to 42 nM 12 for binding to GD1a and GT1b. Unlike with O1, O4, and rHIgM12, the antigen or antigens recognized by rHIgM22, are not known. However, genetic knockout studies indicate that this IgM likely binds a molecule(s) in the sulfatide synthesis pathway though not sulfatide itself. 16 Although a SLB on a biosensor surface is often a suitable system to study interaction of various biomolecules with lipids 17,18, here, due to the unknown nature of the antigen or antigens, we determine the binding kinetics of rHIgM22 to whole myelin by immobilizing mouse brain-derived myelin particles on the SPR sensor. Analysis of binding kinetics and affinity by SPR is traditionally accomplished using pairs of known analytes and receptors. This necessitates the purification and surface immobilization of the receptor molecules, both of which can be a challenge when dealing with membraneassociated species, even when the receptors are known. In the case of lipid receptors, SLBs, as discussed earlier, or liposomes have proved to be suitable schemes for surface immobilization 19– 26

. Intact proteoliposomes or SLBs derived from proteoliposomes can be employed when

membrane proteins are the receptors of interest. New methods for functionalization of surfaces with native cell-derived membranes have emerged recently, but they require unilamellar membrane vesicles

27,28

. Myelin vesicles are multilamellar structures with the lamellae held

together tightly by the proteins that maintain the multilamellar in vivo structure of myelin 29. The structure of native myelin, as well as the particles derived from it necessitate the development of new protocols presented here.


4

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


EXPERIMENTAL SECTION Reagents sHIgM22 and sHIgM39 are human serum IgM isotype antibodies identified during a screening of sera from patients with benign monoclonal gammopathies during an attempt to isolate monoclonal autoantibodies with remyelination-promoting capabilities 3. sHIgM22 binds to myelinated tracts in the CNS and mature oligodendrocytes and promotes remyelination in several murine models of demyelination. sHIgM39 does not bind to any mammalian cells or tissues and does not promote remyelination in vivo. It has been used as an isotype control throughout much of the preclinical development of rHIgM22. Both IgMs were isolated from serum using a 24 hour dialysis (14 kD cut off) against a 6 liter volume distilled water. Precipitated proteins trapped within the dialysis tubing were solubilized in normal saline. IgM was quantified by separating the resulting samples on denaturing and non-denaturing PAGE. rHIgM22 is the recombinant version of a human monoclonal IgM remyelination-promoting autoantibody that was derived from the serum of a patient with Waldenstӧm’s macroglobulinemia 3. The cell line was generated at Mayo Clinic in Rochester, Minnesota 30. rHIgM22 replicates the key characteristics of sHIgM22. It binds myelinated tracts and mature oligodendrocytes, and induces calcium flux in CNS glial cells. rHIgM22 contains human-derived heavy and light chains and a murine derived J chain. rHIgM22 was produced from the master cell bank (Cell Culture Company, Minneapolis, MN) in hollow fiber bioreactors. Purification to greater than 97% with minimal hexamers was accomplished using a 3 column process

31

.

Bioactivity potency assays were performed at Mayo Clinic.


5


Page 6 of 25

Myelin was isolated from pooled whole brains of adult C57BL6/J mice using a series of sucrose gradients, similar to the method described 32. Briefly, frozen brains were homogenized in lysis buffer (1.0 M Sucrose, 150 mM NaCl, 10 mM TES, pH 7.5), followed by centrifugation at 85,000 × g for 28 minutes, to collect the myelin fraction. The myelin fraction was homogenized again and loaded on top of a linear sucrose gradient followed by centrifugation at 85,000 × g for 17 hours. The myelin fraction was collected from between the 15 – 25% sucrose fractions, washed and resuspended in water. Myelin concentration was measured by using a BCA protein assay kit (ThermoFisher Scientific) and the presence of known myelin markers CNPase, MBP and MOG was verified by Western blot analysis. Myelin was stored at -80˚C. Phosphate buffered saline (PBS) buffer 10mM (NaCl 138mM, KCl 2.7mM, pH 7.4) was purchased from Sigma Aldrich, USA. SPR sensing A surface plasmon resonance sensor (Plasmon 6, Institute of Photonics and Electronics, Czech Republic) based on wavelength interrogation and attenuated total reflection geometry (Kretschmann configuration) was used for all presented SPR experiments. The liquid samples were pumped through the acrylic flow-cell with six independent microfluidic channels (1×0.06×5 mm3, width × height × length) in contact with the vertically oriented surface of the sensor chip using a peristaltic pump (IPC, Cole-Palmer GmbH, Germany). The temperature at the sensing area was stabilized at 25˚C. Fabrication of plasmonic chips Plasmonic chips were prepared by deposition of a 1.5 nm titanium adhesion layer and a 50 nm gold layer by thermal evaporation in vacuum on custom polished BK7 glass substrates (Schott, Malaysia). Then, 10 ± 1 nm of silica film was grown on top of gold film using atomic layer deposition (Savannah ALD system, Cambridge NanoTech, USA) from ozone precursor at 180˚C.


6

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


These chips were treated with O2 plasma (100 W, 2 minutes) to clean and oxidize the silica surface and allow efficient and reproducible binding of the myelin particles.

RESULTS AND DISCUSSION Binding kinetics of both the serum (sHIgM22) and recombinant (rHIgM22) forms of human IgM22, as well as a reference serum control antibody, sHIgM39, to myelin particles were measured repeatedly using a flow-through direct SPR assay (Figure 1a). Changes of the refractive index induced by molecular binding were probed by the surface plasmon propagating along the gold-dielectric interface. The penetration depth Lp (Figure 1a) of the propagating surface plasmon field into the probed dielectric is about 350 nm at λ=800 nm 33. Thus, the system is sensitive to events happening within a few hundred nanometers from the silica-coated gold surface. In the first step of the SPR experiments, 30 µg of myelin particles in 300 µL of PBS was flowed through each channel of the sensor flow-cell. The myelin particles likely have a multilamellar structure with a random distribution of extracellular and cytosolic membrane components on their surface 29. The flow-rate was set to 10 µL/min per channel. During this step, myelin particles were passively adsorbed on a hydrophilic silica surface 34 and stayed anchored after injection of clean PBS into the flow-cell (Figure 2). In all experiments, the SPR shift was ~20 nm after 30 min myelin immobilization step. When considering the refractive index of myelin lipid membrane to be 1.46 35 and the calculated (using a rigorous coupled wave analysisbased Matlab script 36) surface sensitivity at the silica-buffer interface about 29 nm/RIU per 1 nm of added continuous thin film, we can estimate the surface coverage of myelin to be about


7


Page 8 of 25

5.3 × 106 nm3/µm2. This is rather a lower estimate as myelin comprises not only lipid (about 7085% of the dry content and protein (between 15-30%), but also about 40% of water

37

.

Furthermore, the content of water in the myelin particles is not known and may be even higher as a multilamellar myelin particles might encapsulate more liquid than the myelin in its natural form. The silica layer deposited by ALD on top of the gold film was introduced for effective immobilization of myelin particles as they are composed mainly of lipids and silica was reported to have suitable properties to support lipid membrane adhesion to its surface

34

. We have also

measured the distribution of the myelin particles sizes in solution prior to injection to the sensor using dynamic light scattering (DLS; Nanotrac Flex, Microtrac, USA) (Figure 1b). However, in the microfluidic channel with laminar flow, the transport of particles to surface is diffusion limited and thus proportional to the diffusion coefficient D 38. In the case of a spherical particle, the diffusion coefficient is described by the Stokes-Einstein equation D=kT/6πηrp, i.e. reciprocally proportional to the particle radius rp. Here, kT represents thermal energy and η is the viscosity of the liquid. Using these assumptions, smaller particles will be more likely to reach the surface and the distribution of surface-bound particle sizes will be biased towards smaller particles. To account for this effect and estimate the size distribution of surface-bound particles, we have divided the relative count for each size group of the distribution measured by DLS in solution by the particle radius. According to this estimate, the size of the majority of particles captured on the sensor surface is within the 101-102 nm range (Figure 1c). This enables the observation of the IgM binding on top of the myelin-covered sensor surface by SPR as the average particle size is smaller than the surface plasmon penetration depth Lp. Detection of IgM molecules binding to myelin was performed shortly after immobilization of the myelin on the surface (Figure 2). The flow-rate was increased to 20 µL/min and PBS


8

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


solutions with different concentrations of antibodies (1, 3 and 10 nM) were flowed across the myelin-rich surface for 10 minutes. Then, the dissociation of IgM from myelin was observed for at least 20 minutes after injection of IgM-free PBS buffer. Sensor responses to 1-10 nM of sHIgM22 and rHIgM22 were measured in triplicate. Typical reference-compensated (response to pure PBS was subtracted from all datasets) sensorgrams for both types of HIgM22 are depicted in Figure 2. As a positive control we also injected the O4 antibody over myelin particles, and it strongly bound indicating the presence of sulfatide on the surface of the myelin particles (Figure S1). In SPR biosensors, a real-time observation of amount of molecules bound to or detached from the sensor surface is possible. This enables observation of binding kinetics and calculation of rate constants by fitting the SPR sensorgrams with suitable kinetic models

39

. When the

receptors in solution interact reversibly with the immobilized ligand through a single binding site (1:1 interaction) and the mass transport effects are neglected, the association and dissociation phases of the sensorgram, RA and RD, respectively, follows simple exponential behavior 40: RA (t ) =

{

}

Ckon Rmax 1 − exp  −(Ckon + koff )t  Ckon + koff

,

RD (t ) = R 0 exp (− koff (t − t0 ) 

(1) (2)

Here, Rmax is the maximum SPR signal obtained when all ligands on the surface are paired with their corresponding receptor molecule, R0 is the level of SPR signal at t0 when the solution with receptors (concentration C) is replaced with a receptor-free buffer (C=0) and kon and koff are on- and off-rate constants, respectively. An IgM molecule is a pentamer with 10 binding sites capable of interactions with multiple antigens. Furthermore, rHIgM22 binds smoothly over the entire surface of oligodendrocytes rather than in punctate spots 4. This suggests that the antigen is distributed over the entire cell


9


Page 10 of 25

surface with relative uniformity. While the diffusive mobility of the antigen in the myelin membrane is unknown, its high density on the cell surface suggests that a single antibody will have ample opportunity to bind multiple antigens. The combined strength of more than one interaction (avidity) has been reported to affect the shape of the measured binding curves 8,41. This is especially true for the dissociation phase of the sensorgram, where the mean bound lifetime of molecules attached through a single binding site will be shorter than for those attached through multiple interactions 42,43. For multiple-bound molecules all binding interactions have to dissociate in order to be released from the surface. This heterogeneity in number of interactions between IgM molecule and accessible antigens is manifested as rapid dissociation of weakly-bound molecules shortly after injection of IgM-free buffer into the flow-cell, followed by a slower dissociation phase corresponding to detachment of more stable multiple-bound IgMs. Indeed, the monophasic (1:1) model described by Eq. 2 was found not to follow the dissociation phase data well (Figure S2) suggesting a more complex kinetic model has to be used. We have employed a biphasic model

42

to discriminate between slow and fast components of the

dissociation: RD (t ) = R 0, slow exp (−koff ,slow (t − t0 ) + R 0, fast exp (−koff , fast (t − t0 )

(3)

As an approximation, we assume only two groups of interacting molecules: the first group consisting of IgMs interacting through a single binding site and a second group encompassing all the more stable cases, i.e. bivalent or multivalent interactions. As the multiply-bound molecules are expected to stay on the sensor surface for much longer time than those captured through a single interaction, these two groups will be represented by distinct rate constants, i.e. koff,slow (multiple-bound) and koff,fast (single-bound), Eq. 3. Also it is possible to determine the relative size of these two groups at t=t0 by comparing the amplitudes R0,slow and R0,fast. This analysis


10

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


revealed that majority of captured molecules at t=t0 (10 minutes after injection of IgMs) are multiple-bound and only (18±6) percent of sHIgM22 and (27±10) percent of rHIgM22 molecules are captured through single interaction. In the association phase on the other hand, the SPR signal is dominated by association events where the single and multiple-bound molecules will produce the same local change of refractive index translated into the same level of SPR response and thus are indistinguishable from each other. Assuming this, we believe it is reasonable to fit the association phase of the SPR data with the monophasic (1:1) interaction model described as RA in Eq. 1. Although both kon and koff rate constants could be in principle obtained from this fit, it is more robust to fit the dissociation phase of the sensorgram first, determine the koff and insert it into the equation used to fit the association phase. Another approach to describe multivalent interactions was published recently

44

. In that work, bivalent antibodies interacting with antigens in the vicinity of solid

nanopores are also separated into two groups – strongly and weakly bound. The transitions between unbound, strongly and weakly bound states are then modeled using differential equations by two distinct models, sequential and parallel and the avidity-related features observed in the binding kinetic curves are successfully reproduced in the fitted data by introducing additional parameter n to the equations. The comparison of dissociation constant of monovalent and multivalent ligands across literature is not straightforward due to the effect of avidity and different kinetic models employed to determine the rate constants. Depending on the molecular system under study, different definitions of dissociation constants are used to describe the interaction of molecules exhibiting avidity

45

. Here, we employ a model that allows us to

determine the dissociation constant of multiple-bound molecules from the measured SPR data and we define the dissociation constant of multiple-bound molecules as KD,M-B=koff,slow/kon. As


11


Page 12 of 25

was shown in the previous paragraph, the majority of IgMs were bound to more than one antigen presented in surface-immobilized myelin. We may speculate that in vivo, the diffusion of both the antigen in myelin and IgM would not be limited by the surface and even more antigens would be accessible to bind to IgM resulting in stronger multivalent interactions. For this reason we consider multivalent dissociation constant KD,M-B with koff,slow to be more suitable to describe the combined strength of multiple interaction than koff,fast, which would describe affinity of a single interaction

45

. Slow and fast off-rates calculated using the biphasic model from the

dissociation phase, on-rate constants determined from the association phase of the sensorgrams and equilibrium constants are summarized (Table 1). All rate constants were determined as a mean value from three independent SPR experiments. In each experiment, sensorgrams for three concentrations (1 nM, 3nM and 10nM) were measured and obtained data was globally fitted using Eq.3 (dissociation phase) and Eq.1 (association phase), see Figure 3. Binding of HIgM22 to surface-immobilized myelin particles was corroborated by immunoassays (Figure 4). The rHIgM22 binds strongly to the central white matter of the cerebellar folia (Figure 4A), and at higher magnification individual myelinated fibers and oligodendrocytes are visible (Figure 4B). Reference mAb sHIgM39 does not bind to any CNS structure (Figure 4C) confirming the specificity of the interaction of the HIgM22 with the CNS myelin 14. Specificity of the HIgM22-myelin binding was verified also using SPR (Figure 5). The SPR signal induced by reference antibody sHIgM39 was measured three times using the same concentrations (1-10 nM) of IgM and identically prepared myelin-coated surfaces as in experiments with sHIgM22 and rHIgM22. The average value of the maximum SPR signal (wavelength shift) at t = 10 min from these three experiments was 0.12 nm for 10 nM sHIgM39,


12

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


which is less than 20-times lower than both types of HIgM22. Furthermore, series of experiments where all three types of IgM molecules were flowed across the silica surface without myelin were performed to assess the non-specific binding to silica. For all three molecules, the maximum response was larger than 4 nm (Figure 4). In particular, for the reference molecule sHIgM39, the difference between binding to myelin-coated surface vs. bare silica surface is 0.12 nm vs. 6.2 nm, respectively. This clearly shows, that in our experiments, 30 µg of myelin particles per channel was sufficient to cover the silica surface and shield it from non-specific interaction with IgM and the response we observed originates from the interaction with myelin and no additional passivation or blocking of the exposed silica surface was needed. CONCLUSIONS In summary, we showed that HIgM22 in both the serum-derived and recombinant forms binds myelin with extraordinary affinity (KD in 100 pM range). The strong binding between rHIgM22 and CNS myelin may explain how the IgM can be therapeutic using a single low dose (25 µg/kg) in animal models of neurodegenerative diseases 4. rHIgM22 is presently in clinical trials for safety in humans with multiple sclerosis. A Phase I single escalating dose trial evaluated the tolerability of rHIgM22 in 72 adult MS patients with chronic disease (NCT01803867). No major adverse effects were reported and the self-reported patient global impression of change trended toward improvement. rHIgM22 was detected by mass spectrometry in the cerebrospinal fluid (CSF) in patients enrolled in the study, definitively demonstrating that the IgM crossed the blood brain barrier. In addition, analysis of CSF one month after dosing detected the accumulation of the antibody in relation to the serum, demonstrating a very long half-life in the central nervous system

46

. The binding kinetic data presented in this work supports the hypothesis that the

surprisingly high affinity of rHIgM22 to myelin, compared to what is described in the literature


13


Page 14 of 25

for natural antibodies, is one aspect of its therapeutic efficacy. On the technology side, our SPR protocols can benefit other studies involving membrane-bound receptors

47

. While liposome

capture chips for commercial SPR instruments typically rely on proprietary carboxymethylated dextran surface coatings 7, here we show that for vesicles that do not rupture, which is the case for many cell-derived vesicles, a simple one-step physisorption scheme with silica surfaces works well and could be broadly applied to other systems. Although the detection of IgM bound on top of myelin-covered planar SPR surface was possible in this study due to the small size of the myelin particles, utilization of planarized natural membranes 28 with nanoplasmonic sensors with more advanced architectures 48–53 and possibly higher sensitivity is anticipated for future work. ASSOCIATED CONTENT Supporting Information Additional information about binding of positive control O4 antibody to myelin and discussion about the selected kinetic models used to determine the rate constants can be found in Supporting Information. The Supporting Information is available free of charge on the ACS Publications website.

AUTHOR INFORMATION Corresponding Authors *E-mail: [email protected], [email protected], and [email protected]

Conflict of interest


14

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Mayo Clinic has been issued patents for the use of rHIgM22 in central nervous system disorders such as multiple sclerosis. Therefore if the antibody proves to be effective, then some of the authors as well as the Mayo Clinic may receive royalties from this discovery.

ACKNOWLEDGMENTS This work was supported by grants from the NIH (Grant No. R01 GM092993 to A.E.W., M.R., S.-H.O.) and Minnesota Partnership for Biotechnology and Medical Genomics (M.V., L.R.J., N.J.W., L.J.M., A.E.W., M.R., S.-H.O.). N.J.W. acknowledges support from Lehigh University. S.-H.O. further acknowledges the Sanford P. Bordeau Endowed Chair in Electrical Engineering.


15


Page 16 of 25

Figure 1 (a) Scheme of the SPR detection of IgM antibodies binding to mouse myelin-derived particles immobilized on a gold and silica-coated glass chip. Binding of IgM causes a change of the local refractive index and shift of spectral dip in the spectrum of reflected light. Size distribution of the suspension of myelin particles in PBS solution measured by dynamic light scattering (b) and estimated distribution of particles immobilized on the sensor surface (c).


16

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2 Typical SPR sensorgram showing immobilization of myelin particles (0.1 mg/ml in 10 mM PBS) followed by binding of recombinant IgM 22 molecules at four different concentrations (1nM, 3nM, 10nM and 30nM) in four parallel sensor channels.


17


Page 18 of 25

Figure 3 Typical SPR sensor response (solid black lines) to injection of three different concentrations of sHIgM22 (top) and rHIgM22 (bottom) antibodies into the flow-cell at t=0 min followed by pure PBS buffer at t=10 min. Interaction between IgM and myelin was modeled using monophasic and biphasic models for association and dissociation phases, respectively (dashed green lines).


18

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4 Slices of unfixed cerebellum from 8 week old SJL mice immunolabeled with 10 µg/ml rHIgM22 (A, B) and sHIgM39 (C). Scale bar A: 200 µm, B and C: 100 µm.


19


Page 20 of 25

Figure 5 Comparison of the binding of reference sHIgM39 to myelin and non-specific binding to silica surface without myelin (red lines). Non-specific binding of serum and recombinant HIgM22 to myelin-free silica is shown for comparison (green and black lines). IgM concentration was 10 nM in all experiments.


20

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1 Calculated rate and equilibrium constants of IgM antibodies binding to myelin particles. * mean value from 3 independent experiments ± standard deviation.

1:1 model kon (M-1s-1) * rHIg M22 sHIg M22

2:1 model koff,slow (s-1) *

koff,fast (s-1) *

KD,M-B (nM)

(3.84±2.02)×10-5

(4.37±1.82)×10-3

0.22±0.13

(4.81±1.73)×10-5

(5.30±0.85)×10-3

0.21±0.10

(1.78±0.46)×1 05 (2.64±0.83)×1 0

5


21


Page 22 of 25

REFERENCES (1)

Noseworthy, J. H.; Lucchinetti, C.; Rodriguez, M.; Weinshenker, B. G. Multiple Sclerosis. N. Engl. J. Med. 2000, 343, 938–952.

(2)

Wootla, B.; Watzlawik, J. O.; Stavropoulos, N.; Wittenberg, N. J.; Dasari, H.; Abdelrahim, M. A.; Henley, J. R.; Oh, S.-H.; Warrington, A. E.; Rodriguez, M. Recent Advances in Monoclonal Antibody Therapies for Multiple Sclerosis. Expert Opin. Biol. Ther. 2016, 16, 827–839.

(3)

Warrington, A. E.; Asakura, K.; Bieber, A. J.; Ciric, B.; Van Keulen, V.; Kaveri, S. V.; Kyle, R. A.; Pease, L. R.; Rodriguez, M. Human Monoclonal Antibodies Reactive to Oligodendrocytes Promote Remyelination in a Model of Multiple Sclerosis. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 6820–6825.

(4)

Warrington, A. E.; Bieber, A. J.; Ciric, B.; Pease, L. R.; Van Keulen, V.; Rodriguez, M. A Recombinant Human IgM Promotes Myelin Repair after a Single, Very Low Dose. J. Neurosci. Res. 2007, 85, 967–976.

(5)

Homola, J. Surface Plasmon Resonance Sensors for Detection of Chemical and Biological Species. Chem. Rev. 2008, 108, 462–493.

(6)

Maynard, J. A.; Maassen, C. B. M.; Leppla, S. H.; Brasky, K.; Patterson, J. L.; Iverson, B. L.; Georgiou, G. Protection against Anthrax Toxin by Recombinant Antibody Fragments Correlates with Antigen Affinity. Nat. Biotechnol. 2002, 20, 597–601.

(7)

Maynard, J. A.; Lindquist, N. C.; Sutherland, J. N.; Lesuffleur, A.; Warrington, A. E.; Rodriguez, M.; Oh, S.-H. Surface Plasmon Resonance for High-Throughput Ligand Screening of Membrane-Bound Proteins. Biotechnology Journal 2009, 4, 1542–1558.

(8)

Wittenberg, N. J.; Im, H.; Xu, X.; Wootla, B.; Watzlawik, J.; Warrington, A. E.; Rodriguez, M.; Oh, S.-H. High-Affinity Binding of Remyelinating Natural Autoantibodies to Myelin-Mimicking Lipid Bilayers Revealed by Nanohole Surface Plasmon Resonance. Anal. Chem. 2012, 84, 6031–6039.

(9)

Notkins, A. L. Polyreactivity of Antibody Molecules. Trends Immunol. 2004, 25, 174– 179.

(10) Zhou, Z.-H.; Tzioufas, A. G.; Notkins, A. L. Properties and Function of Polyreactive Antibodies and Polyreactive Antigen-Binding B Cells. J. Autoimmun. 2007, 29, 219–228. (11) Lutz, H. U.; Binder, C. J.; Kaveri, S. Naturally Occurring Auto-Antibodies in Homeostasis and Disease. Trends Immunol. 2009, 30, 43–51. (12) Xu, X.; Denic, A.; Jordan, L. R.; Wittenberg, N. J.; Warrington, A. E.; Wootla, B.; Papke, L. M.; Zoecklein, L. J.; Yoo, D.; Shaver, J.; et al. A Natural Human IgM That Binds to Gangliosides Is Therapeutic in Murine Models of Amyotrophic Lateral Sclerosis. Dis. Model. Mech. 2015, 8, 831–842. (13) Watzlawik, J. O.; Kahoud, R. J.; Ng, S.; Painter, M. M.; Papke, L. M.; Zoecklein, L.; Wootla, B.; Warrington, A. E.; Carey, W. A.; Rodriguez, M. Polysialic Acid as an Antigen for Monoclonal Antibody HIgM12 to Treat Multiple Sclerosis and Other Neurodegenerative Disorders. J. Neurochem. 2015, 134, 865–878.


22

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(14) Warrington, A. E.; Bieber, A. J.; Van Keulen, V.; Ciric, B.; Pease, L. R.; Rodriguez, M. Neuron-Binding Human Monoclonal Antibodies Support Central Nervous System Neurite Extension. J. Neuropathol. Exp. Neurol. 2004, 63, 461–473. (15) Denic, A.; Macura, S. I.; Warrington, A. E.; Pirko, I.; Grossardt, B. R.; Pease, L. R.; Rodriguez, M. A Single Dose of Neuron-Binding Human Monoclonal Antibody Improves Spontaneous Activity in a Murine Model of Demyelination. PLoS One 2011, 6 (10) e26001. (16) Wright, B. R.; Warrington, A. E.; Edberg, D. D.; Edberg, D. E.; Rodriguez, M. Cellular Mechanisms of Central Nervous System Repair by Natural Autoreactive Monoclonal Antibodies. Arch. Neurol. 2009, 66, 1456–1459. (17) Castellana, E. T.; Cremer, P. S. Solid Supported Lipid Bilayers: From Biophysical Studies to Sensor Design. Surf. Sci. Rep. 2006, 61, 429–444. (18) Richter, R. P.; Bérat, R.; Brisson, A. R. Formation of Solid-Supported Lipid Bilayers: An Integrated View. Langmuir 2006, 22, 3497–3505. (19) Bieri, C.; Ernst, O. P.; Heyse, S.; Hofmann, K. P.; Vogel, H. Micropatterned Immobilization of a G Protein-Coupled Receptor and Direct Detection of G Protein Activation. Nat. Biotechnol. 1999, 17, 1105–1108. (20) Dahlin, A.; Zäch, M.; Rindzevicius, T.; Käll, M.; Sutherland, D. S.; Höök, F. Localized Surface Plasmon Resonance Sensing of Lipid-Membrane-Mediated Biorecognition Events. J. Am. Chem. Soc. 2005, 127, 5043–5048. (21) Jonsson, M. P.; Jönsson, P.; Dahlin, A. B.; Höök, F. Supported Lipid Bilayer Formation and Lipid-Membrane-Mediated Biorecognition Reactions Studied with a New Nanoplasmonic Sensor Template. Nano Lett. 2007, 7, 3462–3468. (22) Im, H.; Wittenberg, N. J.; Lesuffleur, A.; Lindquist, N. C.; Oh, S.-H. Membrane Protein Biosensing with Plasmonic Nanopore Arrays and Pore-Spanning Lipid Membranes. Chem. Sci. 2010, 1, 688–696. (23) Im, H.; Lee, S. H.; Wittenberg, N. J.; Johnson, T. W.; Lindquist, N. C.; Nagpal, P.; Norris, D. J.; Oh, S.-H. Template-Stripped Smooth Ag Nanohole Arrays with Silica Shells for Surface Plasmon Resonance Biosensing. ACS Nano 2011, 5, 6244–6253. (24) Bruzas, I.; Unser, S.; Yazdi, S.; Ringe, E.; Sagle, L. Ultrasensitive Plasmonic Platform for Label-Free Detection of Membrane-Associated Species. Anal. Chem. 2016, 88, 7968– 7974. (25) Granéli, A.; Benkoski, J. J.; Höök, F. Characterization of a Proton Pumping Transmembrane Protein Incorporated into a Supported Three-Dimensional Matrix of Proteoliposomes. Anal. Biochem. 2007, 367, 87–94. (26) Granéli, A.; Rydström, J.; Kasemo, B.; Höök, F. Formation of Supported Lipid Bilayer Membranes on SiO2 from Proteoliposomes Containing Transmembrane Proteins. Langmuir 2003, 19, 842–850. (27) Hsia, C.-Y.; Chen, L.; Singh, R. R.; DeLisa, M. P.; Daniel, S. A Molecularly Complete Planar Bacterial Outer Membrane Platform. Sci. Rep. 2016, 6, 32715.


23


Page 24 of 25

(28) Liu, H.-Y.; Grant, H.; Hsu, H.-L.; Sorkin, R.; Bošković, F.; Wuite, G.; Daniel, S. Supported Planar Mammalian Membranes as Models of in Vivo Cell Surface Architectures. ACS Appl. Mater. Interfaces 2017, 9, 35526–35538. (29) Sedzik, J.; Blaurock, A. E. Myelin Vesicles: What We Know and What We Do Not Know. J. Neurosci. Res. 1995, 41, 145–152. (30) Mitsunaga, Y.; Ciric, B.; Van Keulen, V.; Warrington, A. E.; Paz Soldan, M.; Bieber, A. J.; Rodriguez, M.; Pease, L. R. Direct Evidence That a Human Antibody Derived from Patient Serum Can Promote Myelin Repair in a Mouse Model of Chronic-Progressive Demyelinating Disease. FASEB J. 2002, 16, 1325–1327. (31) Wootla, B.; Denic, A.; Warrington, A. E.; Rodriguez, M. A Monoclonal Natural Human IgM Protects Axons in the Absence of Remyelination. J. Neuroinflammation 2016, 13, 94. (32) Norton, W. T.; Poduslo, S. E. Myelination in Rat Brain: Method of Myelin Isolation. J. Neurochem. 1973, 21, 749–757. (33) Homola, J.; Piliarik, M. Surface Plasmon Resonance (SPR) Sensors. In Surface Plasmon Resonance Based Sensors; Springer Series on Chemical Sensors and Biosensors; Springer, Berlin, Heidelberg, 2006; pp 45–67. doi: 10.1007/b100321 (34) Cremer, P. S.; Boxer, S. G. Formation and Spreading of Lipid Bilayers on Planar Glass Supports. J. Phys. Chem. B 1999, 103, 2554–2559. (35) Pusterla, J. M.; Malfatti-Gasperini, A. A.; Puentes-Martinez, X. E.; Cavalcanti, L. P.; Oliveira, R. G. Refractive Index and Thickness Determination in Langmuir Monolayers of Myelin Lipids. Biochim. Biophys. Acta - Biomembranes 2017, 1859, 924–930. (36) Kwiecien, P.; Richter, I.; Čtyroký, J. RCWA/ARCWA - An Efficient and Diligent Workhorse for Nanophotonic/Nanoplasmonic Simulations - Can It Work Even Better? In 2015 17th International Conference on Transparent Optical Networks (ICTON); 2015; pp 1–8. (37) Morell, P.; Quarles, R. H. Characteristic Composition of Myelin. Basic Neurochemistry: Molecular, Cellular and Medical Aspects. 6th edition 1999. (38) Dahlin, A. B. Plasmonic Biosensors: An Integrated View of Refractometric Detection; Ios Press, 2012. doi: 10.3233/978-1-60750-966-0-i (39) Štěpánek, J.; Vaisocherová, H.; Piliarik, M. Molecular Interactions in SPR Sensors. In Surface Plasmon Resonance Based Sensors; Springer Series on Chemical Sensors and Biosensors; Springer, Berlin, Heidelberg, 2006; pp 69–91. (40) O’Shannessy, D. J.; Brigham-Burke, M.; Karl Soneson, K.; Hensley, P.; Brooks, I. [15] Determination of Rate and Equilibrium Binding Constants for Macromolecular Interactions by Surface Plasmon Resonance. In Methods in Enzymology; Part B: Numerical Computer Methods; Academic Press, 1994; Vol. 240, pp 323–349. (41) Munoz, E. M.; Correa, J.; Riguera, R.; Fernandez-Megia, E. Real-Time Evaluation of Binding Mechanisms in Multivalent Interactions: A Surface Plasmon Resonance Kinetic Approach. J. Am. Chem. Soc. 2013, 135, 5966–5969.


24

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(42) Cooper, M. A.; Williams, D. H. Kinetic Analysis of Antibody-Antigen Interactions at a Supported Lipid Monolayer. Anal. Biochem. 1999, 276, 36–47. (43) Bally, M.; Gunnarsson, A.; Svensson, L.; Larson, G.; Zhdanov, V. P.; Höök, F. Interaction of Single Viruslike Particles with Vesicles Containing Glycosphingolipids. Phys. Rev. Lett. 2011, 107, 188103. (44) Emilsson, G.; Sakiyama, Y.; Malekian, B.; Xiong, K.; Adali-Kaya, Z.; Lim, R. Y. H.; Dahlin, A. B. Gating Protein Transport in Solid State Nanopores by Single Molecule Recognition. ACS Cent. Sci., 2018, 4, 1007–1014. (45) Krishnamurthy, V. M.; Estroff, L. A.; Whitesides, G. M. Multivalency in Ligand Design. In Fragment-based Approaches in Drug Discovery; Wiley-Blackwell, 2006; pp 11–53. ISBN: 978-3-527-31291-7 (46) Eisen, A.; Greenberg, B. M.; Bowen, J. D.; Arnold, D. L.; Caggiano, A. O. A doubleblind, placebo-controlled, single ascending-dose study of remyelinating antibody rHIgM22 in people with multiple sclerosis. Mult. Scler. J. Exp. Transl. Clin. 2017, 1-11. DOI: 10.1177/ 2055217317743097 (47) Cooper, M. A. Advances in Membrane Receptor Screening and Analysis. J. Mol. Recognit. 2004, 17, 286–315. (48) Escobedo, C.; Brolo, A. G.; Gordon, R.; Sinton, D. Optofluidic Concentration: Plasmonic Nanostructure as Concentrator and Sensor. Nano Lett. 2012, 12, 1592–1596. (49) Junesch, J.; Sannomiya, T.; Dahlin, A. B. Optical Properties of Nanohole Arrays in MetalDielectric Double Films Prepared by Mask-on-Metal Colloidal Lithography. ACS Nano 2012, 6, 10405–10415. (50) Brolo, A. G. Plasmonics for Future Biosensors. Nat. Photon. 2012, 6, 709–713. (51) Jackman, J. A.; Ferhan, A. R.; Cho, N.-J. Nanoplasmonic Sensors for Biointerfacial Science. Chem. Soc. Rev. 2017, 46, 3615–3660. (52) Li, X.; Soler, M.; Szydzik, C.; Khoshmanesh, K.; Schmidt, J.; Coukos, G.; Mitchell, A.; Altug, H. Label-Free Optofluidic Nanobiosensor Enables Real-Time Analysis of SingleCell Cytokine Secretion. Small 2018, 14, 1800698. (53) Ferhan, A. R.; Jackman, J. A.; Malekian, B.; Xiong, K.; Emilsson, G.; Park, S.; Dahlin, A. B.; Cho, N.-J. Nanoplasmonic Sensing Architectures for Decoding Membrane CurvatureDependent Biomacromolecular Interactions. Anal. Chem. 2018, 90, 7458–7466.


25

A Remyelination-Promoting Human Antibody ... - ACS Publications

Recommend Documents